MR imaging of the abdomen and pelvis 9783131455918, 3131455918

Citation preview

MR Imaging of the Abdomen and Pelvis Bernd Hamm Gabriel Paul Krestin Michael Laniado Volkmar Nicolas Matthias Taupitz

(tThieme

MR Imaging of the Abdomen and Pelvis Bernd Hamm, MD

Professor of Radiology and Chairman Department of Radiology, Campus Mitte Clinic of Radiology, Campus Virchow-Klinikum Charité – Universitätsmedizin Berlin Berlin, Germany

Gabriel Paul Krestin, MD

Professor of Radiology and Chairman Department of Radiology Erasmus MC Rotterdam, the Netherlands

Volkmar Nicolas, MD

Professor of Radiology and Chairman Department of Diagnostic and Interventional Radiology and Nuclear Medicine Berufsgenossenschaftliches University Hospital Bergmannsheil Bochum, Germany

Matthias Taupitz, MD

Professor of Radiology and Physicist Department of Radiology Charité – Universitätsmedizin Berlin Berlin, Germany

Michael Laniado, MD

Professor of Radiology and Chairman Department of Diagnostic Radiology Carl Gustav Carus University Hospital Technical University of Dresden Dresden, Germany

With contributions by P. Asbach, D. Beyersdorf, F. Dammann, H.B. Gehl, C.R. Habermann, B. Hamm, C.M. Heyer, C. Klessen, C. Kluener, G.P. Krestin, G. Krupski-Berdien, R.A. Kubik-Huch, M. Laniado, M. Lorenzen, W. Luboldt, A.E. Mahfouz, U.G. Mueller-Lisse, M.R. Muehler, V. Nicolas, W. Pennekamp, P. Reimer, M. Reuter, B. Stoever, M. Taupitz, R. Vosshenrich 1060 illustrations

Thieme Stuttgart · New York

IV Library of Congress Cataloging-in-Publication Data MRT von Abdomen und Becken. English. MR imaging of the abdomen and pelvis / Bernd Hamm ... [et al.] ; with contributions by P. Asbach ... [et al.] ; [translator, Bettina Herwig]. p. ; cm. Includes bibliographical references and index. ISBN 978-3-13-145591-8 (alk. paper) 1. Abdomen--Magnetic resonance imaging. 2. Pelvis--Magnetic resonance imaging. I. Hamm, Bernd, Prof. Dr. II. Title. [DNLM: 1. Abdomen--pathology. 2. Magnetic Resonance Imaging. 3. Pelvis--pathology. WI 900 M9395 2009a] RC944.M6613 2009 617.5'507548--dc22 2009018513 This book is an authorized and revised translation of the 2nd German edition published and copyrighted 2007 by Georg Thieme Verlag, Stuttgart, Germany. Title of the German edition: MRT von Abdomen und Becken Translator: Bettina Herwig, Berlin, Germany Illustrators: Stefanie Gay and Bert Sender, Bremen, Germany

© 2010 Georg Thieme Verlag, Rüdigerstrasse 14, 70469 Stuttgart, Germany http://www.thieme.de Thieme New York, 333 Seventh Avenue, New York, NY 10001, USA http://www.thieme.com Cover design: Thieme Publishing Group Typesetting by primustype Hurler, Notzingen, Germany Printed in Germany by APPL aprinta druck, Wemding ISBN 978-3-13-145591-8

123456

Important note: Medicine is an ever-changing science undergoing continual development. Research and clinical experience are continually expanding our knowledge, in particular our knowledge of proper treatment and drug therapy. Insofar as this book mentions any dosage or application, readers may rest assured that the authors, editors, and publishers have made every effort to ensure that such references are in accordance with the state of knowledge at the time of production of the book. Nevertheless, this does not involve, imply, or express any guarantee or responsibility on the part of the publishers in respect to any dosage instructions and forms of applications stated in the book. Every user is requested to examine carefully the manufacturers’ leaflets accompanying each drug and to check, if necessary in consultation with a physician or specialist, whether the dosage schedules mentioned therein or the contraindications stated by the manufacturers differ from the statements made in the present book. Such examination is particularly important with drugs that are either rarely used or have been newly released on the market. Every dosage schedule or every form of application used is entirely at the user’s own risk and responsibility. The authors and publishers request every user to report to the publishers any discrepancies or inaccuracies noticed. If errors in this work are found after publication, errata will be posted at www.thieme.com on the product description page.

Some of the product names, patents, and registered designs referred to in this book are in fact registered trademarks or proprietary names even though specific reference to this fact is not always made in the text. Therefore, the appearance of a name without designation as proprietary is not to be construed as a representation by the publisher that it is in the public domain. This book, including all parts thereof, is legally protected by copyright. Any use, exploitation, or commercialization outside the narrow limits set by copyright legislation, without the publisher’s consent, is illegal and liable to prosecution. This applies in particular to photostat reproduction, copying, mimeographing, preparation of microfilms, and electronic data processing and storage.

V

Preface

Magnetic resonance imaging has developed dramatically in recent years, and the technical advances have expanded the range of diagnostic applications. In the abdomen and pelvis, MRI has long developed beyond the experimental stage and is now the method of choice for many indications. With increasing experience, refined and highly specialized imaging techniques have emerged for the different organ systems, making MRI of the abdomen and pelvis a complex field of its own. This book provides a comprehensive overview of the technically demanding, interesting, and clinically relevant field of abdominopelvic MRI. Each organ or organ system is treated in a separate chapter that can be read independently and begins with a concise description of the specific imaging technique. The individual chapters can be read as a cookbook with regard to the specific imaging technique and indications for MRI. The presentation of pathologic entities is lavishly illustrated and is completed by a description of the MRI findings and differential diagnoses. Unique to MRI, especially of the hepatobiliary system, is the availability of a large number of contrast agents. MRI of the liver can be performed with different nonspecific contrast media, which are available all over the world, or with one of several tissue-specific agents. Readers should

be aware that the specific contrast media may not have been approved for all countries or continents and that their approval status is subject to change. Working on this English edition has made us aware of how difficult it is to keep abreast of technical developments in such a dynamic field as MRI. Thus, new insights at the time of publication might already have outpaced the state of the art at the time of writing. Many radiologists directly involved in the scientific development of abdominal and pelvic MRI, either in cooperation with us or as independent researchers, have contributed to this book as authors or co-authors. We thank all of them for their dedicated, supportive, and inspiring cooperation. Our thanks also go to the staff of Thieme Publishers for making possible this English edition and to Bettina Herwig for the translation. We hope that this book will be a useful resource to our colleagues interested in abdominal and pelvic MRI and help them make use of MRI to the benefit of our patients. B. Hamm G.P. Krestin M. Laniado V. Nicolas M. Taupitz

VI

Contributors

Patrick Asbach, MD Assistant Professor of Radiology Department of Radiology Charité – Universitätsmedizin Berlin Berlin, Germany Dirk Beyersdorff, MD Associate Professor Radiology Department of Radiology Charité – Universitätsmedizin Berlin Berlin, Germany Florian Dammann, MD Professor of Radiology Department of Radiology Klinik am Eichert Göppingen, Germany Hans-Bjoern Gehl, MD Professor of Radiology Department of Radiology Bielefeld-Mitte City Hospitals Bielefeld, Germany Christian R. Habermann, MD Associate Professor of Radiology Diagnostic Center, Department of Diagnostic and Interventional Radiology University Medical Center Hamburg-Eppendorf Hamburg, Germany Christoph M. Heyer, MD Associate Professor of Radiology Department of Diagnostic and Interventional Radiology and Nuclear Medicine Berufsgenossenschaftliches University Hospital Bergmannsheil Bochum, Germany Christian Klessen, MD Center for Diagnostic Radiology and Minimally Invasive Therapy The Jewish Hospital Berlin Berlin, Germany

Claudia Kluener, MD Department of Radiology and Neuroradiology Oldenburg Evangelical Hospital Oldenburg, Germany Gerrit Krupski-Berdien, MD Professor of Radiology Department of Diagnostic and Interventional Radiology Reinbek St. Adolf-Stift Hospital Reinbek, Germany Rahel A. Kubik-Huch, MD Professor of Radiology Department of Radiology Kantonsspital Baden AG Baden, Switzerland Maren Lorenzen, MD Radiology Practice Heegbarg Hamburg, Germany Wolfgang Luboldt, MD, MSc Associate Professor of Radiology and Nuclear Medicine Physicist Department of Radiology University Hospital Frankfurt Frankfurt a. M., Germany Ahmed-Emad Mahfouz, MD Professor of Radiology Hamad Medical Corporation Radiology Department Doha, Quatar Matthias R. Muehler, MD Assistant Professor of Radiology Department of Radiology Charité – Universitätsmedizin Berlin Berlin, Germany Ullrich Gerd Mueller-Lisse, MD Associate Professor of Radiology Department of Clinical Radiology Munich University Hospital, Innenstadt Munich, Germany

Contributors

Werner Pennekamp, MD Assistant Professor of Radiology Department of Diagnostic and Interventional Radiology and Nuclear Medicine Berufsgenossenschaftliches University Hospital Bergmannsheil Bochum, Germany Peter Reimer, MD Professor of Radiology Department of Radiology Karlsruhe Municipal Hospital Academic Teaching Hospital of the University of Freiburg Karlsruhe, Germany

VII

Michael Reuter, MD Professor of Radiology Department of Radiology and Interventional Treatment Vivantes Hospital Neukölln Berlin, Germany Brigitte Stoever, MD Professor of Radiology Department of Pediatric Radiology Charité – Universitätsmedizin Berlin Berlin, Germany Rolf Vosshenrich, MD Professor of Radiology MRI Practice Friederikenstift Hanover, Germany

VIII

Abbreviations

1D, 2D, 3D

One-dimensional, two-dimensional, threedimensional

ACTH ADPKD

Adrenocorticotropic hormone Autosomal dominant polycystic kidney disease Array spatial sensitivity encoding technique

ASSET

BPH BCS

Benign prostatic hyperplasia Budd–Chiari syndrome

CCA CCC CNR CSF CSI CT CTA CTAP D&C

Cholangiocarcinoma Cholangiocellular carcinoma Contrast-to-noise ratio Cerebrospinal fluid Chemical shift imaging Computed tomography Computed tomography angiography Computed tomography during arterial portography Dilatation and curettage

DER DRG DSA DWI

Digital rectal examination Diagnosis-related group Digital subtraction angiography Diffusion-weighted imaging

EPI ERCP

Echo planar imaging Endoscopic retrograde cholangiopancreatography Echo train length

ETL FAST FE FFE FID FISP FLASH FNH FOV FS FSE

Fourier-acquired steady state Field echo Fast field echo Free induction decay Fast imaging with steady state free precession Fast low angle shot Focal nodular hyperplasia Field of view Fat suppression, fat-suppressed Fast spin echo

Gd

Gadolinium

Gd-BOPTA

Gadobenic acid (relaxation-enhancing component of the MR contrast medium Multihance) Gd-DOTA Gadoterate (relaxation-enhancing component of the MR contrast medium Dotarem) Gd-DTPA Gadopentetate (relaxation-enhancing component of the MR contrast medium Magnevist) Gd-DTPA-BMA Gadodiamide (relaxation-enhancing component of the MR contrast medium Omniscan) Gd-EOB-DTPA Gadoxetic acid (relaxation-enhancing component of the MR contrast medium Primovist) Gd-HP-DO3A Gadoteridol (relaxation-enhancing component of the MR contrast medium Multihance) GI Gastrointestinal GIST Gastrointestinal stromal tumor GMR Gradient moment rephasing GRASE Gradient spin echo GRASS Gradient-recalled acquisition in a steady state GRE Gradient echo HASTE HCC HF IMH IMHN

Half Fourier-acquired single shot turbo spin echo Hepatocellular carcinoma High frequency Intramural hematoma Intraductal mucin-hypersecreting neoplasia

IP iPAT IPMT IR IUD IV

In-phase Integrated parallel acquisition techniques Intraductal papillary mucinous tumor Inversion recovery Intrauterine device Intravenous

MEN MFH MIP Mn-DPDP

Multiple endocrine neoplasia Malignant fibrous histiocytoma Maximum intensity projection Mangafodipir (relaxation-enhancing component of the MR contrast medium Teslascan)

Abbreviations

MPR MP-RAGE

MRE MRI MRS MRSI MRU ms MSCT NASH NEX NHL NSF

Multiplanar reconstruction/reformation Magnetization-prepared rapid acquisition gradient echo Mononuclear phagocyte system Magnetic resonance angiography Magnetic resonance colonography Magnetic resonance cholangiopancreatography Magnetic resonance elastography Magnetic resonance imaging Magnetic resonance spectroscopy Magnetic resonance spectroscopic imaging Magnetic resonance urography Millisecond Multislice spiral computed tomography Nonalcoholic steatohepatitis Number of excitations Non-Hodgkin lymphoma Nephrogenic systemic fibrosis

OP

Opposed-phase

PACE PAU PBC PBS PC PCOS PD PET PID PIN PNET ppm PRESS PSA PSC PSIF PTC

Prospective acquisition correction Penetrating atherosclerotic ulcer Primary biliary cirrhosis Primary biliary sclerosis Phase contrast Polycystic ovary syndrome Proton density Positron emission tomography Pelvic inflammatory disease Prostatic intraepithelial neoplasm Primitive neuroectodermal tumor Parts per million Point-resolved spectroscopy Prostate-specific antigen Primary sclerosing cholangitis Inverted FISP Percutaneous transhepatic cholangiography Perfusion-weighted imaging

MPS MRA MRC MRCP

PWI RARE RAS RCC ROI ROPE RPF

Rapid acquisition with relaxation enhancement Renal artery stenosis Renal cell carcinoma Region of interest Respiratory ordering of phase encoding Retroperitoneal fibrosis

IX

s SAR SCT SE SENSE SI SIRS SNR SPGR SPIO SPIR SR SSD SSFP SSFSE STEAM STIR

Second Specific absorption rate Spiral computed tomography Spin echo Sensitivity encoding Signal intensity Systemic inflammatory response syndrome Signal-to-noise ratio Spoiled GRASS Superparamagnetic iron oxide Spectral presaturation by inversion Saturation recovery Shaded surface display Steady state free precession Single shot fast spin echo Stimulated echo acquisition mode Short tau inversion recovery

T T1 w, T2 w T2* TCC TE TEE TFE TI TIM TIPS TIRM TME TOF TR TRUS TVUS TSE TUR TURP

Tesla T1-weighted, T2-weighted Effective T2 relaxation time (T2 star) Transitional cell carcinoma Time to echo (echo time) Transesophageal echocardiography Turbo field echo Time to inversion (inversion time) Total imaging matrix Transjugular intrahepatic portosystemic shunt Turbo inversion recovery magnitude Total mesorectal excision Time of flight Time to repetition (repetition time) Transrectal ultrasound Transvaginal ultrasound Turbo spin echo Transurethral resection Transurethral resection of the prostate

UAE US USPIO

Uterine artery embolization Ultrasound Ultra-small superparamagnetic iron oxide

VIBE

Volume-interpolated breath-hold examination Vasoactive intestinal polypeptide Volume of interest

VIP VOI

X

Contents

1

The Liver

M. Taupitz, P. Asbach, A.E. Mahfouz, and B. Hamm

Focal Liver Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . MRI Appearance of Normal Anatomy . . . . . . . . . . . MRI Appearance of Pathologic Entities . . . . . . . . . . Use of Tissue-Specific Contrast Media . . . . . . . . . . Role of MRI in Focal Liver Lesions . . . . . . . . . . . . . .

1 1 1 1 11 11 27 30

Diffuse Liver Disease . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . MRI Appearance of Pathologic Entities . . . . . . . . . .

34 34 34 34 35

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patient Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . Recommended Imaging Protocol . . . . . . . . . . . . . . .

47 47 48 48 49

MRI Appearance of Pathologic Entities . . . . . . . . . . Ductal Strictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammatory Conditions . . . . . . . . . . . . . . . . . . . . . Primary Biliary Cirrhosis . . . . . . . . . . . . . . . . . . . . . . Primary Sclerosing Cholangitis . . . . . . . . . . . . . . . . Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53 53 55 55 55 61 61

MRI Appearance of Normal Anatomy . . . . . . . . . . .

52

MRI Appearance of Pathologic Entities . . . . . . . . . . Congenital Anomalies and Diseases . . . . . . . . . . . . Inflammatory Disease . . . . . . . . . . . . . . . . . . . . . . . . Pancreatic Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . Differentiation of Inflammatory Pseudotumors and Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pancreas Transplant . . . . . . . . . . . . . . . . . . . . . . . . . .

71 71 72 76

2

3

The Bile and Pancreatic Ducts P. Asbach and H.B. Gehl

The Pancreas

P. Asbach, W. Luboldt, and H.B. Gehl

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

65

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

65

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66 66 66 66

MRI Appearance of Normal Anatomy . . . . . . . . . . .

70

80 83

Contents

4

The Spleen

M. Laniado and F. Dammann

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87 87 87 87

MRI Appearance of Normal Anatomy . . . . . . . . . . . .

91

5

MRI Appearance of Pathologic Entities . . . . . . . . . . Benign Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . Infectious and Noninfectious Conditions . . . . . . . . . Trauma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Splenic Infarction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diffuse Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93 93 99 102 102 103 103

MRI Appearance of Pathologic Entities . . . . . . . . . . Stomach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Small Intestine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Colon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

112 112 113 119

The Gastrointestinal Tract W. Luboldt and M. Laniado

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patient Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6

XI

109 109 110 111

Functional MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

The Rectum and Anal Canal C. Klessen and M. Laniado

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

MRI Appearance of Normal Anatomy . . . . . . . . . . . . 128

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

MRI Appearance of Pathologic Entities . . . . . . . . . . Rectal Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anal Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammatory Bowel Disease . . . . . . . . . . . . . . . . . . . Functional Disorders . . . . . . . . . . . . . . . . . . . . . . . . . .

131 131 136 136 140

MRI Appearance of Pathologic Entities . . . . . . . . . . Benign Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Renal Tumors . . . . . . . . . . . . . . . . . . . . . . Functional Renal Imaging . . . . . . . . . . . . . . . . . . . . . . Evaluation of Living Kidney Donors . . . . . . . . . . . . . Renal Transplants . . . . . . . . . . . . . . . . . . . . . . . . . . . .

150 150 161 174 174 175

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patient Preparation and Positioning . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

127 127 127 127 128 128

The Kidneys and Upper Urinary Tract M. Taupitz and R. A. Kubik-Huch

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

143 143 144 148

MRI Appearance of Normal Anatomy . . . . . . . . . . . . 148

XII

8

Contents

The Adrenal Glands

M. Taupitz and G.P. Krestin

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

Anatomy and Normal MRI Appearance . . . . . . . . . . 184

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

MRI Appearance of Pathologic Entities . . . . . . . . . . 184 Benign Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Malignant Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Image Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9

181 181 182 182

Rational Approach to the MRI Differentiation of Adrenal Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

The Retroperitoneum

G. Krupski-Berdien, C.R. Habermann, and V. Nicolas

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

Classification of Soft-Tissue Tumors . . . . . . . . . . . . 197 MRI in Initial Diagnostic Work-up . . . . . . . . . . . . . . 200 MRI in Tumor Recurrence . . . . . . . . . . . . . . . . . . . . . 206

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10

195 195 196 196

The Urinary Bladder

V. Nicolas and D. Beyersdorff

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209

MRI Appearance of Normal Anatomy . . . . . . . . . . . 210

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209

MRI Appearance of Pathologic Entities . . . . . . . . . . Anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benign Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . .

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

209 209 209 210

210 210 211 211 211

The Prostate and Seminal Vesicles

V. Nicolas, D. Beyersdorff, U.G. Mueller-Lisse, W. Pennekamp, and C.M. Heyer

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

219 219 220 220

MRI Appearance of Normal Anatomy . . . . . . . . . . . 220

(1H) MR Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . 222 MR Spectroscopy of the Prostate: Biochemical Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 MRI Appearance of Pathologic Entities . . . . . . . . . . Anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammatory Conditions . . . . . . . . . . . . . . . . . . . . . Benign Prostatic Hyperplasia . . . . . . . . . . . . . . . . . . Malignant Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . .

223 223 224 225 226

Contents

12

The Uterus and Vagina C. Kluener and B. Hamm

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Imaging Strategy . . . . . . . . . . . . . . . . . . . . . .

241 242 242 244 244

MRI Appearance of Normal Anatomy . . . . . . . . . . . . 245

13

XIII

MRI Appearance of Pathologic Entities . . . . . . . . . . Congenital Uterine Anomalies . . . . . . . . . . . . . . . . . . Acquired Benign Uterine Disorders . . . . . . . . . . . . . Malignant Tumors of the Uterus . . . . . . . . . . . . . . . . Vaginal Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . MRI Appearance after Dilatation and Curettage . . . MRI Appearance after Radiotherapy. . . . . . . . . . . . . Residual Tumor and Recurrence after Radiotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumor Recurrence after Surgery . . . . . . . . . . . . . . . .

252 252 255 262 274 276 276

MRI Appearance of Pathologic Entities . . . . . . . . . . Pelvic Inflammatory Disease . . . . . . . . . . . . . . . . . . . Ectopic Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ovarian Torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benign Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . .

288 288 288 289 289 294

278 278

The Adnexa

M. Reuter and M. Lorenzen

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

285 285 285 286

MRI Appearance of Normal Anatomy . . . . . . . . . . . . 287 Basic Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 MRI Appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

14

Magnetic Resonance Pelvimetry M.R. Muehler

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

Definition of Pelvic Measures . . . . . . . . . . . . . . . . . . . 302

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

Imaging Technique and Findings . . . . . . . . . . . . . . . . 303

15

Magnetic Resonance Angiography of the Abdomen R. Vosshenrich and P. Reimer

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time-of-Flight MRA . . . . . . . . . . . . . . . . . . . . . . . . . . . Phase-Contrast MRA . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast-Enhanced MRA . . . . . . . . . . . . . . . . . . . . . .

305 305 306 306

MRI Appearance of Normal Anatomy . . . . . . . . . . . . 311 Abdominal and Pelvic Arteries . . . . . . . . . . . . . . . . . 311

Portal Venous System . . . . . . . . . . . . . . . . . . . . . . . . . 312 Abdominal and Pelvic Veins . . . . . . . . . . . . . . . . . . . 312 MRI Appearance of Pathologic Entities . . . . . . . . . . Abdominal and Pelvic Arteries . . . . . . . . . . . . . . . . . Portal Venous System . . . . . . . . . . . . . . . . . . . . . . . . . Abdominal and Pelvic Veins . . . . . . . . . . . . . . . . . . .

313 313 321 325

XIV

16

Contents

Intra-abdominal Lymph Nodes M. Taupitz and D. Beyersdorff

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331

MRI Appearance of Normal Lymph Nodes . . . . . . . 334

Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331

MRI Appearance of Abnormal Lymph Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17

332 332 332 333

Use of Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . 336

Abdominal MRI in Children B. Stoever

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 Imaging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coils and Pulse Sequences . . . . . . . . . . . . . . . . . . . . Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sedation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

341 341 341 341 342

Contraindications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

MRI Appearance of Pathologic Entities . . . . . . . . . . Gastrointestinal Tract . . . . . . . . . . . . . . . . . . . . . . . . Hepatobiliary System. . . . . . . . . . . . . . . . . . . . . . . . . Pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spleen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abdominal Vascular Malformations . . . . . . . . . . . . Retroperitoneum . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kidneys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adrenal Glands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pelvis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Systemic Neoplastic Disease . . . . . . . . . . . . . . . . . . .

342 342 345 348 350 350 352 352 356 360 363

1

1

The Liver M. Taupitz, P. Asbach, A.E. Mahfouz, and B. Hamm

Focal Liver Lesions

· Evaluation of indeterminate liver lesions in patients

Introduction

· Preoperative identification and localization of liver

Continuous developments and numerous improvements over the last 15–20 years have made MRI a robust modality for imaging of the liver. The improvements pertain to technical advances and the advent of new intravenous contrast media. High-performance gradient systems and body or torso phased-array coils enable the use of fast imaging techniques with good image quality and ideally allow breath-hold imaging of the entire liver with both T1- and T2-weighted sequences. In terms of contrastenhanced MRI, the liver is unique because it is the only organ for which several recently developed tissue-specific contrast media are available besides conventional, nonspecific Gd-based agents. Thanks to these advances, MRI has evolved into a powerful imaging tool for the detection and characterization of focal liver lesions. In this chapter we first outline the major technical aspects of liver MRI and give a brief overview of contrast media for liver imaging, then describe the MRI appearance of the most important benign and malignant focal liver lesions. We conclude by discussing the role of MRI in the detection and characterization of focal liver lesions and comparing its performance with that of CT, its fiercest competitor.

Indications The liver is the most common site of metastatic disease from a variety of primary tumors and is therefore examined in virtually all patients undergoing tumor staging. Patients with liver metastases, especially from colorectal cancer, benefit from improved surgical approaches. Resection of single or multiple liver metastases has been shown to significantly improve the 5-year survival rate if all hepatic metastases are removed and enough normal liver tissue is left (usually > 30 %) and patients have no additional extrahepatic metastases.1,2 On the other hand, up to 20 % of the general population have noncystic benign liver tumors,3 which must be differentiated from malignant lesions. MRI is highly accurate in detecting and characterizing focal liver lesions and can therefore meet this challenge. The major indications for MRI of the liver are (Table 1.1):

with cancer.

·

·

metastases in candidates for surgical resection (hemihepatectomy, segmental or atypical resection) and exclusion of metastases in the remaining liver. Characterization of incidentally detected liver lesions that are not definitely cystic or benign in patients with no known primary malignancy. MRI can be used as the second imaging test in patients with suspicious ultrasound findings and, because it does not involve radiation exposure, should be preferred to CT, especially in younger patients. Follow-up of malignant liver tumors.

Imaging Technique Patient Preparation, Positioning, and Imaging Planes Liver MRI is performed with the patient in a comfortable supine position. Most patients find it helpful if their knees are elevated with a small foam pad. Unless a body phasedarray coil is used, an abdominal belt should be employed to reduce respiratory artifacts by limiting respiratory excursions if the examination is performed at high magnetic field strength. In addition, patients in whom free-breathing sequences will be acquired should be carefully instructed to breathe shallowly and regularly to minimize respiratory motion of the abdominal wall (Fig. 1.1). Careful breathing instructions are also important if respiratory-triggered sequences are planned. Finally, in patients undergoing breath-hold imaging, the breathing commands that will be given during the examination need to be explained beforehand. If intravenous contrast administration is planned, a flexible indwelling cannula is inserted, preferably into an antecubital vein, before the patient is positioned in the magnet. The cannula is connected via an extension tube to a saline-filled syringe or, if available, an MR-compatible injection pump. The axial plane is the bedrock of liver MRI. Axial images enable good evaluation and delineation of both normal anatomic structures and hepatic abnormalities in most cases, and can be directly compared with CT scans. Additional coronal or sagittal images can be obtained, for

2

1 The Liver

Table 1.1 MRI techniques for different indications Indication

Sequence

Plane

Unenhanced/Contrast medium

Comment

Indeterminate liver lesions

T1w GRE IP or T1w TSE T1w GRE OP

Axial Axial

Unenhanced

T2w TSE Heavily T2w TSE

Axial Axial

Unenhanced Unenhanced

Lesion detection Demonstration of hepatic steatosis (diffuse or focal); identification of intracellular lipid for differential diagnosis Lesion detection and characterization Single-shot technique or during breath-hold; improved differentiation of solid tumors from cysts or hemangiomas Dynamic study after bolus injection of nonspecific contrast medium; arterial and portal venous phase images and additional delayed images; detection of hypervascular lesions, characterization

T1w GRE IP or T1w 3D GRE Axial (e. g., VIBE)

Preoperative evaluation

Follow-up

Unenhanced protocol as above, supplemented by T1w GRE IP or T1w TSE

Axial

T1w or T2*w GRE

Axial

T2w TSE or T2*w GRE

Axial

T1w GRE or TSE T2w TSE

Axial

Nonspecific contrast medium

Hepatobiliary contrast Dynamic study as described above; delayed images for medium, injected as bolus differentiation of hepatocellular and nonhepatocellular tumors Hepatobiliary contrast Improved detection immediately after infusion of medium, given as infusion contrast medium; delayed images for differentiation of hepatocellular and nonhepatocellular tumors Dynamic study as described above; delayed images for SPIO, injected as bolusa improved detection; differentiation of tumor entities based on Kupffer cell content SPIO, given as infusion Improved detection immediately after infusion; delayed images, see above Unenhanced Determination of tumor number and size

Note: For all indications, additional coronal and sagittal sequences may be obtained as deemed necessary using fast imaging techniques. IP = in-phase, OP = opposed-phase. aCurrently not commercially available.

example, to better assess the relationship of a mass to hepatic vessels or to differentiate an intrahepatic from a subphrenic lesion. With the fast pulse sequences available today, several slices can be obtained during one breathhold, and only little extra time is needed to acquire images in a further plane. Coronal images provide a good topographic overview and are also helpful when explaining findings to clinical colleagues.

Coils Nearly all manufacturers of MR imagers offer body or torso phased-array coils for abdominal applications. These coils provide a better signal-to-noise ratio (SNR) than the integrated whole-body resonator. The higher signal yield is especially beneficial in conjunction with fast sequences. A relative disadvantage of these coils is the very high signal intensity of subcutaneous fat, which may accentuate motion artifacts when images are acquired with the patient breathing freely. Phased-array coils should therefore only be employed for breath-hold imaging or, if freebreathing sequences are acquired, in combination with a fat suppression technique (see below). A phased-array body coil not only improves SNR but is also necessary to perform parallel imaging (usually 4–8 elements; e. g.,

SENSE, sensitivity encoding). The gain in speed achieved with parallel imaging can be used either to shorten the acquisition time of a given T1w or T2w sequence or to improve resolution without a penalty in scan time.4–6 However, SNR is reduced in both cases.7

Pulse Sequences The standard liver protocol consists of unenhanced T1w and T2w pulse sequences. Imaging techniques and parameters are different for intermediate field strengths around 0.5 T, which are rarely used today, and high field strengths in the range of 1.0–1.5 T. High soft-tissue contrast is crucial for the reliable identification and characterization of liver lesions. However, contrast must always be weighed against good image quality, which results mainly from the absence of motion artifacts and high SNR. Good image quality is often equated with good visualization of anatomic detail or spatial resolution (Fig. 1.2). T1-Weighted Imaging T1w MR images of the liver are predominantly acquired using breath-hold multislice GRE sequences (e. g., FLASH, FFE), especially when imaging is performed at high mag-

Focal Liver Lesions

netic field strength (Table 1.2). Breath-hold GRE sequences eliminate respiratory artifacts, and the extremely short echo times provide excellent T1 contrast. When a high-performance gradient system is available, up to ca. 25 slices can be acquired during a breath-hold of 15–25 s. The entire liver can thus be imaged with high contrast and good image quality in 1–3 breath-holds, depending on slice thickness and number of slices. Such multislice GRE sequences are also available for lower field strengths but will generate images with poorer SNR. Instead of a multislice GRE sequence with a TR of ca. 150 ms and a TE of ca. 5 ms, one can acquire a series of single-slice sequences with very short TRs and TEs (10–20 ms and 2–7 ms, respectively). The low intrinsic contrast resulting from such short TRs and TEs is improved by applying an inversion pulse at the beginning of each sequence (e. g., TurboFLASH or TFE). These 2D GRE sequences can be replaced with a fat-suppressed 3D GRE sequence (e. g., VIBE, volume-interpolated breath-hold examination).8 A 3D GRE sequence has the advantage of enabling acquisition of thinner slices. It is possible, for example, to reconstruct 64 2.5-mm slices by interpolation from a set of 32 slices acquired with 5.0mm slice thickness. However, unenhanced 3D GRE sequences have a lower signal yield, resulting in poorer contrast compared with their 2D counterparts. The 3D GRE technique is therefore most suitable for dynamic contrast-enhanced studies with intravenous injection of Gd-based contrast medium or delayed imaging after intravenous injection of a hepatobiliary contrast medium (see below). If the available equipment does not allow fast imaging with adequate image quality and good contrast, the liver can be imaged in 4–8 min at intermediate field strength using conventional T1w SE/TSE or GRE sequences and multiple signal averaging. Multiple averages reduce motion artifacts by increasing SNR and improve contrast and image quality with good anatomic resolution.9,10 Due to the short T1 relaxation times at intermediate field strengths, both TSE and GRE sequences are acquired with a short TR of 250–350 ms to achieve high T1 contrast. TE should be as short as possible (10–15 ms for SE and TSE, 5–10 ms for GRE). At higher field strength, SE or TSE sequences are acquired with longer TRs (400–500 ms) because T1 relaxation times are longer. The number of signal averages should not exceed four, corresponding to a scan time of ca. 8 min. Again, TE should be as short as possible (10–15 ms). GRE sequences are not suitable for conventional free-breathing imaging with multiple signal averaging at high field strength because artifacts are accentuated. If non-breath-hold pulse sequences are acquired with a body phased-array coil, use of a fat suppression technique is recommended because the very bright subcutaneous fat near the coil would otherwise cause excessive motion artifacts. In-Phase and Opposed-Phase Imaging The GRE technique can be used to acquire images with inphase (IP) and opposed-phase (OP) echo times to obtain

3

3

4

1

2

Fig. 1.1 Diagrammatic axial view of the upper abdomen showing motion artifacts due to respiration (1), cardiac pulsation (2), vessel pulsation (3), and bowel peristalsis (4) (according to Stark and coworkers10).

supplementary information for tissue characterization. IP and OP TEs depend on the field strength (Table 1.3). The different effects result from the addition or subtraction of the signal contributions from water and fat (in general: lipids) within a volume element (Fig. 1.3). The interface between tissues with different water or fat content typically appears dark on OP images (Fig. 1.4a, b). OP images enable sensitive detection of fatty infiltration and some lipid-containing tumors.11 Compared with IP images, fatty liver parenchyma has low signal intensity on OP images (Fig. 1.4c, d). Note that if images are acquired with a T1w GRE sequence, focal lesions in a fatty liver may be missed on OP images, and IP images are needed for reliable detection (Fig. 1.5). T2-Weighted Imaging T2w imaging techniques are largely the same at intermediate and high field strengths. The liver is nearly exclusively imaged with turbo or fast SE (TSE or FSE) pulse sequences, either for classic TSE imaging with the patient breathing freely or single-shot imaging during breath-holding (single-shot TSE, e. g., HASTE). The whole liver can be imaged in 2–4 min with the breathing technique or during a single breath-hold in ca. 23 s using the single-shot technique (Table 1.2). Note, however, that lesions with only slightly longer T2 relaxation times than the liver may be more difficult to detect on TSE images compared with standard T2w SE images, especially when they are small.12 Imaging

4

1 The Liver

a

b

c

d

e

f Fig. 1.2a–g MRI anatomy of the liver and comparison of different pulse sequences (1.5 T). a–c Images acquired with 2D T1w GRE sequence during breath-holding illustrate good visualization of anatomic detail and high T1 contrast (FLASH 199/4.1;90°; 23 slices in 21 s). Images through the liver at three different levels (from top to bottom) show the right, middle, and left hepatic veins (a, arrows), the right main portal branch (b), and the lower part of the right hepatic lobe (c) (see Fig. 1.7). d Fat-suppressed 3D T1w image (VIBE 4.9/2.4;10°; 32 acquired slices, interpolated to 64 slices; scan time, 22 s. e–g T2w images acquired with single-shot TSE sequence (HASTE ∞/80; 23 slices in 21 s), respiratory-triggered, fatsuppressed TSE sequence (TSE 5900/77; ETL, 21; 31 slices in ca. 5 min), and echo-planar sequence (EPI ∞/59; 31 slices in 5 s).

g

T2

T2 T2

Slice distance, 10–20 % of slice thickness (distance factor, 0.1–0.2) T1w GRE: TEs for acquisition of in-phase and opposed-phase images are given in Table 1.3. Note: The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time (for sequences with one signal average) but may come with a penalty in SNR.

Yes 23 s 7 23 60–80 –

–

> ca. 80

Yes/no

116 × 256

300 (75 %)

1

No No 4–7 min 5–7 min 7 4 21 48 80–100 80

TSE TSE with respiratory trigger HASTE

5000 2500

– –

7–15 7–15

Yes Yes

128 × 256 168 × 320

300 (75 %) 300 (75 %)

3 2

Yes Yes No 23 s 20–24 s 4–8 min 7 2.5 8 1 1 4 23 64 19 300 (75 %) 300 (75 %) 300 (75 %) 2.2–7 2.2–2.6 10–15 2D GRE 3D GRE (VIBE) SE or TSE

Axial Axial Axial (alternative) Axial Axial (alternative) Axial T1 T1 T1

ca. 170–200 5–7 500

90 10 –

– – –

No Yes No

116 × 256 116 × 256 128 × 256

No. of slices TE (ms) TR (ms) Sequence type Weight- Plane ing

Table 1.2 Recommended pulse sequences and imaging parameters

Flip (°)

ETL

FS

Matrix

FOV (mm)

No. of acquisitions

Slice thickness (mm)

Scan time

Breathhold

Focal Liver Lesions





5

 

      

      

Fig. 1.3 Diagram of opposed-phase (OP) and in-phase (IP) imaging. Because water and fat protons have slightly different precession frequencies, the amplitude of the echo in a GRE sequence is modulated (“beats”) in an environment consisting of a finely dispersed mixture of water and fat (as in hepatic steatosis). Under OP conditions, the signal contributions of water and fat are subtracted in a volume element, resulting in low SI (left). Under IP conditions, the signal contributions are added together, resulting in high SI (right). See Fig. 1.4.

is typically performed with TEs in the range of ca. 60–110 ms; longer TEs improve the differentiation of solid liver tumors from hemangiomas and cysts.13 Conventional TSE sequences can be performed as multiecho sequences with readout of several echoes, including very late echoes. Conventional TSE images can be acquired with a TR of 1600–2000 ms at low field strength, while a TR of at least 2500 ms is required at high field strength because of the longer T1 relaxation time. The fastest technique, echoplanar imaging (EPI), is demanding on hardware and software, but most MR systems in use today have echo-planar capability. EPI is basically a T2w technique and enables imaging of the liver in less than 5 s. Another fast sequence for T2w imaging is based on the steady-state principle during free precession (e. g., TrueFISP). This technique is employed for MRI of the heart and other rapidly moving organs but appears to be inferior to other fast techniques (e. g., HASTE) for T2w imaging of the liver.14 Because they improve liver–tumor contrast and overall image quality of T2w images, fat suppression techniques are more important for T2w imaging than T1w imaging. Different techniques of fat suppression are presented below. Motion Artifacts and Artifact Reduction Motion artifacts are a particular nuisance in liver imaging and must be effectively eliminated or at least reduced to obtain diagnostic images with good anatomic detail. In this section we therefore briefly outline the typical sources of motion artifacts that may degrade abdominal MR images and discuss the most important strategies currently available to control them. Artifacts in the upper abdomen can be caused by breathing (the liver moves up and down ca. 3–5 cm with respiration), blood flow and vessel pulsation, cardiac pulsation transmitted below the diaphragm, and gastrointestinal peristalsis (Fig. 1.1). Movement of anatomic structures during the image acquisition process causes blurring

6

1 The Liver

a

b

c

d Fig. 1.4a–d OP and IP imaging with a T1 w GRE sequence at 1.5 T (TR, 199 ms; flip, 90 °) in two patients without focal liver lesions, one with fatty liver (c, d) and one without (a, b). a, c OP images acquired with a TE of 2.4 ms. b, d IP images acquired with a TE of 4.8 ms. In the patient without fatty liver, only the OP image shows the char-

acteristic dark lines, which occur at organ-fat and muscle-fat interfaces (a, arrows). In the patient with pronounced fatty liver, the liver parenchyma is homogeneous and very hypointense (dark) on the OP image (c). Both livers are of normal high SI on IP images (b, d).

Table 1.3 In-phase (IP) and opposed-phase (OP) echo times at different field strengths (approximate values in ms)

specific measures can be distinguished. Software-based solutions are now available ready for use on standard clinical MR scanners, but the radiologist needs to be familiar with them and their effects in order to choose the most appropriate strategy, which is why we give a brief outline here (see Fig. 1.2). The simplest general measure to reduce motion artifacts is to increase the number of signal averages. The most important specific measures are presaturation techniques, breath-hold imaging, gradient moment nulling (also known as flow compensation), respiratory triggering or gating, phase-reordering methods (e. g., respiratory ordering of phase encoding—ROPE), and suppression of the signal from fat. Presaturation is done by applying large presaturation bands to the vessels above and below the volume of interest so that spins entering the volume with the blood produce no signal. The blood appears black, and pulsation and inflow artifacts are eliminated. Inflow artifacts mainly interfere with T1w imaging, especially when fast GRE

0.5 T 1.0 T 1.5 T

IP

OP

Multiple of 14 Multiple of 6.6 Multiple of 4.4

7 (plus multiple of 14) 3.3 (plus multiple of 6.6) 2.2 (plus multiple of 4.4)

or ghosting in the phase-encoding direction regardless of the direction of motion. Ghosts—e. g., a faint duplicate image of the high-signal-intensity abdominal wall that may obscure structures of interest—are due to phase misregistration resulting from respiratory motion during spatial encoding by the magnetic field gradients. Pulsation artifacts—the repeated depiction of the cross-section of large vessels in the phase-encoding direction—result from blood flow in the imaging volume (Fig. 1.1). A variety of hardware and software solutions exist for reducing motion artifacts in liver imaging. General and

Focal Liver Lesions

7

a

b

c

d Fig. 1.5a–d OP and IP imaging with a T1 w GRE sequence at 1.5 T (TR, 199 ms; flip, 90 °) in a patient with hepatic steatosis and a focal liver lesion. a OP image (TE, 2.4 ms) fails to demonstrate the liver metastasis due to low SI of fatty liver. b IP image (TE, 4.8 ms) clearly reveals the metastasis. c Liver–tumor contrast is also low on the

single-shot T2 w TSE image because TSE sequences with long echo trains and short echo spacing exaggerate the signal from fat-containing structures. d Good visualization of the metastasis on conventional, fat-suppressed T2 w TSE image.

sequences are used after intravenous contrast administration, which is why presaturation pulses are most beneficial in this setting. Use of presaturation pulses may result in longer TRs and thus increase scan time. Fast breath-hold sequences completely eliminate respiratory artifacts but do not eliminate artifacts due to transmitted cardiac pulsation, which may obscure the left lobe of the liver, or artifacts due to intestinal peristalsis. The latter can be eliminated by using ultrafast, subsecond imaging techniques (single-shot TSE, TurboFLASH, or TFE). Gradient moment nulling, or flow compensation, corrects for signal loss and spatial misregistration due to motion, thereby also reducing ghost artifacts.15 However, the shortest possible TE is prolonged, which is why gradient moment nulling is primarily used in conjunction with T2w imaging. Note that blood in intrahepatic vessels may appear bright on T2w images acquired with flow compensation, and vessels depicted in cross-section may mimic small focal lesions.

Different techniques are available to reduce or eliminate the signal from fat. Spectral fat saturation uses a frequency-selective pulse to saturate the spectral peak of fat before imaging. Effective fat saturation relies on good separation of the spectral peaks of water and fat, which is achieved at high field strength (1.0–1.5 T) and good magnetic field homogeneity. State-of-the-art scanners have automatic programs for individual patient shimming. These programs ensure maximum magnetic field homogeneity for uniform elimination of fat signal over a large volume when a fat saturation sequence is employed. A second method is short tau inversion recovery (STIR), in which the time of inversion (TI), i. e., the time between application of the 180° inversion pulse and the start of the imaging sequence, is chosen such that the longitudinal magnetization of fat is at zero and no signal is obtained from fat. The inversion delay to null the fat signal is 100–170 ms. STIR sequences are far less susceptible to magnetic field inhomogeneities and can be used at all field strengths. STIR increases scan time but this draw-

8

1 The Liver

back can be compensated for, to some extent, by combining it with TSE or FSE sequences (so-called turbo inversion recovery sequence, TIR). Spectral presaturation with inversion recovery (SPIR) is an inversion recovery technique that combines elements of fat saturation and STIR. Effective fat suppression is absolutely essential to eliminate motion artifacts when free-breathing T2w imaging is performed using a body phased-array coil (Fig. 1.6a, b). Respiratory triggering is almost exclusively used for T2w imaging. It requires special equipment and a pulse sequence that allows selective data acquisition at a specific phase of the respiratory cycle, typically the phase of least motion after expiration. With respiratory triggering, the examination will take at least twice as long. Several techniques are available to monitor respiratory motion. Typically, a belt with an extension sensor is placed around the patient’s abdomen, or a cushion with an integrated pressure sensor is positioned between the abdominal wall and the anterior element of a body phased-array coil. Another efficient and easy-to-handle tool is the navigator echo technique for monitoring diaphragmatic motion, which is known from cardiac imaging.16,17 If adequate elimination of motion artifacts and blurring is achieved with respiratory triggering, the basic matrix resolution can be increased, for instance to 320. With additional administration of a spasmolytic agent, images with high detail resolution of both intrahepatic and extrahepatic structures are obtained (Fig. 1.6c, d). Phase-ordering techniques (e. g., ROPE) improve image quality without a penalty in scan time. With this technique, the data that are most critical for creation of the MR image are acquired during the phase of least respiratory motion. Administration of an antispasmodic agent is usually not necessary to reduce gastrointestinal motion artifacts in liver MRI, especially when fast imaging techniques are used. However, an antispasmodic will improve image quality, particularly in conjunction with respiratorytriggered sequences.

Contrast Media Contrast-enhanced MRI of the liver can be performed with nonspecific contrast media or one of several tissuespecific agents that have become available more recently. The MR contrast media for liver imaging after intravenous injection belong to different classes with different properties and effects on MR signal intensities (alteration of relaxation times, organ distribution, pharmacokinetics). Radiologists need to be familiar with these to choose the most appropriate imaging protocols and make optimal use of their benefits. Contrast media development is a very dynamic field, with frequent approval of new agents or new indications for existing agents and the withdrawal of agents from the market. Before administering any intravenous contrast medium, radiologists should always check the most up-to-date information regarding approval status in their own country.

Nonspecific Contrast Media Nonspecific MR contrast media for intravenous injection that have been in clinical use for some time are Magnevist (Gd-DTPA) and Dotarem (Gd-DOTA). More recently, several so-called nonionic, low-osmolar contrast media have become available, including Omniscan (gadodiamide, GdDTPA-BMA), Prohance (gadoteridol, Gd-HP-DO3A), and Optimark (gadoversetamide). Gadovist (gadobutrol) has become available recently and is the only Gd-based contrast agent with a concentration of 1.0 mmol/mL Gd (twice that of the other Gd compounds). Like iodine-based X-ray contrast media used for liver CT, nonspecific Gd chelates are rapidly distributed in the extracellular space and are excreted by the kidneys. They act mainly by markedly shortening T1 relaxation times, which is best appreciated on heavily T1-weighted sequences. Nonspecific Gd-based MR contrast media are typically used for rapid, dynamic image acquisition after intravenous bolus injection (Table 1.1). The total amount of contrast medium to be injected is ca. 10–20 mL when the standard dose of 0.1 mmol/kg body weight Gd is used. Such a small amount of contrast medium is advantageous for all applications requiring a bolus technique. Tissue-Specific Contrast Media Two types of tissue-specific MR contrast media are used for liver imaging: superparamagnetic iron oxide particles (SPIO particles, magnetites; e.g., Endorem, Sinerem, Resovist), which predominantly accumulate in the liver and spleen, and paramagnetic low-molecular-weight compounds, which specifically target the hepatocytes. The latter include Teslascan (Mn-DPDP), Multihance (GdBOPTA), and Primovist (Gd-EOB-DTPA). All of these contrast media are in clinical use, except for Sinerem, which is in the clinical trial phase and for which approval is sought only for intravenous administration in MR lymphography. Magnetites are very small particles that, after intravenous injection, are selectively taken up by cells of the mononuclear phagocyte system (MPS), in particular Kupffer cells in the liver and macrophages in the spleen. The particles cause marked signal loss in these tissues by shortening their T2 relaxation times,18,19 thereby enhancing contrast between normal liver parenchyma and nonhepatocellular lesions, whose signal intensity remains unchanged because they lack MPS cells.20,21 Only very small amounts of magnetites, typically between 8 and 15 µmol/kg body weight Fe, are needed to achieve the desired effects, which persist for several hours to days and are most conspicuous on moderately T2-weighted pulse sequences, i. e., sequences with a long TR and a not too long TE (< ca. 80 ms). In addition to their T2-shortening effects, SPIO particles also shorten T1 relaxation time while moving freely in blood or interstitial fluid.19 The resulting increase in signal intensity of blood and tissues with a large blood volume is made visible on T1w images acquired with very short TE. This signal-enhancing effect is especially pronounced for ultra-small SPIO particles (USPIO) such as Sinerem (see lymph node imaging, Chap-

Focal Liver Lesions

9

a

b

c

d Fig. 1.6a–d Techniques for reducing motion artifacts on images acquired with a free-breathing T2w sequence. a, b Role of fat suppression in combination with a body phased-array coil (1.5 T). T2w images without (a) and with (b) fat suppression. Pronounced motion artifacts over the liver in a (arrows) due to signal accentua-

tion near the coil; these artifacts are markedly reduced by fat suppression. c, d Effect of respiratory triggering (1.5 T). Fatsuppressed T2w TSE images acquired without (c) and with (d) respiratory triggering (navigator technique). Respiratory triggering markedly improves image quality and detail resolution (d vs c).

ter 16). With their long blood half-life, USPIO particles might be suitable as blood pool agents and cause prolonged signal changes in highly vascularized tumors (not approved for clinical use).22 Hepatocellular MR contrast media are low-molecularweight compounds that accumulate in liver cells and increase the signal intensity of normal liver parenchyma by shortening T1. The signal increase persists for up to 2 h or longer after intravenous injection. Hepatic lesions that do not take up the contrast medium have unchanged signal intensity and are thus rendered more conspicuous.23–25 However, the effect depends on hepatobiliary elimination, which is not the same for all contrast media of this class. For Multihance, hepatobiliary elimination is low and therefore its signal-enhancing effect in the liver is only moderate, while Teslascan and Primovist, with ca. 50 % hepatic elimination, cause a strong selective signal increase in healthy liver tissue. The recommended doses per kg body weight are 0.01 mmol Gd for Teslascan, 0.1 mmol Gd for Multihance, and 0.025 mmol Gd for Primovist. Not all tissue-specific MR contrast media can be given by intravenous bolus injection. Resovist is the only ap-

proved SPIO contrast medium that can be injected in bolus form; Endorem must be given as a slow infusion. Among the hepatobiliary contrast media, Primovist and Multihance can be administered in intravenous bolus form while Teslascan is infused. The contrast media that can be given as a bolus enable dynamic postcontrast MR studies in addition to the acquisition of static images at different phases following contrast administration.

MRI Protocols Unenhanced MRI The MR pulse sequences recommended for unenhanced MR assessment of focal liver lesions are listed in Tables 1.1 and 1.2. Dynamic MRI A dynamic contrast-enhanced MRI study of the liver can be done with nonspecific contrast media or those hepatobiliary agents that can be given as an intravenous bolus. Multiphasic imaging is performed using a T1w multislice IP GRE sequence or a fat-suppressed 3D GRE sequence. To

10

1 The Liver



a

15s

 

60 s

b

c

d Fig. 1.7a–e a Diagrammatic representation of the temporal course of contrast inflow into the liver via the hepatic artery (arterial phase) and portal vein (portal venous phase). b–e Dynamic MR study performed with a T1w volumetric 3D GRE sequence (VIBE) and nonspecific Gd-based contrast medium: precontrast image (b); arterial phase image shows enhancement only of the arteries (c); portal venous phase image shows enhancement of the portal vein and its branches and beginning enhancement of liver parenchyma, but no contrast in the liver veins (d, arrow); parenchymal phase image shows enhancement of all intrahepatic vessels and liver parenchyma (e) (same patient as in Fig. 1.2).

e

capture the dual contrast inflow via the hepatic artery and portal vein (Fig. 1.7), the first acquisition begins 15 s after the start of bolus injection of the contrast medium into a peripheral vein (arterial phase) and the second acquisition at 55–60 s (portal venous phase). Delayed images (equilibrium phase) can be acquired at 2.5 min and 10 min. Determination of individual circulation time is not necessary. An MR-compatible power injector is useful for contrast injection but not a prerequisite for the success of a dynamic MR study. The contrast bolus should be followed by a 20-mL saline flush. Arterial and portal venous phase MRI is similar to biphasic spiral CT in terms of timing and image interpretation. When dynamic liver MRI is per-

formed with Resovist, the signal-enhancing effect during the blood pool phase can be exploited by obtaining a T1w series as with use of nonspecific Gd-based contrast medium. Alternatively, a T2*w GRE sequence can be used, which will show a temporary signal loss in highly vascularized tissues. Static MRI Using Tissue-Specific Contrast Media The relaxivity-altering effects of specific MR contrast media are best depicted on postcontrast images acquired with heavily T1-weighted sequences in conjunction with hepatobiliary contrast media and with moderately T2w sequences after administration of an SPIO contrast agent.

Focal Liver Lesions

11

!"'#%$

) ''#%$

"% ) ''#%

!"'#%

  %&$ (&& '#%          !"$%  Segment VII      !"# ) '#

Fig. 1.8 Segmental anatomy of the liver.

If a hepatobiliary contrast medium is used to differentiate hepatocellular tumors (focal nodular hyperplasia, welldifferentiated hepatocellular carcinoma), additional delayed images should be obtained no earlier than 2 h after contrast injection.

MRI Appearance of Normal Anatomy The normal liver is of uniform signal intensity, which is higher than that of the spleen on T1w images and lower on T2w images. Hepatic vessels appear dark on T1w images due to flow-related signal loss (see Fig. 1.2). They are also dark on T2w images unless gradient moment nulling (flow compensation) is used, in which case intrahepatic vessels are bright. The branches of the portal vein and intrahepatic veins serve to identify the liver segments (Fig. 1.8). Differentiation of the liver from other structures in the upper abdomen is usually straightforward.

MRI Appearance of Pathologic Entities Most benign and malignant liver tumors are hypointense on T1w images and hyperintense on T2w images. Cysts, hemangiomas, and metastases from neuroendocrine tumors are visualized with high contrast, as are intratumoral necrosis and abscesses. The vast majority of other liver metastases and cholangiocarcinoma (CCA) are less conspicuous on both T1w and T2w images. Liver tumors that are isointense or nearly isointense to normal liver parenchyma are focal nodular hyperplasia (FNH) as well as an occasional adenoma or hepatocellular carcinoma (HCC) (Table 1.4). T1w sequences contribute little to lesion characterization, except for cysts, which are clearly identified as sharply demarcated lesions of very low signal intensity on T1w images. In all other cases, characterization of focal liver lesions is primarily based on T2w sequences, which enable good evaluation of tumor margins and internal structures. Use of different T2 weightings, including images generated from late echoes, allows fur-

12

1 The Liver

a

b Fig. 1.9a, b Multiple liver cysts in a patient with autosomal dominant polycystic kidney disease (1.5 T). a Single-shot T2w TSE image. b T1w GRE image. Both images were acquired during breath-hold.

The liver cysts are depicted as homogeneous lesions of very high SI on the T2w image and very low SI on the T1w image. Also seen are multiple cysts in both kidneys.

Table 1.4 Signal intensities of different liver lesions relative to liver tissue on unenhanced images. Note: opposed-phase signal intensities (T1w OP) for patients without hepatic steatosis

ther characterization and improves accuracy in differentiating cysts and hemangiomas from solid liver tumors. Intralesional hemorrhage and fatty components as well as melanotic liver metastases from malignant melanoma exhibit atypical signal behavior with hyperintensity on both T1w and T2w images. Additional information for characterizing focal liver lesions is provided by dynamic contrast-enhanced MRI, which enables the important differentiation of hypervascular and hypovascular lesions and reveals some typical enhancement patterns such as the progressive centripetal enhancement characteristic of liver hemangioma (Table 1.5). Below we describe the MR appearance of the most important benign and malignant focal liver lesions on unenhanced images and dynamic contrast-enhanced imaging with nonspecific Gd-based contrast media. A separate section illustrates the use of tissue-specific MR contrast media for selected tumor entities.

Benign tumors Cysts Hemangioma FNH Adenoma Malignant tumors Metastases · Melanotic melanoma

T1w IP

T1w OP

T2w

↓↓↓ ↓↓↓ 0–(↓) ↓–↑↑

↓↓↓ ↓↓↓ 0–(↓) ↓↓–↑↑

↑↑↑ ↑↑↑ 0–(↑) 0–↑

↑–↑↑

↑–↑↑

↑–↓

· Neuroendocrine tumors ↓↓ · Other primaries ↓↓

↓↓

↑–↑↑↑

↓↓

↑–↑↑

HCC CCA

↓↓ ↓↓

0–↑ ↑–↑↑

↓↓ ↓↓

Table 1.5 List of typically hypervascular liver tumors Benign liver tumors FNH Adenoma Malignant liver tumors HCC Metastases

· · · · ·

Renal cell carcinoma Breast cancer (may also be hypovascular) Neuroendocrine tumors (e. g., carcinoid, insulinoma) Melanoma Sarcoma

Benign Focal Liver Lesions Cyst Because they contain water, cysts have long T1 and T2 relaxation times, resulting in very low signal intensity on T1w images and uniform high signal intensity on T2w images. A liver cyst is usually sharply demarcated from the surrounding tissue but may appear blurred due to partial volume effects if the plane of section is tangential to it (Fig. 1.9). In cases where cysts and hemangiomas cannot be differentiated because of similar T2 signal intensities, a T1w sequence or dynamic imaging will help distinguish the two (Fig. 1.10). MRI clearly differentiates the cyst fluid from the solid wall and septa in echinoccocal cysts, while the characteristic curvilinear calcifications (eggshell calcifications) are far better revealed with CT.

Focal Liver Lesions

13

a

b

c

d Fig. 1.10a–e Liver hemangioma and cyst in segment II (1.5 T). a Axial T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 2 min (d), and 10 min (e) after IV injection of Gdbased contrast medium. Before contrast medium administration, both lesions have very high T2 SI (a) and low T1 SI (b). On dynamic contrast-enhanced images, the hemangioma is characterized by initial nodular peripheral enhancement (c) with progressive fill-in (d), resulting in complete hyperintensity of the lesion 10 min after contrast administration (e). The cyst has unchanged low SI on all postcontrast images.

e

Hemangioma Hemangiomas are the most common benign hepatic tumors. They are incidentally detected in ca. 5 % of patients undergoing abdominal imaging for other reasons. Hemangiomas are mesenchymal tumors characterized by densely packed, dilated vessels with an endothelial lining. Regressive changes such as scars and central hyalinization may be present in large hemangiomas. Although they have no malignant potential, hemangiomas may pose a diagnostic challenge because they must be differentiated

from liver metastases. This is especially important in patients with multiple hemangiomas (up to 35 % of cases). Hemangiomas are well-circumscribed masses of uniform signal intensity on T1w and T2w images and may occasionally be lobular. They are moderately hypointense on T1w images and clearly distinct from cysts, which have very low T1 signal intensity, while distinguishing them from other solid liver tumors is more difficult. On T2w images, on the other hand, hemangiomas are of high signal intensity and therefore at times resemble cysts

14

1 The Liver

(Fig. 1.10). Hemangiomas have long T2 values (> 80 ms), which are due to slow-flowing blood and correlate with the collective size of their constituent vascular spaces, but not with overall tumor size.26 Central thrombosis or hyalinization is occasionally seen, especially in large hemangiomas (Fig. 1.11). Various quantitative approaches have been proposed to improve the differentiation of hemangiomas and malignant tumors, including calculation of T2 relaxation time or contrast-to-noise ratio (CNR). However, none of the quantitative methods appears to be superior to visual interpretation of T2w images, taking into account the morphologic features just outlined. A combination of moderately and heavily T2-weighted sequences has been recommended to increase diagnostic confidence in diagnosing hemangioma,27,28 for instance a moderately T2w TSE sequence and a single-shot TSE sequence, which has inherently stronger T2 weighting. On heavily T2-weighted images, hemangiomas should be of high signal intensity similar to that of cerebrospinal fluid (CSF) or fluid in the gallbladder or stomach. MRI has an accuracy of > 90 % in characterizing hemangioma.13,27 Nevertheless, there are still cases where accurate characterization poses a challenge, e. g., when confronted with an atypical hemangioma with reduced signal intensity and an inhomogeneous appearance due to degenerative changes, or in patients in whom hemangiomas may be confused with hepatic metastases from neuroendocrine tumors, which also have very long T2 relaxation times.29 If unenhanced imaging fails to ascertain the presumptive diagnosis of hemangioma, a dynamic MR study is done, usually with nonspecific contrast medium. On dynamic contrast-enhanced imaging, most hemangiomas show initial peripheral enhancement with slow progressive fill-in, ultimately resulting in uniform high signal intensity on delayed images (ca. 10 min after contrast administration). Nodular peripheral enhancement on early dynamic MR images is pathognomonic for hemangioma (Figs. 1.10 and 1.11). Small and medium-sized hemangiomas are usually of homogeneous high signal intensity on delayed images. Hemangiomas with central thrombosis or hyalinization also show peripheral enhancement with centripetal progression but persistent central hypointensity (Fig. 1.11). Large hemangiomas often undergo degeneration, seen as irregular, nonenhancing areas on delayed images. Very small hemangiomas may show immediate uniform enhancement on arterial phase images, which persists or equilibrates with liver on later images. Focal Nodular Hyperplasia Focal nodular hyperplasia (FNH) is a relatively uncommon benign tumor and has been associated with the use of oral contraceptives. It is assumed that oral contraceptives promote the growth of FNH already present, which is why the tumors become larger and are more commonly detected in women.30 Unlike hepatocellular adenoma, FNH has no malignant potential. It consists of hepatocytes, Kupffer

cells, and proliferating, blind-ending bile ducts and may have a central stellate scar. FNH is nearly isointense or slightly hypointense to surrounding liver on T1w images and isointense or slightly hyperintense on T2w images (Fig. 1.12). The imaging appearance is attributable to the fact that FNH consists of almost the same tissue as normal liver. T1 hypointensity is more commonly seen on GRE images obtained with a very short TE (e. g., FLASH).31 A typical imaging feature is the extremely uniform appearance of the lesion on T1w images, which is observed in 57–94 % of patients.31,32 A central scar has been reported in 35–50 % of FNHs. The scar is always of low signal intensity on T1w images, but is revealed as a hyperintense structure on T2w imaging in only ca. 65 % of cases (Fig. 1.12).32 An occasional FNH is pedunculated (Fig. 1.13) with the risk of torsion and subsequent infarction of the tumor. Being hypervascular, FNH is characterized on dynamic MRI by intense homogeneous enhancement in the arterial phase, which rapidly fades to isointensity on later images (Figs. 1.12 and 1.13). The typical central scar is characterized by nonenhancement on early postcontrast images, becoming hyperintense only after 2–4 min (Fig. 1.12). Hepatocellular Adenoma Hepatocellular adenoma is an uncommon primary liver tumor with the risk of malignant transformation. The tumor resembles hepatocellular carcinoma in that fatty degeneration may be present (Fig. 1.14). Hemorrhage is also quite common. The morphologic features seen on MRI are variable because of the multifarious histologic composition of the tumors.33 On IP T1w images, about half of all hepatocellular adenomas appear heterogeneous with hyperintensities indicating fatty degeneration or hemorrhage. Fatty components can be differentiated from hemorrhage by loss of signal on OP images (Fig. 1.14). The T2 appearance is also predominantly heterogeneous with areas of high and low signal intensity. MRI demonstrates a peritumoral pseudocapsule in ca. 30 % of cases.34 Although common in FNH, a central scar is very rare in hepatocellular adenoma. Overall, the MR appearance of hepatocellular adenoma resembles that of hepatocellular carcinoma, rarely that of FNH. Typical contrast enhancement features are also lacking. Most hepatocellular adenomas are hypervascular and the diagnosis may be made with a high degree of confidence when MRI shows an inhomogeneous hypervascular tumor with signs of hemorrhage.35 However, given the diagnostic problems and malignant potential, surgical resection may be required in unclear cases.

Focal Liver Lesions

15

a

b

c

d Fig. 1.11a–e Hemangioma with central hyaline degeneration in liver segment VII (1.5 T). a Axial image obtained with single-shot T2w TSE sequence. b–e Axial T1w GRE images before (b) and 15 s (c), 2 min (d), and 10 min (e) after IV injection of Gd-based contrast medium. Thrombosis of the hemangioma is indicated by markedly higher SI on T2w image (a) and markedly lower SI on T1w image (b) relative to the periphery of the lesion. The immediate postcontrast image shows nodular peripheral enhancement (c). Subsequent images show progressive peripheral enhancement with persistent central hypointensity.

e

16

1 The Liver

a

b

c

d Fig. 1.12a–e Focal nodular hyperplasia (FNH) in liver segments III and IV (1.5 T). a Axial image obtained with single-shot T2w TSE sequence. b–e Axial T1w GRE images before (b) and 15 s (c), 55 s (d), and 10 min (e) after IV injection of Gd-based contrast medium. Both FNHs are slightly hyperintense and homogeneous on T2w image (arrows in a) and slightly hypointense on T1w image (b). The central scar is seen only on the T1w image. On the arterial phase image (c) both masses show intense enhancement sparing the central scars. Enhancement fades rapidly, while there is delayed enhancement of the central scars (most conspicuous in the FNH indicated by the arrow in e).

e

Focal Liver Lesions

17

a

b

c

d Fig. 1.13a–e Large pedunculated FNH of left hepatic lobe (arrow) (1.5 T). a T2w axial single-shot TSE image. b T2w coronal single-shot TSE image. c–e Axial T1w GRE images obtained before (c) and 15 s (d) and 10 min (e) after IV injection of Gd-based contrast medium. FNH is nearly isointense to surrounding liver on unenhanced images (a–c) and shows hypervascularity on arterial phase image (d).

e

18

1 The Liver

a

b

c

d

e

f Fig. 1.14a–f Large liver adenoma in segment VIII (1.5 T). a Axial T2w TSE image. b IP axial T1w GRE image. c OP axial T1w GRE image. d–f Dynamic contrast-enhanced axial T1w GRE images (IP) obtained 15 s (d), 55 s (e), and 5 min (f) after IV injection of Gdbased contrast medium. Adenoma is of slightly higher SI than surrounding liver on T2w image (a). Presence of lipid in the lesion

is suggested by moderate hyperintensity of the adenoma on IP image (b) and marked hypointensity on OP image (c). In the dynamic contrast-enhanced examination, high SI of the adenoma on arterial phase image (d) indicates hypervascularity. Also seen is FNH in the left hepatic lobe.

Focal Liver Lesions

19

a

b

c

d Fig. 1.15a–e Multiple hypovascular metastases from colorectal adenocarcinoma (1.5 T). a Axial single-shot T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 55 s (d), and 10 min (e) after IV injection of Gd-based contrast medium. T2w image (a) reveals multiple lesions with a hyperintense center (necrosis) and less intense periphery (target sign). On T1w image (b) the metastases are revealed as hypointense lesions with ill-defined margins. Arterial phase image (c) of the dynamic contrast-enhanced examination shows characteristic peripheral enhancement of large metastases (straight arrow), while smaller lesions are obscured due to diffuse enhancement (curved arrow). Later images show peripheral washout of the metastases, seen as a hypointense halo relative to the center of the lesion and surrounding liver.

e

Malignant Focal Liver Lesions Metastases The MR appearance of liver metastases is as heterogeneous as their morphology and also varies with the type of primary tumor (see page 11 and Table 1.4). Dynamic contrast-enhanced MR imaging differentiates hypovascular and hypervascularized metastases and thus provides some clues about the primary tumor (Table 1.5). Most liver metastases are from colorectal cancer. They have a fairly consistent MRI appearance with moderate T1 hypointensity and moderate T2 hyperintensity. If central

necrosis is present, the signal intensity of the center is even lower than that of the surrounding viable tumor on T1w images (doughnut sign) and higher on T2w images (target sign)36 (Fig. 1.15). This appearance strongly suggests metastasis, especially if multiple lesions are present. Calcifications within a metastasis are low in signal intensity on both T1w and T2w MR images, but are less obvious than on CT scans. Intratumoral hemorrhage is highly conspicuous as hyperintensity on T1w images. Liver metastases from colorectal adenocarcinoma are usually hypovascular. In general, hypovascular metastases show thin rim enhancement on early-phase contrast-enhanced images

20

1 The Liver

a

b

c

d Fig. 1.16a–e Hypovascular metastasis from breast cancer (straight arrow, segments V/VI) and small hemangioma (curved arrow, segment VI). a Axial T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 55 s (d), and 10 min (e) after IV injection of Gd-based contrast medium. The metastasis is clearly delineated from surrounding liver on both T1w and T2w images. The dynamic contrast-enhanced examination shows slow and mild inhomogeneous enhancement of the metastasis. Characteristic enhancement pattern of the small hemangioma (see Fig. 1.9).

e

(1–2 min), while their overall signal intensity remains below that of liver (Fig. 1.15). In ca. 35 % of cases, there will be peripheral washout on delayed images (Fig. 1.15), which has 100 % specificity for malignancy.37 Metastases from breast cancer also show moderate hypointensity on unenhanced T1w images and moderate hyperintensity on T2w images but may be either hyper- or hypovascular (Fig. 1.16). Metastases from renal cell carcinoma, neuroendocrine tumors such as carcinoids, melanoma, and sarcoma tend to be hypervascular. Their intense enhance-

ment on arterial phase dynamic images often allows detection of very small lesions (Fig. 1.17). Note, however, that these tumors fade to become isointense to liver by the portal venous phase and will escape detection if the arterial phase of enhancement is missed. On unenhanced images melanotic metastases from malignant melanoma have high T1 signal intensity due to the presence of paramagnetic melanin and high or low T2 signal intensity, depending on the amount of melanin present. Multiple melanoma metastases in the liver may vary substantially

Focal Liver Lesions

21

a

b

c

d Fig. 1.17a–e Hypervascular metastases from functioning neuroendocrine pancreatic carcinoma (1.5 T). a Axial T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 55 s (d), and 5 min (e) after IV injection of Gd-based contrast medium. Only one metastasis is seen on the unenhanced T2 and T1w images, while the arterial phase image of the dynamic contrast-enhanced study reveals a second, very small metastasis (arrow in c). On portal venous phase and delayed images, the small metastasis is obscured again due to enhancement of the liver parenchyma (d, e).

e

in signal features (Fig. 1.18). Intratumoral hemorrhage can be identified by increased T1 signal intensity or the presence of sedimentation effects (Fig. 1.19). Hepatocellular Carcinoma Hepatocellular carcinoma (HCC) may be expansive (unilocular, multilocular) or infiltrative. Like liver adenomas, HCCS have a variable MR appearance due to their heterogeneous morphology. HCCs may be surrounded by a pseudocapsule, which consists largely of compressed liver tissue and blood vessels and has low signal intensity on T1w images (Figs. 1.20 and 1.21). A pseudocapsule is more

common in East Asians: it has been found in 27 % of cases in the USA,38 while Japanese investigators reported a pseudocapsule in 42 % of HCCs.39 Internal structures with high signal intensity on T1w images are typical of HCC. They have been found to be attributable to fatty metamorphosis, which is present in ca. 50 % of HCCs.38 More recently, it has been proposed that such hyperintensities may be due to intratumoral copper.40 About 50 % of HCCs have a mosaic pattern of low and high signal intensities, which is more obvious on T2w images than T1w images.41

22

1 The Liver

a

b Fig. 1.18a, b Multiple metastases from malignant melanoma (1.5 T). a Axial single-shot T2w TSE image. b Axial T1w GRE image. Metastases are of high SI on unenhanced T1w image due to presence of melanin (melanotic metastases).

a

b Fig. 1.19a, b Multiple large metastases from malignant gastrinoma (1.5 T). a Axial single-shot T2w TSE image. b Axial T1w GRE image. Hyperintensity of some of the metastases on T1w image is due to

intralesional hemorrhage. On T2w image, sedimentation effects are seen in part of the lesions.

HCC may be hypervascular or hypovascular, with well differentiated HCC tending to be hypervascular and poorly differentiated HCC tending to be hypovascular42 (see Figs. 1.20 and 1.21). The peritumoral pseudocapsule is of low signal intensity on early dynamic contrast-enhanced images and becomes hyperintense on later-phase images. This pattern of enhancement is more obvious in larger tumors, which also tend to be more inhomogeneous on delayed postcontrast images.

hypovascular to moderately hypervascular. A rather typical feature is inhomogeneous high signal intensity on latephase images acquired after intravenous injection of nonspecific contrast medium.

Cholangiocarcinoma Hepatic cholangiocarcinoma (CCA) can grow along the branches of the portal vein or occur in the form of a solid mass (see Chapter 2). The first type is difficult to identify by imaging, while the focal type has the same MR appearance as liver metastases (low T1 signal intensity and moderately high T2 signal intensity). CCA is suggested if MRI demonstrates an ill-defined mass, satellite lesions, and atrophy of the affected hepatic lobe (Fig. 1.22).43 On dynamic contrast-enhanced imaging, CCA is moderately

Other Focal Liver Lesions Hemorrhagic Liver Lesions Hemorrhagic lesions have a typical MR pattern, which varies with the age of the hemorrhage. Following lysis of erythrocytes and liquefaction, hematoma has high signal intensity on both T1w and T2w images. Increasing uptake of hemosiderin by macrophages in the periphery of a hematoma is seen as a rim of low signal intensity around the lesion on T2w images (see Fig. 1.19).

Focal Liver Lesions

23

a

b

c

d Fig. 1.20a–e Small hypervascular HCC in liver segment IV (arrow) (1.5 T). a Axial T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 55 s (d), and 10 min (e) after IV injection of Gd-based contrast medium. HCC is moderately hyperintense on T2w image and hypointense on T1w image, which also reveals a low-SI pseudocapsule. In the dynamic contrast-enhanced examination, the hypervascular tumor is of high SI on arterial phase image (c) and becomes isointense to liver on later images. The pseudocapsule is seen as a thin rim of slightly higher SI. T2w image also shows small cyst in left hepatic lobe.

e

24

1 The Liver

a

b

c

d Fig. 1.21a–e Large hypovascular HCC in liver segment III (arrows) (1.5 T). a Axial T2w TSE image. b–e Axial T1w GRE images obtained before (b) and 15 s (c), 55 s (d), and 10 min (e) after IV injection of Gd-based contrast medium. The HCC is only slightly higher in SI than surrounding liver on T2w and T1w images and has a pseudocapsule that is clearly depicted as a thin rim of low SI on T1w image. In the dynamic contrast-enhanced examination, there is only little enhancement of the mass, consistent with a hypovascular lesion (c). Delayed phase shows isointense, inhomogeneous tumor surrounded by a hyperintense pseudocapsule (e).

e

Focal Liver Lesions

25

a

b Fig. 1.22a–c Cholangiocarcinoma in the left hepatic lobe (1.5 T). a Axial T2w TSE image. b, c Axial T1w GRE images obtained before (b) and 5 min (c) after IV injection of Gd-based contrast medium. There is characteristic atrophy of the left hepatic lobe. Inhomogeneous appearance of the mass on postcontrast image.

c

a

b

c

d Fig. 1.23a–d Multiple Candida abscesses of the liver in an immunosuppressed patient undergoing antimycotic treatment (1.5 T). a Axial single-shot T2w TSE image. b–d Axial T1w GRE images obtained before (b) and 15 s (c) and 5 min (d) after IV injection of Gd-based contrast medium. On the T2w image (a), only the

central portions of the lesions have high SI, which is why they appear smaller than on T2w imaging. Dynamic contrast-enhanced images demonstrate rim enhancement, which is very intense on delayed image.

26

1 The Liver

a

b

c

d Fig. 1.24a–d Focal fatty liver lesions and focal fatty sparing (1.5 T). a, b Patient with focal hepatic steatosis in segment IV. OP image, TE = 2.4 ms (a) and IP image, TE = 4.8 ms (b) obtained with T1w GRE sequence. c, d Diffuse hepatic steatosis with focal fatty sparing in segment IV in a second patient. OP image, TE = 2.4 ms (c) and IP image, TE = 4.8 ms (d) obtained with T1w GRE sequence. Focal fatty

infiltration is indicated by low SI relative to liver on OP image (a) and isointensity on IP image (b). In the patient with diffuse hepatic steatosis (c, d), focal fatty sparing is indicated by an area of high SI within the low-SI fatty liver on OP image (c) (see Figs. 1.4 and 1.5).

Abscess An intrahepatic abscess is hypointense on T1w images and hyperintense on T2w images. The hypointense lesion seen on T1w images corresponds to the abscess cavity, while the area of hyperintensity seen on T2w images also comprises the perifocal edema. Following initiation of medical therapy, imaging will demonstrate concentric rings representing the surrounding granulation and connective tissue (Fig. 1.23) and resolution of inflammatory edema (on T2w images).

GRE images. Well-defined focal fatty lesions are uncommon and are characterized by high signal intensity on T1w and T2w SE sequences (without fat suppression). On TSE or FSE sequences, which exaggerate the signal from fat, such focal fatty lesions may be of very high signal intensity and thus mimic metastases (e. g., from melanoma), HCC, or hemorrhagic lesions, all of which may be hyperintense on both T1w and T2w images.31,44 Both nondiffuse fatty infiltration in the form of geographic or focal changes and fatty sparing can be diagnosed with a high degree of confidence by comparing OP and IP GRE images (Fig. 1.24). Multifocal fatty infiltration of the liver is uncommon but must be considered because the appearance may well resemble multiple metastases on CT scans or when looking only at OP MR images. The true nature of the lesions will be revealed by taking into account both OP and IP images45 (Fig. 1.25).

Focal Fatty Liver Lesions Hepatic steatosis or fatty infiltration of the liver may be diffuse, geographic, or focal. Another pattern is focal fatty sparing in the setting of diffuse fatty infiltration of the liver. Fatty liver may be caused by external factors such as certain foods, drugs, or toxins and is associated with porphyria and other metabolic disorders. Mild fatty infiltration of the liver is not seen on conventional T1w TSE or IP

Focal Liver Lesions

27

a

b Fig. 1.25a–c Multifocal hepatic steatosis (1.5 T). a Axial single-shot T2w TSE image. b, c Axial T1w GRE images obtained with IP echo time (TE = 4.1 ms) (b) and OP echo time (TE = 2.2 ms) (c). Fatty lesions in the liver are moderately hyperintense on T2w image (a) and IP T1w image (b) and hypointense on OP image (c).

c

Use of Tissue-Specific Contrast Media Superparamagnetic Iron Oxide Particles Superparamagnetic iron oxide particles (SPIO) contrast media significantly increase the detection of liver metastases, particularly small ones, compared with unenhanced MRI46–48 (Fig. 1.26) and can also improve lesion characterization. When Endorem (which can only be given as an infusion) is used, only delayed images can be acquired, and these will show a signal loss in tumors containing MPS cells (see Fig. 1.28). However, this information has limited value for lesion characterization because MPS cells are present in well-differentiated HCC and benign tumors such as FNH or adenomatous hyperplasia. Therefore, the only possible distinction is that between well-differentiated tumors of hepatocellular origin on the one hand and dedifferentiated hepatocellular tumors or nonhepatocellular lesions such as metastases on the other hand. SPIO formulations that can be administered as a bolus, such as Resovist, can be used for perfusion studies analogous to nonspecific contrast media. The signal-enhancing effect of Resovist is low in the blood pool phase but usable to identify hemangiomas, which will

show similar enhancement patterns as after administration of nonspecific Gd-based contrast media (Fig. 1.27). While nonspecific contrast media rapidly diffuse into the extravascular space after intravenous injection, Resovist particles have a longer blood half-life and are still intravascular when early dynamic images are acquired, which is why the strong arterial phase enhancement of hypervascular tumors such as FNH known from nonspecific contrast media is absent when Resovist is used (Fig. 1.28). If dynamic Resovist-enhanced imaging is performed with a T2*-weighted sequence, highly vascularized tumors will show a temporary signal loss.

Hepatobiliary Contrast Media As with iron oxide particles, hepatobiliary contrast media significantly improve the detection of small liver lesions during the phase of hepatocellular uptake23,24,49 (Fig. 1.29). Hepatobiliary contrast media accumulate in tumors of hepatocellular origin, which can thus be differentiated from nonhepatocellular tumors with an accuracy of ca. 94 %. A hyperintense rim on delayed images is nearly 100 % specific for malignancy.50

28

1 The Liver

a

b

c

d

e

f

g

h

Focal Liver Lesions

29

a

b

c

d

e

v

f Fig. 1.27a–f Dynamic T1w MRI of liver hemangioma after bolus injection of iron oxide particles. a–d Axial images obtained with T1w GRE sequence before (a) and 1 min (b), 5 min (c), and 10 min (d) after IV injection of Resovist. e, f Fat-suppressed, respiratorytriggered T2w TSE images obtained before (e) and 15 min after (f) contrast injection. Blood pool effect of iron oxide particles results in nodular peripheral enhancement of hemangioma on early postcon-

trast image (b) with progressive fill-in on later images (c and d; see Fig. 1.10). At the same time, normal liver parenchyma gradually loses SI as a result of contrast uptake. Intravascular signal increases on postcontrast T1w images (b–d). Compare T2w images (e, f). Prolonged blood pool effect of iron oxide particles markedly reduces SI of hemangioma on postcontrast T2w image (f).

Fig. 1.26a–h Examples of improved liver lesion detection in two patients using iron oxide particles (1.5 T). a–e Patient with multiple metastases from colon cancer. Axial T2w images obtained with single-shot TSE sequence and fat-suppressed, respiratory-triggered TSE sequence before (a, c) and 15 min after IV injection of Resovist (b, d). Iron oxide particles improve contrast and conspicuity of liver metastases by selectively reducing the SI of liver parenchyma. Respiratory-triggered TSE sequence is superior to single-shot TSE sequence in terms of image quality and conspicuity of liver meta-

stases. Unenhanced T1w image for comparison (e). f–h Patient with metastasis from islet cell carcinoma. Axial T2w images obtained with fat-suppressed, respiratory-triggered TSE sequence before (f) and 15 min after (g) IV injection of Resovist. Precontrast image does not allow reliable detection of the liver lesion despite excellent image quality. Selective signal reduction of the liver parenchyma on image enhanced with iron oxide particles improves conspicuity of the very small metastasis (arrow in g). Unenhanced respiratorytriggered T1w GRE image fails to demonstrate the metastasis (h).

30

1 The Liver

a

b

c

d

e

f Fig. 1.28a–f Dynamic T1w MRI of FNH after bolus injection of iron oxide particles. a–d Axial images acquired with T1w GRE sequence before (a) and 15 s (b), 55 s (c), and 5 min after (d) IV injection of Resovist. e, f Fat-suppressed, respiratory-triggered T2w TSE images obtained before (e) and 15 min after (f) contrast injection. Iron

oxide particles produce only a negligible increase in SI of FNH on arterial phase image; FNH and liver show a continual signal decrease. Intravascular signal increases on T1w images (b–d). Compare T2w images (e, f). There is marked signal reduction of FNH, indicating Kupffer cell activity.

For those hepatobiliary contrast media that can be administered as a bolus, the experience with dynamic MRI using nonspecific contrast media is directly transferable. Hypovascular metastases will show the wellknown rim enhancement on early phase images (Fig. 1.29), while hemangiomas are characterized by progressive centripetal enhancement51 (Fig. 1.30). FNH displays perfusion-related enhancement in the early phase and high signal intensity due to contrast uptake on delayed images (Fig. 1.31).

Role of MRI in Focal Liver Lesions Detection of Focal Liver Lesions This section discusses the role of MRI in detecting focal liver lesions with regard only to malignant lesions such as metastases and HCC. The degree of vascularization of a lesion determines whether it is best detected during the arterial phase of a dynamic MR study after intravenous injection of a nonspecific Gd-based contrast medium or during the late phase after injection of a tissue-specific

Focal Liver Lesions

31

a

b

c

d Fig. 1.29a–e Liver metastases from rectal cancer on dynamic T1w MRI with delayed imaging after IV bolus injection of hepatobiliary contrast medium (Primovist) (1.5 T). a Axial image obtained with fat-suppressed, respiratory-triggered T2w TSE sequence. b–e Axial images obtained with T1w GRE sequence before (b) and 15 s (c), 55 s (d), and 2 h after (e) contrast injection. T2w and unenhanced T1w images (a, b) reveal one large and two small metastases. Dynamic images show peripheral enhancement of the liver lesions, consistent with the enhancement pattern of hypovascular metastases after administration of nonspecific Gd-based contrast medium (c, d). The two small metastases are most conspicuous on delayed image (arrows in e) as a result of selective enhancement of liver parenchyma (e, arrows). Variable appearance of the metastases is due to different liver positions during free breathing (a) and breathholding (b–e); slice positioning optimized for visualization of the two small metastases in segment IV (images courtesy of Dr. A. Huppertz, Berlin).

e

contrast medium (SPIO particles or hepatobiliary contrast media) (see Table 1.5). Late postcontrast images after administration of a tissue-specific contrast medium appear to be most suitable to detect hypovascular metastases from colorectal cancer. Several studies have demonstrated that SPIO injection significantly improves detection of hypovascular metastases on late images compared with both unenhanced MRI and dynamic MRI using nonspecific contrast media.52–54 Similar results were obtained with

hepatobiliary contrast media. All three hepatobiliary formulations currently approved for clinical use (Multihance, Teslascan, Primovist) have been shown to improve detection of hypovascular metastases on delayed postcontrast images compared with both unenhanced MRI24,54,55 and dynamic contrast-enhanced MRI using nonspecific contrast media.54 The nonspecific Gd contrast media do not greatly improve detection of hypovascular metastases compared with unenhanced MRI. In some cases, e. g., ex-

32

1 The Liver

a

b

c

d

e

f Fig. 1.30a–f Liver hemangioma on dynamic T1w MRI with delayed imaging after IV bolus injection of hepatobiliary contrast medium (Primovist) (1.5 T). a Axial image obtained with single-shot T2w TSE sequence. b–f Axial images obtained with T1w GRE sequence before (b) and 15 s (c), 55 s (d), 5 min (e), and 15 min after (f) contrast injection. Early dynamic images show nodular peripheral enhance-

ment of FNH with subsequent progressive centripetal fill-in (c, d). Complete homogeneous enhancement of hemangioma after 5 min with loss of lesion-liver contrast due to simultaneous enhancement of normal liver tissue (e). Increased lesion–liver contrast on delayed image due to elimination of contrast medium from blood and persistent enhancement of liver (f) (see Fig. 1.10).

tremely hypovascular malignant tumors, lesion–liver contrast may be improved on images acquired immediately after intravenous injection. In general, however, hypovascular liver lesions, in particular small ones, tend to be less conspicuous on dynamic imaging with Gd-based agents.56 The situation is different for hypervascular liver lesions such as metastases from carcinoids or HCC. The detection of these lesions is improved by arterial phase imaging after intravenous injection of nonspecific contrast media

compared with unenhanced MRI. Moreover, arterial phase images are at least equal, if not superior, to delayed images obtained after injection of a tissue-specific agent (SPIO particles or hepatobiliary contrast media).57,58 However, most studies have investigated the detection of hypervascular liver lesions in patient populations with HCC and cirrhosis. Uptake of both SPIO particles and hepatobiliary agents is reduced in the cirrhotic liver. Compared with biphasic spiral CT, delayed MR images acquired after

Focal Liver Lesions

33

a

b

c

d Fig. 1.31a–e Large FNH on dynamic MRI with delayed imaging after IV bolus injection of hepatobiliary contrast medium (Primovist) (1.5 T). a Axial image obtained with T2w SE sequence. b–e Axial images obtained with T1w GRE sequence before (b) and 15 s (c), 2 min (d), and 2 h (e) after contrast injection. FNH is isointense to liver on unenhanced T2w and T1w images and becomes hyperintense on arterial phase image of dynamic study because it is a hypervascular lesion (c). Subsequent uptake of contrast medium by hepatocytes and FNH results in isointensity after 2 min (d). FNH retains contrast medium longer, resulting in higher SI compared with liver on delayed image (e).

e

administration of a liver-specific contrast medium have been found to be equal or even slightly superior in detecting both hypo- and hypervascular focal liver lesions.55,59 It must be noted, however, that these studies did not use state-of-the-art multislice CT scanners. In another study, the sensitivity of SPIO-enhanced MRI was found to be

equal to or slightly better than that of CT during arterial portography (CTAP), the traditional gold standard for detection of focal liver lesions among the cross-sectional imaging modalities, and its rate of false-positive findings was lower.60

34

1 The Liver

Characterization of Focal Liver Lesions Unenhanced MRI The diagnostic information obtained with unenhanced MRI already enables good differentiation of focal liver lesions, particularly of nonsolid (cysts, hemangiomas) and solid tumors. Solid and nonsolid lesions have different T2 relaxation times61 and morphologic features, which are predominantly identified with T2w imaging. These features were established with the advent of liver MRI and are still valid today.36,62 The most common benign liver tumors, such as hemangioma and FNH, can be differentiated from metastases and malignant tumors of hepatocellular origin using the qualitative morphologic criteria described in the sections on the different entities above. Contrast-Enhanced MRI Serial dynamic imaging after administration of an extracellular nonspecific Gd-based contrast medium such as Magnevist or Dotarem improves lesion characterization by providing information on perfusion patterns. Overall, dynamic MRI has become an important component of the diagnostic procedure in patients with indeterminate liver lesions and significantly improves lesion characterization compared with unenhanced MRI.46,63,64 The experience gained from nonspecific contrast media can be extended to tissue-specific contrast media except that hypervascular tumors lack the typical “blush” in the arterial phase after SPIO administration. In the late or accumulation phase, the tissue-specific contrast media provide functional information on hepatocytes (hepatobiliary agents) and Kupffer cells (SPIO particles),65 thereby improving discrimination of well-differentiated HCC with hepatocellular function or Kupffer cell activity from poorly differentiated HCC and metastases.66–68 Preliminary results suggest that the hepatobiliary agent Teslascan is superior to SPIO particles in this respect.69 The ability to assess cell function is an advantage of MRI over CT in characterizing focal liver lesions.

Diffuse Liver Disease Introduction Diffuse liver disease refers to any destructive or infiltrative process that involves the liver parenchyma, the bile ducts, or the vascular structures.70 Although it can be caused by a wide spectrum of underlying disorders ranging from inflammatory, toxic, and metabolic conditions to neoplastic disease, the reaction of the liver is rather uniform and nonspecific, mainly comprising a limited number of changes such as fatty infiltration, fibrosis, and cholestasis.70,71 Diffuse liver disease is a slowly progressive condition, and morphologic changes detectable by imaging may only become apparent in advanced disease. Blood tests and histologic examination therefore continue to

play an important role in the routine clinical setting. Nevertheless, MRI can often provide important clues to the etiology (iron storage disease) and stage (cirrhosis) of the disease as well as to associated complications (hepatocellular carcinoma, cholestasis). Diagnostic imaging may become more important in the future in light of recent scientific evidence suggesting that even advanced liver fibrosis and cirrhosis may be reversible. These new insights have also spurred the development of new imaging techniques such as liver elastography.71

Indications An overview of the underlying causes of diffuse liver disease is given in Table 1.6. In most instances, laboratory tests and histologic examination are the first-line diagnostic procedures with MRI providing only supplementary information. The foremost role of MRI is to identify associated complications of diffuse liver disease, above all in the presence of cirrhosis. MRI is used to demonstrate hepatic collateral circulation secondary to portal hypertension, rule out HCC, and characterize focal fatty infiltration or focal fatty sparing. The latter may resemble neoplasms and their true nature may not be apparent with other diagnostic modalities.

Imaging Technique Routine MRI for morphologic evaluation in patients with suspected diffuse liver disease can be performed using the same pulse sequences as for focal liver lesions (see Table 1.2). In-phase (IP) and opposed-phase (OP) GRE images are an integral part of the protocol and should be obtained in all patients as they enable sensitive detection of fatty infiltration and iron deposition. All contrast media used for MRI of focal liver lesions can also be used to assess diffuse disease (see Table 1.1), but contrast-enhanced imaging plays only a secondary role in diffuse liver disease. Encouraging results have been obtained with more recent MR techniques, such as diffusion and perfusion imaging, but they are not yet part of the routine diagnostic repertoire.72,73 MR elastography (MRE)74–77 is a rapidly developing new technique with great promise for the diagnostic evaluation of liver fibrosis. MRE provides information on the stiffness of biological tissues by measuring their shear elasticity using a time-resolved motionsensitive MRI sequence that detects the propagation of externally applied mechanical waves in the liver. The shear elasticity of the liver has been shown to change with the progression of fibrosis and can thus be exploited to grade fibrosis. MRE is noninvasive and, unlike liver biopsy, provides information on the entire organ.

Diffuse Liver Disease

35

Table 1.6 Overview of conditions causing diffuse liver disease Etiology

Diseases

Presence of fibrosis / cirrhosis

Key imaging findings

Metabolic and storage disorders

Diabetes/obesity/NASH Hemochromatosis

Rare Yes

Thalassemia Wilson disease

Fat accumulation Signal loss of liver parenchyma most obvious on in-phase GRE images due to iron deposition

Yes

None—copper is not ferromagnetic. Increased attenuation on CT

Acute hepatitis (autoimmune, viral, infectious)

No

Increased signal intensity on T2w images

Chronic viral hepatitis (B,C,D)

Yes

Infectious and inflammatory diseases

Heterogeneous enhancement

Sarcoidosis PBC/PSC

Rare Yes

Alcohol

Yes

Expanded gallbladder fossa sign and enlargement of hilar periportal space in early disease Fibrous septa and regenerative nodules in advanced disease Usually none—scattered nodules might be present Biliary duct abnormalities in PSC Cirrhosis in PBC and end-stage PSC

Toxic

Fat accumulation

Medications Enlargement of the liver (very unspecific!)

Radiation Chemicals Vascular

Budd–Chiari syndrome

Possible

Portal vein thrombosis

Occluded vessels Enlargement of caudate lobe in Budd–Chiari syndrome

Hepatic artery occlusion Neoplastic

Metastatic disease

No

Multiple focal liver lesions

Diffuse HCC/CCC Lymphoma

No

Tumor infiltration may be present along portal tracts

CCC = cholangiocellular carcinoma; HCC = hepatocellular carcinoma; NASH = nonalcoholic steatohepatitis; PBC = primary biliary cirrhosis; PSC = primary sclerosing cholangitis.

MRI Appearance of Pathologic Entities As already mentioned, the parenchymal changes occurring in diffuse diseases of the liver are rather nonspecific despite the heterogeneity of underlying etiologies (Table 1.6). The remainder of this chapter is therefore organized by type of morphologic and structural change rather than disease entity. Fibrosis and cirrhosis are of special interest because they can develop in nearly all diseases of the liver parenchyma and are associated with high morbidity and mortality unless proper treatment is initiated.

Fatty Liver Accumulation of fat in the liver is a very common finding in a wide range of clinical conditions, including alcoholic liver disease, diabetes mellitus, nonalcoholic steatohepatitis (NASH), and obesity. The mere diagnosis of fatty liver or steatosis is of little clinical consequence. Generalized or focal fatty infiltration of the liver is detected using a combination of IP and OP GRE images78–81 (Fig. 1.32). In the presence of fatty infiltration, liver SI is markedly lower

than that of the spleen on OP GRE images. Spectral fat saturation or an inversion recovery sequence will be most helpful in identifying focal fatty infiltration in an otherwise normal liver or focal sparing in diffuse fatty liver.82 The aforementioned MRI techniques enable reliable characterization of these focal lesions, whereas they may be confused with focal neoplasms on CT or ultrasonography78,82 (see Figs. 1.4, 1.24, 1.25).

Iron Deposition In patients with iron storage disease, MRI can be used to detect, quantify, and follow up iron overload in the liver and diagnose potential complications,83,84 but is limited in identifying the underlying cause. The magnetic susceptibility of iron stored in the liver results in a decreased SI of the hepatic parenchyma on all MR pulse sequences.85,86 GRE sequences are more sensitive to T2* effects and are therefore superior to SE sequences in detecting iron.86 The iron overload in the liver can be quantified by calculating transverse relaxation time using a multiecho sequence.87 Noninvasive quantification of the iron overload by MRI is

36

1 The Liver

a

b Fig. 1.32a–c Severe geographic fatty infiltration of the liver (1.5 T). a Axial T2w single-shot TSE image. b, c IP (b) and OP (c) T1w GRE images. Normal appearance on T2w image (a) and IP image (b). On OP image (c), fatty infiltration is indicated by very low SI of most of the parenchyma of the right lobe. There is only moderate fatty infiltration of the left lobe (see Fig. 1.4).

c

an alternative to biopsy86 and can be used to monitor the outcome of therapeutic phlebotomy.83 Iron overload of the liver occurs in many different disorders. Primary or hereditary hemochromatosis is a genetic defect of iron metabolism characterized by excessive intestinal absorption of dietary iron (Fig. 1.33). Excess iron enters the liver, where it is taken up into hepatocytes, but not into cells of the mononuclear phagocyte system (MPS).88 In more advanced disease, excess iron is also stored in the joints, gonads, pancreas, pituitary gland, and myocardium and can cause joint disease, insulindependent diabetes mellitus, hypogonadism, and cardiac dysfunction. In primary hemochromatosis, it is of crucial importance to identify abnormal hepatic iron accumulation before cirrhosis develops (precirrhotic stage) because treatment by phlebotomy can prevent progression to cirrhosis. In secondary hemochromatosis, the iron overload is primarily due to accelerated destruction of red blood cells (Fig. 1.34). Causes are ineffective erythropoiesis or external factors (e. g., increased iron supply due to multiple transfusions). The extra iron is initially stored in the MPS cells of the liver, spleen, and bone marrow.84 Once

their storage capacity is exceeded, iron accumulates in hepatocytes, the pancreas, and myocardium.89 MR signal changes in the spleen distinguish secondary hemochromatosis from the primary form. In primary hemochromatosis, the spleen is usually spared, while other organs, notably the pancreas, show signal alterations. The presence of cirrhosis is highly indicative of primary hemochromatosis. Paroxysmal nocturnal hemoglobinuria is characterized by intravascular hemolysis with release of hemoglobin into the blood, where it initially binds to plasma proteins. After exhaustion of the protein-binding capacity, part of the free hemoglobin is excreted in the urine and part is absorbed.90 Ceruloplasmin deficiency is a genetic disorder in which the lack of ceruloplasmin, which catalyzes the oxidation of ferrous iron to ferric iron, leads to iron deposits in the liver.91 In porphyria cutanea tarda, disturbed heme biosynthesis causes iron accumulation in periportal hepatocytes.92

Diffuse Liver Disease

37

a

b Fig. 1.33a, b Hereditary hemochromatosis (1.5 T). a Axial breath-hold T2w single-shot TSE image. b Axial breath-hold T1w GRE image. The liver has very low SI on both images due to iron deposition and shows signs of cirrhosis. The spleen is enlarged but of normal SI.

a

b Fig. 1.34a, b Secondary hemochromatosis in a man with chronic renal failure and hemodialysis (1.5 T). a Axial breath-hold T2w singleshot TSE image (effective TE, 66 ms). b Axial breath-hold T1w GRE

(TE, 4.6 ms) image. Low SI of the liver and spleen due to iron deposition, which is best seen on T1w image (b). No signs of cirrhosis and normal SI of the pancreas.

Acute Infectious and Inflammatory Disease

by an increased T2 SI of the irradiated area, which gradually returns to normal after the end of radiotherapy.95,96 The main role of MRI in acute inflammation of the liver is to identify complications such as abscess formation or biliary obstruction. Parasitic liver infections should also be briefly mentioned. In hepatic schistosomiasis mansoni, eggs deposited in the periportal area cause inflammation and ultimately fibrosis, leading to portal hypertension. Affected periportal areas are isointense on T1w images and hyperintense on T2w images and enhance intensely after injection of Gd-based contrast medium.97

Acute hepatitis is characterized by the presence of edema, which is typically segmental or periportal in distribution but may occasionally involve the entire liver. Edema is best identified by increased SI on T2w images. In addition, there will be inhomogeneous enhancement of the liver parenchyma after intravenous administration of a nonspecific extracellular Gd-based contrast medium (Fig. 1.35). Severe acute inflammation can cause enlargement of the liver and thickening of the gallbladder wall. It has been shown that uptake of hepatobiliary contrast media such as Mn-DPDP93 and MPS-specific contrast media such as SPIO particles94 is decreased in acute hepatitis. However, this is not a reliable criterion for the diagnosis of hepatitis.93 Radiation-induced hepatitis is characterized

38

1 The Liver

a

b

c

d Fig. 1.35a–d Acute hepatitis (1.5 T). a Axial breath-hold T2w singleshot TSE image. b–d Axial breath-hold T1w GRE images obtained before contrast administration (b) and in the arterial (c) and portal venous (d) phases after IV injection of nonspecific Gd-based con-

trast medium. The liver is enlarged and of high T2 SI, consistent with intrahepatic edema. The parenchyma appears heterogeneous on T1w image (b) and shows heterogeneous enhancement, best seen in the arterial phase (c).

Fibrosis and Cirrhosis

the diagnosis, but new MR techniques, most notably MRE, have the potential to provide a noninvasive alternative in the future.74–77 Definitive imaging changes will be present once the stage of cirrhosis has been reached. Below we outline the course of fibrosis progression from the first manifestations that can be detected by MRI to the imaging features of cirrhosis, finally focusing on the development of HCC in cirrhotic livers. The first MRI signs of cirrhosis are due to early atrophy of the medial segment of the left lobe (segment IV). Enlargement of the hilar periportal space (i. e., the space anterior to the right portal vein) with fat thickness ≥ 1 cm is highly specific for early cirrhosis.98 Another change resulting from atrophy of the left medial segment

Fibrosis and cirrhosis are characterized by the loss and regeneration of liver tissue in the setting of chronic hepatocellular damage. Cirrhosis is the common endpoint of a variety of conditions, primarily including viral hepatitis, autoimmune disorders, toxic damage, and metabolic defects. Nevertheless, the pathophysiologic mechanism underlying the development of cirrhosis is fairly uniform and is termed fibrosis progression.71 It must be noted that MRI will detect changes only if there is fibrosis progression, and very early fibrotic changes are not detectable by any of the currently available MR techniques. Biopsy therefore remains the method of choice for confirming

Diffuse Liver Disease

39

a

b Fig. 1.36a, b Cirrhosis with multiple small siderotic nodules in the liver (1.5 T). a Axial image obtained with respiratory-triggered, fatsuppressed T2w TSE sequence. b Axial image obtained with breathhold T1w GRE sequence. T1w image (b) shows nearly normal liver

contour, normal-sized caudate lobe, and multiple small foci of low SI. Siderotic nodules and high-SI interlobular septa are more conspicuous on T2w image (a).

b

a Fig. 1.37a, b Cirrhosis with mild ascites (1.5 T). a Axial image obtained with T2w single-shot TSE sequence. b Axial breath-hold T1w GRE image. Atrophy of right lobe and left medial segment;

hypertrophy of left lateral segment. High-SI ascites on T2w image (a). Multiple high-SI regenerative nodules are apparent on T1w image (b).

is the expanded gallbladder fossa sign. In the further course of fibrosis progression, narrowing of the hepatic veins occurs and the morphologic changes become more conspicuous. In advanced cirrhosis, there will be additional atrophy of the right hepatic lobe with simultaneous hypertrophy of the left lateral segment and the caudate lobe. The combination of cirrhotic changes may result in nodularity of the liver surface (Figs. 1.36 and 1.37). Other features are presence of fibrous septa and confluent intrahepatic fibrosis (Fig. 1.38), regenerative nodules, recanalization of the umbilical vein in the falciform ligament, and intrahepatic portosystemic venous shunts. MRI will also show extrahepatic manifestations of cirrhosis such as extrahepatic portosystemic venous shunts, splenomegaly, iron deposits in the spleen (Gamna–Gandy bodies), ascites, and edema of the gallbladder wall77,99–102 (Figs.

1.39, 1.40, 1.41; see Fig. 1.33). These features are rather specific for cirrhosis but occur late in the course of disease. While MR contrast media are predominantly used for evaluating focal liver lesions, they may also be useful when imaging diffuse liver disease.103 In cirrhosis, the heterogeneous SI of the liver parenchyma will also persist after intravenous injection of a nonspecific extracellular Gd-based contrast medium and reflects altered perfusion and a change in the profile of the interstitial space in cirrhotic livers. The uptake of cell-specific contrast media such as SPIO particles is reduced, which was attributed to structural changes that possibly alter the distribution and activity of Kupffer cells.94 Diminished uptake was also observed for hepatobiliary contrast media and attributed to impairment of hepatocellular function.104 The characterization of focal liver lesions in the cirrhotic liver, most notably the diagnosis of HCC, is one of

40

1 The Liver

a

b Fig. 1.38a–c Cirrhosis with confluent fibrosis (1.5 T). a Axial breathhold T2w single-shot TSE image. b, c Axial breath-hold T1w GRE images obtained before (b) and after (c) IV injection of nonspecific Gd-based contrast medium. Confluent fibrosis extending from the liver hilum to the capsule and retraction of the capsule. Intense heterogeneous signal enhancement on postcontrast image (c) indicates inflammatory activity of fibrosis.

c

Fig. 1.39 Cirrhosis with portal vein thrombosis and severe gastric fundal varices (1.5 T). Axial image obtained with breath-hold T2w single-shot TSE sequence. The markedly dilated fundal varices are of low SI due to flow-induced signal loss (arrow).

the most important indications for MRI in patients with cirrhosis. HCC must always be a consideration when there are focal lesions in the cirrhotic liver; other liver tumors, especially metastases, are less common in the presence of cirrhosis. Advanced cirrhosis is characterized by the formation of hepatocellular nodules. Histologically, these nodular lesions are classified into regenerative lesions and dysplastic or neoplastic lesions. Regenerative lesions are consid-

ered benign, while dysplastic lesions are regarded as either premalignant or malignant. Accurate classification of hepatic nodules is important but not always possible by MRI or histology because the lesions represent successive stages, which are not always distinct, of a carcinogenic pathway along which regenerative nodules develop into dysplastic nodules, high-grade HCC, and ultimately lowgrade HCC. Radiologists additionally regard siderotic nodules as a distinct entity. They are of low SI on T2w images

Diffuse Liver Disease

41

a

b Fig. 1.40a, b Macronodular cirrhosis with mild ascites (1.5 T). a Axial fat-suppressed respiratory-triggered T2w TSE image. b Breath-hold T1w GRE image.

a

b Fig. 1.41a, b Cirrhosis with marked liver atrophy and severe ascites (1.5 T). Images obtained with breath-hold T2w single-shot TSE sequence. a Axial image. b Coronal image. Air–fluid level (a) secondary to ascites puncture.

(Fig. 1.36) and are more conspicuous on GRE sequences, which are more susceptible to iron deposits and therefore make the lesions appear larger. On T1w GRE images, the nodules are hypointense or hyperintense (Fig. 1.36). Siderotic nodules have virtually no malignant potential.105 In summary, the following focal lesions need to be differentiated in the cirrhotic liver105: · regenerative lesions: micronodular (< 3 mm) or macronodular (≥ 3 mm) · dysplastic lesions: low-grade and high-grade dysplastic nodules (≥ 1 mm) · neoplastic lesions: well, moderately, and poorly differentiated HCC. Several criteria have been proposed to discriminate these lesions: size, T2 signal intensity, vascularity, hepatocellular function (uptake of hepatocellular contrast medium),

and Kupffer cell density (uptake of SPIO-based contrast medium). None of these criteria alone enables reliable characterization. Size. The larger a lesion, the more likely it is HCC, especially if a capsule can be demonstrated. However, regenerative nodules may also reach a size of several centimeters. T2 signal intensity. Hyperintensity suggests a malignant tumor since virtually all regenerative nodules are isointense or hypointense on T2w images. Vascularity. Dedifferentiation is associated with an increase in the proportion of arteries and a decrease in veins. Early intense enhancement on arterial-phase images after contrast administration indicates a high-grade dysplastic nodule or HCC. If there is additional washout on portal venous and/or equilibrium phase images, HCC is likely.106

42

1 The Liver

a

b Fig. 1.42a, b Budd–Chiari syndrome (1.5 T). a Axial breath-hold T2w single-shot TSE image. b Axial breath-hold T1w GRE image. Peripheral edema is indicated by high T2 SI and low T1 SI, predom-

inantly in the right lobe. The hypertrophied medial areas are of nearly normal SI and have a tumorlike appearance with compression of the superior vena cava. Ascites is present.

Hepatocellular function. Regenerative nodules have nearly normal hepatocyte function and their signal enhancement following administration of a hepatocellular contrast medium therefore parallels that of normal liver parenchyma. Hepatocellular function and the corresponding signal enhancement decrease with progressive dedifferentiation at premalignant and malignant stages; welldifferentiated HCCs can have some residual hepatocellular function. Kupffer cell density. Regenerative nodules have the same density of Kupffer cells as normal liver parenchyma and therefore show the same signal decrease on T2w images after administration of SPIO-based contrast medium. Kupffer cell density decreases with dedifferentiation, indicated by a less pronounced signal decrease on SPIO-enhanced images. A slight signal decrease can indicate a well differentiated HCC with residual Kupffer cells. Histologically, focal HCC has been demonstrated within atypical adenomatous hyperplasia and can be seen on T2w images as a high-signal-intensity cancer in a lowsignal-intensity adenomatous hyperplasia (nodule-innodule). This phenomenon is especially conspicuous after injection of SPIO contrast medium.

vena cava with or without secondary occlusion of the hepatic veins; type II is occlusion of the major hepatic veins; and type III is defined as occlusion of the small centrilobular veins. BCS is identified on MRI by the absence of blood flow in the occluded veins. In type II BCS, the inferior vena cava is patent but compressed by the hypertrophied caudate lobe (Fig. 1.42). Absence of blood flow is at times difficult to identify on SE images because flowing blood may cause a signal void or flow-related increase in signal, depending on the pulse sequence parameters used and the plane of imaging relative to the direction of blood flow. Venous thrombi have variable MR signal intensity depending on age and composition. MR angiography is helpful in demonstrating vascular occlusion and collaterals in those cases where the cross-sectional images do not yield a definitive diagnosis.107 Besides direct demonstration of venous occlusion, there are numerous morphologic features that are indicative of BCS, including compensatory hypertrophy of the caudate lobe, intra- and extrahepatic systemic-portal and systemic–systemic venous anastomoses, ascites, inhomogeneous signal intensity, and occasional intraparenchymal hemorrhage107–111 (Fig. 1.42). MRI will also detect some rare causes of congestion such as membranous occlusion of the inferior vena cava110 or leiomyosarcoma of the inferior vena cava.108

Vascular Disease Budd–Chiari syndrome (BCS) is of special clinical relevance among the vascular diseases of the liver, along with thrombosis of the portal vein or hepatic artery, because it often has a fulminant clinical cause. BSC is a congestive liver disease caused by obstruction of hepatic venous outflow with subsequent thrombosis of affected vessels. Three types of BCS are distinguished according to the site of obstruction: type I is occlusion of the inferior

Other Diffuse Liver Diseases Wilson disease is characterized by the abnormal accumulation of copper in hepatocytes. Copper toxicity leads to parenchymal necrosis and scarring of the liver with subsequent cirrhosis. Early stages of the disease cause no MR signal changes because copper is not ferromagnetic and

Diffuse Liver Disease

consequently does not affect the MR signal. Morphologic changes first become apparent on MRI with the onset of cirrhosis.112,113 Hypertrophy of the caudate lobe, a typical feature of other forms of cirrhosis, has been reported to be absent in Wilson disease.114 Sarcoidosis is a chronic granulomatous inflammation that primarily affects mediastinal lymph nodes and the lung with advanced disease occasionally involving the liver. MRI findings include hepatomegaly and scattered nodules in the liver (and spleen). Hepatic granulomas tend to occur along portal tracts and are suggested by increased periportal SI on T2w images.115 Enlarged lymph nodes may be demonstrated in the liver hilum.

17.

18. 19. 20. 21.

References

22.

1. Curley SA, Izzo F, Abdalla E, Vauthey JN. Surgical treatment of colorectal cancer metastasis. Cancer Metastasis Rev 2004;23 (1-2):165–182 2. Junginger T, Kneist W, Seifert JK. [Surgical treatment of colorectal liver metastases]. Zentralbl Chir 2003;128(11):911–919 3. Karhunen PJ. Benign hepatic tumours and tumour like conditions in men. J Clin Pathol 1986;39(2):183–188 4. Dobritz M, Radkow T, Nittka M, Bautz W, Fellner FA. [VIBE with parallel acquisition technique - a novel approach to dynamic contrast-enhanced MR imaging of the liver]. Rofo 2002;174(6): 738–741 5. Kim YK, Kim CS, Chung GH, Jeon SB, Lee JM. Feasibility of application of sensitivity encoding to the breath-hold T2weighted turbo spin-echo sequence for evaluation of focal hepatic tumors. AJR Am J Roentgenol 2005;184(2):497–504 6. Yoshioka H, Sato J, Takahashi N, et al. Dual double arterial phase dynamic MR imaging with sensitivity encoding (SENSE): which is better for diagnosing hypervascular hepatocellular carcinomas, in-phase or opposed-phase imaging? Magn Reson Imaging 2004;22(3):361–367 7. Vogt FM, Antoch G, Hunold P, et al. Parallel acquisition techniques for accelerated volumetric interpolated breath-hold examination magnetic resonance imaging of the upper abdomen: assessment of image quality and lesion conspicuity. J Magn Reson Imaging 2005;21(4):376–382 8. Rofsky NM, Lee VS, Laub G, et al. Abdominal MR imaging with a volumetric interpolated breath-hold examination. Radiology 1999;212(3):876–884 9. Hamm B, Fischer E, Hopfenmüller W, Sander B. [Improvement of MR tomography of the liver using a multi-slice gradient echo sequence]. Rofo 1989;150(3):307–315 10. Stark DD, Hendrick RE, Hahn PF, Ferrucci JT Jr. Motion artifact reduction with fast spin-echo imaging. Radiology 1987;164(1): 183–191 11. Mitchell DG. Chemical shift magnetic resonance imaging: applications in the abdomen and pelvis. Top Magn Reson Imaging 1992;4(3):46–63 12. Catasca JV, Mirowitz SA. T2-weighted MR imaging of the abdomen: fast spin-echo vs conventional spin-echo sequences. AJR Am J Roentgenol 1994;162(1):61–67 13. Kim YH, Saini S, Blake MA, et al. Distinguishing hepatic metastases from hemangiomas: qualitative and quantitative diagnostic performance through dual echo respiratory-triggered fast spin echo magnetic resonance imaging. J Comput Assist Tomogr 2005;29(5):571–579 14. Herborn CU, Vogt F, Lauenstein TC, Goyen M, Debatin JF, Ruehm SG. MRI of the liver: can True FISP replace HASTE? J Magn Reson Imaging 2003;17(2):190–196 15. Haacke EM, Lenz GW. Improving MR image quality in the presence of motion by using rephasing gradients. AJR Am J Roentgenol 1987;148(6):1251–1258 16. Asbach P, Klessen C, Kroencke TJ, et al. Magnetic resonance cholangiopancreatography using a free-breathing T2-weighted turbo spin-echo sequence with navigator-triggered prospective

23. 24.

25. 26. 27.

28.

29. 30.

31. 32. 33.

34. 35.

36. 37.

43

acquisition correction. Magn Reson Imaging 2005;23(9): 939–945 Zech CJ, Herrmann KA, Huber A, et al. High-resolution MRimaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Reson Imaging 2004;20(3):443–450 Saini S, Stark DD, Hahn PF, et al. Ferrite particles: a superparamagnetic MR contrast agent for enhanced detection of liver carcinoma. Radiology 1987;162(1 Pt 1):217–222 Taupitz M, Schmitz S, Hamm B. [Superparamagnetic iron oxide particles: current state and future development]. Rofo 2003;175(6):752–765 Stark DD, Weissleder R, Elizondo G, et al. Superparamagnetic iron oxide: clinical application as a contrast agent for MR imaging of the liver. Radiology 1988;168(2):297–301 Weissleder R, Elizondo G, Stark DD, et al. The diagnosis of splenic lymphoma by MR imaging: value of superparamagnetic iron oxide. AJR Am J Roentgenol 1989;152(1):175–180 Saini S, Edelman RR, Li W, Petersein J, Hahn PF. Clinical evaluation of ultrasmall superparamagnetic iron oxide particles for liver imaging. Acad Radiol 1996;3(Suppl 2):S409–S412 Hamm B, Vogl TJ, Branding G, et al. Focal liver lesions: MR imaging with Mn-DPDP – initial clinical results in 40 patients. Radiology 1992;182(1):167–174 Huppertz A, Balzer T, Blakeborough A, et al; European EOB Study Group. Improved detection of focal liver lesions at MR imaging: multicenter comparison of gadoxetic acid-enhanced MR images with intraoperative findings. Radiology 2004;230(1):266–275 Reimer P, Rummeny EJ, Shamsi K, et al. Phase II clinical evaluation of Gd-EOB-DTPA: dose, safety aspects, and pulse sequence. Radiology 1996;199(1):177–183 Tung GA, Vaccaro JP, Cronan JJ, Rogg JM. Cavernous hemangioma of the liver: pathologic correlation with high-field MR imaging. AJR Am J Roentgenol 1994;162(5):1113–1117 Ito K, Mitchell DG, Outwater EK, Szklaruk J, Sadek AG. Hepatic lesions: discrimination of nonsolid, benign lesions from solid, malignant lesions with heavily T2-weighted fast spin-echo MR imaging. Radiology 1997;204(3):729–737 Soyer P, Dufresne AC, Somveille E, Lenormand S, Scherrer A, Rymer R. Differentiation between hepatic cavernous hemangioma and malignant tumor with T2-weighted MRI: comparison of fast spin-echo and breathhold fast spin-echo pulse sequences. Clin Imaging 1998;22(3):200–210 Lombardo DM, Baker ME, Spritzer CE, Blinder R, Meyers W, Herfkens RJ. Hepatic hemangiomas vs metastases: MR differentiation at 1.5 T. AJR Am J Roentgenol 1990;155(1):55–59 Kreitner KF, Thelen M, Schild H, Heintz A, Störkel S. [Epidemiological and clinical aspects of focal nodular hyperplasia of the liver. An assessment of 886 cases]. Dtsch Med Wochenschr 1987;112(22):891–896 Lee MJ, Saini S, Hamm B, et al. Focal nodular hyperplasia of the liver: MR findings in 35 proved cases. AJR Am J Roentgenol 1991;156(2):317–320 Shamsi K, De Schepper A, Degryse H, Deckers F. Focal nodular hyperplasia of the liver: radiologic findings. Abdom Imaging 1993;18(1):32–38 Paulson EK, McClellan JS, Washington K, Spritzer CE, Meyers WC, Baker ME. Hepatic adenoma: MR characteristics and correlation with pathologic findings. AJR Am J Roentgenol 1994;163(1):113–116 Arrivé L, Fléjou JF, Vilgrain V, et al. Hepatic adenoma: MR findings in 51 pathologically proved lesions. Radiology 1994;193(2):507–512 Chung KY, Mayo-Smith WW, Saini S, Rahmouni A, Golli M, Mathieu D. Hepatocellular adenoma: MR imaging features with pathologic correlation. AJR Am J Roentgenol 1995;165(2): 303–308 Wittenberg J, Stark DD, Forman BH, et al. Differentiation of hepatic metastases from hepatic hemangiomas and cysts by using MR imaging. AJR Am J Roentgenol 1988;151(1):79–84 Mahfouz AE, Hamm B, Wolf KJ. Peripheral washout: a sign of malignancy on dynamic gadolinium-enhanced MR images of focal liver lesions. Radiology 1994;190(1):49–52

44

1 The Liver

38. Rummeny E, Weissleder R, Stark DD, et al. Primary liver tumors: diagnosis by MR imaging. AJR Am J Roentgenol 1989;152(1): 63–72 39. Itoh K, Nishimura K, Togashi K, et al. Hepatocellular carcinoma: MR imaging. Radiology 1987;164(1):21–25 40. Ebara M, Watanabe S, Kita K, et al. MR imaging of small hepatocellular carcinoma: effect of intratumoral copper content on signal intensity. Radiology 1991;180(3):617–621 41. Kadoya M, Matsui O, Takashima T, Nonomura A. Hepatocellular carcinoma: correlation of MR imaging and histopathologic findings. Radiology 1992;183(3):819–825 42. Vogl TJ, Stupavsky A, Pegios W, et al. Hepatocellular carcinoma: evaluation with dynamic and static gadobenate dimeglumineenhanced MR imaging and histopathologic correlation. Radiology 1997;205(3):721–728 43. Fan ZM, Yamashita Y, Harada M, et al. Intrahepatic cholangiocarcinoma: spin-echo and contrast-enhanced dynamic MR imaging. AJR Am J Roentgenol 1993;161(2):313–317 44. Yates CK, Streight RA. Focal fatty infiltration of the liver simulating metastatic disease. Radiology 1986;159(1):83–84 45. Kröncke TJ, Taupitz M, Kivelitz D, et al. Multifocal nodular fatty infiltration of the liver mimicking metastatic disease on CT: imaging findings and diagnosis using MR imaging. Eur Radiol 2000;10(7):1095–1100 46. Hamm B, Reichel M, Vogl T, Taupitz M, Wolf KJ. [Superparamagnetic iron particles. The clinical results in the MR diagnosis of liver metastases]. Rofo 1994;160(1):52–58 47. Kim YK, Ko SW, Hwang SB, Kim CS, Yu HC. Detection and characterization of liver metastases: 16-slice multidetector computed tomography versus superparamagnetic iron oxideenhanced magnetic resonance imaging. Eur Radiol 2006; 16(6):1337–1345 48. Ros PR, Freeny PC, Harms SE, et al. Hepatic MR imaging with ferumoxides: a multicenter clinical trial of the safety and efficacy in the detection of focal hepatic lesions. Radiology 1995; 196(2):481–488 49. Wang C, Ahlström H, Ekholm S, et al. Diagnostic efficacy of MnDPDP in MR imaging of the liver. A phase III multicentre study. Acta Radiol 1997;38(4 Pt 2):643–649 50. Rofsky NM, Weinreb JC, Bernardino ME, Young SW, Lee JK, Noz ME. Hepatocellular tumors: characterization with Mn-DPDPenhanced MR imaging. Radiology 1993;188(1):53–59 51. Reimer P, Rummeny EJ, Daldrup HE, et al. Enhancement characteristics of liver metastases, hepatocellular carcinomas, and hemangiomas with Gd-EOB-DTPA: preliminary results with dynamic MR imaging. Eur Radiol 1997;7(2):275–280 52. del Frate C, Bazzocchi M, Mortele KJ, et al. Detection of liver metastases: comparison of gadobenate dimeglumine-enhanced and ferumoxides-enhanced MR imaging examinations. Radiology 2002;225(3):766–772 53. Kim MJ, Kim JH, Chung JJ, Park MS, Lim JS, Oh YT. Focal hepatic lesions: detection and characterization with combination gadolinium- and superparamagnetic iron oxide-enhanced MR imaging. Radiology 2003;228(3):719–726 54. Kim YK, Lee JM, Kim CS, Chung GH, Kim CY, Kim IH. Detection of liver metastases: gadobenate dimeglumine-enhanced threedimensional dynamic phases and one-hour delayed phase MR imaging versus superparamagnetic iron oxide-enhanced MR imaging. Eur Radiol 2005;15(2):220–228 55. Bartolozzi C, Donati F, Cioni D, et al. Detection of colorectal liver metastases: a prospective multicenter trial comparing unenhanced MRI, MnDPDP-enhanced MRI, and spiral CT. Eur Radiol 2004;14(1):14–20 56. Hamm B, Mahfouz AE, Taupitz M, et al. Liver metastases: improved detection with dynamic gadolinium-enhanced MR imaging? Radiology 1997;202(3):677–682 57. Kwak HS, Lee JM, Kim YK, Lee YH, Kim CS. Detection of hepatocellular carcinoma: comparison of ferumoxides-enhanced and gadolinium-enhanced dynamic three-dimensional volume interpolated breath-hold MR imaging. Eur Radiol 2005;15(1): 140–147 58. Youk JH, Lee JM, Kim CS. MRI for detection of hepatocellular carcinoma: comparison of mangafodipir trisodium and gadopentetate dimeglumine contrast agents. AJR Am J Roentgenol 2004;183(4):1049–1054

59. Hori M, Murakami T, Kim T, et al. Detection of hypervascular hepatocellular carcinoma: comparison of SPIO-enhanced MRI with dynamic helical CT. J Comput Assist Tomogr 2002;26(5): 701–710 60. Ba-Ssalamah A, Heinz-Peer G, Schima W, et al. Detection of focal hepatic lesions: comparison of unenhanced and SHU 555 Aenhanced MR imaging versus biphasic helical CTAP. J Magn Reson Imaging 2000;11(6):665–672 61. Fenlon HM, Tello R, deCarvalho VL, Yucel EK. Signal characteristics of focal liver lesions on double echo T2-weighted conventional spin echo MRI: observer performance versus quantitative measurements of T2 relaxation times. J Comput Assist Tomogr 2000;24(2):204–211 62. Lüning M, Koch M, Abet L, et al. [The accuracy of the imaging procedures (sonography, MRT, CT, angio-CT,nuclear medicine) in characterizing liver tumors]. Rofo 1991;154(4):398–406 63. Whitney WS, Herfkens RJ, Jeffrey RB, et al. Dynamic breath-hold multiplanar spoiled gradient-recalled MR imaging with gadolinium enhancement for differentiating hepatic hemangiomas from malignancies at 1.5 T. Radiology 1993;189(3):863–870 64. Yamashita Y, Hatanaka Y, Yamamoto H, et al. Differential diagnosis of focal liver lesions: role of spin-echo and contrast-enhanced dynamic MR imaging. Radiology 1994;193(1):59–65 65. Jung G, Poll L, Cohnen M, et al. [Differential diagnosis of focal liver lesions using contrast-enhanced MRI with SHU 555 A in comparison with unenhanced MRI and multidetector spiral-CT]. Rofo 2005;177(11):1571–1577 66. Grazioli L, Morana G, Kirchin MA, Schneider G. Accurate differentiation of focal nodular hyperplasia from hepatic adenoma at gadobenate dimeglumine-enhanced MR imaging: prospective study. Radiology 2005;236(1):166–177 67. Kacl GM, Hagspiel KD, Marincek B. Focal nodular hyperplasia of the liver: serial MRI with Gd-DOTA, superparamagnetic iron oxide, and Gd-EOB-DTPA. Abdom Imaging 1997;22(3):264–267 68. Scharitzer M, Schima W, Schober E, et al. Characterization of hepatocellular tumors: value of mangafodipir-enhanced magnetic resonance imaging. J Comput Assist Tomogr 2005; 29(2):181–190 69. Kim MJ, Kim JH, Lim JS, et al. Detection and characterization of focal hepatic lesions: mangafodipir vs. superparamagnetic iron oxide-enhanced magnetic resonance imaging. J Magn Reson Imaging 2004;20(4):612–621 70. Danet I-M, Semelka RC, Braga L. MR imaging of diffuse liver disease. Radiol Clin North Am 2003;41(1):67–87 71. Friedman SL. Liver fibrosis – from bench to bedside. J Hepatol 2003;38(Suppl 1):S38–S53 72. Asbach P, Hein PA, Stemmer A, et al. Free-breathing echo-planar imaging based diffusion-weighted magnetic resonance imaging of the liver with prospective acquisition correction. J Comput Assist Tomogr 2008;32(3):372–378 73. Hagiwara MH, Rusinek H, Lee VS, et al. Advanced liver fibrosis: diagnosis with 3D whole-liver perfusion MR imaging – initial experience. Radiology 2008;246(3):926–934 74. Asbach P, Klatt D, Hamhaber U, et al. Assessment of liver viscoelasticity using multifrequency MR elastography. Magn Reson Med 2008;60(2):373–379 75. Huwart L, Sempoux C, Salameh N, et al. Liver fibrosis: noninvasive assessment with MR elastography versus aspartate aminotransferase-to-platelet ratio index. Radiology 2007; 245(2):458–466 76. Klatt D, Asbach P, Rump J, et al. In vivo determination of hepatic stiffness using steady-state free precession magnetic resonance elastography. Invest Radiol 2006;41(12):841–848 77. Rouvière O, Yin M, Dresner MA, et al. MR elastography of the liver: preliminary results. Radiology 2006;240(2):440–448 78. Hamer OW, Aguirre DA, Casola G, Lavine JE, Woenckhaus M, Sirlin CB. Fatty liver: imaging patterns and pitfalls. Radiographics 2006;26(6):1637–1653 79. Levenson H, Greensite F, Hoefs J, et al. Fatty infiltration of the liver: quantification with phase-contrast MR imaging at 1.5 T vs biopsy. AJR Am J Roentgenol 1991;156(2):307–312 80. Merkle EM, Nelson RC. Dual gradient-echo in-phase and opposed-phase hepatic MR imaging: a useful tool for evaluating more than fatty infiltration or fatty sparing. Radiographics 2006;26(5):1409–1418

Diffuse Liver Disease 81. Mitchell DG, Kim I, Chang TS, et al. Fatty liver. Chemical shift phase-difference and suppression magnetic resonance imaging techniques in animals, phantoms, and humans. Invest Radiol 1991;26(12):1041–1052 82. Mathieu D, Paret M, Mahfouz AE, et al. Hyperintense benign liver lesions on spin-echo T1-weighted MR images: pathologic correlations. Abdom Imaging 1997;22(4):410–417 83. Olthof AW, Sijens PE, Kreeftenberg HG, Kappert P, van der Jagt EJ, Oudkerk M. Non-invasive liver iron concentration measurement by MRI: Comparison of two validated protocols. Eur J Radiol 2008;19: e-pub ahead of print 84. Wood JC, Ghugre N. Magnetic resonance imaging assessment of excess iron in thalassemia, sickle cell disease and other iron overload diseases. Hemoglobin 2008;32(1-2):85–96 85. Gandon Y, Guyader D, Heautot JF, et al. Hemochromatosis: diagnosis and quantification of liver iron with gradient-echo MR imaging. Radiology 1994;193(2):533–538 86. Guyader D, Gandon Y, Robert JY, et al. Magnetic resonance imaging and assessment of liver iron content in genetic hemochromatosis. J Hepatol 1992;15(3):304–308 87. Rocchi E, Cassanelli M, Borghi A, et al. Magnetic resonance imaging and different levels of iron overload in chronic liver disease. Hepatology 1993;17(6):997–1002 88. Fillet G, Beguin Y, Baldelli L. Model of reticuloendothelial iron metabolism in humans: abnormal behavior in idiopathic hemochromatosis and in inflammation. Blood 1989;74(2):844–851 89. Siegelman ES, Mitchell DG, Semelka RC. Abdominal iron deposition: metabolism, MR findings, and clinical importance. Radiology 1996;199(1):13–22 90. Mathieu D, Rahmouni A, Villeneuve P, Anglade MC, Rochant H, Vasile N. Impact of magnetic resonance imaging on the diagnosis of abdominal complications of paroxysmal nocturnal hemoglobinuria. Blood 1995;85(11):3283–3288 91. Logan JI, Harveyson KB, Wisdom GB, Hughes AE, Archbold GP. Hereditary caeruloplasmin deficiency, dementia and diabetes mellitus. QJM 1994;87(11):663–670 92. Campo E, Bruguera M, Rodés J. Are there diagnostic histologic features of porphyria cutanea tarda in liver biopsy specimens? Liver 1990;10(3):185–190 93. Tanimoto A, Kreft BP, Baba Y, et al. Evaluation of hepatocytespecific paramagnetic contrast media for MR imaging of hepatitis. J Magn Reson Imaging 1993;3(5):786–793 94. Elizondo G, Weissleder R, Stark DD, et al. Hepatic cirrhosis and hepatitis: MR imaging enhanced with superparamagnetic iron oxide. Radiology 1990;174(3 Pt 1):797–801 95. Unger EC, Lee JK, Weyman PJ. CT and MR imaging of radiation hepatitis. J Comput Assist Tomogr 1987;11(2):264–268 96. Yankelevitz DF, Knapp PH, Henschke CI, Nisce L, Yi Y, Cahill P. MR appearance of radiation hepatitis. Clin Imaging 1992;16(2): 89–92 97. Willemsen UF, Pfluger T, Zoller WG, Kueffer G, Hahn K. MRI of hepatic schistosomiasis mansoni. J Comput Assist Tomogr 1995;19(5):811–813

45

98. Ito K, Mitchell DG, Gabata T. Enlargement of hilar periportal space: a sign of early cirrhosis at MR imaging. J Magn Reson Imaging 2000;11(2):136–140 99. Itai Y, Kurosaki Y, Saida Y, Niitsu M, Kuramoto K. CT and MRI in detection of intrahepatic portosystemic shunts in patients with liver cirrhosis. J Comput Assist Tomogr 1994;18(5):768–773 100. Mergo PJ, Ros PR, Buetow PC, Buck JL. Diffuse disease of the liver: radiologic-pathologic correlation. Radiographics 1994;14(6): 1291–1307 101. Ohtomo K, Baron RL, Dodd GDIII, Federle MP, Ohtomo Y, Confer SR. Confluent hepatic fibrosis in advanced cirrhosis: evaluation with MR imaging. Radiology 1993;189(3):871–874 102. Rofsky NM, Fleishaker H. CT and MRI of diffuse liver disease. Semin Ultrasound CT MR 1995;16(1):16–33 103. Hughes-Cassidy F, Chavez AD, Schlang A, et al. Superparamagnetic iron oxides and low molecular weight gadolinium chelates are synergistic for direct visualization of advanced liver fibrosis. J Magn Reson Imaging 2007;26(3):728–737 104. Murakami T, Baron RL, Federle MP, et al. Cirrhosis of the liver: MR imaging with mangafodipir trisodium (Mn-DPDP). Radiology 1996;198(2):567–572 105. Hanna RF, Aguirre DA, Kased N, Emery SC, Peterson MR, Sirlin CB. Cirrhosis-associated hepatocellular nodules: correlation of histopathologic and MR imaging features. Radiographics 2008;28(3):747–769 106. Roncalli M, Roz E, Coggi G, et al. The vascular profile of regenerative and dysplastic nodules of the cirrhotic liver: implications for diagnosis and classification. Hepatology 1999;30(5): 1174–1178 107. Kane R, Eustace S. Diagnosis of Budd-Chiari syndrome: comparison between sonography and MR angiography. Radiology 1995;195(1):117–121 108. Cacoub P, Piette JC, Wechsler B, et al. Leiomyosarcoma of the inferior vena cava. Experience with 7 patients and literature review. Medicine (Baltimore) 1991;70(5):293–306 109. Miller WJ, Federle MP, Straub WH, Davis PL. Budd-Chiari syndrome: imaging with pathologic correlation. Abdom Imaging 1993;18(4):329–335 110. Park JH, Han JK, Choi BI, Han MC. Membranous obstruction of the inferior vena cava with Budd-Chiari syndrome: MR imaging findings. J Vasc Interv Radiol 1991;2(4):463–469 111. Shapiro RS, Maldjian JA, Stancato-Pasik A, Ramos R. Hepatic mass in Budd-Chiari syndrome: CT and MRI findings. Comput Med Imaging Graph 1993;17(6):457–460 112. Nakakoshi T, Fujita N, Jong-Hon K, Takeichi N, Miyasaka K. Influence of in vivo copper on MR images of the liver in rats. J Magn Reson Imaging 1994;4(4):559–562 113. Vogl TJ, Steiner P, Hammerstingl R, et al. MRT der Leber bei Morbus Wilson. Fortschr Röntgenstr. 1994;160:40–45 114. Akpinar EO, Akhan O. Liver imaging findings of Wilson’s disease. Eur J Radiol 2007;61(1):25–32 115. Warshauer DM, Lee JK. Imaging manifestations of abdominal sarcoidosis. AJR Am J Roentgenol 2004;182(1):15–28

47

2

The Bile and Pancreatic Ducts P. Asbach and H.B. Gehl

Introduction

Indications

Magnetic resonance cholangiopancreatography (MRCP) has evolved into an important noninvasive imaging modality for routine clinical evaluation of the pancreatobiliary tree. MRCP is a very elegant imaging tool, as it is fast and straightforward and has very few adverse effects. While invasive visualization of the bile ducts and pancreatic duct using conventional radiographic techniques requires intravenous injection of a biliary contrast medium (cholangiography) or direct instillation of contrast medium into the ducts (percutaneous transhepatic cholangiography [PTC] or endoscopic retrograde cholangiopancreatography [ERCP]), noninvasive MRCP exploits the very long T2 relaxation times of fluids for visualization of the ductal system. Conventional invasive procedures carry considerable risks, including the potential adverse effects of contrast administration. They are examiner-dependent and stressful for the patient. The most feared complications of PTC include bleeding and injuries to the bile ducts, while ERCP may cause duodenal perforation, pancreatitis,1,2 cholangitis, and duct perforation. Complication rates between 4 % and 11 % have been reported for ERCP.2,3 Another major advantage of MRCP is that it also provides more comprehensive diagnostic information on adjacent anatomic structures and organs, thus contributing to an earlier diagnosis. ERCP, on the other hand, is not only a diagnostic tool but, more importantly, can also be used to perform therapeutic interventions in the same session. Noninvasive MRCP and invasive ERCP therefore supplement each other and together are a mainstay of patient care. A comparative cost-benefit analysis of ERCP and MRCP concludes that primary diagnosis with MRCP is economically advantageous,4 which is an important consideration for health-care providers in the era of cost cutting and reimbursement based on diagnosis-related groups (DRG).

Noninvasive MRCP of the pancreatobiliary tree can contribute valuable diagnostic information in patients with suspected involvement of the bile ducts or pancreatic duct if the clinical symptoms and the first-line imaging test (ultrasound) do not yield a definitive diagnosis. This is primarily the case in patients with suspected biliary pancreatitis, acute or chronic cholangitis, strictures or anomalies of the bile ducts (e. g., Caroli disease, choledochocele, biliary atresia), anatomic variants (pancreas divisum), and cholangiocarcinoma (e. g., Klatskin tumor). In patients scheduled for an interventional procedure, preinterventional localization of strictures and stones and detection of pseudocysts are crucial for planning the approach. Imaging provides the clinician with detailed information on the anatomic situation, thereby facilitating interventions such as laparoscopic surgery.

Imaging Technique Coils High spatial resolution is crucial for detailed anatomic visualization of the target structures (very small bile ducts, side branches of the pancreatic duct), which is not achieved with the integrated whole-body resonator. Surface coils have the essential advantage of providing a better signal-to-noise ratio (SNR), which improves spatial resolution and contrast while also shortening imaging time (breath-hold sequences, see below). However, the most suitable coil is a multielement body or torso phased-array surface coil (usually consisting of 4–8 elements, which, in addition to the aforementioned advantages, enables use of a larger field of view (FOV) and reduces artifacts. When a multielement coil is used, the entire biliary tree and pancreatic duct as well as the liver and pancreas can be imaged in a single session without coil repositioning. Conventional imaging of the liver and pancreas should be part of any routine MRCP protocol.5

48

2 The Bile and Pancreatic Ducts

a

b Fig. 2.1a, b Illustration of the effect of negative oral contrast medium in a 58-year-old man with recurrent pancreatitis and prior biliary–enteric anastomosis who underwent MRCP to rule out cholestasis. a Thick-slab coronal MRCP image from T2w single-slice HASTE sequence obtained before administration of negative oral contrast medium. Due to heavy T2 weighting, the image also depicts the fluid-filled stomach and proximal small intestine

(arrows). The pancreatic duct appears normal. Also seen is a parapelvic renal cyst causing hydronephrosis of the left kidney. Such fluid-filled structures may obscure parts of the pancreatic duct. b Thick-slab coronal MRCP image from T2w single-slice HASTE sequence acquired after negative oral contrast medium. The signal from the stomach and proximal small intestine is effectively suppressed.

Pulse Sequences

ures the position of the diaphragm to synchronize acquisition accordingly (PACE technique: prospective acquisition correction – Siemens) and was originally developed for cardiac imaging. The main advantage of respiratory navigator triggering is that the longer acquisition time allows use of sequences with much better spatial resolution.8

MRCP exploits the long T2 relaxation time of fluids (several seconds for bile) for indirect visualization of ductal structures. At the same time, signal from other tissue structures is effectively suppressed when heavily T2weighted sequences are used because the echo time (TE) of these sequences is a multiple of the T2 relaxation time of stationary tissues (40–100 ms in the abdomen). However, there may be overlap from other fluid-containing structures such as cysts (pancreatic pseudocysts, hepatic and renal cysts), intestinal loops, and the renal collecting system and ureter. Sequences based on the rapid acquisition with relaxation enhancement (RARE) technique,6 such as the halfFourier acquisition single-shot turbo spin echo (HASTE) sequence, are among the most widely used sequences for MRCP. The examination can be performed in one of two complementary ways: acquisition of a single thick-slab projection image and 2D acquisition of a series of thin slices with subsequent reconstruction using the MIP (maximum intensity projection) algorithm.7 While thickslab imaging (which takes < 6 s) is well tolerated during breath-hold, multislice acquisition requires a breath-hold of ca. 20 s, which is beyond the capacity of some patients. Alternatively, MRCP can be performed with the patient breathing freely if acquisition is synchronized with the respiratory cycle. Various techniques are available from different vendors, but the equipment is cumbersome to handle and has found little acceptance. A practical, software-based alternative is respiratory triggering using the navigator echo approach. The navigator technique meas-

Patient Preparation Oral Contrast Media When heavily T2w thick-slab images are acquired at various projections along the course of the pancreatic and biliary ducts, there will invariably be some overlap from other fluid-filled structures, particularly the stomach and duodenum. The signal from gastrointestinal fluids can be suppressed by administering a negative oral contrast medium containing iron oxide particles. Since such contrast media do not distend the ducts, normal anatomy is not distorted.9 However, one must be aware that negative oral contrast media will also obscure anatomic landmarks such as the descending duodenum (Fig. 2.1). There is controversy in the literature as to whether the diagnostic quality of MRCP can be improved most effectively by suppressing or enhancing the gastrointestinal (GI) signal. Positive oral contrast media cause the GI lumen to appear bright, thereby improving delineation of the liver and pancreas from intestinal loops if MRCP is supplemented by conventional axial MRI of these two organs. The latter is desirable in all patients undergoing

Imaging Technique

MRCP because it will identify possible external causes of ductal narrowing as well as other pathology not affecting the ducts (see below). Finally, the decision whether or not to use oral contrast medium is also affected by increasing pressure to reduce health-care costs. In this context, blueberry juice, which has the same effects on T2w images as negative oral contrast medium, has been proposed as an inexpensive alternative.10 Some investigators dispense with oral contrast administration altogether; in this case, MRCP should be performed in the morning after fasting when only a small amount of fluid is present in the upper GI tract.

Other Medication Motion artifacts due to respiration and upper GI tract peristalsis are a major problem for MRCP, given the small size of some of the ductal structures. Most patients can hold their breath for thick-slab imaging, which has an acquisition time of less than 10 s. Peristalsis should be suppressed by intramuscular injection of an antispasmodic at the beginning of the examination unless a patient has contraindications. Visualization of the pancreatic ductal system (especially side branches and the accessory duct) can be improved by secretin stimulation. Intravenous secretin injection enhances pancreatic exocrine function, which improves filling of the ducts and thus allows more accurate evaluation of ductal morphology. In addition, pancreatic exocrine function can be assessed by quantitative determination of subsequent duodenal filling.11,12 Imaging 5 min after intravenous injection yields the best results. Secretin stimulation is contraindicated in patients with acute pancreatitis or acute on chronic pancreatitis.

Recommended Imaging Protocol General Preparation MRCP is best performed in the morning when secretion in the upper GI tract is minimal. Patients should fast for at least 6 h and abstain from even small amounts of fluid before the examination.

Oral Contrast Media We usually perform MRCP with an oral contrast medium solution, which the patient starts drinking approximately 30 min before the examination (at least 300 mL; 600 mL creates optimal imaging conditions). The contrast solution can be mixed 1:1 with water to prevent susceptibility artifacts that would result from sedimentation of iron oxide particles in the immobilized GI tract.

49

Imaging of Upper Abdominal Organs An MRCP protocol should comprise conventional plain MRI of the liver and pancreas. If these images reveal a lesion requiring further characterization after intravenous contrast administration (Gd-based, low-molecular-weight contrast medium), MRCP is followed by a dynamic contrast-enhanced MRI study of the respective organ (for further details see the chapter on the liver or pancreas). Intravenous contrast media do not impair the quality of the MRCP examination, and their T2-shortening effect even reduces interference from overlying contrast-enhancing structures on thick-slab images, especially the renal collecting systems and ureters.

MRCP of the Bile and Pancreatic Ducts The MRCP sequences we use are summarized in Table 2.1. The examination begins with acquisition of a fast axial T2w multislice sequence (e. g., T2w HASTE) that covers the entire upper abdomen and is used to plan the subsequent MRCP sequences. Next, a coronal thick-slab projection image (120 mm) is acquired with a fast, fat-saturated T2w sequence and a FOV adapted to the imaging plane. A single projection image acquired in this way will depict the central bile ducts and most of the pancreatic duct (Fig. 2.2a). Such an image has a low SNR but nevertheless provides a good anatomic overview of the biliary and pancreatic duct systems. Additional thick-slab acquisitions are then performed with projections and slice thicknesses (30–80 mm) selected according to the duct segments imaged. Optimal orientations for targeted visualization of specific parts of the ducts depend on the patient’s anatomy, which is why only some very general suggestions can be made here (in terms of a clockface): parallel to the pancreatic duct (projection between 8 and 9 o’clock); parallel to the common bile duct (projection at approximately 10 o’clock); parallel to the right and left hepatic ducts (highly variable). Additional imaging planes may be acquired as deemed necessary depending on the presence of normal anatomic variants and site of pathology (Fig. 2.2b–d). Thick-slab imaging is followed by multislice, thin-slice acquisitions (3 mm, no interslice gap) for evaluating very small ductal structures. The individual series of thin slices are acquired at the same angles as the corresponding thick-slab images. The acquisition time for one such series is ca. 20–23 s. We perform thin-slice MRCP with a T2w multislice HASTE sequence (Fig. 2.3a–c). The source data can be reconstructed for 3D display using the MIP algorithm (Fig. 2.3c); however, the source images are sufficient for evaluation. Figure 2.3 d shows an MIP reconstruction of a fat-saturated T2w multislice TSE sequence (1.5-mm slice thickness) acquired with respiratory navigator triggering (PACE). This sequence has become the mainstay of any MRCP examination because it yields

Weighting

Plane

Sequence type

TR (ms)

TE (ms) Flip (°)

50

Table 2.1 Recommended pulse sequences and imaging parameters for MRCP ETL

FS

Matrix

No. of acquisitions

Slice thickness (mm)

Scan time

Breathhold

23

1

7

ca. 18 s

Yes

1

1

120

ca. 6 s

Yes

1

1

30–80

ca. 6 s

Yes

15

1

3

ca. 20 s

Yes

40

1

1.5

ca. 4–8 min No

Note: Use of a multielement body or torso phased-array surface coil is recommended for all sequences. * TR here is a technical parameter referring to the intervals between slice acquisitions; physical TR = ∞ since only one excitation pulse is applied per slice.

2 The Bile and Pancreatic Ducts

Scout sequence for planning subsequent MRCP acquisitions 800* 63 150 115 No 115 × 256 T2 Axial Single-shot TSE with halfFourier acquisition (e. g., HASTE) Thick-slab MRCP overview of ductal anatomy T2 Coronal Single-shot TSE with half– 1100 150 256 Yes 256 × 256 Fourier acquisition (e. g., HASTE) Thick-slab MRCP acquisition of multiple projections adjusted to individual ducts of interest – 1100 150 256 Yes 256 × 256 T2 Paracoronal Single-shot TSE with halfRun 3 × Fourier acquisition (e. g., HASTE) Multislice sequence (with MIP reconstruction as needed), same projections as for thick-slab acquisitions T2 Paracoronal Single-shot TSE with half1100* 87 150 218 No 218 × 256 Run 3 × Fourier acquisition (e. g., HASTE) Multislice sequence with respiratory navigator (and MIP reconstruction) T2 Paracoronal TSE (FSE) 1910 832 180 145 Yes 384 × 384

No. of slices

Imaging Technique

51

a

b

c

d Fig. 2.2a–d Normal MRCP findings after negative oral contrast medium in a 38-year-old woman. a Thick-slab (120-mm) coronal MRCP image from T2w single-slice HASTE sequence. b–d Thick-slab

(50 mm) paracoronal MRCP (same sequence as a) aligned along pancreatic duct (b), common bile duct (c), and right hepatic duct (d).

images with higher resolution and fewer artifacts, and the data set lends itself to multiplanar reconstruction.8 Depending on the findings, additional axial images may be acquired with the T2w multislice HASTE sequence. Axial images will even show very small intraductal filling defects, since the ducts are nearly perpendicular to the

axial plane, and allow differentiation of air bubbles (located in the nondependent portion of the duct) and small stones (located in the dependent portion). In patients with a stent, it is sometimes possible to differentiate stent obstruction (low signal intensity) from a patent stent (intraluminal fluid signal) (Fig. 2.4).

52

2 The Bile and Pancreatic Ducts

a

b

c

d Fig. 2.3a–d MRCP without negative oral contrast medium in a 62-year-old woman. a, b (Para-)coronal fat-suppressed T2w multislice HASTE sequences aligned along pancreatic duct (a) and com-

MRI Appearance of Normal Anatomy MRCP plays an important role in the preoperative evaluation of pancreaticobiliary ductal anatomy. Detailed anatomic information is especially important prior to minimally invasive surgery because the small incisions for instrument insertion have to be planned beforehand and cannot easily be changed during the intervention if an unexpected anatomic variant is encountered.10 Normal anatomic variants of the pancreatic and biliary ducts are quite common, the most important being pancreas divisum (see Chapter 3, p. 71). Surgical and imaging studies have demonstrated pancreas divisum in 1.5–10 % of indi-

mon bile duct (b). c MIP image created from the slab (15 slices) in a. d Coronal MIP image from T2w multislice TSE sequence acquired with respiratory navigator (PACE; 40 slices).

viduals.10,13,14 Pancreas divisum may be symptomatic and can cause (chronic) pancreatitis.15 While in the normal pancreas the secretions produced in the body and tail enter the duodenum through the major papilla (papilla of Vater), pancreas divisum is drained via an accessory duct (Santorini duct) that lies anteriorly in the head of pancreas and terminates in the minor papilla. The lumen of the accessory duct and/or opening of the minor duodenal papilla is often small, obstructing proper drainage of pancreatic exocrine secretions, which may leak into the pancreatic tissue and ultimately cause chronic pancreatitis. In some patients, pancreas divisum is associated with cystic dilatation of the terminal portion of the accessory duct, a condition known as santorinicele.16 A pancreatic

MRI Appearance of Pathologic Entities

53

a

b

c

d Fig. 2.4a–d Caroli syndrome in a 29-year-old woman. Massive cystic dilatation of the intra- and extrahepatic bile ducts. Patient presented with recurrent symptoms of cholestasis after stenting of the common bile duct. MRCP for evaluation of the anatomic situation and stent position. a Coronal MRCP image from T2w single-slice HASTE sequence. b, c Coronal (b) and axial (c) T2w multislice (3-mm)

HASTE sequences. There is massive dilatation of the bile ducts; the choledochal stent is seen as a hypointense round structure (arrow in b). d Axial T2w multislice HASTE sequence. Stent occlusion is suggested by in-stent signal void and was confirmed after stent replacement.

duct crossing the common bile duct and terminating in the duodenum proximal to the major papilla seen on coronal MRCP images is pathognomonic of pancreas divisum (Fig. 2.5; see Fig. 2.10). For comparison, the normal configuration is shown in Fig. 2.6, and another normal variant with separate duodenal openings of the common bile duct and pancreatic duct in Fig. 2.7b. Other common variants include an aberrant right hepatic duct, which needs to be identified in patients who undergo liver surgery, and variable insertion of the cystic duct.

MRI Appearance of Pathologic Entities Ductal Strictures One of the most common indications for MRCP of both the bile ducts and the pancreatic duct is to locate strictures or stenoses and evaluate ductal structures proximal to the narrowing. Relevant stenoses can be excluded if MR images show a small-caliber duct without abrupt narrowing (Fig. 2.6). The causes of strictures or stenoses (inflammation, stone, tumor) can be differentiated on conventional,

54

2 The Bile and Pancreatic Ducts

a

b

c

d Fig. 2.5a–e MRCP findings 2 months after acute pancreatitis in a 53-year-old man with pancreas divisum. MRCP for evaluation of ductal structures and pancreatic parenchyma. a Coronal MIP image from T2w multislice TSE sequence (acquired with respiratory navigator, PACE). The pancreatic duct crosses the distal common bile duct. b Paracoronal MIP image from T2w multislice TSE sequence. The ducts cross each other and drain separately into the duodenum. c, d Axial T2w multislice TSE sequence (PACE). The pancreatic duct (arrow) courses anterior to the common bile duct. e ERCP image shows nonopacification of the pancreatic duct during cannulation of the major papilla and normal appearance of the common bile duct.

e

MRI Appearance of Pathologic Entities

55

a

b Fig. 2.6a, b Normal MRCP findings in a 56-year-old man. a Coronal MIP image from T2w multislice TSE sequence (acquired with respiratory navigator, PACE). Small bile ducts and small pancreatic duct. Note incomplete coverage of the intrahepatic bile ducts in the right lobe, which must not be mistaken for pathology. b Paracoronal MIP image from T2w multislice TSE sequence (PACE). Slightly tortuous

course of the cystic duct (normal appearance). The major duodenal papilla is not depicted on the T2w images because the sphincter of Oddi was contracted and there was no fluid in the papilla at the time of imaging. This appearance is normal and must not be misinterpreted as a papillary stone, which is why the papilla should additionally be evaluated on axial images.

axial or coronal, sequences (including a dynamic contrastenhanced series if necessary) (Figs. 2.7 and 2.8).

chronic pancreatitis (see Chapter 3, “The Pancreas”), while demonstration of a communication between the cyst and the duct may not always be possible. When evaluating a patient for postinflammatory strictures, great care is necessary to differentiate true luminal narrowing from incomplete depiction of the pancreatic duct on paracoronal images (see Fig. 2.6a). Differentiation of inflammatory changes from a solid malignancy often presents a diagnostic challenge (see Chapter 3). Imaging evaluation of the pancreatic duct may be helpful in such cases. Demonstration of an unobstructed pancreatic duct in an area where malignancy is suspected makes inflammation the more likely cause (so-called duct-penetrating sign)21 (see Chapter 3, in particular Fig. 3.18). In contrast, a solid malignant tumor is typically associated with complete obstruction of the duct.

Stones Ultrasonography is the primary imaging modality for diagnosing stones in the gallbladder but has limited accuracy in evaluating choledocholithiasis and will miss up to 80 % of asymptomatic stones in the common bile duct.17 In contrast, MRCP has a sensitivity of over 95 % for detecting stones in the common bile duct18,19 (Figs. 2.9 and 2.10) or pancreatic duct (Figs. 2.7 and 2.10) and is superior to ERCP in detecting intrahepatic stones.20 MRCP also plays a role in diagnosing gallbladder stones in cases where small gallbladder polyps must be considered in the differential diagnosis. Gallbladder stones are shown in Figs. 2.11 and 2.12, polyps in Fig. 2.13. A rare differential diagnosis is adenomyomatosis of the gallbladder (Fig. 2.14). In patients who have undergone cholecystectomy, stones in the cystic duct stump occasionally cause upper abdominal symptoms. The common bile duct can have a diameter of up to 10 mm after cholecystectomy (Fig. 2.15).

Inflammatory Conditions MRCP is commonly indicated for the evaluation of the pancreatic duct proximal to a stenosis or stricture in patients with incomplete ERCP (Figs. 2.7 and 2.8). Strictures frequently occur secondary to inflammation. T2w sequences also reliably depict pseudocysts associated with

Primary Biliary Cirrhosis Primary biliary cirrhosis (PBC) is a chronic granulomatous inflammation of the peripheral intrahepatic bile ducts. Early macroscopic changes are not seen on MRCP images, which is why the indication for MRCP in suspected PBC is to rule out other causes. Extrahepatic bile ducts are not involved. Lymphadenopathy in the liver hilum is an indirect, nonspecific sign of PBC. The cause of the disease is unknown, but an autoimmune etiology is likely. It slowly progresses to liver cirrhosis, resulting in morphologic changes of the bile ducts that are visualized on MR images.

56

2 The Bile and Pancreatic Ducts

a

b

c

d

e

f Fig. 2.7a–f Recurrent acute pancreatitis in a 41-year-old man. a Coronal MIP image from T2w multislice TSE sequence (acquired with respiratory navigator, PACE). Cutoff of the pancreatic duct in the pancreatic body (arrow) and marked dilatation in the tail. Also seen in the tail are dilated side branches. b Paracoronal MIP reconstruction from T2w TSE multislice sequence (PACE). Normal variant with separate duodenal openings of the common bile duct and pancreatic duct. Normal appearance of intra- and extrahepatic bile ducts. c Axial T2w multislice HASTE sequence. Cutoff of the pan-

creatic duct in the body of the gland (arrow). d Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). Stone in the pancreatic duct (arrow). Inflammatory changes of the pancreatic parenchyma and fluid collections around the pancreatic head and body. e ERCP image before contrast administration already shows multiple calcifications projected over the pancreas. f ERCP image after instillation of contrast medium into the pancreatic duct confirms ductal discontinuity and stone.

MRI Appearance of Pathologic Entities

57

a

b

c

d Fig. 2.8a–e Suspected pancreatic mass in a 57-year-old man. a Axial T1w 2 D multislice FLASH sequence. b Axial T2w multislice HASTE sequence. There is dilatation of the pancreatic duct beginning at the junction of the body and tail (arrow). c Coronal T2w multislice HASTE sequence. Low-SI structure in the pancreatic duct (arrow), consistent with intraductal detritus. d MIP image created from the T2w multislice HASTE sequence. e MIP image from a T2w multislice TSE sequence acquired with respiratory navigator (PACE). There is mild dilatation of the common bile duct, central intrahepatic bile ducts, and pancreatic duct in the head and body of the gland caused by papillary sclerosis. The suspected pancreatic mass was not confirmed.

e

58

2 The Bile and Pancreatic Ducts

a

b

c

d Fig. 2.9a–d Choledocholithiasis and episodes of biliary colic in a 55-year-old woman. a, b Axial T2w multislice TSE sequence. Stones are seen as round structures of low SI in the common bile duct

(arrow in a ). Arrow in b indicates a prepapillary stone. c, d (Para-) coronal T2w multislice HASTE sequence. Numerous stones in the bile duct at the level of the pancreatic head.

Fig. 2.10a–f MRCP without oral contrast administration in a 66-year-old man presenting with acute right upper abdominal pain. Abdominal ultrasound evaluation was incomplete due to large amounts of air in the colon. Elevated cholestasis and inflammatory markers. MRCP to rule out choledocholithiasis. a Axial T2w multislice HASTE sequence. There is a large stone in the distal common bile duct (arrow), which was subsequently extracted during ERCP. b, c Axial T2w multislice HASTE sequence. Pancreas divisum. The

pancreatic duct crosses the common bile duct (arrow) and enters the duodenum via the minor papilla (arrow). d Coronal MIP image generated from the T2w multislice HASTE sequence in b. The pancreatic duct appears normal and crosses the common bile duct at the level of the stone. e, f Thick-slab coronal (e) and paracoronal (f) MRCP images from T2w single-slice HASTE sequence show multiple stones in the gallbladder and cystic duct. e

MRI Appearance of Pathologic Entities

59

a

b

c

d

e

f

60

2 The Bile and Pancreatic Ducts

a

b Fig. 2.11a–c Incidental finding of gallbladder stones in a 58-yearold woman with suspected hepatocellular carcinoma. a Axial T1w 2 D multislice FLASH sequence. b Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). The stone has low T1 and T2 SI. c Axial T2w multislice TSE sequence (PACE). Rounded, 2-mm structure of low SI in the dependent portion of the descending duodenum (arrow), consistent with a stone passed into the duodenum.

c

a

b Fig. 2.12a, b Incidental finding of numerous small stones in the dependent portion of the gallbladder in a 75-year-old man who underwent MRI work-up of a focal liver lesion. a Axial T2w multislice

HASTE sequence. b Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE).

MRI Appearance of Pathologic Entities

61

a

b

c

d Fig. 2.13a–d Small lesions contiguous with the gallbladder wall noted on ultrasound in a 52-year-old man. MRI for differential diagnosis—gallbladder polyps vs cholesterol sediments. a Axial T2w multislice HASTE sequence. b Coronal T2w multislice HASTE sequence. c, d T1w multislice 2 D FLASH sequence (VIBE) 55 s after

IV contrast bolus injection (0.1 mmol/kg Gd). The gallbladder contains numerous rounded lesions attached to the wall and up to 2 mm in size. Enhancement of the lesions after contrast injection establishes the diagnosis of polyps.

Primary Sclerosing Cholangitis

wall irregularities of the intra- and extrahepatic bile ducts as the disease progresses22 (Fig. 2.16).

Primary sclerosing cholangitis (PSC) is an idiopathic chronic inflammation of the bile ducts, leading to cholestasis and, in advanced disease, secondary biliary cirrhosis.22 There is a close association with chronic inflammatory bowel disease, in which case the condition is called secondary sclerosing cholangitis. PSC is considered a risk factor for the development of cholangiocarcinoma. MRCP will show a “beaded” appearance of the central bile ducts due to strictures alternating with normal or dilated ductal segments in early disease and multifocal strictures and

Masses Compression by a tumor should be considered in the differential diagnosis of nearly all cases of luminal narrowing or (sub-)total occlusion of a duct. Since tumors are best excluded in the axial plane, additional axial images should be obtained whenever MRCP images show ductal discontinuity. MRCP allows excellent evaluation of the

62

2 The Bile and Pancreatic Ducts

a

b

c

d

e

Fig. 2.14a–e Diffuse adenomyomatosis of the gallbladder (hyperplastic cholecystosis) in a 43-year-old man presenting with colicky symptoms. a Axial T2w multislice HASTE sequence. b Axial fat-suppressed T2w multislice HASTE sequence. There is characteristic thickening of the gallbladder wall (due to overgrowth of the mucosa and hypertrophy of the muscular layer). Increased intraluminal pressure leads to formation of intramucosal fluid collections resembling diverticula known as Rokitansky–Aschoff sinuses, which are pathognomonic of adenomyomatosis of the gallbladder. c Axial T1w 2 D multislice FLASH sequence 15 s after IV contrast bolus injection (0.1 mmol/kg Gd). Marked uniform mucosal enhancement. d Coronal T2w multislice HASTE sequence. e Thick-slab coronal MRCP image from T2w single-slice HASTE sequence. The etiology of adenomyomatosis of the gallbladder is unknown; the pattern of involvement may be diffuse (as in the case presented here), segmental, or localized. Segmental adenomyomatosis is considered a premalignant stage of carcinoma of the gallbladder.

MRI Appearance of Pathologic Entities

63

a

b Fig. 2.15a, b MRCP in a 62-year-old woman presenting with acute upper abdominal pain 8 years after cholecystectomy for symptomatic cholecystolithiasis. Ultrasound suggested a stone in the common bile duct. The patient was scheduled for ERCP but underwent MRCP for further work-up. MRCP did not confirm cholelithiasis. A stomach ulcer was finally diagnosed as the cause of abdominal pain.

a

b

a Coronal T2w multislice HASTE sequence after negative oral contrast medium. b MIP image created from T2w multislice TSE sequence (acquired with respiratory navigator, PACE). Normal appearance after cholecystectomy. The common bile duct has a maximum diameter of 10 mm. No remnant of the cystic duct is seen. No intrahepatic cholestasis. Small pancreatic duct.

c

Fig. 2.16a–c Primary sclerosing cholangitis in a 26-year-old woman; advanced disease with multiple strictures of the bile ducts. a Axial T2w multislice HASTE sequence. b Coronal T2w multislice HASTE sequence. c MIP image from coronal T2w multislice TSE sequence

(acquired with respiratory navigator, PACE). Beaded appearance of the intra- and extrahepatic ducts due to strictures alternating with dilated segments.

degree of outflow obstruction (also after stent placement, see Fig. 2.4). MRCP is especially useful in evaluating the extent of cholangiocarcinoma in the bifurcation of the common hepatic duct, a cholangiocellular carcinoma known as Klatskin tumor. This tumor spreads intraluminally along the duct wall (Fig. 2.17b) with secondary infiltration of the liver parenchyma at the liver hilum (Fig. 2.17a, c).

Klatskin tumors are staged according to the Bismuth classification, which distinguishes four types: intraluminal tumor below the bifurcation of the common hepatic duct (type I); tumor at the level of the bifurcation (type II); tumor extending into the right or left hepatic duct (type IIIa and IIIb, respectively); and tumor extending into both hepatic ducts (type IV). In these patients, the MRCP findings are also useful for intervention planning.23

64

2 The Bile and Pancreatic Ducts

a

b

Fig. 2.17a–c Klatskin tumor (Bismuth type IV) in a 72-year-old man. a Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). b Coronal T2w multislice HASTE sequence. c T1w VIBE sequence 5 min after IV contrast bolus injection (0.1 mmol/kg

c Gd). Coronal image shows intraductal extent in the hepatic bifurcation (arrow, b). Infiltration of the liver (arrow, c) is best appreciated on the contrast-enhanced parenchymal phase image.

References 1. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for postERCP pancreatitis: a prospective, multicenter study. Gastrointest Endosc 2001;54(4):425–434 2. Murray B, Carter R, Imrie C, Evans S, O’Suilleabhain C. Diclofenac reduces the incidence of acute pancreatitis after endoscopic retrograde cholangiopancreatography. Gastroenterology 2003;124(7):1786–1791 3. Loperfido S, Angelini G, Benedetti G, et al. Major early complications from diagnostic and therapeutic ERCP: a prospective multicenter study. Gastrointest Endosc 1998;48(1):1–10 4. Carlos RC, Scheiman JM, Hussain HK, Song JH, Francis IR, Fendrick AM. Making cost-effectiveness analyses clinically relevant: the effect of provider expertise and biliary disease prevalence on the economic comparison of alternative diagnostic strategies. Acad Radiol 2003;10(6):620–630 5. Takehara Y, Ichijo K, Tooyama N, et al. Breath-hold MR cholangiopancreatography with a long-echo-train fast spin-echo sequence and a surface coil in chronic pancreatitis. Radiology 1994;192(1):73–78 6. Hennig J, Nauerth A, Friedburg H. RARE imaging: a fast imaging method for clinical MR. Magn Reson Med 1986;3(6):823–833 7. Tang Y, Yamashita Y, Arakawa A, et al. Pancreaticobiliary ductal system: value of half-Fourier rapid acquisition with relaxation enhancement MR cholangiopancreatography for postoperative evaluation. Radiology 2000;215(1):81–88 8. Asbach P, Klessen C, Kroencke TJ, et al. Magnetic resonance cholangiopancreatography using a free-breathing T2-weighted turbo spin-echo sequence with navigator-triggered prospective acquisition correction. Magn Reson Imaging 2005;23(9): 939–945 9. Petersein J, Reisinger W, Mutze S, Hamm B. [Value of negative oral contrast media in MR cholangiopancreatography (MRCP)]. Rofo 2000;172(1):55–60 10. Neuhaus H, Ungeheuer A, Feussner H, Classen M, Siewert JR. [Laparoscopic cholecystectomy: ERCP as standard preoperative diagnostic technique]. Dtsch Med Wochenschr 1992;117(49): 1863–1867 11. Fukukura Y, Fujiyoshi F, Sasaki M, Nakajo M. Pancreatic duct: morphologic evaluation with MR cholangiopancreatography after secretin stimulation. Radiology 2002;222(3):674–680 12. Hellerhoff KJ, Helmberger HIII, Rösch T, Settles MR, Link TM, Rummeny EJ. Dynamic MR pancreatography after secretin ad-

13.

14. 15.

16.

17. 18.

19.

20.

21. 22.

23.

ministration: image quality and diagnostic accuracy. AJR Am J Roentgenol 2002;179(1):121–129 Kim HJ, Kim MH, Lee SK, et al. Normal structure, variations, and anomalies of the pancreaticobiliary ducts of Koreans: a nationwide cooperative prospective study. Gastrointest Endosc 2002;55(7):889–896 Morgan DE, Logan K, Baron TH, Koehler RE, Smith JK. Pancreas divisum: implications for diagnostic and therapeutic pancreatography. AJR Am J Roentgenol 1999;173(1):193–198 Warshaw AL, Simeone JF, Schapiro RH, Flavin-Warshaw B. Evaluation and treatment of the dominant dorsal duct syndrome (pancreas divisum redefined). Am J Surg 1990;159(1):59–64, discussion 64–66 Manfredi R, Costamagna G, Brizi MG, et al. Pancreas divisum and “santorinicele”: diagnosis with dynamic MR cholangiopancreatography with secretin stimulation. Radiology 2000;217(2): 403–408 Stott MA, Farrands PA, Guyer PB, Dewbury KC, Browning JJ, Sutton R. Ultrasound of the common bile duct in patients undergoing cholecystectomy. J Clin Ultrasound 1991;19(2):73–76 Laubenberger J, Büchert M, Schneider B, Blum U, Hennig J, Langer M. Breath-hold projection magnetic resonance-cholangio-pancreaticography (MRCP): a new method for the examination of the bile and pancreatic ducts. Magn Reson Med 1995;33(1):18–23 Topal B, Van de Moortel M, Fieuws S, et al. The value of magnetic resonance cholangiopancreatography in predicting common bile duct stones in patients with gallstone disease. Br J Surg 2003;90(1):42–47 Kim TK, Kim BS, Kim JH, et al. Diagnosis of intrahepatic stones: superiority of MR cholangiopancreatography over endoscopic retrograde cholangiopancreatography. AJR Am J Roentgenol 2002;179(2):429–434 Ichikawa T, Sou H, Araki T, et al. Duct-penetrating sign at MRCP: usefulness for differentiating inflammatory pancreatic mass from pancreatic carcinomas. Radiology 2001;221(1):107–116 Vitellas KM, Keogan MT, Freed KS, et al. Radiologic manifestations of sclerosing cholangitis with emphasis on MR cholangiopancreatography. Radiographics 2000;20(4):959–975, quiz 1108–1109, 1112 Hintze RE, Abou-Rebyeh H, Adler A, Veltzke-Schlieker W, Felix R, Wiedenmann B. Magnetic resonance cholangiopancreatography-guided unilateral endoscopic stent placement for Klatskin tumors. Gastrointest Endosc 2001;53(1):40–46

65

3

The Pancreas P. Asbach, W. Luboldt, and H.B. Gehl

Introduction The pancreas has two very different functions in the regulation of metabolism, which is why functional disorders present with a variety of clinical symptoms. Treatment is complex and may include pancreas transplant as a final option in some patients. Insofar as specialized cells (grouped in the islets of Langerhans) secrete several important hormones—insulin, glucagon, somatostatin, and pancreatic polypeptide—the pancreas is an endocrine organ. It is also an exocrine organ, producing several purely serous secretions with important digestive functions in the gastrointestinal tract: proteases (trypsin, chymotrypsin, and others), esterases (including lipase), carbohydrases, and nucleases. The pancreas is therefore made up of glandular tissue and a system of ducts. It is embedded between the liver, stomach, and duodenum (Fig. 3.1) and is a highly vascularized organ with multiple blood supply. Since the pancreas lacks a capsule and is in close vicinity to the vessels and organs of the upper abdomen, pancreatic disease often involves these structures. The clinical presentation is chronic to hyperacute.

Indications The high intrinsic soft-tissue contrast justifies the use of MRI as the primary imaging modality in the diagnosis and staging of pancreatic tumors (including identification of nodal and distant metastases) and evaluation of patients for normal anatomic variants (Table 3.1). MRI using fast breath-hold sequences with high spatial resolution has been shown to be superior to CT and ultrasound in the detection of pancreatic lesions.1–3 Moreover, MRI offers a wider range of diagnostic options because it enables noninvasive 3D imaging of the pancreatic duct system using MR cholangiopancreatography (MRCP; see Chapter 2) in the same session. By contrast, initial results suggest that multiplanar reconstruction of multislice spiral CT data does not greatly increase the sensitivity of CT in detecting pancreatic tumors.4 However, results are expected to improve given the ongoing rapid development of multislice spiral CT technology. CT is currently the preferred imaging

modality in the acute setting (acute pancreatitis, trauma) because it is faster and more readily available than MRI. One of the most challenging tasks for radiologists is to differentiate pancreatic malignancy from benign lesions, in particular chronic inflammatory changes with a mass effect. The distinction can be made with a high degree of accuracy by using a combination of MR strategies (axial imaging, MRCP, MR angiography).5 In patients with chronic pancreatitis, MRI is useful in assessing and following up the extent of the disease process and possible effects on adjacent structures (splenic vein thrombosis, perforation of pseudocysts into neighboring organs). In contrast, CT is still superior in identifying pancreatic calcifications and abscess-related air inclusions. As open MR scanners are becoming more widely available, the option of performing interventional procedures (biopsy, drainage, celiac plexus block) under real-time multiplanar MRI guidance is also expected to expand the therapeutic applications of pancreatic MRI.

Vena cava Portal vein

Celiac trunk

Superior mesenteric artery

Fig. 3.1 Anatomic relationships and lymphatic drainage of the pancreas.

66

3 The Pancreas

Table 3.1 Indications and imaging techniques for MRI of the pancreas Indication

Sequence

Plane

Comment

Pancreas divisum

Single-shot TSE with half-Fourier acquisition (e. g., HASTE)

Axial and coronal

Annular pancreas

Single-shot TSE with half-Fourier acquisition (e. g., HASTE) T2w TSE, T2*w GRE

Axial and coronal

Additional MRCP, T2-positive oral contrast medium (e. g., water) if necessary to delineate the duodenum With positive oral contrast medium to delineate the duodenum

Hemochromatosis (primary and secondary) Pancreatitis (> 72 h)

Pancreatic tumors

Single-shot TSE with half-Fourier acquisition (e. g., HASTE); dynamic contrast-enhanced MRI (3D GRE, e. g., VIBE) Dynamic contrast-enhanced MRI (3D GRE, e. g., VIBE)

Imaging Technique High spatial resolution and good soft-tissue contrast are essential when imaging the pancreas. Therefore, a high signal-to-noise ratio (SNR) is needed. This can be achieved by imaging at high field strength (at least 1 T) and using a body or torso phased-array surface coil.6 Fat saturation techniques are also more effective at higher field strengths. Motion artifacts can be minimized by using fast breath-hold sequences, which require high-performance gradients, or by using respiratory triggering.7 Respiratory-triggered sequences take much longer to acquire but improve spatial resolution and are therefore an integral part of pancreatic MRI. Application of presaturation bands above and below the field of view (FOV) is recommended to reduce artifacts from inflowing blood.

Imaging Planes Axial images through the pancreas are usually sufficient to establish a diagnosis. An additional coronal or sagittal sequence may help delineate the pancreatic head from the duodenum and predict the resectability of pancreatic cancer.

Pulse Sequences An MRI protocol for the pancreas is presented in Table 3.2. We start by obtaining a three-plane localizer of the upper abdomen using a breath-hold GRE sequence. Next, fast axial multislice breath-hold sequences (e. g., T1w FLASH, T2w HASTE; Fig. 3.2a, b) are acquired of the pancreas, the entire liver, the spleen, and the upper kidney poles (each within one breath-hold, acquisition time < 23 s). Alternatively, T2w images can be obtained without breath-holding using a respiratory-triggered TSE sequence (e. g., PACE)7 (Fig. 3.2c). The non-breath-hold sequence has higher spatial resolution but is more susceptible to peri-

Axial Axial

Axial

Should include full coverage of the liver and spleen Without/with fat suppression; additional MRCP; MR angiography if necessary to evaluate for vascular complications Additional planes on an individual basis; liver staging if necessary

staltic artifacts because of the longer acquisition time. Since the high signal intensity of fat improves the conspicuity of the pancreatic contours (Fig. 3.2b), a fat-suppression technique is not applied unless peripancreatic fluid collections are suspected (Fig. 3.2c). Additional axial (coronal slices if necessary) T1w and T2w thin-slice sequences (3 mm) should be acquired for evaluation of ductal structures and identification of very small parenchymal lesions (Fig. 3.2 d, e). A dynamic MR examination after intravenous bolus injection of Gd-based contrast medium (e. g., Omniscan, Magnevist) is recommended for more detailed morphologic evaluation of the pancreatic parenchyma. Arterial phase images obtained 15 s after the contrast bolus (at a standard dose of 0.1 mmol/kg body weight Gd) are most useful. Additional images should be acquired at 55 s, 2 min, and 5 min. The dynamic study is best performed using a thin-slice T1w 3D GRE sequence with high spatial resolution, such as a VIBE sequence (volume-interpolated breath-hold examination) (Fig. 3.2 f–j). Postcontrast images obtained with this GRE sequence at the above times are well suited for detecting pancreatic lesions6,8,9 and can replace imaging with conventional T1w sequences, which are inferior in terms of spatial resolution and slice thickness. Since MR cholangiopancreatography (MRCP; see Chapter 2) is fast to perform and often provides useful additional information, it should be done routinely in all patients undergoing MRI of the pancreas. Dynamic contrast-enhanced MRI can be dispensed with in favor of MR angiography to rule out splenic vein thrombosis or vascular infiltration in patients with a pancreatic tumor or severe pancreatitis.

Contrast Media Dynamic contrast-enhanced MRI for evaluating perfusion of the pancreas at different times can be performed after intravenous injection of a nonspecific, Gd-based contrast medium (e. g., Magnevist, Omniscan) at a dose of

Table 3.2 Recommended pulse sequences and imaging parameters for MRI of the pancreas Weighting

Plane

T1 T1 T2

Axial Axial Axial

Sequence type

TR (ms) TE (ms)

Flip (°)

ETL

FS

Matrix

2D GRE (e. g., FLASH) 199 4.1 90 No 115 × 256 2D GRE (e. g., FLASH) 192 4.38 90 No 115 × 256 Single-shot TSE with 800* 63 150 115 No 115 × 256 half-Fourier acquisition (e. g., HASTE) 800* 66 150 115 No 115 × 256 T2 Axial Single-shot TSE with half-Fourier acquisition (e. g., HASTE) T2 Axial TSE (FSE) 2340 80 180 21 Yes 168 × 320 T2 Coronal*** Single-shot TSE with 800* 63 150 115 Yes/no 115 × 256 half-Fourier acquisition (e. g., HASTE) Dynamic contrast-enhanced MRI 15 s, 55 s, 2 min, 5 min after IV contrast bolus (0.1 mmol/kg Gd) T1 Axial 3D GRE (e. g., VIBE) 5.2 2.59 10 – Yes 115 × 256 MR angiography (optional for improved vascular assessment, especially to evaluate for vascular involvement T1 Coronal 3D GRE (e. g., FLASH) 3.67 1.21 20 – Yes 250 × 512

No. of slices

No of acquisitions

Slice thickness (mm)

Scan time

Breath-hold

23 17 23

1 1 1

7 3 7

ca. 23 s ca. 22 s ca. 18 s

Yes Yes Yes

23

1

3

ca. 18 s

Yes

42 23

2 1

4 7

ca. 6–9 min ca. 18 s

No** Yes

2.5

ca. 22 s

Yes

2

ca. 24 s

Yes

64 1 in tumor staging) 56 1

* TR as a technical parameter referring to the interval between slice acquisitions; physical TR = ∞ since only one excitation pulse is applied per slice. ** Acquisition with respiratory triggering using the navigator echo technique (PACE, prospective acquisition correction). *** Optional to improve assessment in complex anatomic situations (e. g., pancreas divisum, annular pancreas).

Imaging Technique

67

68

3 The Pancreas

a

b

c

d

f

e Fig. 3.2a–j Illustration of normal findings obtained with standard protocol for MRI of the pancreas (see Table 3.2). a T1w 2D multislice FLASH sequence. b T2w multislice HASTE sequence. c Axial fat-

saturated T2w multislice TSE sequence (acquired with respiratory navigator, PACE). d Thin-slice (3-mm) T1w 2D multislice FLASH sequence. e Thin-slice (3-mm) T2w multislice HASTE sequence.

0.1 mmol/kg body weight (see above). As with CT,10 arterial phase images (15 s after the contrast bolus) are most suitable for parenchymal evaluation and tumor detection.8 Determination of individual circulation time after administering a test bolus (2 mL) may be advisable to optimize timing of the scan delay for arterial phase imag-

ing in patients with reduced cardiac output. Since paramagnetic contrast media are low-molecular-weight substances with rapid distribution in the extracellular space, focal pancreatic lesions may be obscured on equilibrium phase images.11 Current knowledge suggests that Gdbased contrast media are not likely to adversely affect

Imaging Technique

69

g

h

i

j Fig. 3.2a–j f–j T1w 3D GRE sequences (VIBE) before (f) and 15 s (g), 55 s (h), 2 min (i), and 5 min (j) after IV contrast bolus (0.1 mmol/kg Gd).

the course of acute pancreatitis and can therefore be administered at the standard dose.12 Timing of postcontrast image acquisition is less critical when tissue-specific contrast media are used. A tissuespecific contrast medium approved for use in MRI of the pancreas is mangafodipir trisodium (MnDPDP; Teslascan, GE Healthcare). It is administered as a short intravenous infusion, and images are acquired with a delay of ca. 15 min. MnDPDP was originally developed for liver imaging but can also be used to detect focal pancreatic lesions as it markedly enhances the signal intensity of normal pancreas on T1w GRE images.13 Oral contrast media14 are helpful to differentiate the pancreatic head from the duodenum or free fluid from fluid in the intestine (Fig. 3.3). Tap water acts as a positive contrast medium on T2w sequences and is sufficient to improve delineation of the pancreas from the duodenum. A negative oral contrast medium, which reduces or eliminates the signal from the bowel lumen on T2w images, is

needed to distinguish between free and endoluminal fluids. Negative contrast media include blueberry juice, which is rich in manganese,15 and iron oxide particles.16 Use of negative oral contrast medium is recommended especially when additional MRCP is performed (see Chapter 2). One must be aware that T1-positive oral contrast media, much like fat, exacerbate motion artifacts and should only be used in conjunction with breath-hold sequences. Oral contrast media on the basis of superparamagnetic iron oxide particles17 will eliminate this problem but may lead to susceptibility artifacts on GRE images if the concentration is too high or agglutination occurs.18 The oral contrast solution (300–600 mL) should be ingested over a period of 30 min before the examination. On the scanner table, the patient can be briefly turned into the left lateral position to improve contrast filling of the duodenum. Peristaltic artifacts can be reduced by intramuscular injection of a spasmolytic.

70

3 The Pancreas

a

b Fig. 3.3a, b Chronic pancreatitis with an acute episode 2 weeks earlier in a 41-year-old man. MRI with negative oral contrast medium. T2w multislice TSE sequence (acquired with respiratory navigator, PACE). The duodenal content has low SI (curved arrow)

(a descending part, b horizontal part), while the free peripancreatic fluid (straight arrows) is seen as a thin rim of high SI around the head of the pancreas (a) and immediately below the head (b).

Fig. 3.4a, b Incidental finding of extensive lipomatosis of the pancreas in a 71-year-old woman who presented with repeated episodes of acute upper abdominal complaints and underwent MRI for evaluation of the biliary tract. a T1w 2D multislice FLASH sequence. b T2w

multislice HASTE sequence. Lipomatosis is seen throughout the pancreas, and it is very difficult to distinguish the gland from surrounding fatty tissue (arrows).

MRI Appearance of Normal Anatomy

artery and vein with the right transverse colon and gastroduodenal artery lying anterior to it. Posterior to the body of the pancreas are the aorta and the celiac and superior mesenteric arteries arising from it, the superior mesenteric and splenic veins, the left renal vessels, the hilum of the left kidney, and the left adrenal gland; anterior to the body lie the transverse mesocolon and the omental bursa with the posterior gastric wall. The normal pancreatic parenchyma is hyperintense to liver on fat-suppressed T1w SE sequences and isointense on T1w GRE sequences (Fig. 3.2a). It shows a marked and early signal increase on contrast-enhanced arterial phase images, which rapidly returns to almost precontrast intensities (Fig. 3.2 g–j). Partial or complete fatty replacement (lipomatosis) of the pancreas is common in older persons and is characterized by prominent lobulation and a decrease in anteroposterior diameter (Fig. 3.4).

a

b

The pancreas is located in the retroperitoneum at the level of the first and second lumbar vertebrae. It consists of a head with the uncinate process (which hooks behind the superior mesenteric artery), body, and tail. The head is supplied by the gastroduodenal artery (arising from the common hepatic artery) and the superior mesenteric artery and its pancreaticoduodenal branches. The body and tail are supplied by the pancreatic branches of the splenic artery. The veins draining the pancreas empty into the splenic vein or directly into the portal vein. The most important lymph node stations are located in the porta hepatis, along the celiac trunk, and at the mesenteric root (Fig. 3.1). Most of the pancreatic head is surrounded by the duodenum, the so-called duodenal C-loop. Posterior to the head run the inferior vena cava and the right renal

MRI Appearance of Pathologic Entities

71

a

b Fig. 3.5a, b Primary hemochromatosis in a 34-year-old woman. a T1w 2D multislice FLASH sequence. b T2w multislice HASTE sequence. There is marked hypointensity of the parenchyma of the liver and pancreas on the T2w image. The MR signal reduction

caused by iron deposits in affected organs is best appreciated on T2*w images. In primary (congenital) hemochromatosis, iron does not accumulate in the spleen but in the pancreas, which is spared in secondary (acquired) hemochromatosis.

MRI Appearance of Pathologic Entities

Pancreatic Cysts

Congenital Anomalies and Diseases Pancreas Divisum

Pancreatic cysts (which are true cysts and must not be confused with pseudocysts, see below) typically occur in association with cysts in the liver, kidneys, or cerebellum and in patients with von Hippel–Lindau disease. They are clearly identified by their high signal intensity on T2w images and, unlike cystic tumors, do not enhance after contrast administration.

Pancreas divisum results when the dorsal and ventral pancreatic anlagen fail to fuse and there is persistence of two separate ductal systems that do not communicate.19 In affected individuals, the secretions produced in the body and tail of the pancreas drain through the accessory pancreatic duct (duct of Santorini), which courses in the anterior part of the pancreatic head and enters the duodenum through the minor papilla. Pancreas divisum has a reported prevalence of 1.5–10 %.19–21 Although pancreas divisum is a normal anatomic variant, affected individuals may suffer from recurrent episodes of pancreatitis due to functional stenosis obstructing the outflow of exocrine juices through the minor papilla.22,23 Improvement can be achieved by sphincteroplasty, sphincterotomy, or stenting.24 In approximately 42 % of patients with pancreas divisum, the pancreatic head is enlarged and can sometimes mimic a tumor.25 MRCP may be helpful in diagnosing pancreas divisum (see Chapter 2, Fig. 2.5).

All three patterns can be distinguished by MRI. The degree of fatty replacement can be estimated by comparing T1w images acquired without and with fat saturation.

Annular Pancreas

Primary Hemochromatosis

Annular pancreas is an extremely uncommon congenital anomaly characterized by a ring of normal pancreatic tissue encircling and obstructing the descending duodenum. To diagnose this condition, administering oral contrast medium and acquiring coronal images to supplement the standard axial sequence is recommended.26

Primary (hereditary, idiopathic) hemochromatosis is an autosomal recessive disorder of iron storage. Excessive absorption of ingested iron in the jejunum and a limited capacity of the mononuclear phagocyte system (MPS) to cope with the excess iron lead to deposition of hemosiderin in tissue, particularly in the liver, heart, gonads, skin, and pancreas (Fig. 3.5). The iron overload causes the islet cells to atrophy and can ultimately result in endocrine

Cystic Fibrosis Cystic fibrosis (mucoviscidosis) is an autosomal recessive disease causing fibrosis in the pancreas and lungs. Three distinct patterns of pancreatic involvement have been noted27: · Lobulated, enlarged pancreas with complete replacement by fatty tissue · Small atrophic pancreas with partial fatty replacement · Diffuse atrophy without fatty replacement.

72

3 The Pancreas

Fig. 3.6 Acute edematous pancreatitis secondary to alcohol abuse in a 39-year-old man. Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). The pancreatic head is enlarged and shows subtle diffuse hyperintensities.

dysfunction (clinically manifesting as bronze diabetes). Stored iron has higher susceptibility (i. e., is more readily magnetized) and thus distorts the magnetic field, shortening T2 and T2* relaxation times and reducing signal intensity on T2w and T2*w images (Fig. 3.5b). The signal reduction is best appreciated on T2*w GRE images. SE sequences are less sensitive in detecting iron because stored iron interacts directly with the nuclear spins and magnetic field inhomogeneities are compensated for by the 180° refocusing pulse.

Secondary Hemochromatosis In secondary (acquired) hemochromatosis (also known as hemosiderosis), the iron overload is due to increased erythrocyte breakdown and accumulation in the phagocytosing system (e. g., Kupffer cells in the liver). Causes may be internal (ineffective or hypoplastic erythropoiesis) or external (increased iron intake, e. g., through repeat transfusion). The pancreas is usually spared in secondary hemochromatosis and retains its normal signal intensity, while the liver shows the same iron-related signal changes as in primary hemochromatosis.28

Inflammatory Disease Acute Pancreatitis Acute pancreatitis is a common condition with an incidence of 5–10 cases per 100 000 population and a peak between 40 and 60 years. The most common causes are cholelithiasis (60–70 %) and alcohol abuse (20–30 %). Less common causes include autoimmune diseases, genetic defects,29 trauma (including ERCP), drugs, metabolic disorders, normal anatomic variants (pancreas divisum), and

tumors obstructing the pancreatic duct. The underlying pathophysiologic mechanism is premature, intrapancreatic activation of proteolytic enzymes (particularly trypsin) with subsequent autodigestion.30 There are two forms of acute pancreatitis—edematous and necrotizing. Edematous pancreatitis is the more common and less severe form, accounting for 85 % of cases. The remaining 15 % of patients develop systemic inflammatory response syndrome (SIRS) with fulminant necrotizing pancreatitis, ultimately leading to partial or complete necrosis of the organ.30 The diagnosis of acute pancreatitis is based on clinical and laboratory findings. Cross-sectional imaging is indicated if no improvement is seen 72 h after initiation of conservative treatment.31 Since acute inflammatory and normal pancreatic tissue have similar MR signal intensities, the MR diagnosis primarily relies on morphologic criteria, but these are not always specific in edematous pancreatitis (Fig. 3.6). An enlarged gland with loss of definition of its normal boundaries suggests interstitial edema and inflammation of the surrounding tissue. Peripancreatic fluid collections occur in 40 % of patients with acute pancreatitis and are best appreciated on T2w images (Fig. 3.3). However, free fluid is often difficult to differentiate from endoluminal fluid. A negative oral contrast medium may help make the distinction. The peripancreatic fluid contains proteolytic enzymes, which induce an inflammatory reaction in the retroperitoneal fatty tissue. Ascites and thickening of the fascial plane, especially of the Gerota fascia, indicate more severe inflammation. Collections of exudative fluid can extend as far as the true pelvis. The pancreatic contour may become indistinguishable since inflammatory strands and fluid in the peripancreatic fat may have the same signal intensity as the pancreas. Differentiating uninvolved pancreatic tissue from necrotic areas on unenhanced MR images may also be difficult at this stage. Such cases are absolute indications for dynamic contrast-enhanced imaging after intravenous contrast administration. The complications of acute pancreatitis include pseudocysts (Fig. 3.7), fistulas to adjacent organs, splenic vein thrombosis, pseudoaneurysm of the splenic artery with a 37 % risk of bleeding, autodigestion of arterial walls with resultant blood leaks, and bacterial superinfection of necrotic tissue and fluid collections.32 Superinfection is the most severe complication. Since CT is more readily available and easier to perform, it remains the method of choice, especially when following up ICU patients. However, CT does not detect small areas of intrapancreatic necrosis and requires administration of iodine-based contrast medium, which could potentially harm the kidneys. MRI, on the other hand, being highly sensitive to paramagnetic contrast media, enables good differentiation of unperfused (necrosis) and hyperperfused (abscess) areas.12 The paramagnetic effects of blood breakdown products also result in improved hemorrhage detection by MRI (Fig. 3.8). Finally, T2w imaging is superior to CT in differentiating between

MRI Appearance of Pathologic Entities

73

a

d

e

b

c

f Fig. 3.7a–f Chronic pancreatitis and recurrent acute episodes in a 55-year-old man. a Axial T2w multislice HASTE sequence. The pancreatic duct is dilated due to compression by a large pseudocyst in the pancreatic head (string of pearls appearance). There are also multiple pseudocysts in the right renal compartment (some within the anterior perirenal space). b Coronal T2w multislice HASTE sequence. Right-sided pseudocysts below the diaphragm and in the

renal compartment. c–e Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). f MRCP. MIP image from T2w TSE multislice sequence (PACE). There is marked dilatation of the pancreatic duct with normal appearance of the intra- and extrahepatic biliary ducts. Also seen are several pseudocysts in the pancreatic head.

74

3 The Pancreas

a

b

c

d Fig. 3.8a–d Hemorrhagic pseudocyst in the body of the pancreas in a 45-year-old man. a Axial T1w 2 D multislice FLASH sequence. b, c Axial T2w multislice HASTE sequence. d Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). The cyst

contains areas of low and high SI on T1w and T2w images, reflecting different stages of intralesional hemorrhage (arrow). There is a second, small, nonhemorrhagic pseudocyst in the pancreatic tail.

fluid collections (edema) and solid tissue (necrotic fatty tissue). This differentiation is important in patients with bacterial infections because it may affect treatment: while infected fluid collections are amenable to percutaneous drainage in most cases, infected necrotic tissue requires surgical débridement.31,33 To identify vascular complications, such as pseudoaneurysm of the splenic artery or splenic vein thrombosis, full MRI evaluation of acute pancreatitis should include contrast-enhanced MR angiography with a fast 3D GRE sequence.34

the pancreas is common. Imaging typically shows dilatation of the main pancreatic duct with branch duct ectasia, yielding a “chain of lakes” or “string of pearls” appearance (Figs. 3.7 and 3.9). Calcium deposits resulting from the precipitation of proteins with calcium in acini and terminal ducts are present in 50 % of patients with chronic pancreatitis,36 but usually only in late stages.37 Another finding is obstruction of the pancreatic and/or bile duct by stones (see Chapter 2, Fig. 2.7). Pseudocysts develop in ca. 30 % of cases. Other complications of chronic pancreatitis are abscess formation and splenic vein thrombosis. Fibrosis results in less intense enhancement compared with normal pancreatic tissue on fat-suppressed T1w SE and GRE sequences acquired after intravenous contrast administration. MRI therefore appears to be more sensitive than CT in evaluating pancreatic fibrosis, although quantitative data to confirm this observation are still lacking. CT, however, has the advantage of directly visualizing calcifications (Fig. 3.10). With MRI, calcifications are suggested only indirectly by signal voids on all sequences and therefore remain difficult to detect, even on T2*w GRE

Chronic Pancreatitis About 70 % of patients with chronic pancreatitis have a history of alcohol abuse and recurrent episodes of acute pancreatitis.35 Chronic damage to acini results in parenchymal destruction and pancreatic autodigestion with subsequent fibrous replacement of the acinar tissue and loss of both exocrine and endocrine function. Atrophy of

MRI Appearance of Pathologic Entities

75

a

b

c

d

e

f Fig. 3.9a–f MRI findings 2 weeks after an acute episode of pancreatitis in a 41-year-old man with a history of chronic pancreatitis. a, b Axial T1w 2D multislice FLASH sequences acquired at the level of the head (a) and body (b) of the pancreas. c, d Axial T2w multislice HASTE sequences at the level of the head (c) and body (d) of the pancreas. e, f Axial T2w multislice TSE sequences (acquired with

respiratory navigator, PACE) at the level of the head (e) and body (f) of the pancreas. The pancreas is enlarged (pancreatic head measuring 5 cm in axial diameter) with inhomogeneous T1 and T2 SI of the parenchyma. The pancreatic duct is dilated and appears irregular (arrow) in the tail of the pancreas. There is a small peripancreatic fluid collection persisting after an episode of acute pancreatitis.

sequences, where the signal voids appear larger than the actual calcifications (long TE). An advantage of MRI in patients with chronic pancreatitis is that it enables noninvasive evaluation of the ductal system and will confirm or rule out dilatation of the pancreatic duct with high sensitivity (MRCP, see Chapter 2).

Pseudocysts About 30–50 % of pseudocysts develop from fluid collections arising in the setting of pancreatitis (Figs. 3.7 and 3.8). If such fluid collections persist, they may develop a capsule over a period of 4 weeks or longer due to inflammatory reactions. Pancreatic pseudocysts can contain proteinaceous fluid or blood.38 Unlike retention cysts,

76

3 The Pancreas

a

b Fig. 3.10a, b Acute cholangitis developing after replacement of a common bile duct stent in a 50-year-old man with a several-year history of chronic pancreatitis. CT scans obtained using multislice

spiral CT (4-mm reconstructed slice thickness). Multiple calcifications in the head and tail of the pancreas.

Fig. 3.11a, b Pancreatic head carcinoma in a 61-year-old woman. Axial T1w 2D multislice FLASH sequence before (a) and 15 s after (b) IV bolus injection of contrast medium (0.1 mmol/kg Gd). On the precontrast image, the 2-cm tumor (arrow) is of homogeneous low

SI compared with the normal surrounding parenchyma. Following contrast injection, the tumor becomes more conspicuous due to intense parenchymal enhancement and is seen to be directly contiguous with the duodenal wall.

pseudocysts do not have an epithelial lining.39 The cysts can lead to compression of the splenic vein or common bile duct, formation of fistulas to adjacent organs, infection, or obstruction of lymphatic drainage. Rupture of a pseudocyst can cause peritonitis. Pseudocysts can occur anywhere in the abdomen (Fig. 3.7) but rarely outside the abdomen (e. g., in the mediastinum). The cyst fluid is best appreciated on T2w MR images. Pseudocysts do not enhance after intravenous contrast administration. Negative oral contrast medium is helpful to reliably differentiate suspected pseudocysts from fluid in the duodenum.

Pancreatic Neoplasms Pancreatic Cancer Adenocarcinoma is the most common pancreatic neoplasm, accounting for 95 % of malignant tumors of the pancreas. Ninety percent are ductal carcinomas arising from the epithelium of small pancreatic ducts and 10 % originate in the acinar epithelium (acinar carcinomas). Adenocarcinomas typically occur in the pancreatic head (60 %) (Fig. 3.11), less commonly in the body (15 %) or tail (5 %) (Fig. 3.12). About 20 % are diffuse tumors involving the entire pancreas.37 Pancreatic head tumors are defined

MRI Appearance of Pathologic Entities

as tumors located to the right of the left border of the superior mesenteric vein (the uncinate process is part of the head), tumors of the body lie between the left border of the superior mesenteric vein and the left border of the aorta, and tumors of the tail are those to the left of the left border of the aorta. While tumors located in the head of the gland tend to present early with symptoms of common bile duct obstruction, most pancreatic carcinomas are diagnosed at an advanced stage and are unresectable because of liver and/or lymph node metastases, vascular invasion, or peritoneal or retroperitoneal infiltration.40 Complete surgical resection is the only potentially curative treatment but is only possible if the tumor is confined to the pancreas (T classification in Table 3.3). Most pancreatic carcinomas have lower signal intensity than normal pancreatic tissue on T1w SE and GRE images, which reflects the low protein content of malignant tissue. On T2w images, they are of variable signal intensity, typically slightly hyperintense to normal glandular tissue (Fig. 3.12). Unlike endocrine tumors, adenocarcinomas are hypovascular because of their desmoplastic, fibrotic composition, which is why they have lower signal intensity than normal pancreas on dynamic contrast-enhanced images (Fig. 3.12 d–g). Depending on its location, pancreatic carcinoma will cause dilatation of the pancreatic and/or bile duct.41 Ductal dilatation can be confirmed or ruled out using T2w thin-slice HASTE sequences (axial and coronal), MRCP, or contrast-enhanced GRE sequences. The lack of a pancreatic capsule allows early spread of cancer to local tissues and invasion of the large vessels, particularly the superior mesenteric artery and vein and the portal vein. Patients with vascular invasion have a poorer prognosis, as have patients with perineural and lymphatic invasion. More than 50 % of patients with pancreatic cancer have lymph node metastases at the time of surgery. A distinction is made between involvement of lymph nodes near the pancreas and involvement of a second nodal group along the superior mesenteric, gastroduodenal, common hepatic, and splenic arteries and celiac trunk (see Fig. 3.1). Because the lymphatics are very short, there is also early invasion of para-aortic, paracaval, and porta hepatis nodes (see Fig. 3.1). Pancreatic cancer initially spreads to the liver (66 % of cases) and lymph nodes (22 % of cases). Lung metastases occur relatively late. Other metastatic sites are the bones, intestine, and peritoneal cavity. Advanced pancreatic cancer is characterized by contiguous invasion of adjacent structures including the duodenum, inferior vena cava, splenic vein, stomach, colon, spleen, adrenal, and kidney. Surgical resection is an option in patients without hepatic or nodal metastases or vascular encasement.42 Vascular encasement is suggested by loss of surrounding fat planes, which is more obvious on T1w MR images43 than on CT.44 MRA using a GRE sequence is helpful for evaluating the lumen of the superior mesenteric vein and the hepatic portal system. MRI is slightly better than CT in assessing

77

Table 3.3 T classification of pancreatic and ampullary cancer

T1

T2

T3

T4

Pancreatic cancer

Ampullary cancer

Tumor limited to the pancreas, ≤ 2 cm in greatest dimension Tumor limited to the pancreas, > 2 cm in greatest dimension Tumor extends beyond the pancreas (but does not involve the celiac trunk or superior mesenteric artery) Tumor invades the celiac trunk or superior mesenteric artery

Tumor limited to the ampulla of Vater or sphincter of Oddi Tumor invades the duodenal wall Tumor invades the pancreas

Tumor invades peripancreatic soft tissues and/or other adjacent organs or structures

tumor resectability,3 for which it has a positive predictive value well over 80 %.1 Nevertheless, exploratory laparotomy is necessary to assess resectability in all critical cases.41

Ampullary Cancer Because of its more favorable prognosis, ampullary cancer (carcinoma of the ampulla of Vater) must be differentiated from tumors arising in the pancreatic head. Other periampullary neoplasms to be considered in the differential diagnosis are tumors of the terminal pancreatic and common bile ducts. While, histologically, they are also adenocarcinomas, true ampullary carcinomas have a better 5-year survival rate after Whipple operation. A separate T classification system exists for ampullary cancer (Table 3.3). Ampullary cancer presents early with obstructive jaundice. MRCP may be helpful in distinguishing ampullary cancer from pancreatic head masses. Usually, the cancer is most conspicuous on contrast-enhanced GRE images. Use of negative oral contrast medium is indicated to facilitate differentiation from the duodenum.

Cystic Pancreatic Neoplasms Although very rare, cystic neoplasms present a challenge for imaging because they may be confused with pseudocysts or congenital cysts of the pancreas.38,45 The primary role of imaging is to differentiate cystic neoplasms from these cysts because, unlike cysts, most cystic tumors require resection or at least histologic confirmation. Four major types of cystic neoplasms can be distinguished (Table 3.4): serous cystadenoma (micro- and macrocystic forms) and mucinous cystadenoma/cystadenocarcinoma, which are most common, and the less common intraductal papillary mucinous tumor (IPMT).46,47 Very rarely,

78

3 The Pancreas

a

b

c

d

e

f Fig. 3.12a–g Pancreatic tail carcinoma (arrow) incidentally detected at MRI in a 64-year-old woman with focal liver lesions and a history of breast cancer. The liver lesions are most likely metastases from the pancreatic carcinoma. a Plain axial T1w 2D multislice FLASH sequence. b Axial T2w multislice HASTE sequence. c–g Dynamic MRI using a thin-slice T1w 3D FLASH sequence (VIBE). Images acquired before (c) and 15 s (d), 55 s (e), 2 min (f), and 5 min (g) after IV contrast bolus injection (0.1 mmol/kg Gd). The tumor is most conspicuous in the arterial phase due to intense enhancement of normal parenchyma.

g

MRI Appearance of Pathologic Entities

79

a

b Fig. 3.13a, b Microcystic serous cystadenoma at the junction of the body and tail of the pancreas. a Axial T1w 2D multislice FLASH sequence. b Axial T2w multislice HASTE sequence. Images show multiple clustered cysts (< 2 cm). Also seen is cirrhosis of the liver.

Table 3.4 Cystic pancreatic neoplasms Neoplasm

M:F ratio

Peak age (years)

Cyst architecture

Differential diagnosis

Serous cystadenoma

1:4

65 (82 % > 60)

Microcystic serous

Cystic neuroendocrine tumor

Histologic confirmation (biopsy)

Macrocystic serous

Histologic confirmation (biopsy)

Mucinous cystadenoma/ cystadenocarcinoma IPMT

1:9

40–60

Macrocystic mucinous

Mucinous cystadenoma/ cystadenocarcinoma Macrocystic serous cystadenoma

M>F

?

M=F

30–60

Macrocystic mucinous Microcystic

Cystic neuroendocrine tumor

neuroendocrine pancreatic tumors may have cystic components (see “Islet Cell Tumors” below).48 Serous cystadenomas have a predilection for the pancreatic head. There are basically two variants, microcystic and macrocystic cystadenomas. Some authors assign the macrocystic variant to the group of mucinous cystadenomas and use microcystic cystadenoma synonymously with serous cystadenoma. However, because treatment is different, we recommend distinguishing between microcystic and macrocystic serous adenomas and separating these from mucinous cystadenomas. The microcystic form is characterized by a grapelike cluster of multiple small cysts (< 2 cm) (Fig. 3.13) with nodular margins. A central scar with calcifications is common. The cyst fluid may contain blood, resulting in high signal intensity on T1w images.49 Microcystic serous cystadenomas are considered benign tumors that do not encase vessels;50 however, they cannot be reliably differentiated from the very rare cystic neuroendocrine tumors solely on the basis of their imaging appearance (Fig. 3.14). Macrocystic serous cystadenomas consist of clusters of larger cysts (> 2 cm),

Mucinous cystadenoma/ cystadenocarcinoma Microcystic serous cystadenoma

Management

Surgical resection

Histologic confirmation (biopsy) Surgical resection

and their imaging features may resemble a mucinous cystadenoma/cystadenocarcinoma. Mucinous cystadenomas/cystadenocarcinomas typically arise in the pancreatic body or tail (Fig. 3.15). The tumors produce proteinaceous mucus and always exhibit a macrocystic pattern.45 Mucinous cystadenomas have high malignant potential (mucinous cystadenocarcinomas). They have irregular borders, a capsule, and septa; peripheral calcification is quite common. Intraductal papillary mucinous tumors (IPMT) are usually found in the pancreatic head (uncinate process). They arise in the pancreatic duct or one of its main branches and are characterized by a macrocystic appearance and excessive intraductal mucus production,45 which may obstruct the pancreatic duct and cause recurrent pancreatitis. Differentiation of IPMT from chronic pancreatitis is difficult, unless there is clear evidence of a mucous plug. In patients without signs of pancreatitis, IPMT is also difficult to distinguish from mucinous cystadenoma/cystadenocarcinoma.

80

3 The Pancreas

a

b Fig. 3.14a, b Well-differentiated cystic neuroendocrine carcinoma of the pancreas. a Axial T2w multislice HASTE sequence. b Axial T1w 3D multislice FLASH sequence (VIBE) 2 min after IV bolus injection of contrast medium (0.1 mmol/kg Gd). As in Fig. 3.13, there are

multiple clustered cysts (< 2 cm). The cyst walls enhance, and no solid tumor components are seen. This tumor cannot be reliably distinguished from microcystic serous cystadenoma using morphologic imaging criteria.

Islet Cell Tumors

ized, gastrinomas enhance only moderately compared with insulinomas. Glucagoma, somatostatinoma56 (Fig. 3.17), and VIPoma (vasoactive intestinal polypeptide) are very rare, usually malignant, tumors that tend to metastasize to the liver.57 Unlike insulinomas and gastrinomas, these islet cell tumors are so rare that they are unlikely to be encountered in the routine clinical setting. MRI has evolved into an important imaging modality for evaluating patients with islet cell tumors: preoperatively for precise tumor localization and exclusion of hepatic or nodal metastases and postoperatively for early detection of tumor recurrence, which is seen in 50 % of patients with gastrinoma.

With an incidence of 0.5 per 100 000 population, pancreatic islet cell tumors are fairly uncommon. The majority, 60–85 %, are functioning islet cell tumors51 and present early with symptoms of hormone overproduction, making the diagnosis straightforward. Nonfunctioning tumors only become apparent once they have reached a certain size and cause biliary or intestinal obstruction. Islet cell tumors are hypointense to normal pancreatic tissue on T1w images and markedly hyperintense on T2w images.52,53 Being hypervascular, they enhance intensely on dynamic contrast-enhanced arterial phase images,52,53 as do liver metastases from these tumors.52–54 Short tau inversion recovery (STIR) sequences are also helpful in highlighting islet cell tumors,54,55 but the improved conspicuity is accompanied by a penalty in spatial resolution. Most islet cell tumors are insulinomas, which are usually benign (> 90 %), and only half of them produce insulin. They are hypervascular tumors and are therefore best detected by dynamic imaging. Large insulinomas, as well as their liver and lymph node metastases, are characterized by rim enhancement.52 Gastrinomas are gastrin-producing tumors of the pancreas (80 %) or duodenum (20 %). They secrete large amounts of gastrin, resulting in chronic overproduction of gastric acid (Zollinger–Ellison syndrome). Thirty percent of gastrinomas occur in conjunction with other endocrine neoplasms (pituitary and parathyroid tumors) in patients with multiple endocrine neoplasia type 1 (MEN 1). Most gastrinomas are malignant and typically larger and less well vascularized than insulinomas. Gastrinomas are hypointense to normal pancreatic tissue on fat-suppressed T1w SE images and, regardless of their size, always have high signal intensity on fat-suppressed T2w SE sequences49 (Fig. 3.16). Being less well vascular-

Differentiation of Inflammatory Pseudotumors and Neoplasms Patients with recurrent pancreatitis often have areas of focal enlargement in the gland, which are very difficult to characterize. Differentiating these so-called pseudotumors from malignant tumors remains an interdisciplinary challenge. Neither the patient’s history nor laboratory tests (tumor markers) provide a definitive diagnosis because the tumor marker CA 19-9 may be elevated in both conditions.58 Similarly, morphologic imaging criteria fail to provide a reliable diagnosis. Focal enlargement of the pancreas favors the diagnosis of neoplasm in the absence of classic signs of inflammation such as exudation, peripancreatic fascial thickening, or beading of the pancreatic duct. There is more intense and longer enhancement of the parenchyma in pancreatitis compared with carcinoma.59 The duct-penetrating sign60 may also be helpful in establishing the differential diagnosis (Fig. 3.18)—if narrowing of the pancreatic duct within

MRI Appearance of Pathologic Entities

81

a

b

c

d

e

f Fig. 3.15a–f Macrocystic mucinous cystadenoma in a 78-year-old woman. a Axial T2w multislice HASTE sequence. b Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). c, d Axial T1w 3D FLASH sequence (VIBE) 15 s (c) and 2 min (d) after IV bolus injection of contrast medium (0.1 mmol/kg Gd).

e, f Single-slice MRCP (e) and MIP image (f) from T2w multislice TSE sequence (PACE). The characteristic septa of the macrocystic mass are most conspicuous on the MRCP images. The postcontrast sequences show enhancement of the septa (d) and absence of hypervascular areas within the mass (c).

82

3 The Pancreas

a

b

c

d

e

f Fig. 3.16a–g Gastrinoma. a Axial T1w SE sequence. b Axial T2w fatsuppressed SE sequence. c–f Dynamic MRI with T1w GRE sequence before (c) and 15 s (d), 55 s (e) and 3 min (f) after IV contrast bolus injection (0.1 mmol/kg Gd). g Delayed postcontrast axial T1w fatsuppressed SE sequence.

g

MRI Appearance of Pathologic Entities

83

a

b Fig. 3.17a–c Somatostatinoma. a, b Axial T1w SE images obtained before (a) and after (b) contrast administration (0.1 mmol/kg Gd). c Contrast-enhanced spiral CT scan for comparison.

c

the mass is minimal, inflammation is the more likely cause. Absence of the duct-penetrating sign is highly indicative of malignancy. In addition, the appearance of the pancreatic duct can also serve to distinguish a malignant tumor from an inflammatory process. Gradual narrowing of the duct suggests inflammation, while abrupt obstruction indicates a stone or neoplasm. The duct is best evaluated by performing an additional MRCP study. Intact perivascular fat planes are more common in focal pancreatitis, while vascular encasement is more likely in pancreatic cancer. Note, however, that edema of the fat plane around the mesenteric artery and vein in acute pancreatitis may mimic vascular encasement.61 Complete replacement of the fatty tissue in chronic fibrotic pancreatitis may also be mistaken for vascular encasement. Because of the difficulties just outlined, histologic examination remains the diagnostic standard for differentiating an inflammatory pseudotumor from pancreatic cancer with secondary obstructive pancreatitis, despite ongoing advances in CT and MRI.

Pancreas Transplant Patients with diabetic nephropathy often receive a pancreas transplant in combination with renal transplantation. MRI is therefore the preferred imaging modality for evaluating pancreas grafts since Gd-based contrast media pose little risk29 compared with CT and use of a contrast medium that is potentially harmful to the kidneys. Dynamic contrast-enhanced MRI with acquisition of images in different phases of perfusion (arterial, parenchymal, and interstitial) is most suitable for detecting postoperative complications such as venous thrombosis or abscess. MR angiography allows excellent assessment of the vascular status and can replace conventional digital subtraction angiography (DSA).62

84

3 The Pancreas

a

b

c

d Fig. 3.18a–e Chronic pancreatitis in a 41-year-old man. There is enlargement of the head of the pancreas (5.5 cm in axial diameter) with inhomogeneous T1 and T2 SI. a Axial T1w 2D multislice FLASH sequence. b Axial T2w multislice HASTE sequence. c Axial T2w multislice TSE sequence (acquired with respiratory navigator, PACE). d Axial T1w 2D multislice FLASH sequence 15 s after IV contrast bolus injection (0.1 mmol/kg Gd). The pancreatic head has low SI relative to the body. Carcinoma confined to the head cannot be excluded. e MIP image created from T2w multislice TSE sequence (acquired with respiratory navigator, PACE). There is no compression of the pancreatic duct in the head (arrow), which would be expected if malignancy were present. Carcinoma was ruled out by histologic examination of the pancreatic head after Whipple resection.

e

MRI Appearance of Pathologic Entities

References 1. Fischer U, Vosshenrich R, Horstmann O, et al. Preoperative local MRI-staging of patients with a suspected pancreatic mass. Eur Radiol 2002;12(2):296–303 2. Ichikawa T, Haradome H, Hachiya J, et al. Pancreatic ductal adenocarcinoma: preoperative assessment with helical CT versus dynamic MR imaging. Radiology 1997;202(3):655–662 3. Schima W, Függer R, Schober E, et al. Diagnosis and staging of pancreatic cancer: comparison of mangafodipir trisodium-enhanced MR imaging and contrast-enhanced helical hydro-CT. AJR Am J Roentgenol 2002;179(3):717–724 4. Prokesch RW, Chow LC, Beaulieu CF, et al. Local staging of pancreatic carcinoma with multi-detector row CT: use of curved planar reformations initial experience. Radiology 2002;225 (3):759–765 5. Lopez Hänninen E, Amthauer H, Hosten N, et al. Prospective evaluation of pancreatic tumors: accuracy of MR imaging with MR cholangiopancreatography and MR angiography. Radiology 2002;224(1):34–41 6. Semelka RC, Ascher SM. MR imaging of the pancreas. Radiology 1993;188(3):593–602 7. Klessen C, Asbach P, Kroencke TJ, et al. Magnetic resonance imaging of the upper abdomen using a free-breathing T2weighted turbo spin echo sequence with navigator triggered prospective acquisition correction. J Magn Reson Imaging 2005;21(5):576–582 8. Gabata T, Matsui O, Kadoya M, et al. Small pancreatic adenocarcinomas: efficacy of MR imaging with fat suppression and gadolinium enhancement. Radiology 1994;193(3):683–688 9. Semelka RC, Kroeker MA, Shoenut JP, Kroeker R, Yaffe CS, Micflikier AB. Pancreatic disease: prospective comparison of CT, ERCP, and 1.5-T MR imaging with dynamic gadolinium enhancement and fat suppression. Radiology 1991;181(3):785–791 10. Lu DS, Vedantham S, Krasny RM, Kadell B, Berger WL, Reber HA. Two-phase helical CT for pancreatic tumors: pancreatic versus hepatic phase enhancement of tumor, pancreas, and vascular structures. Radiology 1996;199(3):697–701 11. Brailsford J, Ward J, Chalmers AG, Ridgway J, Robinson PJ. Dynamic MRI of the pancreas—gadolinium enhancement in normal tissue. Clin Radiol 1994;49(2):104–108 12. Johnson CD, Stephens DH, Sarr MG. CT of acute pancreatitis: correlation between lack of contrast enhancement and pancreatic necrosis. AJR Am J Roentgenol 1991;156(1):93–95 13. Gehl HB, Urhahn R, Bohndorf K, et al. Mn-DPDP in MR imaging of pancreatic adenocarcinoma: initial clinical experience. Radiology 1993;186(3):795–798 14. Brown JJ. Gastrointestinal contrast agents for MR imaging. Magn Reson Imaging Clin N Am 1996;4(1):25–35 15. Papanikolaou N, Karantanas A, Maris T, Gourtsoyiannis N. MR cholangiopancreatography before and after oral blueberry juice administration. J Comput Assist Tomogr 2000;24(2):229–234 16. Jacobsen TF, Laniado M, Van Beers BE, et al. Oral magnetic particles (ferristene) as a contrast medium in abdominal magnetic resonance imaging. Acad Radiol 1996;3(7):571–580 17. Rogers J, Lewis J, Josephson L. Use of AMI-227 as an oral MR contrast agent. Magn Reson Imaging 1994;12(4):631–639 18. Briggs RW, Wu Z, Mladinich CR, et al. In vivo animal tests of an artifact-free contrast agent for gastrointestinal MRI. Magn Reson Imaging 1997;15(5):559–566 19. Warshaw AL. Pancreas divisum—really. Surgery 2000;128(5): 832–833 20. Kim HJ, Kim MH, Lee SK, et al. Normal structure, variations, and anomalies of the pancreaticobiliary ducts of Koreans: a nationwide cooperative prospective study. Gastrointest Endosc 2002;55(7):889–896 21. Morgan DE, Logan K, Baron TH, Koehler RE, Smith JK. Pancreas divisum: implications for diagnostic and therapeutic pancreatography. AJR Am J Roentgenol 1999;173(1):193–198 22. Ohshima Y, Tsukamoto Y, Naitoh Y, et al. Function of the minor duodenal papilla in pancreas divisum as determined by duodenoscopy using indigo carmine dye and a pH sensor. Am J Gastroenterol 1994;89(12):2188–2191 23. Warshaw AL, Simeone JF, Schapiro RH, Flavin-Warshaw B. Evaluation and treatment of the dominant dorsal duct syndrome

24.

25. 26. 27. 28. 29. 30. 31. 32. 33.

34. 35. 36. 37. 38. 39.

40. 41. 42.

43.

44.

45.

46. 47.

85

(pancreas divisum redefined). Am J Surg 1990;159(1):59–64, discussion 64–66 Lans JI, Geenen JE, Johanson JF, Hogan WJ. Endoscopic therapy in patients with pancreas divisum and acute pancreatitis: a prospective, randomized, controlled clinical trial. Gastrointest Endosc 1992;38(4):430–434 Zeman RK, McVay LV, Silverman PM, et al. Pancreas divisum: thin-section CT. Radiology 1988;169(2):395–398 Lecesne R, Stein L, Reinhold C, Bret PM. MR cholangiopancreatography of annular pancreas. J Comput Assist Tomogr 1998;22(1):85–86 Tham RT, Heyerman HG, Falke TH, et al. Cystic fibrosis: MR imaging of the pancreas. Radiology 1991;179(1):183–186 Yoon DY, Choi BI, Han JK, Han MC, Park MO, Suh SJ. MR findings of secondary hemochromatosis: transfusional vs erythropoietic. J Comput Assist Tomogr 1994;18(3):416–419 Haustein J, Niendorf HP, Krestin G, et al. Renal tolerance of gadolinium-DTPA/dimeglumine in patients with chronic renal failure. Invest Radiol 1992;27(2):153–156 Mitchell RM, Byrne MF, Baillie J. Pancreatitis. Lancet 2003; 361(9367):1447–1455 Balthazar EJ, Freeny PC, vanSonnenberg E. Imaging and intervention in acute pancreatitis. Radiology 1994;193(2):297–306 Fishman EK, Soyer P, Bliss DF, Bluemke DA, Devine N. Splenic involvement in pancreatitis: spectrum of CT findings. AJR Am J Roentgenol 1995;164(3):631–635 Fernández-del Castillo C, Rattner DW, Makary MA, Mostafavi A, McGrath D, Warshaw AL. Débridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 1998; 228(5):676–684 Leung DA, McKinnon GC, Davis CP, Pfammatter T, Krestin GP, Debatin JF. Breath-hold, contrast-enhanced, three-dimensional MR angiography. Radiology 1996;200(2):569–571 Etemad B, Whitcomb DC. Chronic pancreatitis: diagnosis, classification, and new genetic developments. Gastroenterology 2001;120(3):682–707 Luetmer PH, Stephens DH, Ward EM. Chronic pancreatitis: reassessment with current CT. Radiology 1989;171(2):353–357 Clark LR, Jaffe MH, Choyke PL, Grant EG, Zeman RK. Pancreatic imaging. Radiol Clin North Am 1985;23(3):489–501 Kim YH, Saini S, Sahani D, Hahn PF, Mueller PR, Auh YH. Imaging diagnosis of cystic pancreatic lesions: pseudocyst versus nonpseudocyst. Radiographics 2005;25(3):671–685 Boeve WJ, Kok T, Tegzess AM, et al. Comparison of contrast enhanced MR-angiography-MRI and digital subtraction angiography in the evaluation of pancreas and/or kidney transplantation patients: initial experience. Magn Reson Imaging 2001; 19(5):595–607 Tsuchiya R, Noda T, Harada N, et al. Collective review of small carcinomas of the pancreas. Ann Surg 1986;203(1):77–81 Freeny PC, Traverso LW, Ryan JA. Diagnosis and staging of pancreatic adenocarcinoma with dynamic computed tomography. Am J Surg 1993;165(5):600–606 Megibow AJ, Zhou XH, Rotterdam H, et al. Pancreatic adenocarcinoma: CT versus MR imaging in the evaluation of resectability—report of the Radiology Diagnostic Oncology Group. Radiology 1995;195(2):327–332 Sironi S, De Cobelli F, Zerbi A, Balzano G, Di Carlo V, DelMaschio A. Pancreatic carcinoma: MR assessment of tumor invasion of the peripancreatic vessels. J Comput Assist Tomogr 1995;19: 739–744 Vellet AD, Romano W, Bach DB, Passi RB, Taves DH, Munk PL. Adenocarcinoma of the pancreatic ducts: comparative evaluation with CT and MR imaging at 1.5 T. Radiology 1992;183(1): 87–95 Sahani DV, Kadavigere R, Saokar A, Fernandez-del Castillo C, Brugge WR, Hahn PF. Cystic pancreatic lesions: a simple imaging-based classification system for guiding management. Radiographics 2005;25(6):1471–1484 Lichtenstein DR, Carr-Locke DL. Mucin-secreting tumors of the pancreas. Gastrointest Endosc Clin N Am 1995;5(1):237–258 Procacci C, Graziani R, Bicego E, et al. Intraductal mucin-producing tumors of the pancreas: imaging findings. Radiology 1996;198(1):249–257

86

3 The Pancreas

48. Imaoka H, Yamao K, Salem AA, et al. Pancreatic endocrine neoplasm can mimic serous cystadenoma. Int J Gastrointest Cancer 2005;35(3):217–220 49. Minami M, Itai Y, Ohtomo K, Yoshida H, Yoshikawa K, Iio M. Cystic neoplasms of the pancreas: comparison of MR imaging with CT. Radiology 1989;171(1):53–56 50. Fuhrman GM, Charnsangavej C, Abbruzzese JL, et al. Thin-section contrast-enhanced computed tomography accurately predicts the resectability of malignant pancreatic neoplasms. Am J Surg 1994;167(1):104–111, discussion 111–113 51. Klöppel G, Heitz PU. Pancreatic endocrine tumors. Pathol Res Pract 1988;183(2):155–168 52. Liessi G, Pasquali C, D’Andrea AA, Scandellari C, Pedrazzoli S. MRI in insulinomas: preliminary findings. Eur J Radiol 1992;14(1):46–51 53. Semelka RC, Cumming MJ, Shoenut JP, et al. Islet cell tumors: comparison of dynamic contrast-enhanced CT and MR imaging with dynamic gadolinium enhancement and fat suppression. Radiology 1993;186(3):799–802 54. Tjon A Tham RT, Falke TH, Jansen JB, Lamers CB, Lamers CB. CT and MR imaging of advanced Zollinger-Ellison syndrome. J Comput Assist Tomogr 1989;13(5):821–828 55. Kier R, Kinder B. Insulinomas: MR imaging with STIR sequences and motion suppression. [letter] AJR Am J Roentgenol 1992; 158(2):457–458

56. Tjon A Tham RT, Jansen JB, Falke TH, Lamers CB, Lamers CB. Imaging features of somatostatinoma: MR, CT, US, and angiography. J Comput Assist Tomogr 1994;18(3):427–431 57. Tjon A Tham RT, Jansen JB, Falke TH, et al. MR, CT, and ultrasound findings of metastatic vipoma in pancreas. J Comput Assist Tomogr 1989;13(1):142–144 58. Chung YS, Ho JJ, Kim YS, et al. The detection of human pancreatic cancer-associated antigen in the serum of cancer patients. Cancer 1987;60(7):1636–1643 59. Sittek H, Heuck AF, Fölsing C, Gieseke J, Reiser M. [Static and dynamic MR tomography of the pancreas: contrast media kinetics of the normal pancreatic parenchyma in pancreatic carcinoma and chronic pancreatitis]. Rofo 1995;162(5):396–403 60. Ichikawa T, Sou H, Araki T, et al. Duct-penetrating sign at MRCP: usefulness for differentiating inflammatory pancreatic mass from pancreatic carcinomas. Radiology 2001;221(1):107–116 61. Luetmer PH, Stephens DH, Fischer AP. Obliteration of periarterial retropancreatic fat on CT in pancreatitis: an exception to the rule. AJR Am J Roentgenol 1989;153(1):63–64 62. Klöppel G, Maillet B. Pseudocysts in chronic pancreatitis: a morphological analysis of 57 resection specimens and 9 autopsy pancreata. Pancreas 1991;6(3):266–274

87

4

The Spleen M. Laniado and F. Dammann

Introduction The spleen is a lymphatic organ that can be regarded as a large lymph node integrated into the blood circulatory system. At the same time, with its special vascular architecture, the spleen also affects the nonlymphatic cells in the blood. As a functionally complex organ, the spleen thus has a central, though not vital, role in adult life, which is why many disease processes involve the spleen, while primary splenic disease is fairly uncommon.1,2

Indications There are no established indications for MRI of the spleen because ultrasound and CT allow excellent diagnostic evaluation of this organ. Relative indications for MRI exist in the staging of lymphoproliferative disease (Hodgkin and non-Hodgkin lymphoma) and characterization of focal splenic lesions. MRI is also a suitable imaging modality for planning infradiaphragmatic radiotherapy in patients with lymphoproliferative disease as coronal sequences give an excellent overview of the vascular anatomy in the splenic hilum. Another relative indication is the follow-up of posttraumatic splenic hematoma in pediatric patients with inconclusive sonographic findings after nonsurgical treatment.

Imaging Technique Before the examination, the patient is given a general explanation of the procedure (e. g., duration of breathholds) including information about placement of an intravenous line for contrast injection—either Gd-based extracellular contrast medium or SPIO particles (superparamagnetic iron oxide)—and the need for administering an antispasmodic drug (e. g., butylscopolamine or glucagon). Oral contrast is not required for MRI examinations of the spleen. Coil selection depends on the equipment available (e. g., torso or body phased-array coil). Scanning is performed with the splenic hilum in the isocenter of the magnet. To this end, the patient is positioned supine

with the alignment light of the scanner centered at the level of the xiphoid process.

Imaging Planes The axial plane is the backbone of a dedicated examination of the spleen (Fig. 4.1). Additional coronal imaging is helpful if the patient’s breath-hold capacity and the MR equipment allow acquisition of good-quality images during breath-hold (Fig. 4.2). The spleen is imaged with a slice thickness of 6–8 mm and an interslice gap of 2 mm.

Pulse Sequences The basic protocol consists of unenhanced T1w and T2w sequences. T1w images are acquired with an SE or TSE sequence but preferably with a 2D GRE sequence during breath-hold. All of these sequences are acquired with the shortest possible TE to optimize image contrast. Unenhanced T2w imaging is performed with a breath-hold single-shot TSE sequence (e. g., HASTE). Because these 2D sequences have short acquisition times, basic imaging in the axial plane can be supplemented by a coronal sequence, or even sagittal images if needed, with only a minimum of additional time. A free-breathing TSE sequence with fat suppression (e. g., inversion recovery technique, spectral fat saturation) may improve image quality but takes longer to acquire. The image quality of free-breathing T2w sequences can be improved by using respiratory gating (respiratory bellows, navigator echo technique), which will markedly reduce motion artifacts. (Tables 4.1 and 4.2).

Contrast Media Intravenous bolus injection of extracellular Gd-based contrast medium (e. g., Magnevist, Dotarem) with dynamic acquisition of serial contrast-enhanced T1w images is recommended to detect and characterize focal lesions in the spleen. The dynamic images are acquired with a GRE sequence during breath-holds before and 15 s, 45 s, 90 s, and ca. 3 min after injection of the contrast medium at a dose of 0.1 mmol Gd per kg body weight. Immediate

88

4 The Spleen

a

b

c

d

e

f Fig. 4.1a–f Normal MR appearance of the spleen at 0.2 T (a, b) and 1.5 T (c–f). a, b At 0.2 T, the spleen has lower SI than the liver on T1w image (SE 450/15) (a), while it is hyperintense on T2w image (FSE 3000/102; turbo factor, 13) (b). A solid tumor of low SI is seen in the upper pole of the right kidney. c, d At 1.5 T, the spleen is again hypointense to the liver on T1w image (GRE 126/5;75°) (c) and

hyperintense on T2w image (FSE 2000/128; turbo factor, 23) (d). e On T2w HASTE image, the SI difference between spleen and liver is less pronounced (HASTE ∞/90). f On fat-suppressed T2w image, the spleen has markedly higher SI than the liver (FSE 2000/128; turbo factor, 23).

Imaging Technique

89

Fig. 4.2 Coronal T2w image acquired with HASTE sequence at 1.0 T (HASTE ∞/43). Using this technique, the normal spleen is nearly isointense to liver. Bowel is seen in the splenic hilum.

Table 4.1 Recommended pulse sequences and imaging parameters for MRI of the spleen No. of slices

No. of acquisitions

Scan time

Breathhold

75–90 –

No/(yes)

15–23

1

15–23 s

Yes

15

–

No

3

5–8 min

No

87

180

19 (16–20) < 20

1

< 20 s

Yes

Sequence type

T1

GRE (e. g., FLASH) 127–199 4.1

SE

T1

Unenhanced and dynamic Gd-enhanced Alternative (unenhanced)

T2 T2

Alternative

T2

Alternative

T2

Alternative

PD T1 T2*

Dynamic Gd-enhanced

TR (ms)

FS

Weighting

600

Single-shot TSE (e. g., HASTE) TSE (or FSE)

TE (ms)

Flip (°)

ETL

–

No

2000– 128 2500 FSE (or TSE); may ca. 3500 90 be performed with respiratory gating (belt or navigator echo) IR (e. g., TIRM, ca. 3500 60–90 STIR) TSE double-echo > 3500 22, 90

–

15–28 No + yes

6 (3–8)

1

12–28 min

Yes

–

3–7

No

19 (16–20)

2

6–9 min

No

–

5–9

Yes

2

6–9 min

No

–

3–7

No

2

6–9 min

No

3D GRE (e. g., VIBE) GRE (e. g., FLASH)

2.5

10

–

Yes

19 (16–20) 19 (16–20) 64

1

23 s

Yes

141 18 (77–196)

30

–

No

6 (3–8)

1

16 (16–23) s

Yes

5.2

Use of a torso or body phased-array coil is recommended; matrix, 192 × 256; FOV, 300–400 mm; rectangular FOV, 75 %; slice thickness, 5–8 mm; distance factor, 0.2–0.3; planes – axial, coronal (optionally), sagittal (rarely). Note: All recommendations for imaging at 1.0–1.5 T. The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time but may come with a penalty in SNR.

postcontrast T1w GRE images allow sensitive detection of focal splenic lesions3,4 (Fig. 4.3). Additional late postcontrast images should be acquired after ca. 6–10 min using a fat-suppressed SE or TSE sequence or fat-suppressed GRE sequence. Alternatively, the dynamic T1w series can be

acquired with a 3D GRE sequence (e. g., VIBE), which has higher spatial resolution and produces an angiographic effect while slightly reducing soft tissue contrast. Instead of an extracellular Gd-based contrast medium, an SPIO preparation can be infused or injected. Examples

90

4 The Spleen

a

b Fig. 4.3a–c Splenic lesion in a 28-year-old man on chemotherapy for Hodgkin disease. Unenhanced and dynamic postcontrast images obtained immediately and 5 min after injection of Gd-based contrast medium (0.1 mmol/kg Gd). a Unenhanced T1w image does not allow reliable differentiation of the splenic lesion. b Early postcontrast image clearly reveals the lesion within the inhomogeneously enhancing normal spleen. c Lesion conspicuity is reduced again on 5-min postcontrast image (1.0 T; GRE 64/5;70°).

c

Table 4.2 Recommended sequence protocol for MRI of the spleen

a) b) c) Optional Optional Optional d) Optional

Optional

Sequence

Plane

Breath-hold

Indication

T1w GRE T2w HASTE T2w TSE -/+ fat suppression, may be performed with respiratory gating T1w GRE with dynamic Gd-enhanced series with delayed images T1w SE or TSE TIRM Post-Gd T1w GRE Post-Gd T1w GRE + fat saturation

Axial Axial Axial

Yes Yes No

Axial

Yes

Basic protocol Basic protocol Basic protocol; improves image quality compared with (b) (Suspected) focal lesion

Axial Axial Axial Axial

No No Yes Yes

Axial

No

Post-SPIO T2w HASTE, TSE

If (a) not possible If (c) not possible Basic protocol For example in patients with extrasplenic disease extension or suspected pathology of surrounding structures Improved characterization of focal and diffuse changes

Note: Axial images can be replaced or supplemented by other planes (primarily coronal) depending on the anatomic situation.

of commercially available SPIO preparations are Endorem and Resovist (Europe) and Feridex (e. g., USA), which are approved for liver imaging and can therefore be used in patients whose spleen is imaged as part of an MRI examination of the liver (see Chapter 1, Contrast Media, p. 8 ff.). In this case, the precontrast protocol is supplemented by a proton density (PD) sequence.5,6 Identical T2w, PD, and T1w sequences are obtained before and after SPIO admin-

istration. The pre- and postcontrast PD images are ideally acquired with a double-echo TSE sequence, producing a T2w image with the second echo.

MRI Appearance of Normal Anatomy

MRI Appearance of Normal Anatomy The spleen has longer T1 and T2 relaxation times than the liver, resulting in lower signal intensity on T1w images and higher signal intensity on T2w images compared with the liver7,8 (see Fig. 4.1). However, the two organs may be

91

of similar signal intensity when fast T2w techniques are used (Fig. 4.4). After bolus injection of Gd-based contrast medium, the spleen shows early heterogeneous enhancement (arciform, mottled, or peripheral) and within a few minutes becomes uniformly hyperintense 4,9–12 (Fig. 4.5). Occasionally, there may be immediate uniform enhancement.4,10 Following SPIO injection, ca. 10 % of the dose

a

b Fig. 4.4a, b Effects of imaging parameters on T2 contrast illustrated in a 53-year-old woman with central liver metastasis (segments III/ IVa) and subcapsular liver cyst (segment IVa) but normal spleen (1.0 T). a On T2w FSE image (TR/TE 3000/90; turbo factor, 5), the spleen

a

is of high SI similar to that of the liver metastasis. b On breath-hold image (FSE 2500/138; turbo factor, 29), the spleen and liver are isointense.

b Fig. 4.5a–c Normal dynamic contrast-enhanced MRI of the spleen in a 63-year-old man with portal vein thrombosis (1.0 T). a Shadowing in the unenhanced T1w image is due to the use of a phased-array body coil. b Immediately after bolus injection of Gd-based contrast medium (0.1 mmol/kg Gd), there is arciform enhancement of the spleen. c Uniform enhancement of the spleen after 1 min (GRE 106/ 5;70°).

c

92

4 The Spleen

a

b

c

d

e

f Fig. 4.6a–f Normal spleen before (a–c) and after (d–f) administration of SPIO contrast medium (Resovist, 8 µmol/kg Fe) in a 45-yearold woman with focal nodular hyperplasia in liver segment VII (1.0 T). a–c On precontrast PD image (SE 2500/45) (a) and T2w image (SE 2500/90) (b), the spleen is hyperintense to the liver while on T1w image (c) it is hypointense (GRE 150/5;70°). d, e On post-

contrast PD (d) and T2w (e) images, there is marked signal reduction of the spleen and liver (70 min after SPIO). f Postcontrast T1w image obtained with identical window/center settings shows markedly reduced SI in the liver and slightly decreased SI in the spleen.

reaches the spleen, resulting in a marked signal decrease on PD and T2w images.5 On SPIO-enhanced T1w GRE images, the spleen is slightly hypointense or hyperintense compared with precontrast images (Fig. 4.6).

The neonatal spleen is isointense or slightly hypointense relative to the liver on T2w images in the first week of life and then gradually becomes hyperintense until 8 months of age. On T1w images, the spleen is isointense or slightly hypointense to the liver during the first 2 weeks of life, only later becoming more markedly hypointense.13

MRI Appearance of Pathologic Entities

MRI Appearance of Pathologic Entities Benign Tumors The two most common benign tumors of the spleen are cysts and hemangiomas. Less common are lymphangiomas, hamartomas, and inflammatory pseudotumors. Note that MR data on these entities are still limited because only case reports have been published on the MR appearances of most of the benign splenic lesions presented below. Most splenic cysts are pseudocysts secondary to trauma or splenic infarction (80 % of all splenic cysts), epidermoid cysts (true congenital cysts with an epithelial lining), and hydatid (echinococcal) cysts (splenic echinococcosis < 2 %). In rare cases, splenic pseudocysts may arise in patients with pancreatitis. Cysts appear as well-defined lesions with low signal intensity on T1w images and high signal intensity on T2w images and do not enhance after

93

contrast administration14 (Fig. 4.7). Only intracystic septa, which are common in hydatid cysts, may increase in signal intensity on postcontrast images.15 Calcifications can occur in hydatid, posttraumatic, and epidermoid cysts and are characterized by uniform low signal intensity on all pulse sequences. When the cyst fluid has a high cell or protein content or contains blood, signal intensity is higher on T1w images (Fig. 4.8), which may help differentiate secondary from primary cysts. While common in the liver, hemangiomas are very rare in the spleen but constitute the majority of primary splenic tumors. The most common type in the spleen is cavernous hemangioma, which often has cystic spaces containing serous or hemorrhagic fluid.16 Capillary hemangioma of the spleen is rare. Spontaneous splenic rupture occurs in up to 25 % of patients. If the entire splenic parenchyma is replaced by multiple hemangiomas, the condition is referred to as hemangiomatosis and may be a sign of generalized angiomatosis (Klippel–Trenaunay– Weber syndrome). Hemangiomas are characterized by

a

b

c

d Fig. 4.7a–f Epidermoid cyst in a 34-year-old woman (1.5 T). a, b On T1w images, the cyst has lower SI than the spleen (GRE 126/5;75°). c On the breath-hold T2w image, the cyst is seen as a well circumscribed lesion of uniform high SI (FSE 2500/138; turbo factor, 23). d The cyst is even more conspicuous on fat-suppressed image (FSE

5000/108). e, f There is no signal enhancement of the cyst after injection of Gd-based contrast medium (0.1 mmol/kg Gd; GRE 126/ 5;75°). Note wraparound artifacts in c and f. Fig. 4.7e, f e

94

4 The Spleen

e

f Fig. 4.7e, f

a

b

c

d Fig. 4.8a–e Splenic cyst after septic embolism in a 62-year-old woman (1.0 T). a On unenhanced T1w image, the cyst has higher SI than the spleen (GRE 112/5;70°). b–d PD (SE 3000/22) (b) and T2w images (SE 3000/90; FSE 2500/128; turbo factor, 15) (c, d)

show inhomogeneities within the cyst, probably due to cell debris. The low SI of the spleen on PD and T2w images suggests iron storage in the MPS. Fig. 4.8e e

MRI Appearance of Pathologic Entities

95

Fig. 4.8 e There is no enhancement of the cyst after injection of Gdbased contrast medium (0.1 mmol/kg Gd; GRE 112/5;70°).

e

a

b

c

d Fig. 4.9a–d Hemangiomas in the spleen of a 59-year-old man (1.5 T). a, b Three hyperintense lesions are revealed on PD image (SE 2500/15) (a) and T2w image (SE 2500/90) (b). c The lesions are

not visible on the unenhanced T1w image (SE 600/15). d Delayed image after injection of Gd-based contrast medium shows intense, uniform enhancement of the hemangiomas (SE 600/15).

uniform high signal intensity on T2w images and typically have smooth margins17–19 (Fig. 4.9). An inhomogeneous appearance suggests the presence of cystic and solid components (e. g., hemangiomatosis) or infarction within a large hemangioma17,19,20 (Fig. 4.10). On T1w images, they have the same or lower signal intensity compared with the spleen but may be hyperintense in the presence of intralesional hemorrhage.20–22 On contrast-enhanced

dynamic images acquired after injection of extracellular Gd-based contrast medium, hemangiomas characteristically show peripheral enhancement with progression to uniform enhancement, resulting in hyperintensity relative to the spleen on late images (Fig. 4.9).20 Other enhancement patterns are immediate and persistent homogeneous enhancement and initial peripheral enhancement with centripetal progression but persistent nonen-

96

4 The Spleen

a

b Fig. 4.10a–c Splenic hemangiomatosis in a 63-year-old woman (1.5 T). a T2w image shows an inhomogeneous spleen with both hypointense and hyperintense lesions. The sediment-fluid level in one of the two hyperintense lesions in the posterior aspect of the spleen is due to sedimentation after hemorrhage (SE 2100/90). b Only the hyperintense hemorrhagic lesion is seen on unenhanced T1w image (SE 600/15). c Delayed image obtained after injection of Gd-based contrast medium (0.1 mmol/kg Gd) reveals multiple lesions in the spleen (SE 600/15).

c

hancement of the center on delayed images, consistent with a central fibrous scar. Mixed high and low signal intensity is seen in hemangiomatosis19,23 (Fig. 4.10) and littoral cell angioma, an extremely rare vascular tumor of the spleen24,25 (Fig. 4.11). Lymphangiomas of the spleen resemble hemangiomas in that they also have epithelium-lined spaces, which, however, contain lymph fluid rather than blood.16 Capillary, cavernous, and cystic lymphangiomas are distinguished.26 Splenic lymphangiomas are single or multiple (lymphangiomatosis) and can occur alone or in association with hemangiomas and lymphangiomas in other organs, so-called systemic cystic angiomatosis, which is a progressive condition with a poor prognosis. Lymphangiomas have the same MR appearance as proteinaceous cysts.16 Hamartomas are very rare benign splenic tumors that are composed of an anomalous mixture of normal red pulp elements with or without pulp cells. Most hamarto-

mas are rather small, but larger ones can occur and may be associated with hypersplenism. Splenic hamartomas are isointense on T1w images and usually markedly hyperintense on T2w images. Their appearance is heterogeneous, and they may contain areas of fibrosis and calcification, which are hypointense on T2w images.27 On postcontrast images following administration of Gd-based contrast medium, hamartomas are characterized by slow enhancement with a clear signal increase on more delayed images, which may be inhomogeneous.18,20 Inflammatory pseudotumor is another very rare splenic mass. Histologically, the tumor is composed of a fibroblastic stroma and contains inflammatory cells. Its MR signal pattern is not uniform. Inflammatory pseudotumor has been reported to be isointense to the spleen on T1w images and heterogeneously hypointense on T2w images with delayed enhancement after contrast administration.28 However, the tumor may also show early, intense enhancement (Fig. 4.12).

MRI Appearance of Pathologic Entities

97

a

b

c

d

e

f Fig. 4.11a–f Littoral cell angioma in a 51-year-old woman (1.0 T). a, b T2w images show multiple focal lesions of high SI in the spleen (FSE 2500/128; turbo factor, 32). c Only some of the lesions are seen as hypointense foci on unenhanced T1w image (GRE 112/5;70°). d Immediate postcontrast image obtained after injection of Gd-

based contrast medium (0.1 mmol/kg Gd) depicts numerous hypointense lesions (GRE 63/5;70°). e The lesions are poorly demarcated from the surrounding spleen ca. 2 min after contrast injection (GRE 63/5;70°). f On 10-min postcontrast image, the lesions are hyperintense to the spleen (GRE 112/5;70°).

98

4 The Spleen

a

b

c

d Fig. 4.12a–e Inflammatory pseudotumor in a 72-year-old woman (1.5 T). a T1w image shows the tumor with slightly lower SI than the spleen (GRE 127/4;80°). b On T2w image, the tumor is slightly hyperintense and inhomogeneous (TrueFISP 4/2;80°). c Immediately after bolus injection of Gd-based contrast medium

(0.1 mmol/kg Gd), there is intense enhancement of the tumor (GRE 127/4;80°). d The tumor becomes isointense to the spleen within 1 min (GRE 127/4;80°). Fig. 4.12e e

MRI Appearance of Pathologic Entities

99

Malignant Tumors

e Fig. 4.12 e The pseudotumor remains isointense on delayed image with fat saturation (GRE 86/4;60°). There are two small cysts in the left kidney.

a

Lymphoma is the most common malignancy of the spleen. When first diagnosed with the disease, 23–34 % of patients with Hodgkin lymphoma and 30–40 % of those with nonHodgkin lymphoma have splenic involvement. Splenic lymphoma can manifest as diffuse involvement without circumscribed masses, small focal lesions (< 2 cm), or large tumor nodes. If there is necrosis, cystic components will be present. Rare primary splenic lymphoma without nodular manifestation (accounting for 1 % of non-Hodgkin lymphomas, more common in AIDS-associated lymphoma) can extend through the capsule and infiltrate adjacent organs. Diffuse infiltration of the spleen and lesions < 1 cm (45–70 % of cases) are difficult to detect by MRI and may only be suggested by splenomegaly. However, absence of the signal loss normally observed after SPIO administration has been reported to be a sensitive predictor of diffuse splenic lymphoma.6 Alternatively, diffuse involvement can be identified as patchy enhancement of the spleen on dynamic MRI after administration of extracellular Gd-based contrast medium4 (Fig. 4.13). Even some larger focal lesions may be difficult to detect with unen-

b Fig. 4.13a–c Micronodular non-Hodgkin lymphoma of the spleen in a 58-year-old woman (1.5 T). a T1w image reveals splenomegaly and a lymphoma in the splenic hilum (GRE 105/5;75°). b T2w image depicts multiple small lesions that are slightly hypointense (FSE 3800/90; turbo factor, 5). c The small nodules are more conspicuous following injection of Gd-based contrast medium (0.1 mmol/kg Gd) (GRE 105/5;75°).

c

100

4 The Spleen

a

b Fig. 4.14a–c Focal non-Hodgkin lymphoma of the spleen in a 48-year-old man (1.0 T). a Unenhanced T1w image shows an enlarged spleen with small foci of low SI (GRE 127/5;70°). b On T2w image, the lesions also have low SI but more lesions are visible (FSE 2000/138; turbo factor, 29). c Injection of Gd-based contrast medium (0.1 mmol/kg Gd) leads to marked heterogeneous enhancement of the spleen and numerous focal lesions are revealed (GRE 127/5;70°).

c

hanced sequences because contrast to surrounding spleen is minimal due to similar T1 and T2 relaxation times (Fig. 4.14).8,16,22,29 Such lesions are easier to identify if they contain necrotic areas, which have higher signal intensity than the spleen on T2w images.8 Focal lesions are much more conspicuous on dynamic MR images.4 Experience suggests that the use of an SPIO contrast medium will also improve detection of focal splenic lymphoma compared with unenhanced pulse sequences. Metastases in the spleen are seen in approximately 7 % of patients with metastatic tumors. The most common primary tumor spreading to the spleen is melanoma (50 %); less common are metastases from cancer of the breast, lungs, colon, ovaries, endometrium, and prostate. Splenic metastases have low signal intensity on T1w images and increased signal intensity on T2w images8 (Fig. 4.15). SPIO contrast media improve the detection of splenic metastases on PD and T2w images.5 Metastases are also more conspicuous on T1w images obtained im-

mediately after bolus injection of extracellular Gd-based contrast medium. Splenic metastases from melanotic melanoma may be sensitively detected on unenhanced T1w images as a result of their high signal intensity. On T2w images they may have low signal intensity relative to the spleen30 (Fig. 4.16). Hemorrhagic metastases from other primary tumors may have the same signal characteristics.31 Angiosarcoma is a very rare splenic neoplasm but constitutes the most common nonlymphoid primary tumor of the spleen. Metastatic spread is typically to the liver, and patients are at risk of spontaneous splenic rupture. Lesions are single or multiple and may be made up of cystic and solid components. Thorotrast storage has been shown to induce hepatic angiosarcoma but there is no evidence that it also causes splenic angiosarcoma. If intratumoral hemorrhage is present, angiosarcoma is hypointense to the spleen on T1w and T2w images.32,33

MRI Appearance of Pathologic Entities

101

a

b

c

d Fig. 4.15a–d Multiple liver metastases and a single splenic metastasis from breast cancer in a 61-year-old woman (1.0 T). a Unenhanced T1w image reveals metastases in the liver but not the metastasis in the low-SI spleen (GRE 127/5;70°). b T2w image demonstrates both the hyperintense metastases in the liver and the hyperintense metastasis in the spleen (FSE 2500/128; turbo

factor, 23). c On fat-suppressed T2w image, the splenic metastasis is poorly delineated because it has the same high SI as the spleen (FSE 3142/22; turbo factor, 5). d Contrast-enhanced T1w image obtained 2 min after injection of Gd-based contrast medium (0.1 mmol/kg Gd) clearly reveals the splenic metastasis, while the small liver metastases are difficult to see (GRE 127/5;70°).

Fig. 4.16a–c Metastases from melanotic melanoma in the liver and spleen of a 64-year-old man (1.0 T). a Unenhanced T1w image reveals multiple high-SI metastases in the liver and both metastases in the spleen (GRE 112/5;70°). b The metastases have low SI on T2w

image and are clearly delineated only in the spleen because of its higher normal SI compared with the liver (FSE 2500/128; turbo factor, 23). Fig. 4.16c e

a

b

102

4 The Spleen Fig. 4.16 c Following injection of Gd-based contrast medium (0.1mmol/kg Gd), the metastases in the liver and spleen are less conspicuous but still demarcated from the enhancing surrounding tissue as slightly hyperintense lesions (GRE 127/5;70°).

c

Infectious and Noninfectious Conditions Histoplasmosis, tuberculosis, and echinococcosis are the most common nonviral infections involving the spleen in immunocompetent individuals. Viral infections are most commonly caused by Epstein–Barr virus, varicella, and cytomegalovirus and result in splenomegaly. Immunocompromised patients have an increased risk of splenic infection with Candida albicans, Aspergillus fumigatus, and Cryptococcus neoformans (multiple microabscesses); Mycobacterium tuberculosis (usually in a miliary form); Mycobacterium avium-intracellulare (MAI); and Pneumocystis carinii (characterized by early disseminated calcifications). There are five mechanisms leading to formation of splenic abscesses: · hematogenous dissemination (e. g., endocarditis, sepsis) · direct spread from adjacent organs (e. g., pancreatitis, perinephric abscess) · superinfection secondary to thromboembolic splenic infarction · trauma · immunocompromised states. A pyogenic abscess has variable signal intensity on T1w and T2w images. It is hypointense to normal spleen on T1w images but may also be hyperintense (pus containing many granulocytes). On contrast-enhanced images, a splenic abscess is visualized as an irregular, cystic lesion. Multiple microabscesses due to fungal infection of the spleen are seen as rounded lesions with diminished signal intensity on T1w images and increased signal intensity on T2w images.22,34 Even very small abscesses are sensitively detected by dynamic MRI.34 The fact that small fungal abscesses are clearly visualized even in patients with storage of excess iron in the mononuclear phagocyte system (MPS) of the spleen35 indicates that administration of SPIO contrast medium will improve the detection of microabscesses by MRI.

Sarcoidosis is a multiorgan granulomatous disease that can also involve the spleen. In splenic sarcoidosis, the spleen has heterogeneous low SI, and/or hypointense lesions are seen on T2w images36 (Fig. 4.17). In Niemann–Pick disease, a rare lipidosis due to a congenital enzyme defect with autosomal recessive inheritance, there is generalized abnormal accumulation of sphingomyelin in various organs including the spleen. The lipid deposits cause a slight increase in splenic signal intensity on T1w and T2w images. Localized lesions are characterized by delayed uptake of extracellular Gd-based contrast medium and are markedly hyperintense on T2w images.37 Focal splenic lesions of variable size and signal intensity (typically isointense on T1w images and hypointense on T2w images) in an enlarged spleen may also be seen in patients with Gaucher disease (19–30 % of cases).

Trauma Injury to the spleen can result from blunt or penetrating abdominal trauma or be caused iatrogenically (by surgical or other medical procedures). Acute hematoma has low signal intensity on T1w images and high signal intensity on T2w images and, by day 7, becomes hyperintense on T1w images as well. At higher field strength (e. g., 1.5 T), chronic hematoma is characterized by a low-signal-intensity rim, which is due to hemosiderin deposits and is best appreciated on T2w images.38 Calcifications in older hematomas have low signal intensity with all sequences. A posttraumatic pseudocyst has the same signal characteristics as cysts of other origin. Splenic hematomas do not enhance.

MRI Appearance of Pathologic Entities

103

a

b

c

d Fig. 4.17a–d Sarcoidosis involving the spleen in a 44-year-old woman (1.0 T). a Unenhanced T1w image depicts the spleen with uniform, low SI (GRE 139/7;70°). b T2w image reveals hypointense lesions in the spleen (FSE 2000/128; turbo factor, 23). c The lesions

are even more conspicuous on the fat-suppressed image (FSE 2000/ 128; turbo factor, 23). d Marked enhancement of the splenic granulomas 3 min after injection of Gd-based contrast medium (0.1 mmol/kg Gd) (GRE 139/7;70°).

Splenic Infarction

histiocytosis), hematologic disorders (e. g., polycythemia vera, myelofibrosis, leukemia), inflammatory conditions (e. g., infectious mononucleosis), and neoplastic disease (e. g., lymphoma, leukemia). Some causes of splenomegaly can be diagnosed by MRI (e. g., portal hypertension with fundal varices). In most cases, however, the underlying condition is suggested by the clinical findings (Fig. 4.20). A very small spleen is seen in thorotrastosis and the end stage of splenic atrophy in homozygous sickle cell anemia (autosplenectomy).22 In patients with sickle cell anemia, hemosiderin deposits and calcifications in the spleen result in low signal intensity of the organ on T1w and T2w images.7 Low signal intensity on all pulse sequences is also seen when there is extensive fibrosis in thorotrastosis. Splenic Iron Deposition. The two most common causes of iron storage in the spleen are acute sequestration of red blood cells in the spleen, e. g., in patients having received multiple transfusions (extravascular hemolysis), and portal hypertension. A less common cause is rhabdomyolysis. In the setting of rhabdomyolysis and extravascular hemolysis, there is diffuse iron deposition in the MPS of the spleen, while portal hypertension is associated with focal

Splenic infarction is most commonly caused by thromboembolism in patients with endocarditis, atrial fibrillation, or left ventricular thrombus (ischemic infarction). Venous infarction can occur as a result of thrombosis in the splenic sinusoids due to reduced blood flow in massive splenomegaly. Complications of splenic infarction are subcapsular hematoma, splenic rupture, and superinfection. The morphologic appearance and signal intensity of splenic infarcts vary with the stage. In acute infarction, the infarcted area is typically wedge-shaped (Fig. 4.18). Subacute infarcts may be seen as cystlike lesions (Fig. 4.19), while the presence of fibrosis determines the MR appearance in chronic infarction.16,22

Diffuse Disease Splenomegaly. Many conditions can cause enlargement of the spleen. These include venous outflow obstruction (e. g., portal hypertension, splenic vein occlusion or thrombosis), infiltrative conditions (e. g., Gaucher disease,

104

4 The Spleen

a

b Fig. 4.18a, b Acute splenic infarction in a 37-year-old man (1.0 T). a The lesions in the spleen are not visible on unenhanced T1w image (GRE 128/5;70°). b On contrast-enhanced T1w image (0.1 mmol/kg

Gd), the lesions are seen as perfusion defects, some of which are wedge-shaped (GRE 128/5;70°).

a

b

c

d Fig. 4.19a–d Splenic infarction with hemorrhage in a 6-year-old boy with cirrhosis and splenomegaly (1.0 T). a Unenhanced T1w image shows a slightly hyperintense, arciform area in the enlarged spleen, which likely corresponds to a hemorrhagic rim. There is hypointense ascites in the right paracolic gutter (SE 6000/12). b, c On PD (SE

3000/22) (b) and T2w (SE 3000/90) (c) images, the infarcted area has markedly increased SI. The presumed hemorrhagic rim has low SI, consistent with hemosiderin deposits. d There is no enhancement of the infarcted area after injection of Gd-based contrast medium (0.1 mmol/kg Gd) (SE 600/12).

MRI Appearance of Pathologic Entities

105

hemosiderin deposition in so-called Gamna–Gandy bodies, which are made up of thickened collagen bundles. Congenital (autosomal recessive) and acquired hemochromatosis (e. g., in sideroblastic anemia) with excess iron absorption in the intestine is characterized by iron deposits in hepatocytes, pancreas, endocrine glands, and heart, while the MPS cells of the liver and spleen are spared. Paroxysmal nocturnal hemoglobinuria (intravascular hemolysis) is another condition in which excess iron is stored in other organs (liver, kidneys) but not in the spleen. When excess iron is stored in the MPS, the spleen has the same MR signal characteristics as after administration of SPIO-based contrast medium (Fig. 4.21). The resulting diffuse signal reduction is most sensitively detected using PD and T2*w GRE sequences.7,37 Large

Fig. 4.20 Marked splenomegaly in a 41-year-old man with cirrhosis w (1.0 T; coronal plane; GRE 128/5;70°).

a

b

c

d Fig. 4.21a–d Iron storage in the MPS of the spleen in a 23-year-old man with autoimmune hemolysis (1.5 T). a On unenhanced T1w image, the spleen is markedly hypointense relative to the liver (GRE 75/5;72°). b The hypointensity of the spleen is even more pronounced on T2w image (FSE 3000/138; turbo factor, 28). c T2*w

image shows not only hypointensity of the spleen but also reduced SI of the liver (GRE 96/18;30°). d Postcontrast image after injection of Gd-based contrast medium (0.1 mmol/kg Gd) with fat saturation reveals only little enhancement of the spleen (GRE 73/5;60°).

106

4 The Spleen

a

b Fig. 4.22a–c Normal SI of the spleen and hypointensity of the liver due to iron accumulation in hepatocytes in a 76-year-old man with hemochromatosis and cirrhosis, hepatocellular carcinoma (not shown), and splenomegaly (0.2 T). a T1w image demonstrates abnormally low SI of the liver and isointensity of the spleen (SE 520/15). b, c On PD image (FSE 3500/26; turbo factor, 5) (b) and T2w image (FSE 3500/106; turbo factor, 5) (c), the enlarged spleen has normal high SI, while liver SI is reduced. The hypointensity to the left of the spleen represents the hypertrophic left hepatic lobe.

c

amounts of iron also cause signal voids on T1w GRE images. In patients with hemochromatosis, the signal intensity is reduced in the cirrhotic liver, while the spleen may be enlarged but has normal signal intensity (Fig. 4.22). Gamna–Gandy bodies, or siderotic nodules, are typically diffusely distributed throughout the spleen. The nodules are small, with a diameter of a few millimeters. They are most conspicuous as lesions of low signal intensity on GRE images acquired at high field strength.40–42 On T1w images, the lesions are more conspicuous after administration of Gd-based contrast medium.40 When SPIO particles are used, the nodules become isointense to the spleen or remain visible (Fig. 4.23). Gamna–Gandy bodies are most commonly caused by portal hypertension and portal or splenic vein thrombosis or are idiopathic. Rarely, they occur in the setting of hemolytic anemia, leukemia, and lymphoma.

References 1. Dachman AH, Friedman AC. Radiology of the Spleen. St. Louis: Mosby; 1993 2. Semelka RC, Shoenut JP. The spleen. In: Semelka RC, JP Shoenut. MRI of the Abdomen with CT Correlation. New York: Raven Press; 1993: pp. 53–58 3. Mirowitz SA, Brown JJ, Lee JK, Heiken JP. Dynamic gadoliniumenhanced MR imaging of the spleen: normal enhancement patterns and evaluation of splenic lesions. Radiology 1991; 179(3):681–686 4. Semelka RC, Shoenut JP, Lawrence PH, Greenberg HM, Madden TP, Kroeker MA. Spleen: dynamic enhancement patterns on gradient-echo MR images enhanced with gadopentetate dimeglumine. Radiology 1992;185(2):479–482 5. Weissleder R, Hahn PF, Stark DD, et al. Superparamagnetic iron oxide: enhanced detection of focal splenic tumors with MR imaging. Radiology 1988;169(2):399–403 6. Weissleder R, Elizondo G, Stark DD, et al. The diagnosis of splenic lymphoma by MR imaging: value of superparamagnetic iron oxide. AJR Am J Roentgenol 1989;152(1):175–180 7. Adler DD, Glazer GM, Aisen AM. MRI of the spleen: normal appearance and findings in sickle-cell anemia. AJR Am J Roentgenol 1986;147(4):843–845 8. Hahn PF, Weissleder R, Stark DD, Saini S, Elizondo G, Ferrucci JT. MR imaging of focal splenic tumors. AJR Am J Roentgenol 1988;150(4):823–827 9. Hamed MM, Hamm B, Ibrahim ME, Taupitz M, Mahfouz AE. Dynamic MR imaging of the abdomen with gadopentetate dimeglumine: normal enhancement patterns of the liver, spleen, stomach, and pancreas. AJR Am J Roentgenol 1992;158(2): 303–307

MRI Appearance of Pathologic Entities

107

a

b

c

d Fig. 4.23a–d Gamna–Gandy bodies in the enlarged spleen of a 35-year-old man with cirrhosis and hepatocellular carcinoma in the right hepatic lobe (1.0 T). a–c T1w (GRE 150/5;70°) (a), PD (SE 2500/45) (b), and T2w (SE 2500/90) (c) images before contrast administration show multiple hypointense lesions in the spleen.

d T1w image (GRE 150/5;70°) obtained 40 min after injection of SPIO contrast medium (Resovist, 8 µmol/kg Fe) shows improved lesion conspicuity (arrows).

10. Ito K, Mitchell DG, Honjo K, et al. Gadolinium-enhanced MR imaging of the spleen: artifacts and potential pitfalls. AJR Am J Roentgenol 1996;167(5):1147–1151 11. Ito K, Mitchell DG, Honjo K, et al. MR imaging of acquired abnormalities of the spleen. AJR Am J Roentgenol 1997;168(3): 697–702 12. Mirowitz SA, Gutierrez E, Lee JK, Brown JJ, Heiken JP. Normal abdominal enhancement patterns with dynamic gadoliniumenhanced MR imaging. Radiology 1991;180(3):637–640 13. Donnelly LF, Emery KH, Bove KE, Bissett GSIII. Normal changes in the MR appearance of the spleen during early childhood. AJR Am J Roentgenol 1996;166(3):635–639 14. Shirkhoda A, Freeman J, Armin AR, Cacciarelli AA, Morden R. Imaging features of splenic epidermoid cyst with pathologic correlation. Abdom Imaging 1995;20(5):449–451 15. von Sinner WN, Stridbeck H. Hydatid disease of the spleen. Ultrasonography, CT and MR imaging. Acta Radiol 1992;33(5): 459–461 16. Urrutia M, Mergo PJ, Ros LH, Torres GM, Ros PR. Cystic masses of the spleen: radiologic-pathologic correlation. Radiographics 1996;16(1):107–129 17. Disler DG, Chew FS. Splenic hemangioma. AJR Am J Roentgenol 1991;157(1):44 18. Ohtomo K, Fukuda H, Mori K, Minami M, Itai Y, Inoue Y. CT and MR appearances of splenic hamartoma. J Comput Assist Tomogr 1992;16(3):425–428

19. Peene P, Wilms G, Stockx L, Rigauts H, Vanhoenacker P, Baert AL. Splenic hemangiomatosis: CT and MR features. J Comput Assist Tomogr 1991;15(6):1070–1072 20. Ramani M, Reinhold C, Semelka RC, et al. Splenic hemangiomas and hamartomas: MR imaging characteristics of 28 lesions. Radiology 1997;202(1):166–172 21. Harris RD, Simpson W. MRI of splenic hemangioma associated with thrombocytopenia. Gastrointest Radiol 1989;14(4): 308–310 22. Rabushka LS, Kawashima A, Fishman EK. Imaging of the spleen: CT with supplemental MR examination. Radiographics 1994; 14(2):307–332 23. Teufl F, Duda SH, Horny HP, Xiac JC, Busch FW, Schareck W. Hämangiomatose der Milz. Erscheinungsbild in der Magnetresonanztomographie. Radiol Diagn (Berl) 1992;33:193–196 24. Oliver-Goldaracena JM, Blanco A, Miralles M, Martin-Gonzalez MA. Littoral cell angioma of the spleen: US and MR imaging findings. Abdom Imaging 1998;23(6):636–639 25. Schülen V, Horny P, Busch F-W. Two cases of vascular tumors of the spleen; imaging results with pathologic correlation. Radiologie 1994;14:61–62 26. Ito K, Murata T, Nakanishi T. Cystic lymphangioma of the spleen: MR findings with pathologic correlation. Abdom Imaging 1995;20(1):82–84 27. Pinto PO, Avigado P, Garcia H, Alves EC, Marques C. Splenic hamartoma: a case report. Eur Radiol 1995;5:93–95

108

4 The Spleen

28. Irie H, Honda H, Kaneko K, et al. Inflammatory pseudotumors of the spleen: CT and MRI findings. J Comput Assist Tomogr 1996;20(2):244–248 29. Hess CF, Griebel J, Schmiedl U, Kurtz B, Koelbel G, Jaehde E. Focal lesions of the spleen: preliminary results with fast MR imaging at 1.5 T. J Comput Assist Tomogr 1988;12(4):569–574 30. Premkumar A, Sanders L, Marincola F, Feuerstein I, Concepcion R, Schwartzentruber D. Visceral metastases from melanoma: findings on MR imaging. AJR Am J Roentgenol 1992;158(2): 293–298 31. Torres GM, Terry NL, Mergo PJ, Ros PR. MR imaging of the spleen. Magn Reson Imaging Clin N Am 1995;3(1):39–50 32. Ha HK, Kim HH, Kim BK, Han JK, Choi BI. Primary angiosarcoma of the spleen. CT and MR imaging. Acta Radiol 1994;35(5):455–458 33. Kaneko K, Onitsuka H, Murakami J, et al. MRI of primary spleen angiosarcoma with iron accumulation. J Comput Assist Tomogr 1992;16(2):298–300 34. Semelka RC, Shoenut JP, Greenberg HM, Bow EJ. Detection of acute and treated lesions of hepatosplenic candidiasis: comparison of dynamic contrast-enhanced CT and MR imaging. J Magn Reson Imaging 1992;2(3):341–345 35. Cho J-S, Kim EE, Varma DGK, Wallace S. MR imaging of hepatosplenic candidiasis superimposed on hemochromatosis. J Comput Assist Tomogr 1990;14(5):774–776

36. Kessler A, Mitchell DG, Israel HL, Goldberg BB. Hepatic and splenic sarcoidosis: ultrasound and MR imaging. Abdom Imaging 1993;18(2):159–163 37. Omarini LPA, Frank-Burkhardt SE, Seemayer TA, Mentha G, Terrier F. Niemann-Pick disease type C: nodular splenomegaly. Abdom Imaging 1995;20(2):157–160 38. Hahn PF, Saini S, Stark DD, Papanicolaou N, Ferrucci JTJr. Intraabdominal hematoma: the concentric-ring sign in MR imaging. AJR Am J Roentgenol 1987;148(1):115–119 39. Siegelman ES, Mitchell DG, Rubin R, et al. Parenchymal versus reticuloendothelial iron overload in the liver: distinction with MR imaging. Radiology 1991;179(2):361–366 40. Minami M, Itai Y, Ohtomo K, et al. Siderotic nodules in the spleen: MR imaging of portal hypertension. Radiology 1989; 172(3):681–684 41. Roubidoux MA. MR of the kidneys, liver, and spleen in paroxysmal nocturnal hemoglobinuria. Abdom Imaging 1994;19(2): 168–173 42. Sagoh T, Itoh K, Togashi K, et al. Gamna-Gandy bodies of the spleen: evaluation with MR imaging. Radiology 1989;172(3): 685–687

109

5

The Gastrointestinal Tract W. Luboldt and M. Laniado

Introduction

Indications

For many decades, imaging of the gastrointestinal (GI) tract was the domain of radiographic procedures, notably double-contrast techniques. Today, with endoscopy having evolved into a serious competitor, even the small intestine is no longer the undisputed territory of radiographic techniques. New perspectives are also opened up by the advances made in CT, which offers good visualization of the bowel wall and provides more diagnostic information than conventional X-ray examinations and endoscopy because it also allows direct evaluation of the tissues outside the GI tract. MRI, the other major crosssectional imaging modality, was of little use for GI imaging in its early years because of long acquisition times. The situation has changed with the development of fast MR pulse sequences and parallel imaging techniques. Nevertheless, CT is still preferred to MRI for most diagnostic applications in the GI tract because it is superior in terms of speed, resolution, multiplanar image reformation, imaging of the entire abdomen in a single scan, and absence of artifacts. On the other hand, MRI has the crucial advantage of not involving ionizing radiation and is therefore generally preferred in young persons and pregnant women, for follow-up examinations, and in dynamic studies. With its high intrinsic contrast, MRI has great potential for evaluating inflammatory and neoplastic processes of the bowel.

MR enteroclysis1–6 has been used since 1997 to evaluate and follow up complications of Crohn disease that require surgery. MR colonography (MRC) is still under investigation for colorectal screening7–12 (Table 5.1). MRI is also a promising modality for diagnosing appendicitis in cases where ultrasound findings are inconclusive and imaging does not delay surgery. Finally, MRI enables functional assessment in esophageal reflux, gastric emptying disorders, adhesions, pelvic floor hernias, and defecation disorders.

Imaging Technique Patient Preparation Fasting for 12 h is sufficient before MRI of the stomach or small intestine, but active bowel preparation is needed for imaging of the colon. Bowel cleansing for MR colonography is the same as for colonoscopy: using, for example, a polyethylene glycol electrolyte solution (PEG-ES; Golytely, Klean-Prep).13 For adequate cleansing, patients should start drinking the solution at 3 p. m. on the day before the examination and restrict themselves to clear liquids. The question of whether less strict bowel preparation or dietary measures alone may be sufficient before intestinal MRI7 needs to be investigated in further studies. An ad-

Table 5.1 Indications for MRI of the GI tract Indication (examination)

Sequence (planes)

Comment

Crohn disease (MR enteroclysis)

HASTE, TrueFISP (?), contrast-enhanced VIBE (abdomen: coronal; pelvis: axial) HASTE, TrueFISP (?), contrast-enhanced VIBE (abdomen: coronal; pelvis: axial) HASTE, HASTE-IRM, contrast-enhanced VIBE (?) (identification of the appendix with axial and coronal HASTE and HASTE-IRM; if still inconclusive, use additional focused contrast-enhanced axial VIBE for MPR)

To evaluate complications (see Table 5.3) Under investigation

Colorectal screening (MR colonography) Appendicitis

If clinical findings are unclear and MRI does not delay surgery

Start VIBE (volume-interpolated breath-hold examination) sequence ca. 40 s after IV administration of spasmolytic and contrast medium and repeat sequence 2–4 times to have alternative images for interpretation in case of peristalsis artifacts. If peristalsis artifacts are encountered, reduce scan time by decreasing number of slices and increasing slice thickness.

110

5 The Gastrointestinal Tract

a b Fig. 5.1a–c Mechanical ileus due to residual mesenteric lymphoma (arrow). a T2w HASTE used as fast and robust screening sequence. b TrueFISP. c T1w 3D GRE (VIBE) after IV injection of Gd-based contrast medium. The site of obstruction is revealed indirectly by prestenotic dilatation and water-filled bowel on the screening se-

c quence (a). Adjunct TrueFISP sequence (b) acquired in the same plane confirms the obstruction. Postcontrast VIBE sequence (c) yields a high-resolution image of the gastrointestinal wall and differentiates contrast-enhancing viable tissue from scars and residual stool.

vantage of MRI over unenhanced CT is that evaluation of the intestinal wall is not degraded by residues of the cleansing solution.

Spasmolysis

Oral/Rectal Contrast Media Evaluation of the GI tract is possible only after adequate distention (see Fig. 5.2). Although water alone is sufficient for distention and delineation of the gastric or intestinal wall, 2.5 % mannitol needs to be added to prevent premature osmotic absorption of the water when the small bowel is examined. Alternatively, small bowel loops can be distended and delineated with PEG-ES, as used for bowel lavage before colonoscopy.6 When PEG-ES is used, the MRI scans must be timed such that enough of the solution is retained in the small bowel without inducing a desire to defecate. If the contrast solution is administered through a nasojejunal tube (MR enteroclysis), filling of the small bowel can be evaluated in a dynamic MR study.5 Contrast administration through a nasal tube is more efficient, but oral intake is also possible, since the tube cannot be positioned under MRI guidance and some patients refuse intubation. For MR colonography, a rectal water enema administered with a bag suspended 1 m above the patient is theoretically superior to air insufflation because water results in better delineation of the bowel wall on HASTE sequences than air (Fig. 5.1) and possibly also better distention (Fig. 5.2). However, as with CT colonography, air insufflation is more convenient and is preferred as long as there is no clear evidence of its disadvantages.

Unless butylscopolamine is contraindicated, intravenous spasmolysis with 20 mg of Buscopan is recommended to maximize bowel distention and minimize peristalsisrelated artifacts. The maximum effect of butylscopolamine occurs as early as 1 min after intravenous injection and persists for only a few minutes. Butylscopolamine should therefore be saved for bolus administration immediately before the scan for which effective suppression of peristalsis is most important. Imaging should then start 1 min after bolus injection and should be completed in less than 10 min. If butylscopolamine is not approved in a country or is contraindicated in a patient (prostatic hyperplasia, narrow angle glaucoma, tachyarrhythmia, myasthenia gravis), an unenhanced VIBE sequence can be acquired to estimate the extent of peristaltic artifacts. If reduction of persistaltic artifacts is deemed necessary, the patient should be given 1 mg of glucagon.

Pulse Sequences Abdomen As abdominal imaging should be performed with breathholding, there are basically three sequences that can be used: HASTE, TrueFISP, and 3D GRE (VIBE) (see Fig. 5.1). The HASTE sequence can also be obtained with the patient breathing freely and is therefore well suited for orientation and the initial search for pathology. A TrueFISP sequence allows evaluation of vessel patency and differentiation of vessels and lymph nodes. Moreover, it can serve

Imaging Technique

111

a

b Fig. 5.2a, b Pitfall. Coronal T2w HASTE images obtained with inadequate rectal air insufflation (a) and after adequate air insufflation (b). With poor colonic distention (a), a nondistended segment

mimics a tumor, which is shown to be a fold on the image obtained after adequate distention (b).

to identify infiltration of adjacent organs by exploiting chemical shift effects that occur at interfaces between soft tissue and fat. Presence of this effect indicates an intact fat plane between two adjacent organs. A VIBE sequence should be started ca. 1 min after butylscopolamine administration and 40 s after intravenous contrast injection (nonspecific Gd-based contrast medium at a dose of 0.1 mmol Gd per kg body weight; e. g., Magnevist or Dotarem). Recommended practice is to repeat the sequence immediately to have a second data set in case the first one is degraded by motion artifacts. As with CT, the 3D data set can be used for multiplanar reconstruction. In the abdomen, images should be acquired in the coronal plane for three reasons: · It is the most suitable for following the intestine. · Long, continuous segments of the bowel wall can be evaluated. · The regular pattern of opposing haustra in the coronal plane facilitates identification of polyps.

Imaging Strategy

Pelvis Artifacts due to respiratory motion are usually less severe in the pelvis, and breath-hold imaging is not necessary in most instances. The pelvis can therefore be imaged using a TSE or SE sequence with scan times longer than 25 s. STIR imaging is the method of choice for evaluating fistulas. With the improved coils available today, most diagnostic questions can be answered using a fast T2w sequence, which markedly reduces scan time and does not require contrast administration (see Fig. 5.1). The rectum and sigmoid colon are best evaluated on axial images.

Gastrointestinal imaging is performed with the patient in a comfortable supine position using a surface coil combined with the coil elements integrated in the table. As with CT, an overview of the abdomen and pelvis should be obtained in the axial plane, for which a HASTE sequence is used. While coronal imaging is preferred for further evaluation of the abdominal portion of the GI tract, the pelvic part is primarily imaged in the axial plane. Double angulation along the terminal ileum is helpful in patients with colitis involving this bowel segment. Data sets acquired with a VIBE sequence allow multiplanar reconstruction images to be retrospectively generated in any desired plane. If peristaltic artifacts are present and it is not possible to reduce the scan volume, the number of slices can be reduced by increasing the slice thickness (1.5–2 mm). Since there is not much motion in the pelvic region, spatial resolution can be improved by acquiring the VIBE sequence during shallow breathing. Higher resolution improves multiplanar reconstruction for image interpretation; the rectum can then be evaluated on sagittal reconstructions instead of performing primary sagittal imaging of this region. To maintain adequate SNR, the number of acquisitions should also be increased if the slice thickness is decreased (minimum of 1 mm) and the matrix increased. When a VIBE sequence is used, intravenous contrast administration is needed to evaluate the bowel wall (see Fig. 5.1). Timing of postcontrast imaging is not very critical as signal enhancement of colorectal masses persists for a few minutes. However, there should be a delay of at

112

5 The Gastrointestinal Tract

least 40 s after the start of contrast administration to ensure that the contrast medium has reached the target site when the central k-space lines are sampled. It is good practice to repeat the VIBE sequence as a dynamic series to have backup images if others are degraded by artifacts.

MRI Appearance of Pathologic Entities Stomach Adenocarcinoma. Adenocarcinoma is the most common malignant tumor of the stomach. Nearly half of all gastric carcinomas occur in the region of the gastroesophageal junction. Three growth patterns are distinguished: localized, circular, and diffusely infiltrating (linitis plastica, “leather bottle” stomach). Thickening of the gastric wall may be minimal to very pronounced (> 4 cm) and correlates with transmural tumor extension. A wall thickness > 2 cm is highly indicative of extragastric spread. Proximal cancer spreads along the gastrosplenic ligament to the left lobe of liver, the diaphragm, and the spleen, while distal cancer may invade the pancreas. Lymph node metastases occur in perigastric, celiac trunk, hepatoduodenal ligament, retropancreatic, mesenteric root, and para-aortic nodes. Advanced disease is characterized by tumor spread to the major omentum, peritoneum, and ovaries (KrukenTable 5.2 TNM staging system for gastric cancer Invasion level

Lymph node metastases

Distant metastases

T1

N1: 1–6 regional lymph nodes N2: 7–15 regional lymph nodes N3: > 15 regional lymph nodes

M1*

T2 T3 T4

Lamina propria or submucosa Muscularis propria or subserosa Serosa (visceral peritoneum) Adjacent structures

* E.g., liver (45–55 %), lung (6–30 %), peritoneum, ovaries.

Fig. 5.3 Gastric carcinoma. The tumor has high SI (arrow) on T2w TSE image.

berg tumor). TNM staging of gastric cancer is summarized in Table 5.2. Gastric Lymphoma. The stomach is the most common site of lymphoma in the GI tract (especially in non-Hodgkin lymphoma), with gastric lymphoma accounting for 3–5 % of malignant stomach tumors. While there may be isolated lymphoma of the stomach, concomitant manifestation in other parts of the GI tract is more common. Since lymphoma is primarily submucosal, it may be difficult to diagnose by endoscopy. Gastric lymphoma can directly infiltrate perigastric fat and adjacent organs, be associated with regional lymphadenopathy, and/or disseminate into the peritoneal cavity. Unlike adenocarcinoma, lymphoma often involves the entire stomach and may be confined to the wall, even if the wall is > 4 cm thick. Leiomyosarcoma. This rare tumor accounts for 1–3 % of malignant gastric neoplasms. The primary tumor is usually large and round to ellipsoid in shape. In contrast to other gastric tumors, wall thickening may be absent because the bulk of the tumor is often outside the stomach. Leiomyosarcomas metastasize by contiguous extension to adjacent organs, peritoneal seeding, or hematogenous spread (liver, lungs, bone). Metastasis to perigastric and para-aortic lymph nodes is uncommon. Gastric Metastasis. The most common primaries are malignant melanoma, breast cancer, and lung cancer. Gastric metastases from breast cancer occasionally manifest as diffuse wall thickening, which cannot be differentiated from primary adenocarcinoma. Squamous cell carcinoma of the distal esophagus involves the cardia in 15 % of cases. Cancer of the transverse colon and pancreas can directly invade the stomach wall along the gastrocolic and gastrosplenic ligament, respectively. Leiomyoma. This is the most common benign tumor of the stomach and arises in the submucosa. It is usually asymptomatic. Leiomyomas > 5 cm frequently undergo ulceration and hemorrhage. They can then be diagnosed on T1w images by the presence of high-signal-intensity blood (between 2 days and 4 weeks) or susceptibility artifacts produced by iron-containing blood products such as hemosiderin. Other conditions that can cause thickening of the gastric wall are Ménétrier disease (typically of the fundus and corpus) and Crohn disease (typically of the antrum). MRI The patient is asked to drink as much water as possible 5 min before the examination and is given an intravenous injection of butylscopolamine to ensure maximum distention of the stomach. Most gastric cancers are hyperintense relative to the stomach wall on T2w images (Fig. 5.3). However, they may also be isointense, which is why a contrast-enhanced study is recommended in patients with suspected gastric cancer (Fig. 5.4). Chemical

MRI Appearance of Pathologic Entities

113

shift imaging appears well suited for local cancer staging (Fig. 5.5). The destructive interaction between fat and water signals on opposed-phase (OP) images, which are obtained at specific TEs (2.1 ms or an odd-numbered multiple at 1.5 T), produces a dark line at the interface between the gastric wall and surrounding fat. Interruption of this line on OP images is a fairly reliable indicator of tumor extension to adjacent organs (Fig. 5.5).

Small Intestine Tumors of the small intestine are rare, accounting for ca. 6 % of gastrointestinal neoplasms. One-third are found in the duodenum, and most of the remaining two-thirds are located in the terminal ileum. The most common benign tumors of the small intestine are leiomyomas, adenomas, and lipomas. Malignant small-bowel tumors include adenocarcinoma, leiomyosarcoma, lymphoma, neurofibrosarcoma, carcinoids, gastrointestinal stromal tumors (GIST), and metastases (e. g., from the lungs, breasts, or malignant melanomas). Neoplasms of the small intestine may cause intussusception.

a

Leiomyoma and Leiosarcoma. These are highly vascular tumors, which often undergo central necrosis and commonly present with profuse bleeding. Since their growth is eccentric, they typically have an extraluminal component. Leiomyomas and leiosarcomas are often difficult to distinguish on the basis of their imaging appearance. Adenocarcinoma and Lymphoma. These tumors cause concentric thickening of the bowel wall. While adenocarcinomas have a propensity for the proximal small intestine (50 % occur in the duodenum) and are associated with clinical signs of obstruction, lymphomas (20 % of all smallbowel tumors) are most frequently found in the terminal ileum with concomitant para-aortic or mesenteric lymphomas. Mesenteric lymphomas are common in nonHodgkin disease (50 %) and rare in Hodgkin disease (4 %). Carcinoid is the most common primary tumor of the small bowel (ileocecal region) and appendix. Regional lymph node enlargement is associated with stellate retraction of the mesentery. Gastrointestinal Stromal Tumor. A rare entity, gastrointestinal stromal tumor (GIST) arises from the wall of the small intestine (20–35 %) or stomach (39–70 %). Rarely, the tumor originates in the large bowel (5–15 %), omentum or mesentery (9 %), or esophagus (< 5 %). GIST was defined as a new tumor entity in 1998 based on the expression of a CD117 (c-kit proto-oncogene) receptor on the cell surface.14 The mutation of this receptor is associated with constant activation of tyrosine kinase, which in turn results in uncontrolled cell proliferation and protection from apoptosis. The diagnosis of GIST has important therapeutic implications because the tumor can be effectively treated by selective inhibition of tyro-

b

c Fig. 5.4a–c Gastric cancer. a T2w HASTE. b STIR. c T1w GRE image after IV injection of Gd-based contrast medium. The tumor shows low SI on T2w and STIR images (a, b) and marked enhancement on the postcontrast image (c).

114

5 The Gastrointestinal Tract

a

b

c

d Fig. 5.5a–d Gastric cancer. a T2w HASTE. b T2w TSE. c T2w HASTE. d OP GRE. The tumor is of low SI on T2w images (arrows in a–c). On the OP image (d), loss of the fat plane indicates invasion of the pancreas (arrowhead).

sine kinase (imatinib, currently at a dose of 400 mg/day). GISTs are typically large, well-circumscribed, and heterogeneous lesions with central necrosis.15 Despite their size and metastatic spread to the liver, GISTs are not associated with obstruction (Fig. 5.6). GISTs have an increased glucose metabolism, which can be exploited to monitor therapy using 18F-FDG PET/CT. Given the excellent response rate, tumor size alone appears to be an adequate parameter to monitor the response to therapy. Crohn Disease. Crohn disease is the most important inflammatory condition of the small intestine. Although the disease can involve any part of the small bowel, it has a propensity for the terminal ileum. There is characteristic thickening of the bowel wall involving all layers and resulting in luminal narrowing with or without dilatation

proximal to the stricture (Fig. 5.7). Extraintestinal manifestations comprise fistula and abscess formation, phlegmons, lymph node enlargement, and excess mesenteric fat. Currently, the patient’s prognosis and choice of optimal therapy are primarily determined on the basis of clinical and laboratory findings, while MRI is used to corroborate indications for surgery (Table 5.3) and localize pathology. The role of MRI remains to be defined, either as a competing diagnostic tool or a supplementary procedure. Available data suggest that MRI findings appear to correlate with the Crohn Disease Activity Index (CDAI) (http://www.ibdjohn.com/cdai/). This index comprises the following items: number of liquid or very soft stools/ week, abdominal pain, general well-being, associated conditions (arthritis/arthralgia, iritis/uveitis, erythema nodosum, pyoderma gangrenosum, aphthous stomatitis,

MRI Appearance of Pathologic Entities

115

Fig. 5.6 Mesenteric GIST. The inherent bright blood effect of the TrueFISP sequence permits evaluation of vascular invasion. In the patient shown, the large tumor encases long segments of the mesenteric vessels. No thrombosis in the portal vein and confluence.

anal fissures, fistulas, abscess, fever), intake of medications for diarrhea, hematocrit, and body weight. Other Small-Intestinal Pathology. A much less common cause of postinflammatory wall thickening of the small bowel is radiation enteritis. Other small-bowel conditions potentially amenable to assessment by MRI include duodenal diverticula, which are typically found on the medial side of the duodenal C-loop or in the papillary region and may cause dilatation of the pancreatic and bile ducts. MRI The search for an inflammatory process should begin with a coronal fat-suppressed HASTE sequence (Fig. 5.8). This initial sequence can be supplemented by axial and coronal

HASTE imaging without fat suppression since inflammatory processes are less conspicuous on T2w images when a bright-lumen contrast medium such as water is used rather than a dark-lumen negative contrast medium, and since wall thickness is difficult to evaluate when the wall is not outlined by fat. Non-fat-suppressed sequences thus help identify wall thickening or at least the terminal ileum as the preferred site of Crohn disease. Coronal and axial contrast-enhanced VIBE sequences can then be acquired to continue the search or to assess the severity of inflammation (Fig. 5.8). With its better resolution and high contrast of the bowel wall after intravenous contrast administration, the T1w VIBE sequence is superior to T2w HASTE and TrueFISP sequences in both the colon and the small intestine (see Fig. 5.1) as long as artifacts due to bowel

116

5 The Gastrointestinal Tract

Table 5.3 Surgical indications for complications of Crohn disease Complication Fistula

Surgical indication Interenteric Enterocutaneous Blind-ending Enterovesical Enterovaginal

Abscess Stricture

Dysplasia Fulminant episode

In patients with functional short bowel syndrome (bypass); in all other cases only in the setting of surgery performed for other reasons For “high” enterocutaneous fistulas with skin irritation; limited indication for distal enterocutaneous fistulas Absolute surgical indication because a blind-ending fistula may lead to severe, insidious sepsis, which is typically not controllable Absolute surgical indication in all patients with proven or strongly suspected enterovesical fistula (e. g., recurrent urinary tract infection) In patients with severe symptoms (daily secretions, recurrent vaginal infection, recurrent urinary tract infection) Interventional drainage of all intra-abdominal abscesses; surgical drainage if not amenable to intervention or very superficial Stricture causing postprandial pain; subileus not responding to conservative management (including endoscopic dilatation); and colonic stricture including asymptomatic stricture of unclear etiology Proven diagnosis of dysplasia and repeat work-up required to rule out ulcerative colitis Bleeding cannot be stopped by conservative measures and > 2 units of erythrocyte concentrate are required per day; patient does not respond to intensified immunosuppressive therapy

a

b

c

d Fig. 5.7a–f Acute episode of Crohn disease. a CT after IV contrast administration. b T2w HASTE. c–f T1w GRE images obtained before and at various time points after IV injection of Gd-based contrast medium. Characteristic findings are present: free fluid (b), homogeneous (edematous) thickening of long wall segments with luminal

narrowing, prestenotic dilatation, distancing of bowel loops, and hyperemia (strong enhancement after IV contrast) of the affected bowel wall (see Table 5.4). Fig. 5.7e, f e

MRI Appearance of Pathologic Entities

117

e

f Fig. 5.7e, f

motion can be effectively suppressed by administration of an antispasmodic.2 The presence of edema and contrast enhancement can help differentiate inflammatory stricture from stricture due to scar formation, which is important because the latter may require surgery. The length of involvement, mural thickening, and postcontrast enhancement can be quantified and serve as objective measures of inflammatory activity16,17 (Table 5.4). However, the optimal therapeutic regimen is currently prescribed on the basis of clinical and laboratory findings. A STIR sequence is the sequence of choice for the pelvis, especially for evaluating fistulas (Fig. 5.8).

Table 5.4 MRI criteria for assessing severity in inflammatory bowel disease5

Fig. 5.8a–j Crohn disease with enterocutaneous fistula. a, b Percutaneous (retrograde) fistulography combined with Sellink. c–j On MRI the fistula opening is marked with a nifedipine capsule. Search for inflammatory changes with a fat-saturated HASTE sequence (c, d). Affected bowel segments are then further evaluated using higher-resolution breath-hold TrueFISP sequence (2-mm slice thick-

ness; e, f) or STIR sequence (4-mm slice thickness; g, h). 3D GRE images after IV injection of Gd-based contrast medium (1.2-mm slice thickness; i, j) show concomitant inflammation of fatty tissue and edematous wall thickening (target sign). Fig. 5.8c–j e

MRI criterion

Inflammatory activity Mild

Wall thickening 4–5 mm Length of diseased < 5 cm segment Contrast enhancement < 50 %

Moderate

Severe

5–10 mm –

> 10 mm –

≤ 100 %

> 100 %

a

b

118

5 The Gastrointestinal Tract

c

d

e

f

g

h

i

j Fig. 5.8c–j

MRI Appearance of Pathologic Entities

119

a

b Fig. 5.9a–c Small-bowel lymphoma. a Axial T1w GRE image. b Axial T2w fat-suppressed TSE image. c Sagittal T2w TSE image. There is marked homogeneous wall thickening, which is characteristic of small-bowel lymphoma.

c

Lipoma can be diagnosed with a high degree of accuracy using MRI. It is suggested by high signal intensity on T1w images with a typical signal drop on fat-suppressed images. Most other small-bowel tumors are not depicted by T1w sequences unless intravenous contrast medium is given. Lymphomas show only moderate enhancement and are primarily identified as wall thickening (Fig. 5.9) in association with enlarged mesenteric lymph nodes. No data are available in the literature on the T2 signal intensities of small-bowel tumors except for lymphomas, which are isointense to fat and hyperintense to muscle on T2w images.18

Colon Carcinoma Colorectal cancer is the second most common malignant tumor in men and women.19 About 70–95 % of colorectal tumors arise from adenomas.20 The typical growth pattern is circumferential (“apple core” appearance), ultimately resulting in bowel obstruction. About 50 % of all colorectal carcinomas are found in the rectum, 20 % in the sigmoid colon, 6 % in the descending colon, 8 % in the transverse colon, 6 % in the ascending colon, and 10 % in the cecum (Fig. 5.10). Multiple colorectal carcinomas are seen in 2–5 % of patients. The TNM staging system for colorectal cancer is summarized in Table 5.5. Colorectal tumors spread locally to perirectal or pericolic fat, via the lymphatics, and through the bloodstream if there is vascular invasion. Primary spread is lymphatic in most cases. Resection of draining lymph nodes along the supplying blood vessels is crucial

120

5 The Gastrointestinal Tract

a

b Fig. 5.10a–c Malignant polyp (stage 1 cancer) in the proximal transverse colon. MR colonography with water enema. a T2w HASTE. b TrueFISP. c T1w 3 D GRE (VIBE) after IV injection of Gd-based contrast medium. This example clearly illustrates that the HASTE sequence is less susceptible to artifacts than the TrueFISP sequence. On the postcontrast VIBE image (c), enhancement of the suspected lesion indicates vascularization and permits differentiation from residual stool.

c Table 5.5 TNM staging system for colorectal cancer Invasion level

Lymph node metastases

Distant metastases

T1

Submucosa

M1*

T2

Muscularis propria

N1: 1–3 regional lymph nodes N2: ≥ 4 regional lymph nodes

T3

Subserosa, nonperitonealized pericolic/perirectal tissues Adjacent structures/visceral peritoneum

T4

* Colon cancer: liver (69–80 %), lung (12–37 %), peritoneum (17–32 %), bones, adrenals, brain. Rectal cancer: liver (59–66 %), lung (19–47 %), peritoneum (12 %), bones, adrenals, brain.

for preventing local recurrence and secondary distant metastases. The lymphatic drainage of the colon follows its arterial supply along the ileocolic artery, the right colic artery, the two branches of the middle colic artery, and the short trunk of the inferior mesenteric artery with the left colic artery and the sigmoid arteries (Fig. 5.10). In contradistinction to other parts of the colon, the rectum has relatively sparse lymphatic drainage, which is why lymphatic metastatic spread does not occur until a rectal tumor has invaded the muscularis mucosae and submucosa. These tumors may then spread through the lymphatics following the superior rectal artery to the inferior mesenteric artery and the inferior rectal arteries and internal iliac vessels to the pelvic sidewall. Spread to distal and inguinal lymph nodes, as in anal cancer, occurs only in case of very low rectal tumors extending to the level of the levator ani muscle or if proximal lymphatic

MRI Appearance of Pathologic Entities

drainage is blocked. Metastatic distal lymph nodes are usually not more than 2–3 cm away from the tumor margin. If gelatinous peritoneal carcinosis is present, the differential diagnosis includes mucinous adenocarcinoma of the colon and ovarian cancer but also the rare pseudomyxoma peritonei (PMP, “jelly belly”), which typically arises as a mucinous tumor of the appendix that ruptures into the peritoneum.

Distant Metastatic Disease Since the venous drainage of the colon and upper rectum empties into the hepatic portal system, the liver is the first and most common site of organ metastases from colorectal cancer (Table 5.5). Second in frequency are pulmonary metastases. The lungs may also be the primary site of metastatic spread, especially in low rectal carcinoma with drainage via the pelvic and paravertebral veins. Less common are metastases to the bones, adrenals, and brain. In advanced disease, there will be peritoneal seeding, possibly with ovarian metastases. The likelihood of synchronous and metachronous metastases is determined by different features of the primary tumor. The incidence increases not only with the T and N stages but also with decreasing tumor differentiation. Other contributory but less well-established factors are ploidy status, growth fraction, and loss of adhesion molecules. Colorectal cancer typically recurs locoregionally. Recurrence at the anastomotic site is especially common after low anterior resection of tumors of the proximal rectum or distal sigmoid. Locally recurrent tumors may involve the abdominal wall, pancreas, ureters, and pelvic bones. Presacral recurrence of rectal carcinoma typically involves the piriform muscle and may also invade the sciatic nerve. Colorectal cancer recurs in up to 50 % of patients, depending on the tumor stage at diagnosis.

MRI in Colon Cancer Since surgical resection is the method of choice regardless of the T stage, there is no need for preoperative T staging of colon cancer by imaging. CT is more accurate than MRI in detecting enlarged lymph nodes due to its higher spatial resolution, which improves the quality of multiplanar reconstruction. Screening for distant metastases is also more practical with CT because it can cover the lungs and abdomen with one 10-s scan and is superior to MRI in detecting pulmonary metastases. PET/CT is superior to MRI in searching for a primary tumor or ruling out a tumor. CT colonography and endoscopy appear to be equally suitable for detecting clinically relevant colorectal adenomas.21 Because of the radiation exposure involved, CT is limited as a screening test, and MR colonography may be considered as an alternative for this indication.11,12 The option of combining MR colonog-

121

raphy with imaging of the breast or prostate might be exploited in the screening setting. However, MR colonography is a relatively new technique and is limited by the fact that it is still difficult to visualize the entire colon with high resolution and no artifacts. MR Colonography (MRC) After bowel preparation, water or air is introduced rectally according to the individual patient’s tolerance. Adequacy of colonic distention is monitored by acquisition of a coronal HASTE sequence with thick slices and corresponding interslice gaps. An axial HASTE sequence is acquired to determine the anteroposterior extent of the colon. The coronal and axial HASTE images serve to prescribe the scan volume for the subsequent HASTE and VIBE sequences. The FOV must be large enough to encompass the entire colon and planning must therefore take into account that water filling in conjunction with a spasmolytic may further expand the colon. All subsequent sequences are acquired during breath-holds. The breathhold intervals, slice thickness, and number of slices are tailored to the individual colon size. Next, an intravenous bolus of 20 mg of butylscopolamine is given, followed by adjustment of the hydrostatic pressure to the patient’s tolerance level by varying the height at which the enema bag is hung from the IV stand (50-cm water column). A coronal HASTE sequence and three successive postcontrast VIBE sequences are acquired 40 s after a bolus of Gdbased contrast medium (0.2 mmol per kg body weight) (Fig. 5.10). Since the maximum effect of intravenous butylscopolamine is reached ca. 1 min after administration, the contrast medium for the subsequent VIBE sequence can already be given during acquisition of the HASTE sequence. TrueFISP images can be acquired before the VIBE sequence, but their diagnostic benefit remains to be determined (Figs. 5.10 and 5.11). The T2 signal intensity of colorectal carcinomas is very low. It is highest on STIR images (Fig. 5.11), but abdominal STIR sequences are very susceptible to motion artifacts, especially when accentuated by water as positive T2 contrast medium. If the colon cannot be imaged in its entirety in a single acquisition due to coil or magnetic field limitations, two scans can be obtained, one to image the abdomen in the coronal plane and one to image the pelvis in the axial plane (Fig. 5.12).

Inflammatory Disease Diverticulitis secondary to diverticulosis, ulcerative colitis, and Crohn disease are the most common inflammatory conditions affecting the colon. Much less common are radiation-induced changes of the colon and rectum. While Crohn disease is characterized by transmural involvement, ulcerative colitis predominantly involves the mucosa but may nevertheless cause marked wall thickening in severe inflammation. Ulcerative colitis typically begins in the rectum and continuously progresses to involve

122

5 The Gastrointestinal Tract Fig. 5.11a–d Carcinoma of the ascending colon. MR colonography after air insufflation. a T2w HASTE. b T2w TrueFISP. c STIR. d T1w 3D GRE (VIBE) after IV injection of Gd-based contrast medium. The colon tumor (arrow in a) has rather low SI on HASTE, TrueFISP, and STIR images. The contrast-enhanced VIBE image (d) allows good tumor delineation 40–50 s after contrast administration.

a

b

c

d

b

a

c Fig. 5.12a–c Sigmoid carcinoma. 3D GRE images (VIBE) obtained after IV injection of Gd-based contrast medium in coronal (a) and axial (b, c) planes. There is pronounced signal enhancement of the tumor (arrows). Enhancement persists for at least 5 min and enables

serial acquisitions for evaluation of the tumor in different planes. Note prestenotic dilatation in an otherwise fully distended colon as an indirect sign of cancer.

MRI Appearance of Pathologic Entities

123

a

b Fig. 5.13a, b Acute ulcerative colitis in the sigmoid colon. T1w SE images before (a) and after (b) IV injection of Gd-based contrast medium. Contiguous extension of the disease process from the

rectum is typical of ulcerative colitis. Wall thickening, the length of the involved colon segment, and signal enhancement indicate the severity of inflammation. No prestenotic dilatation.

Fig. 5.14a, b Radiation enteritis in the sigmoid colon. T1w GRE images before (a) and after (b) IV injection of Gd-based contrast medium. The inflammatory process is revealed only after contrast

administration and is confined to the mucosa and submucosa (distinguishing it from Crohn disease). No prestenotic dilatation.

more proximal colon segments. In Crohn disease, on the other hand, the inflammation is discontinuous and can produce skip lesions throughout the bowel. Fistulas, abscesses, and locoregional lymph node enlargement characterize Crohn disease, but they are rare in ulcerative colitis, though sigmoidovesical fistulas do occur. Diverticulosis, which is found in the sigmoid colon in two thirds of cases, will also cause wall thickening as the disease evolves. Diverticulitis may be complicated by strictures, peridiverticular infiltration of fatty tissue, covered perforations, abscesses, and fistula formation. Acute appendicitis occurs in ca. 6 % of the population. Uncomplicated appendicitis is associated with wall thickening, inflammatory changes in surrounding fat (fluid collection around the appendix; 58–88 %), and fecaliths (appendicoliths; in up to 23 % of cases). The reported incidence of perforated acute appendicitis is 25 %. Perforated appendicitis may be associated with phlegmons or perityphlitic abscesses. Appendicoliths are much more common than in uncomplicated appendicitis.

MRI in Inflammatory Disease MR colonography (see above) can also be used for assessing inflammatory disease of the colon. In ulcerative colitis, imaging will show loss of haustra. The colon wall (mucosa and submucosa) has increased signal intensity on T1w and T2w images.22 In inactive ulcerative colitis, the disease is more likely to be recognized by thickening of the colon wall. MRI is not superior to endoscopy in differentiating Crohn disease and ulcerative colitis.23 Transmural enhancement and skip lesions that involve the terminal ileum but spare the rectum are characteristic findings in Crohn disease, whereas a disease process extending continuously from the rectum suggests ulcerative colitis (Fig. 5.13). Whereas Crohn disease is characterized by enhancement of all wall layers on postcontrast GRE images, the serosa usually does not enhance in chronic ulcerative colitis or unspecific enteritis (Fig. 5.14). Submucosal thickening due to edema and lymphangiectasia is seen as rather low T1 signal intensity and occurs in chronic ulcerative colitis already after one year.

a

b

124

5 The Gastrointestinal Tract

a

b Fig. 5.15a, b Sigmoid diverticulitis. a T2w TSE. b 3D GRE image (VIBE) obtained after IV injection of Gd-based contrast medium. MRI provides information on the extent of diverticulitis and, in conjunc-

tion with the clinical findings, thus helps identify those patients who require surgical management. But MRI cannot exclude perforation, for which CT is the imaging modality of choice.

a

b Fig. 5.16a, b Sigmoidovesicular fistula. a Fat-suppressed T2w HASTE. b 3D GRE image (VIBE) after 1 % Gd and water enema. The fistula is revealed as a hyperintense structure (arrow in b) after rectal contrast medium has entered the tract.

Complications of Crohn disease are fistulas, abscesses, strictures, and conglomerate tumors (Table 5.3), whereas ulcerative colitis may be complicated by bleeding and toxic megacolon. Patients with ulcerative colitis have a higher risk of developing colorectal cancer than those with Crohn disease, in whom cancer is a very rare late complication. The risk of carcinoma development correlates with the extent of colonic involvement and disease duration (40 % in patients with involvement of the entire colon and a 25-year history of disease), which must be borne in mind in the follow-up of patients with inflammatory bowel disease. CT is preferred to MRI in suspected diverticulitis because it will demonstrate free air as a sign of perforation. Moreover, with its higher spatial resolution, CT appears to

be more sensitive in identifying inflammation of surrounding fat (peridiverticulitis). However, MRI can equally well be used to diagnose diverticulitis (Fig. 5.15) and spares the patient an intravenous contrast injection, which may be necessary with CT. Fat-suppressed T2w HASTE images allow evaluation of the inflammatory process and exclusion of abscess. A VIBE sequence acquired after a 2 % Gd and water enema will identify fistulas between the sigmoid and the urinary bladder either directly (Fig. 5.16) or indirectly by the presence of contrast medium in the bladder. However, contrast medium in the bladder can serve as indirect evidence only in patients who have not received intravenous Gd-based contrast medium.

Outlook

125

Intense enhancement of the appendix wall on fat-suppressed T1w images after intravenous contrast administration was found to be a sensitive and specific sign of appendicitis,24 as was marked hyperintensity of the lumen, a slightly hyperintense, thickened wall (Fig. 5.17), and marked hyperintensity of surrounding tissue on T2w TSE images.25 These are preliminary results, but they suggest that MRI may also have the potential to diagnose appendicitis.

Functional MRI MRI is very well suited for functional studies because it does not involve radiation exposure and thus allows repeated sequential imaging. Hence, MRI can be used to quantify esophageal reflux26 and gastric motility and emptying.27 Causes of defecation disorders such as rectocele or rectal invagination (enteroceles with or without prolapse of the urinary bladder, vagina, uterus, or colon) can be identified and evaluated by means of dynamic MR defecography.28 Dynamic MRI requires rectal filling. If water is used, in the form of either ultrasound gel or a cellulose mixture, imaging can be performed with a TrueFISP or HASTE sequence. If volumetric analysis is necessary, imaging can be performed using a 3D GRE sequence after a 1 % Gd and water enema.

Outlook In the gastrointestinal tract, MRI competes not only with other cross-sectional imaging modalities (CT, ultrasound) and conventional projection radiography (double-contrast barium enema, Sellink) but also with endoscopy and PET/CT imaging. Unlike colonoscopy, both CT and MRI yield digital volume data, which are a prerequisite for using computer-aided diagnosis (e. g., CT-density- and/ or shape-based algorithms in colorectal cancer screening29). At present, CT is preferred to MRI for gastrointestinal imaging because it is fast, combines high spatial resolution with a minimum of artifacts, and allows the combination of lung and abdominal imaging for staging in less than 20 s. MRI, on the other hand, theoretically offers some advantages such as high soft-tissue contrast, a higher sensitivity to contrast media, the absence of radiation exposure, the possibility of dispensing with contrast administration, and the option of performing kinematic imaging for functional assessment. PET/CT is increasingly being recommended as the baseline staging modality for oncologic imaging (except for hepatocellular and pancreatic carcinoma) and can be considered the standard of reference for GIST and possibly colorectal cancer, although it is supposed to be inferior to MRI in the detection of small liver metastases due to poorer contrast and additional blurring by partial volume effects caused by liver

Fig. 5.17 Appendicitis. Coronal T2w TrueFISP image acquired in an open 0.2-T scanner during breath-holding (25 s). Appendicits (arrow) is suggested by wall thickening and high SI. At 1.5 T the HASTEIRM sequence should be used to quickly identify the appendix by periappendiceal inflammatory stranding and free fluid. Once the appendix has been found the examination can be supplemented by an axial 3D contrast-enhanced VIBE sequence if the results are still inconclusive. This VIBE sequence allows optimal angulation in the multiplanar reconstruction mode.

movement. Ga-68-DOTATOC-PET/CT using the somatostatin analogue DOTATOC as tracer to replace somatostatin receptor scintigraphy (octreotid-scan) is emerging as the standard of reference for the detection of gastroenteropancreatic neuroendocrine tumors (GEP-NETs).

References 1. Umschaden HW, Szolar D, Gasser J, Umschaden M, Haselbach H. Small-bowel disease: comparison of MR enteroclysis images with conventional enteroclysis and surgical findings. Radiology 2000;215(3):717–725 2. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, Maris T, Prassopoulos P. MR enteroclysis protocol optimization: comparison between 3 D FLASH with fat saturation after intravenous gadolinium injection and true FISP sequences. Eur Radiol 2001;11(6):908–913 3. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, Prassopoulos P. MR enteroclysis: technical considerations and clinical applications. Eur Radiol 2002;12(11):2651–2658 4. Low RN, Sebrechts CP, Politoske DA, et al. Crohn disease with endoscopic correlation: single-shot fast spin-echo and gadolinium-enhanced fat-suppressed spoiled gradient-echo MR imaging. Radiology 2002;222(3):652–660 5. Rohr A, Rohr D, Kühbacher T, Schreiber S, Heller M, Reuter M. [Radiological assessment of small bowel obstructions: Value of conventional enteroclysis and dynamic MR-enteroclysis]. Rofo 2002;174(9):1158–1164 6. Laghi A, Borrelli O, Paolantonio P, et al. Contrast enhanced magnetic resonance imaging of the terminal ileum in children with Crohn’s disease. Gut 2003;52(3):393–397

126

5 The Gastrointestinal Tract

7. Luboldt W, Bauerfeind P, Steiner P, Fried M, Krestin GP, Debatin JF. Preliminary assessment of three-dimensional magnetic resonance imaging for various colonic disorders. Lancet 1997; 349(9061):1288–1291 8. Luboldt W, Bauerfeind P, Wildermuth S, Marincek B, Fried M, Debatin JF. Colonic masses: detection with MR colonography. Radiology 2000;216(2):383–388 9. Luboldt W, Luz O, Vonthein R, et al. Three-dimensional doublecontrast MR colonography: a display method simulating double-contrast barium enema. AJR Am J Roentgenol 2001; 176(4):930–932 10. Luboldt W, Fletcher JG, Vogl TJ. Colonography: current status, research directions and challenges. Update 2002. Eur Radiol 2002;12(3):502–524 www.screening.info 11. Luboldt W, Hoepffner N, Holzer K, et al. [Early detection of colorectal tumors: CT or MRI?]. Radiologe 2003;43(2):136–150 www.screening.info 12. Lauenstein TC, Goehde SC, Ruehm SG, Holtmann G, Debatin JF. MR colonography with barium-based fecal tagging: initial clinical experience. Radiology 2002;223(1):248–254 13. Ell C, Fischbach W, Keller R, et al; Hintertux Study Group. A randomized, blinded, prospective trial to compare the safety and efficacy of three bowel-cleansing solutions for colonoscopy (HSG-01*). Endoscopy 2003;35(4):300–304 14. Hohenberger P, Reichardt P, Stroszczynski C, Schneider U, Hossfeld DK. Gastrointestinale Stromatumoren–Tumorentität und Therapie mit Imatinib. Dtsch Arztebl 2003;100:A1612–A1618 15. Burkill GJ, Badran M, Al-Muderis O, et al. Malignant gastrointestinal stromal tumor: distribution, imaging features, and pattern of metastatic spread. Radiology 2003;226(2):527–532 16. Koh DM, Miao Y, Chinn RJ, et al. MR imaging evaluation of the activity of Crohn’s disease. AJR Am J Roentgenol 2001;177(6): 1325–1332 17. Pauls S, Kratzer W, Rieber A, et al. [Quantifying the inflammatory activity in Crohn’s disease using CE dynamic MRI]. Rofo 2003;175(8):1093–1099 18. Negendank WG, al-Katib AM, Karanes C, Smith MR. Lymphomas: MR imaging contrast characteristics with clinical-pathologic correlations. Radiology 1990;177(1):209–216

19. Surveillance: Epidemiology and End Results (SEER). http://seer. cancer.gov/csr/1973_1999/sections.html 20. Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations during colorectal-tumor development. N Engl J Med 1988; 319(9):525–532 21. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003;349(23):2191–2200 22. Giovagnoni A, Misericordia M, Terilli F, Brunelli E, Contucci S, Bearzi I. MR imaging of ulcerative colitis. Abdom Imaging 1993;18(4):371–375 23. Shoenut JP, Semelka RC, Magro CM, Silverman R, Yaffe CS, Micflikier AB. Comparison of magnetic resonance imaging and endoscopy in distinguishing the type and severity of inflammatory bowel disease. J Clin Gastroenterol 1994;19(1):31–35 24. Incesu L, Coskun A, Selcuk MB, Akan H, Sozubir S, Bernay F. Acute appendicitis: MR imaging and sonographic correlation. AJR Am J Roentgenol 1997;168(3):669–674 25. Hörmann M, Paya K, Eibenberger K, et al. MR imaging in children with nonperforated acute appendicitis: value of unenhanced MR imaging in sonographically selected cases. AJR Am J Roentgenol 1998;171(2):467–470 26. Knippig C, Fass R, Malfertheiner P. Tests for the evaluation of functional gastrointestinal disorders. Dig Dis 2001;19(3): 232–239 27. Schwizer W, Fox M, Steingötter A. Non-invasive investigation of gastrointestinal functions with magnetic resonance imaging: towards an “ideal” investigation of gastrointestinal function. Gut 2003;52(Suppl 4):iv34–iv39 28. Bertschinger KM, Hetzer FH, Roos JE, Treiber K, Marincek B, Hilfiker PR. Dynamic MR imaging of the pelvic floor performed with patient sitting in an open-magnet unit versus with patient supine in a closed-magnet unit. Radiology 2002;223(2): 501–508 29. Luboldt W, Tryon C, Kroll M, et al. Automated mass detection in contrast-enhanced CT colonography: an approach based on contrast and volume. Eur Radiol 2005;15(2):247–253 www. screening.info

127

6

The Rectum and Anal Canal C. Klessen and M. Laniado

Introduction

Imaging Technique

With the advent of powerful gradients and high-resolution surface and endorectal coil systems, MRI is increasingly used to evaluate inflammatory and neoplastic diseases of the rectum and anal canal. MRI is superior to endoscopic and endosonographic techniques in that it enables noninvasive evaluation of the rectal lumen and wall and additionally provides detailed information on surrounding structures in the true pelvis. Accurate information on the extent of a disease process and possible involvement of adjacent structures is crucial for tailoring surgical or medical treatment to the individual patient’s needs. The superior tissue contrast provided by MRI is an advantage over CT, but ongoing technical developments may redefine the role of CT in assessing the local extent of rectal tumors. Evaluation for nodal and distant metastases in patients with colorectal cancer is still the domain of CT.

Patient Preparation and Positioning

Indications Currently, MRI has the following indications in the diagnostic assessment of the rectum, anal canal, and pelvic floor: · Determination of the activity and extent of acute and chronic inflammatory bowel disease in the rectum and anal canal and involvement of surrounding organs. · Evaluation of the extent and course of perianal and perirectal fistulas and abscesses, either in isolation or as a complication of chronic inflammatory bowel disease, for devising the most suitable therapeutic strategy. · Tumor staging in patients with anorectal malignancy, especially if an advanced tumor stage (T3, T4) is suspected on clinical grounds or endosonography is inconclusive or not tolerated by the patient due to pain. · Detection of recurrent tumor after surgical, conservative, or combined treatment, especially in patients with suspected extraluminal or presacral tumor extension. · Morphologic and functional assessment of the pelvic floor in patients with fecal incontinence or other defecation disorders.

Before the examination, a careful explanation of the procedure is given, and patients are informed about the risks of administration of an antispasmodic agent and asked about possible contraindications. A venous cannula for contrast injection is placed before image acquisition in those cases where additional contrast-enhanced pulse sequences need to be acquired. Unless contraindicated, an antispasmodic agent (butylscopolamine or glucagon) should be given to reduce artifacts caused by bowel motion. Antispasmodics have very short half-lives and an intravenous injection should therefore be given immediately before the start of image acquisition. The patient should be in a comfortable supine position, with the knees supported by a foam pad if necessary. Before the examination, the patient is instructed to lie still during acquisitions.

Contrast Media Most MRI examinations of the rectum can be performed without giving an enteral contrast medium. Some authors advocate the use of a positive or negative enteral contrast agent, but there appears to be general agreement in the current literature that this is not necessary. No intraluminal contrast is needed for MRI in patients with inflammatory rectal disease. Some clinical indications require intravenous injection of an extracellular, Gd-based contrast medium such as Magnevist (Gd-DTPA) or Dotarem (Gd-DOTA). The standard dose is 0.1 mmol Gd per kg body weight, and the contrast medium is automatically injected during the examination.

Coils Ideally, phased-array surface coils should be used because they provide a much better signal-to-noise ratio (SNR).1 The coils are positioned according to the target region and

128

6 The Rectum and Anal Canal

kept in place with belts. When first introduced, endorectal coils appeared promising, but they have failed to establish themselves for imaging of the rectum and anal canal because they are cumbersome to handle, expensive (disposable systems), and have a limited field of view (FOV).

Imaging Planes Following the usual three-plane localizer, sequences in the sagittal and axial planes are best suited for imaging the rectum and anal canal and evaluating their relationship to surrounding structures in the true pelvis. When MRI is performed for suspected fistula, additional coronal images should be obtained, and sagittal images may not be needed (depending on the site of the fistula). In patients with rectal cancer, imaging in an oblique axial plane perpendicular to the lumen of the affected bowel segment is necessary to evaluate the wall and identify invasion into perirectal fat. Additional coronal images will improve identification of possible involvement of the anal sphincter in patients with low rectal cancer.

Imaging Protocol After a three-plane localizer, fast breath-hold sequences in at least two planes (e. g., T2w HASTE) should be obtained to define the exact area involved and plan the subsequent sequences. These sequences have extremely short acquisition times and are fairly insensitive to motion artifacts. In patients with inflammatory bowel disease, an inversion recovery sequence with a short inversion time (e. g., TIRM, STIR) is acquired next for detection of fluid in abscess cavities and fistula tracts, making such a sequence especially useful when looking for perianal and perirectal fistulas. Unenhanced images for assessment of local tumor extent in patients with rectal cancer are obtained using high-resolution T2w TSE sequences, which provide a better contrast-to-noise ratio (CNR) and markedly higher spatial resolution than T2w HASTE sequences, but acquisition times are much longer and they are more susceptible to motion artifacts. It is especially with these sequences that imaging quality can be markedly improved by placement of saturation bands over the anterior subcutaneous fat and administration of an antispasmodic agent (see above). In patients with rectal inflammatory disease, intravenous contrast medium will improve evaluation of disease activity and delineation of anorectal fistula tracts and abscesses. Though we present some examples of contrast-enhanced images, investigators seem to agree that no intravenous contrast medium is needed for initial imaging of rectal cancer.2 Intravenous contrast may be helpful, though, in patients with suspected tumor recurrence.

Contrast-enhanced MRI is performed using T1w SE or TSE sequences with spectral fat saturation. The pelvic lymph nodes are best imaged with axial T1w or PD sequences, covering the entire area from the aortic bifurcation to the pelvic floor (see Chapter 16). Parallel imaging techniques (SENSE, iPAT) can be used to shorten acquisition time,3 but this is accomplished at the expense of SNR. Several recent studies suggest that certain diffusion and perfusion parameters may predict the response to radiochemotherapy, which is why the use of diffusionand perfusion-weighted imaging (DWI and PWI sequences) is under clinical investigation.4–6 Details of the recommended pulse sequences and imaging parameters are summarized in Tables 6.1 and 6.2.

MRI Appearance of Normal Anatomy The rectum is ca. 12–15 cm in length and extends from the level of the third sacral vertebra, where the mesentery of the sigmoid colon ends, to the anus (Fig. 6.1). Unlike the colon, the rectum has no teniae, haustra, or omental appendices. When viewed laterally, the rectum has two bends—the upper sacral flexure with posterior convexity and the lower anorectal flexure with anterior convexity. The rectum can be divided into the pelvic rectum or rectal ampulla, a contractile reservoir, and the perineal rectum or anal canal. The latter is ca. 3 cm long and extends from the levator ani muscle to the anus. Surgeons typically refer to the anal canal as the portion of the rectum below the pectinate or dentate (anorectal) line, which marks the transition from the columnar epithelium of the rectal mucosa to the squamous epithelium of the anoderm. Fistulas in this area are therefore designated as perianal fistulas in the radiologist’s report. A second line, the anocutaneous line, demarcates the transition from the hairless skin of the anal canal to the perianal skin. The normal wall layers of the rectum are depicted diagrammatically in Fig. 6.2a. The anterior and lateral surfaces of the upper two thirds of the ampulla are covered by peritoneum. The part of the rectum above the middle transverse rectal fold (Kohlrausch fold) is retroperitoneal, while the portion below is extraperitoneal. The rectal fascia constitutes a cylindrical sheath surrounding the rectum down to the pelvic floor and is referred to as the mesorectal fascia in the clinical and surgical literature7 (Fig. 6.3). Posteriorly, the mesorectal fascia is thickened and overlies the parietal pelvic fascia, which covers the sacrum. So-called total mesorectal excision (TME) of rectal cancer is the en bloc resection of the tumor-bearing rectum within its enveloping fascia including lymphatics, lymph nodes, and mesorectal fat, while preserving the parietal pelvic fascia and pelvic splanchnic nerves (nervi erigentes). Closure of the anal canal is ensured by the external and internal anal sphincters and levator ani muscle in the wall

MRI Appearance of Normal Anatomy

129

Table 6.1 Recommended pulse sequences and imaging parameters for MRI of rectal carcinoma Weighting

Plane

Sequence type

TR (ms)

TE (ms)

Flip (°)

ETL

FS

Matrix

No. of acquisitions

Slice thickness (mm)

Breathhold

PD T2

Axial Axial

1980 800*

10 63

150 150

3 115

–

219 × 512 115 × 256

3 1

5 5

– Yes

T2

Sagittal

800*

63

150

115

–

115 × 256

1

5

Yes

T2** T2** T1*** T1***

Axial Sagittal Axial Sagittal

TSE (FSE) Single-shot TSE with half-Fourier acquisition, e. g., HASTE Single-shot TSE with half-Fourier acquisition, e. g., HASTE TSE (FSE) TSE (FSE) TSE (FSE) TSE (FSE)

3550 3760 575 575

68 68 11 11

180 180 150 150

19 19 3 3

– – + +

179 × 256 179 × 256 230 × 256 230 × 256

5 5 3 3

3 3 3 3

– – – –

Note: Use of a surface phased-array multicoil is recommended. The imaging parameters are only examples and have to be adjusted for use on different brands of scanners. * TR: used here as a technical parameter referring to the intervals between slice acquisitions; physical TR = infinite since only one slice is acquired per excitation. ** FOV: 180 × 180, centered on the tumor. *** Fat-suppressed T1w sequences after IV contrast administration. Rarely necessary for this indication (see text). Table 6.2 Recommended pulse sequences and imaging parameters for MRI of anorectal fistulas and abscesses Weighting

Plane

Sequence type

TR (ms)

TE (ms)

Flip (°)

ETL

FS

Matrix

No. of acquisitions

Slice thickness (mm)

Breathhold

PD IR

Axial Axial

2360 7770

8.5 30

150 150

5 7

– IR

192 × 512 192 × 512

2 1

5 5

– –

IR

Coronal

4990

30

150

7

IR

256 × 512

1

4

–

T1* T1*

Axial Coronal

TSE (FSE) IR (TI, 150 ms), e. g., TIRM IR (TI, 150 ms), e. g., TIRM TSE (FSE) TSE (FSE)

901 901

8.5 8.5

150 150

5 5

+ +

192 × 512 256 × 512

2 2

5 5

– –

Note: Use of a surface phased-array multicoil is recommended. The imaging parameters are only examples and have to be adjusted for use on different brands of scanners. * Fat-suppressed T1w sequences after IV contrast administration.

of the anal canal. Muscular closure is supported by the anal cushions (corpora cavernosa recti) (Figs. 6.4 and 6.5). The rectum above the anal canal is supplied by the superior rectal artery, which is the continuation of the inferior mesenteric artery. The paired middle rectal arteries arising from the internal iliac or internal pudendal artery course laterally through the paraproctium to supply the anal canal and lower portion of the ampulla. At least three layers of the rectal wall can be distinguished on high-resolution T2w MR images: an inner layer of high signal intensity comprising the mucosa and

Fig. 6.1 Normal MR appearance of the rectum in a male on a sagittal T2w TSE image. Sacral flexure (1) anorectal flexure (2), anal canal (3), sacrum (4), urinary bladder (5), prostate (6), and pubic symphysis (7).

130

6 The Rectum and Anal Canal

a

b Fig. 6.2a, b Rectal anatomy. a Diagram of microscopic anatomy. b MR anatomy. Intestinal lumen (1), mucosa (2), submucosa (3), muscular layer (muscularis propria) consisting of inner circular (4), and outer longitudinal layers (5). On the corresponding MR image,

the mucosa and submucosa are depicted as a high-SI band (white arrows), and the muscularis propria appears as a thin line of low SI (open arrows) surrounded by the high-SI mesorectal fat (*).

v Fig. 6.3 Normal appearance on axial T2w TSE image. Note good delineation of the mesorectum (*) and mesorectal fascia (arrowheads). Stool is present in the rectum (arrow).

a

b Fig. 6.4a, b Normal anatomy. a Coronal TSE image. b Axial PD TSE image. Curved arrows indicate the ischiorectal fossa, open arrows the para-anal space, and straight arrows the subcutaneous space.

The coronal image (a) clearly depicts the levator ani muscle on both sides (arrowheads).

MRI Appearance of Pathologic Entities

131

a

b Fig. 6.5a–c Normal MR appearance of the anal sphincter complex. a Coronal PD TSE image. b Axial PD TSE image. c Axial fat-suppressed T2w TSE image. The external anal sphincter muscle (arrowhead) is of the same SI as skeletal muscle, while the internal anal sphincter is markedly hyperintense (arrow). c Contrast is improved on the fatsuppressed image.

c

submucosa; a middle layer of low signal intensity, which corresponds to the muscularis propria; and an outer layer of high signal intensity representing the perirectal fat (Fig. 6.2b). Under optimal conditions, with the bowel relaxed and empty, the mucosa is seen as a thin line of low signal intensity. The mesorectal fascia is consistently depicted as a thin, hypointense line surrounding the rectum and perirectal fatty tissue. On contrast-enhanced T1w images, the mucosa and muscularis mucosae enhance early and intensely, differentiating these two layers from the nonenhancing muscularis propria. Perirectal fat has high signal intensity on non-fat-suppressed T1w images (Fig. 6.5).

MRI Appearance of Pathologic Entities Rectal Carcinoma Colorectal cancer is the third most common cancer worldwide. In the United States, ca. 145 000 new cases and 56 000 deaths were estimated for 2005.8 In all European countries, the incidence of malignant colorectal tumors is on the rise. While colon cancer affects both sexes nearly equally, rectal cancer is more common in men. The incidence of rectal cancer doubles every decade after the age of 40, reaching a peak in the sixth to seventh decade. Nearly all children and young adults who are diagnosed with colon cancer have a predisposing condition. About 90 % of colorectal cancers arise via the so-called

132

6 The Rectum and Anal Canal

Table 6.3 TNM staging system for colorectal carcinoma Stage

Invasion

T1 T2 T3 T4 N1

Submucosa Muscularis propria Subserosa, pericolic/perirectal tissue Other organs or structures/visceral peritoneum Metastasis in 1–3 regional lymph nodes (of at least 12 sampled) Metastasis in 4 or more regional lymph nodes (of at least 12 sampled)

N2

Table 6.4 UICC stage grouping of rectal carcinoma Stage 0 Stage I

Tis

N0

M0

T1

N0

M0

T2

N0

M0

Stage II A

T3

N0

M0

Stage II B

T4

N0

M0

Stage III A

T1, T2

N1

M0

Stage III B

T3, T4

N1

M0

Stage III C Stage IV

Any T Any T

N2 Any N

M0 M1

adenoma–carcinoma sequence. Individuals with benign adenomas have a 2–3 times higher risk of developing colorectal cancer than the normal population. The risk of malignant transformation increases with adenoma size, especially in individuals with the villous type of adenoma. Invasive carcinoma is present in 1 % of all adenomas < 1 cm in diameter, 10 % of those 1–2 cm in size, and 30–50 % of adenomas > 2 cm. Rectal cancer is defined as a tumor located within 16 cm of the anal verge by rigid sigmoidoscopy. The fact that 50 % of colorectal carcinomas arise in the rectum underlines the importance of early detection and accurate tumor staging in this region. Staging is now mostly done using the TNM and UICC9 staging systems, which have largely replaced the older Duke classification (Tables 6.3 and 6.4). Rectal cancer spreads locally through the rectal wall into the perirectal tissue (mesorectum). Initial metastatic spread is predominantly through the lymphatic system.

Lymph Node Metastasis Because the rectum has only sparse lymphatic drainage compared with other colonic segments, lymph node metastases do not occur unless a tumor has infiltrated the muscularis mucosae and submucosa. Lymphatic vessels from the proximal rectum drain into the inferior mesenteric and presacral nodes. The midrectum, which is above the pelvic floor, drains into the internal iliac and sacral nodes. Lymphatic vessels from the perianal skin drain into the superficial inguinal nodes in

the groin or pass through the pelvic floor to terminate in the sacral and internal iliac nodes.

Distant Metastasis Blood from the rectum above the anal canal drains through the superior rectal and inferior mesenteric veins into the portal venous system. The middle and inferior rectal veins drain into the inferior vena cava via the internal iliac vein. Since venous drainage is through the portal system, the liver is the first and most common site of organ metastases from rectal tumors. Another common site is the lung, particularly in patients with low rectal cancer. Next in frequency are metastases to the bones, adrenals, and brain. Advanced rectal cancer may be associated with peritoneal seeding.

MRI Although MRI may not allow exact T staging, the contrast resolution it affords enables precise definition of tumor spread in relation to the mesorectal fascia, which forms the boundary of the surgical excision plane in total mesorectal excision (TME), the standard surgical approach for resection of tumors involving the middle or lower third of the rectum in combination with neoadjuvant or adjuvant therapy.10,11 MRI is currently the only imaging modality that is able to reliably predict whether a tumor-free circumferential resection margin (CRM) is likely to be achieved, thus providing information which is of paramount importance to selecting the most effective therapeutic approach, especially in patients with advanced rectal cancer.12–19 The radiologist’s report must describe the relationship of the tumor to the mesorectal fascia in detail. Since, as already mentioned, MRI does not reveal the mucosa and submucosa as distinct layers in most cases, it is often not possible to discriminate T1 and T2 tumors. Stage T2 rectal cancer is characterized by invasion of the muscularis propria with loss of the interface between the submucosa and circular muscle layer of the rectum, while the outer boundary between the muscle layer and perirectal fat is intact (Fig. 6.6). The most important feature distinguishing T3 tumors from T2 tumors is transmural extension into the mesorectum, seen as obliteration of the interface between muscle and perirectal fat. A problem facing MRI is that inflammatory changes such as desmoplastic reaction, wall edema, and hypervascularization can lead to overstaging of T2 tumors. Therefore, the only reliable criterion for T3 rectal cancer is nodular extension of tumor into perirectal fat (Figs. 6.7 and 6.8). The accurate identification of early T3 tumors, in which either tumor tissue or just fibrous tissue due to desmoplastic reaction extends into the mesorectum, has little impact on patient management.

MRI Appearance of Pathologic Entities

133

a Fig. 6.6a, b Rectal carcinoma, stage T1/2. a Axial T2w TSE image. b Sagittal T2w TSE image. Tumor (arrows) with a broad-based attachment to the left rectal wall. There is no evidence of tumor extension through the wall.

b

a

b Fig. 6.7a, b Rectal carcinoma, stage T3. a Axial T2w TSE image. b Axial T1w fat-suppressed image after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). There is nodular extension of the tumor into the mesorectal fat (arrows), but the tumor does not extend as far as the mesorectal fascia (arrow-

heads). Subtotal intestinal occlusion due to large intraluminal tumor mass. Intense enhancement of the tumor on postcontrast image (b). A lymph node is seen in the mesorectal fatty tissue on the right (open arrow).

T4 tumors invade the visceral peritoneum, surrounding pelvic organs, or the muscle layer of the pelvic sidewall (Fig. 6.9). Despite advances in MRI, identification of metastatic perirectal lymph nodes remains a problem (Fig. 6.10).

Recent data suggest that, with use of high-resolution pulse sequences, signal intensity and border contour may be more reliable predictors of nodal status than size. Metastatic lymph nodes tend to have mixed signal intensity and an irregular border.20

134

6 The Rectum and Anal Canal

b

a

Fig. 6.8a, b Rectal carcinoma, stage T3. a Axial T2w TSE image. b Axial T1w fat-suppressed image after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The tumor (arrows) arises from the right circumference of the rectum. Tumor streaks extending into the perirectal fat correspond to a desmoplastic reaction. In addition, there is nodular extension into the mesorectum (open arrow). The tumor enhances intensely on the postcontrast image. A suspicious mesorectal lymph node is present (arrowhead).

a

b Fig. 6.9a–c Low rectal carcinoma, stage T4. a Axial T2w TSE image. b, c Axial (b) and sagittal (c) T1w fat-suppressed images after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The ulcerating tumor (arrows) invades the external and internal anal sphincter muscles, vagina, and anal canal. There is subtotal intestinal occlusion.

c

MRI Appearance of Pathologic Entities

135

Tumor Recurrence Local recurrence is defined as regrowth of tumor in or around the tumor bed after radical surgical excision (R0 resection) and includes regrowth within the regional lymphatics, surgical scars, and drain sites. A distinction is made between intraluminal recurrence at the anastomotic sites and extraluminal recurrence (Fig. 6.11). Recurrent rectal cancer in the presacral area often involves the piriform muscle and may invade the sacral bone and ischiadic nerve. Rectal cancer has a recurrence rate of up to 50 %, depending on the tumor stage at diagnosis. It may be very difficult to distinguish postoperative or radiation-induced scars from recurrent tumor. Recurrent rectal cancer is of high signal intensity on T2w images and enhances after contrast administration.

Fig. 6.10 Mesorectal lymph nodes in a man with rectal carcinoma. Axial PD image. There is good delineation of two enlarged lymph nodes (arrows) within the mesorectal fatty tissue. MRI provides no reliable criteria for deciding whether these lymph nodes are metastatic or not.

a

b

c

d Fig. 6.11a–g Two cases of rectal carcinoma recurrence. a–d Recurrent rectal carcinoma. Axial T2w TSE images (a, c); axial T1w fatsuppressed images (b, d) after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). There is extraluminal tumor growth on the right side (arrows) with extension (arrowhead)

to the right pelvic sidewall and invasion of the right ureter and the vagina. A stent is present in the ureter (open arrow). The more proximal slices (c, d) show a tumor cavity (*) containing liquefied tumor tissue. Fig. 6.11e–g e

136

6 The Rectum and Anal Canal

e

f Fig. 6.11e–g Recurrent mucinous rectal carcinoma. Axial (e) and sagittal (f) T2w TSE images. Axial (g) T1w fat-suppressed image after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The extraluminal tumor on the left side (arrows) is of high SI on the T2w images due to its high mucinous content. Strong peripheral enhancement of the tumor on the postcontrast image.

g

Anal Carcinoma Anal carcinomas account for approximately 1–2 % of all colorectal carcinomas. The annual incidence is ca. 1 per 100 000 population. A distinction is made between tumors of the anal canal and perianal tumors. The median age at diagnosis is ca. 60 years, and women are affected more commonly. Risk factors are chronic inflammation resulting from benign anal lesions, such as fistulas, fissures, and abscesses, and viruses transmitted by anal intercourse. Histologically, the vast majority are squamous cell carcinomas (ca. 90 %), which, unlike colorectal cancer, do not arise in adenomatous polyps. Anal carcinoma is characterized by local infiltrative growth. Tumors arising in the upper portion of the anal canal typically metastasize to perirectal, iliac, and mesenteric lymph nodes. Distant metastasis is rare. Lower tumors tend to invade inguinal lymph node groups. The standard treatment

is radiochemotherapy; radiotherapy alone may be adequate in patients with small anal tumors. The TNM staging system for anal carcinoma is summarized in Table 6.5.

Inflammatory Bowel Disease Patients with Crohn disease of the small intestine have a 32 % risk of developing perianal complications such as fistulas and abscesses. The risk for these complications increases to 56 % in patients with additional colonic involvement (Fig. 6.12) and is as high as 70 % in isolated Crohn disease of the colon. In 5 % of cases, perianal fistulas and abscesses are the first manifestation of Crohn disease. Other conditions associated with perianal fistulas are tuberculosis and ulcerative colitis. Fistulas and abscesses in Crohn disease differ from their conventional, simple counterparts in that they often form complex branching net-

MRI Appearance of Pathologic Entities

137

works of tracts and tend to recur. It is mainly in the preoperative evaluation of such complex branching fistulas that MRI can provide important supplementary information on the exact extent and course.21–24 As a rule, no rectal contrast medium is needed. The MRI protocol primarily consists of IR sequences (STIR, TIRM) and fat-suppressed T1w sequences after intravenous contrast administration in coronal and axial planes. Additional sagittal sequences may be required if rectovaginal, rectovesical, or rectourethral fistulas are suspected. The aim of preoperative MRI is to accurately characterize fistulous tracts and their relationship to the sphincter complex.25 The types of perianal fistulas that are distinguished on the basis of their course are summarized in Table 6.6. The radiologist’s report should provide information on the internal and external openings of any fistulous tract and its course in relation to the internal and external anal sphincters and the levator ani muscle. In more extensive fistulous disease, the relationship to adjacent pelvic organs (uterus, uterine tubes, ovaries, vagina, prostate, seminal vesicles, urethra) and their possible involvement in the inflammatory process are described. Active fistulous tracts are high in signal intensity on fat-suppressed T2w or IR images because they contain fluid, while they are isointense to fat on non-fat-suppressed T2w images. On T1w imaging, fistulas are of low signal intensity and enhance after contrast administration. Perianal abscesses, which are defined as lesions > 1 cm in size, have the same signal characteristics. A subcutaneous fistula arises low in the anal canal and traverses the internal anal sphincter to then track down

Table 6.5 TNM staging system for anal carcinoma

Fig. 6.12a, b Crohn disease with rectal involvement. Axial (a) and coronal (b) T1w fat-suppressed images after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The wall of the rectum and rectosigmoid junction is markedly thickened due

to severe inflammation (arrows) and enhances intensely. Thickening causes subtotal occlusion of the intestinal lumen (*). Perirectal fat stranding (open arrow) is consistent with an accompanying inflammatory response.

Primary tumor size

Lymph node metastasis

Distant metastasis

T1 T2

≤ 2 cm > 2–5 cm

T3

> 5 cm

N1 – perirectal nodes M1* N2 – unilateral internal iliac/ inguinal node(s) N3 – perirectal and inguinal nodes; bilateral internal iliac/ inguinal nodes

T4

Organ invasion

* E.g., liver (50 %), pelvis and peritoneum (25 %), lung (15 %)

Table 6.6 Anatomic classification of perianal fistulas Fistula

Course

Subcutaneous Intersphincteric

Courses from the dentate line to exit the skin in the perianal area, bypassing the sphincter mechanism Passes from the dentate line through the internal sphincter and intersphincteric space to exit the skin in the perianal area TransExtends from the dentate line through the internal sphincteric and external sphincters, then continues para-anally or through the ischiorectal fossa to exit the skin in the perianal area Ischiorectal Extends from the skin into the ischiorectal fossa, does not communicate with the sphincter or anal canal, often has branching tracts SupraExtends into the supralevator space above the lelevator vator ani muscle, opens into the anal canal or rectum

a

b

138

6 The Rectum and Anal Canal

a

b

c

d Fig. 6.13a–d Subcutaneous fistula tract in Crohn disease. a Axial fat-suppressed IR image. b–d Axial (b, c) and coronal (d) T1w fatsuppressed images after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The cutaneous opening of the fistula seen at the 6 o’clock position contains fluid and is

Fig. 6.14 Transsphincteric fistula on the left (arrow). Axial T2w TSE image. The transsphincteric fistula opens into a pararectal abscess on the left. There is an additional, interspincteric abscess anterior to the anal canal.

therefore shown with high SI on the IR image (arrowhead in a). There is inflammatory soft-tissue edema (* in a) in the area of the gluteal fold on the left. The internal opening is at ca. 3 o’clock (open arrow in c). Nearly the entire course of the fistula is visualized on the coronal T1w fat-suppressed image (arrows in d).

the intersphincteric space to the skin (Fig. 6.13). A fistula that arises at a higher level, or a distal fistula that exclusively or additionally courses upward initially, passes through the intersphincteric space, where an abscess may develop (intersphincteric abscess). More complicated fistulas traverse both layers of the sphincter and extend into the para-anal space (Fig. 6.14), or even further into the ischiorectal fossa, and are often associated with paraanal or ischiorectal abscess formation (Fig. 6.15). An intersphincteric fistula can traverse the levator ani muscle and give rise to a supralevator abscess (Fig. 6.16) and/or track through the levator plate from above (supralevator, translevator fistula, Fig. 6.17). In this way, the inflammatory process spreads to the ischiorectal fossa, where an abscess and/or additional fistulous tracts may form. Scars are of low signal intensity on T1w and T2w images but may have slightly higher signal intensity than fat on fat-suppressed T2w images.

MRI Appearance of Pathologic Entities

139

a

b

c

d Fig. 6.15a–d Para-anal horseshoe abscess. Axial (a) and coronal (b) fat-suppressed IR images. Axial (c) and coronal (d) T1w fat-suppressed images after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The large para-anal abscess (*)

is of high SI on the IR images. On the postcontrast fat-suppressed T1w images, the liquid has low SI while there is marked enhancement of the abscess membrane.

a

b Fig. 6.16a, b Supralevator abscess. a, b Axial (a) coronal (b) T1w fat-suppressed images after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). There is a small

abscess (arrow) above the left levator ani muscle. Low SI in the abscess cavity. Intense enhancement of the abscess membrane. A utricular cyst is present (open arrow).

140

6 The Rectum and Anal Canal

Fig. 6.17 Supralevator (translevator) fistula. Axial T1w fat-suppressed image after IV bolus administration of Gd-based contrast medium (0.1 mmol/kg body weight). The fistula tract (open arrow) extends through the right levator ani muscle.

Functional Disorders Functional disorders of the pelvic floor are a common clinical problem and can severely impair a patient’s quality of life because they may be associated with anal incontinence, stress incontinence, obstipation, organ prolapse, or incomplete defecation. MR defecography is superior to conventional radiographic defecography because it involves no radiation exposure and affords superior softtissue contrast and enables direct image acquisition in any plane. Bowel cleansing with an enema is advisable before MR defecography. Immediately before the MRI examination, a contrast medium—most often ultrasound gel—is intro-

duced into the rectum for bowel opacification. In contrast to conventional defecography, opacification of the bladder and vagina is usually not required because of the high intrinsic contrast of MRI, nor is it necessary to use intravenous contrast medium. Performing MR defecography in an open-configuration, low-field MR imager, which is normally used for research and interventional procedures, appears to offer some advantages over conventional, closed MR scanners because it allows examining the patient in a natural sitting position.26 In a closed MR system, defecography is performed with the patient in the supine position. Before positioning the patient, the table is protected with an absorbent cover and a waterproof sheet. The examination should start with a fast three-plane T2w sequence (e. g., HASTE) to obtain an anatomic overview and rule out other pathology in the true pelvis. The localizer is followed by a TrueFISP cine sequence for acquisition of images in the midsagittal plane at rest and with the patient performing stool evacuation and holding maneuvers. In patients with a suspected lateral rectocele or pelvic floor hernia, the dynamic sequence is repeated in the axial plane. MR defecography will reveal anterior and lateral rectoceles (Fig. 6.18), invaginations, and rectal prolapse.27 Hernias of the pelvic floor can be identified and their severity assessed. Anatomic landmarks for identifying abnormalities are the anorectal angle (angle formed between the central axis of the anal canal and a line parallel to the posterior wall of the distal rectum) and the pubococcygeal line (line extending from the inferior border of the pubic symphysis to the last coccygeal joint).

a

b Fig. 6.18a, b MR defecography in a 65-year-old woman after hysterectomy. Sagittal TrueFISP cine sequences. a At rest. b During defecation. The ultrasound gel introduced using a rectal tube and bladder syringe allows good delineation of the rectum (*) from the

other anatomic structures of the true pelvis. During defecation, there is prolapse of the pelvic floor (arrows) and a large anterior rectocele becomes apparent (open arrows).

MRI Appearance of Pathologic Entities

References 1. Beets-Tan RG, Beets GL, van der Hoop AG, et al. High-resolution magnetic resonance imaging of the anorectal region without an endocoil. Abdom Imaging 1999;24(6):576–581, discussion 582–584 2. Vliegen RF, Beets GL, von Meyenfeldt MF, et al. Rectal cancer: MR imaging in local staging—is gadolinium-based contrast material helpful? Radiology 2005;234(1):179–188 3. Oberholzer K, Junginger T, Kreitner KF, et al. Local staging of rectal carcinoma and assessment of the circumferential resection margin with high-resolution MRI using an integrated parallel acquisition technique. J Magn Reson Imaging 2005; 22(1):101–108 4. Devries AF, Griebel J, Kremser C, et al. Tumor microcirculation evaluated by dynamic magnetic resonance imaging predicts therapy outcome for primary rectal carcinoma. Cancer Res 2001;61(6):2513–2516 5. DeVries AF, Kremser C, Hein PA, et al. Tumor microcirculation and diffusion predict therapy outcome for primary rectal carcinoma. Int J Radiat Oncol Biol Phys 2003;56(4):958–965 6. Dzik-Jurasz A, Domenig C, George M, et al. Diffusion MRI for prediction of response of rectal cancer to chemoradiation. Lancet 2002;360(9329):307–308 7. Grabbe E, Lierse W, Winkler R. The perirectal fascia: morphology and use in staging of rectal carcinoma. Radiology 1983;149(1): 241–246 8. Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005. CA Cancer J Clin 2005;55(1):10–30 9. Sobin LH, Wittekind C. International Union Against Cancer (UICC). TNM Classification of Malignant Tumours. New York: Wiley; 2002 10. Swedish Rectal Cancer Trial. Improved survival with preoperative radiotherapy in resectable rectal cancer. N Engl J Med 1997;336(14):980–987 11. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al; Dutch Colorectal Cancer Group. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001;345(9):638–646 12. Bartram C, Brown G. Endorectal ultrasound and magnetic resonance imaging in rectal cancer staging. Gastroenterol Clin North Am 2002;31(3):827–839 13. Beets-Tan RG, Beets GL, Vliegen RF, et al. Accuracy of magnetic resonance imaging in prediction of tumour-free resection margin in rectal cancer surgery. Lancet 2001;357(9255):497–504

141

14. Beets-Tan RG, Beets GL. Rectal cancer: review with emphasis on MR imaging. Radiology 2004;232(2):335–346 15. Brown G, Radcliffe AG, Newcombe RG, Dallimore NS, Bourne MW, Williams GT. Preoperative assessment of prognostic factors in rectal cancer using high-resolution magnetic resonance imaging. Br J Surg 2003;90(3):355–364 16. Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma: thin-section MR imaging for staging in 28 patients. Radiology 1999;211(1):215–222 17. Cawthorn SJ, Parums DV, Gibbs NM, et al. Extent of mesorectal spread and involvement of lateral resection margin as prognostic factors after surgery for rectal cancer. Lancet 1990; 335(8697):1055–1059 18. Klessen C, Rogalla P, Taupitz M. Local staging of rectal cancer: the current role of MRI. Eur Radiol 2007;17(2):379–389 19. Laghi A, Ferri M, Catalano C, et al. Local staging of rectal cancer with MRI using a phased array body coil. Abdom Imaging 2002;27(4):425–431 20. Brown G, Richards CJ, Bourne MW, et al. Morphologic predictors of lymph node status in rectal cancer with use of high-spatialresolution MR imaging with histopathologic comparison. Radiology 2003;227(2):371–377 21. Beets-Tan RG, Beets GL, van der Hoop AG, et al. Preoperative MR imaging of anal fistulas: Does it really help the surgeon? Radiology 2001;218(1):75–84 22. Laniado M, Makowiec F, Dammann F, Jehle EC, Claussen CD, Starlinger M. Perianal complications of Crohn disease: MR imaging findings. Eur Radiol 1997;7(7):1035–1042 23. Maccioni F, Colaiacomo MC, Stasolla A, Manganaro L, Izzo L, Marini M. Value of MRI performed with phased-array coil in the diagnosis and pre-operative classification of perianal and anal fistulas. Radiol Med (Torino) 2002;104(1–2):58–67 24. Myhr GE, Myrvold HE, Nilsen G, Thoresen JE, Rinck PA. Perianal fistulas: use of MR imaging for diagnosis. Radiology 1994; 191(2):545–549 25. Parks AG, Gordon PH, Hardcastle JD. A classification of fistula-inano. Br J Surg 1976;63(1):1–12 26. Roos JE, Weishaupt D, Wildermuth S, Willmann JK, Marincek B, Hilfiker PR. Experience of 4 years with open MR defecography: pictorial review of anorectal anatomy and disease. Radiographics 2002;22(4):817–832 27. Hilfiker PR, Debatin JF, Schwizer W, Schoenenberger AW, Fried M, Marincek B. MR defecography: depiction of anorectal anatomy and pathology. J Comput Assist Tomogr 1998;22(5): 749–755

143

7

The Kidneys and Upper Urinary Tract M. Taupitz and R. A. Kubik-Huch

Introduction

· Identification of the cause of urinary obstruction, e. g., vascular compression of the ureteropelvic junction.

· Evaluation for renal artery stenosis as a cause of hyperMRI no longer merely provides morphologic information on the kidneys and ureters but has evolved into a versatile imaging technique for use in patients with renal or ureteral tumors and for detailed assessment of the renal vasculature. Moreover, it also allows an insight into the excretory function of the kidneys. A comprehensive MRI protocol with use of intravenous contrast medium combines multiplanar soft-tissue evaluation and qualitative assessment of tissue perfusion with arterial and venous MR angiography (MRA) and also includes excretory MR urography (MRU). In particular, MRI can serve as a “onestop shop” modality, replacing the classic combination of cross-sectional imaging, intravenous urogram, and conventional angiography in the preoperative evaluation of renal and ureteral tumors. Such an MRI protocol can also be used for assessing patients with renal and ureteral anomalies. Two notable advantages—the absence of radiation exposure and the lower nephrotoxicity of MRI due to the need for smaller amounts of contrast material—make this method suitable for younger patients and also for patients with nonmalignant disease. MRI yields the same diagnostic information as CT in patients who have reduced kidney function or do not tolerate iodine-based contrast media. The only disadvantage of MRI compared with CT is its poor visualization of urinary calculi, although it nevertheless allows excellent evaluation of the urinary obstruction secondary to urolithiasis.

Indications Thanks to its technical versatility, MRI enables evaluation of the renal parenchyma, the arteriovenous system, and the renal collecting system and ureters, resulting in a broad spectrum of indications for renal MRI: · Characterization of incidentally detected renal tumors if ultrasound or CT findings cannot exclude malignancy. · Comprehensive evaluation of surgical renal lesions (extent, lymphadenopathy, intra-abdominal metastases) for surgical planning, which includes assessment of renal vascular anatomy and of the upper urinary tract.

tension (discussed in Chapter 15).

· Evaluation of potential living kidney donors. · Post-transplant evaluation for suspected vascular complications, urinary obstruction, parenchymal perfusion defects, or tumor.

Imaging Technique Renal MRI is performed with the patient in a comfortable supine position; a knee support can be offered to improve comfort. As renal MRI regularly comprises sequences in the coronal plane with a left-to-right phase-encoding direction, the patient’s arms may cause wraparound artifacts. Such artifacts can be avoided by placing cushions underneath the arms to elevate them above kidney level.1 The arms must not be placed on the patient’s abdomen as this will cause wraparound on axial images. Patients are carefully instructed to breathe shallowly and regularly during acquisition of non-breath-hold sequences to avoid excessive breathing motion of the abdominal wall. Careful breathing instructions are also necessary if respiratorytriggered acquisition is planned. Finally, it is important that the breathing commands to be given for breath-hold imaging are explained beforehand. For contrast-enhanced imaging, a flexible cannula should be placed, ideally in an antecubital vein, before positioning the patient in the magnet, and connected to a saline-filled syringe or, if available, an MR-compatible injection pump via extension tubing.

Coils Body or torso phased-array coils for abdominal imaging are available from nearly all manufacturers and should be used to improve the signal-to-noise ratio (SNR), especially when fast pulse sequences are acquired2. Phased-array coils (with 4–8 elements) are also needed to employ parallel imaging techniques such as sensitivity encoding (SENSE), which can be used to shorten scan time or to acquire higher-resolution images with both T1w and T2w

144

7 The Kidneys and Upper Urinary Tract

sequences.3–5 However, this gain may come with a penalty in SNR.6

Pulse Sequences The renal MRI protocol can be tailored to different clinical indications (Table 7.1). The basic protocol for soft-tissue evaluation comprises T2w and T1w sequences in multiple planes and should include a T1w sequence acquired with in-phase (IP) and opposed-phase (OP) echo times (TEs) (see Chapter 1). Static MRU of the renal collecting systems and ureters is performed by acquiring a heavily T2w sequence in the coronal plane (comparable to that employed for MRCP; see Chapter 2). The contrast-enhanced series is acquired using a coronal 3D MRA sequence and includes an arterial and venous phase for vascular assessment (see Chapter 15). The protocol is completed by acquiring a delayed axial T1w sequence identical to the corresponding precontrast sequence. To obtain an excretory MR urogram, the 3D MRA sequence is repeated 5 min and 10 min after intravenous contrast injection as well as at later times if contrast medium excretion is delayed. Details of the sequences used for the three main portions of the renal MRI protocol are summarized in Table 7.2.

Precontrast Imaging At our institution, we acquire a localizer followed by a T2w single-shot TSE sequence (e. g., HASTE) in axial, sagittal, and coronal planes. The sagittal images are used to angle the coronal sequence to the long axis of the kidney. An additional respiratory-triggered, fat-suppressed highresolution T2w TSE sequence should be acquired in patients with a small tumor7 (see “T2-Weighted Imaging” in Chapter 1) (Fig. 7.1). An axial 2D T1w GRE sequence (e. g., 2D FLASH) acquired with IP and OP TEs (see Table 1.3) is required for identification of lipid-containing renal tumors such as angiomyolipoma.1 If available, IP and OP images can be acquired using a double-echo sequence. The 2D T1w GRE sequence can be replaced or supplemented by an axial or coronal fat-suppressed 3D GRE sequence (e. g., VIBE) (Fig. 7.1). The 3D GRE sequence improves spatial resolution because thinner slices can be acquired compared with its 2D counterpart; however, this is achieved at the expense of soft-tissue contrast on nonenhanced images.

Contrast-Enhanced Imaging If contrast-enhanced MR images are primarily needed for optimal soft-tissue visualization, and vessel contrast is less important, an axial or coronal 3D GRE sequence em-

Table 7.1 MRI techniques for different indications Indication

Sequence

Plane

Unenhanced/ Contrast medium

Comment

Soft-tissue evaluation for characterization of renal tumors and assessment of tumor extent

T2w single-shot TSE

Axial, sagittal, coronal

Unenhanced

Localization, extent, and characterization of renal tumors; breath-hold acquisition

T1w GRE IP and OP

Axial

Renal vasculature (MRA)

T2w TSE, respiratory- Axial or triggered coronal T1w 3D GRE MRA Coronal sequence

Improved contrast-enT1w 3D GRE hanced soft-tissue eval- (e. g., VIBE) uation (instead of MRA)

Visualization of the collecting systems and ureters (MRU)

Follow-up

Localization, extent, and characterization; breath-hold acquisition Unenhanced Nonspecific contrast medium

Adjunct high-resolution images; free-breathing acquisition Arterial and venous phases for MRA and evaluation of renal perfusion; tumor characterization (hypovascular versus hypervascular); aberrant vessels; anomalies

Axial or coronal

Nonspecific contrast medium

Cortical, parenchymal, and excretory phases for evaluation of tissue perfusion; tumor characterization (hypovascular versus hypervascular)

T1w GRE IP

Axial

Delayed images after contrast administration

Detection of small tumors, characterization

Heavily T2w TSE

Coronal

Unenhanced

Without or with furosmide; morphology of the collecting system; anomalies

T1w 3D GRE sequence (same as for MRA) T1w GRE, T2w TSE

Coronal

Nonspecific contrast medium (administered for MRA) Unenhanced

Without or with furosemide; excretion and outflow, anomalies

Axial, coronal

Number and size of tumors

Note: Preoperative evaluation of renal tumors is performed using all three portions of the renal protocol (soft-tissue evaluation, MRA, and MRU).

Table 7.2 Recommended pulse sequences and imaging parameters Weighting

Plane

Sequence type

T2

Axial, sagittal, HASTE coronal T2 (suppleAxial TSE, respiratorymentary T2w) triggered T1 Axial 2D GRE T1 postcontrast, soft tissue T1 postcontrast, vessels T2, collecting systems T2, collecting systems T2, collecting systems T1, contrast medium excretion

Axial or coronal

3D GRE (VIBE)

Coronal

3D GRE MRA

Coronal, thin HASTE slices Coronal, thick RARE slab Coronal TSE, respiratorytriggered Axial, coroHASTE nal, sagittal

TR (ms)

TE (ms)

∞

60–80

ca. 2500

80

ca. 170–200 5–7

Flip (°)

ETL

FS

Matrix

FOV (mm)

No. of slices

No. of acquisitions

Slice thick- Scan time ness (mm)

Breathhold

Fixed

Yes/no

116 × 256

300 (75 %)

23

1

7

23 s

Yes

–

7–15

Yes

168 × 320

300 (75 %)

48

2

4

5–7 min

No

2.2–7.0

90

–

No

116 × 256

300 (75 %)

23

1

7

23 s

Yes

2.2–2.6

10

–

Yes

116 × 256

300 (75 %)

64

1

2.5

20–24 s

Yes

10–15

–

–

No

128 × 256

300 (75 %)

19

4

8

4–8 min

No

∞

91

–

218

Yes

218 × 256

400 (100 %)

35

1

3

40 s

2800

1100

–

256

Yes

256 × 256

500 (100 %)

1

1

120

5s

Yes, multiple Yes

2000

740

–

129

Yes

384 × 384

380 (100 %)

52

1

1.5

5–7 min

No

∞

60–80

Fixed

Yes/no

116 × 256

300 (75 %)

23

1

7

23 s

Yes

Slice distance, 10–20 % of slice thickness (distance factor, 0.1–0.2). T1w GRE: TEs for acquisition of in-phase and opposed phase images, see Table 1.3. FOV may have to be adjusted based on body habitus. Note: The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time (for sequences with one signal average) but may come with a penalty in SNR.

Imaging Technique

145

146

7 The Kidneys and Upper Urinary Tract

a

b

c

d

e

f Fig. 7.1a–f Renal protocol in normal kidneys. Soft-tissue evaluation using unenhanced axial and coronal images and contrast-enhanced axial images. a Axial T2w TSE image acquired with respiratory triggering. b Coronal breath-hold T2w TSE image (HASTE). c–f Dynamic contrast-enhanced series comprising axial fat-suppressed breath-hold T1w 3D GRE images (VIBE) acquired before (c) and at

different time points after IV injection of Gd-based contrast medium: at 15 s (d), 1 min (e), and 3 min (f). This renal protocol provides high-resolution postcontrast images for optimal soft tissue evaluation with excellent corticomedullary differentiation in the arterial phase (d) and characteristic tiger-stripe pattern of the spleen.

Imaging Technique

147

a

b

c

d Fig. 7.2a–d Renal protocol with coronal contrast-enhanced MRA sequence (3D GRE sequence) instead of the soft-tissue sequence illustrated in Fig. 7.1. a, b MRA images were acquired 15 s (a) and 1 min (b) after IV contrast injection. c MIP reconstruction of a to generate an arteriogram. d MIP reconstruction of a sequence ac-

quired 10 min after contrast injection to generate an excretory MRU. This variant of the renal protocol yields postcontrast images of the renal parenchyma, images for evaluation of renal vessels, and images of the renal collecting systems and ureters for evaluation of urinary outflow.

phasizing soft tissue (e. g., VIBE) should be used. Dynamic contrast-enhanced imaging is performed during the arterial/cortical phase (15 s), venous/corticomedullary phase (1 min), and early excretory phase (at ca. 3 min) (Fig. 7.1). If the contrast-enhanced series is performed for vascular evaluation, a coronal 3D MRA sequence is acquired with a field of view completely including the aorta and vena cava as well as the kidneys. The sequence is angled to the longitudinal axis of the kidney using the sagittal T2w precontrast sequence (see above). Following a timing run, the MRA sequence is acquired during the arterial

phase and 1 min after bolus injection of the contrast medium (Fig. 7.2). The MRA dataset serves to generate multiplanar reformations (MPR) or maximum intensity projections (MIP) of the target vessels and at the same time provides detailed information on renal perfusion. Finally, the precontrast 2D or 3D GRE sequence is repeated to obtain delayed images emphasizing soft tissues. To obtain an excretory MR urogram, the 3D MRA sequence is repeated ca. 10 min after contrast injection (Fig. 7.2).

148

7 The Kidneys and Upper Urinary Tract

MR Urography This can be performed in one of two ways:8,9 · T2w static MRU · T1w excretory MRU. T2w Static MR Urography. This technique involves the use of heavily T2-weighted TSE sequences initially developed for MRCP and yielding images on which stagnant fluid is depicted with a bright signal (hydrography). The slices are acquired in coronal planes angled to the kidneys and ureters using a sagittal localizer. One way to obtain a static MR urogram is to acquire thick slabs in multiple projections during short breath-holds of ca. 3–5 s using a singleshot T2w TSE sequence (e. g., HASTE, RARE). These have limited detail resolution and merely serve to gain a quick overview or as an additional planning scan. Better image quality is obtained by acquiring multiple thin slices during a 20-s breath-hold using a T2w single-shot TSE sequence. The thin-slice datasets can be postprocessed using MPR or MIP algorithms. Image quality can be further improved by respiratory-triggered acquisition of T2w 3D TSE sequences;10 however, the scan time is several minutes (Fig. 7.3). Static MR urograms provide no functional information but allow evaluation of the collecting system in patients with no or only little excretory function. T1w Excretory MR Urography. This technique is based on the renal excretion of an intravenously injected MR contrast medium and its detection with a heavily T1-weighted coronal 3D GRE sequence identical to that employed for MRA. Images are acquired at similar time points as with conventional intravenous urography (Fig. 7.3). However, adequate contrast filling may be delayed in patients with severely dilated collecting systems and ureters or reduced excretory function. This limits the use of excretory MRU in the routine clinical setting as it may not be feasible to repeat imaging at later time points. Both techniques of MRU can be performed with administration of a diuretic (0.05–0.1 mg furosemide) for dilatation and improved visualization of the collecting systems and ureters. Furosemide improves overall image quality of excretory MRU by enhancing the elimination of the contrast medium and its mixing with the urine.8

Contrast Media Renal MRI including MRA and MRU is performed using a nonspecific Gd-based contrast medium such as Gd-DTPA (Magnevist), gadoteridol (Prohance), gadobutrol (Gado-

vist), or Gd-DOTA (Dotarem). These are only examples, and contrast media development is a very dynamic field, which is why radiologists should always check the most up-to-date information regarding approval status in their own country before administering any intravenous contrast medium. The nonspecific Gd-based agents rapidly distribute in the extracellular space after intravenous injection and are eliminated via the kidneys. Their elimination is thus comparable to that of iodine-based X-ray contrast media. Soft-tissue evaluation and MRU are performed with administration of the standard dose of 0.1 mmol Gd per kg body weight. A dose of 0.2 mmol/kg Gd is administered if the protocol also includes MRA. Especially when the higher dose is injected, the high urinary concentration can cause signal voids in the collecting system on delayed images. Note that static T2w MRU cannot be performed after intravenous contrast injection. Patients with severely compromised renal function have a small risk of developing nephrogenic systemic fibrosis (NSF) after intravenous injection of Gd-based contrast medium. Radiologists should therefore use such agents with great caution in these patients, taking into account the most recent official guidelines and manufacturers’ recommendations. In general, patients with severe renal impairment should not receive Gd doses exceeding 0.1 mmol/kg and should not undergo repeated contrast-enhanced MRI at short intervals.

MRI Appearance of Normal Anatomy The kidneys are paired organs located in the lumbar fossae in the retroperitoneum. Grossly, the renal parenchyma is composed of the renal medulla and the renal cortex. With MRI, the corticomedullary differentiation is best appreciated on unenhanced and early contrast-enhanced T1w images (Fig. 7.1). The kidney has a firm capsule, which is not usually seen on imaging, and an outer adipose capsule (or perirenal fat). The kidney, adrenal gland, and adipose capsule are enclosed by the renal fascia. The fascia is most conspicuous on MR images when it is thickened, e. g., due to inflammation.11

MRI Appearance of Normal Anatomy

149

a

b

c

d Fig. 7.3a–d Illustration of normal collecting system on static MRU (hydrography) and excretory MRU. a MIP reconstruction of coronal heavily T2w multislice TSE sequence acquired with respiratory trig-

gering. b–d Coronal 3D GRE images (MRA sequence) acquired 1 min (b), 7 min (c), and 15 min (d) after IV injection of Gd-based contrast medium (augmented by IV injection of 5 mg furosemide).

150

7 The Kidneys and Upper Urinary Tract

MRI Appearance of Pathologic Entities Benign Conditions Anomalies An ectopic kidney, such as a pelvic kidney, results from failure of the developing kidney to ascend to its normal position. The multiplanar capability of MRI, in particular the coronal plane, is helpful for defining the exact position of an ectopic kidney, and an MRA sequence can be performed to assess its blood supply. An ectopic kidney is typically supplied by aberrant arteries arising from the aorta or pelvic arteries. A horseshoe kidney (Fig. 7.4) is the fusion of the inferior poles of the two kidneys by an isthmus of fibrous tissue or renal parenchyma. An ectopic or horseshoe kidney typically has normal function but is more susceptible to trau-

a

matic injury because it is less well protected than a kidney in normal location.12 Ectopic insertion of a ureter results from abnormal migration of a ureteral bud. An ectopic ureter opening into the bladder typically inserts lower than a normal ureter. Other ectopic insertions are the urethra, the vagina, and the uterus in females and the posterior urethra, the seminal vesicles, the ejaculatory duct, and the vas deferens in males. An ectopic ureter inserting below the continence mechanism causes urinary incontinence, which is the case in ca. 50 % of ectopic ureters. About 70 % of ectopic ureteral openings are associated with a duplex kidney and a bifid ureter.13 It is important to identify such anomalies as early as possible; however, the diagnosis is sometimes delayed until adulthood. MRI with the option of static T2w urography is the ideal imaging modality for this purpose because it will also depict a nonfunctioning or poorly functioning renal moiety (Fig. 7.5).14

b

c Fig. 7.4a–d Horseshoe kidney without associated anomalies. a, b Axial T1w (a) and T2w (b) images through the renal hilum. c Axial fat-suppressed T2w TSE image acquired through the parenchymal isthmus at a lower level. d MIP reconstruction of a contrastenhanced MRA showing the parenchymal vessels and beginning excretion of the contrast medium (from test bolus injection) into the renal caliceal system.

d

MRI Appearance of Pathologic Entities

151

a

b Fig. 7.5a, b Ectopic insertion of left ureter draining the upper moiety in duplex kidney. a Coronal T2w HASTE image. b Coronal heavily T2w thick-slab TSE image (projection technique). The rudimentary upper moiety shows hydronephrotic changes (arrow); the

markedly dilated ureter draining the upper moiety inserts into the urethra (curved arrow), and there is high-grade stenosis of the ureteral orifice.

Trauma

Inflammation/Abscess

Abdominal trauma such as that associated with traffic accidents often causes renal contusion or renal vascular injury. Kidneys with prior damage (e. g., due to hydronephrosis) or in an ectopic position are more susceptible to traumatic injury. The role of MRI in emergencies is limited because monitoring of severely injured patients is generally not possible. Multislice CT is the method of choice in acute situations because it enables simultaneous evaluation of the renal parenchyma and renal vessels. Apart from accidents, intra- or perirenal hematoma can be caused by extracorporeal shockwave lithotripsy, diagnostic puncture, clotting disorders, arteriovenous malformations, renal artery aneurysms, renal cell carcinoma, or rupture of a renal cyst. In this setting, MRI will detect small hemorrhage and contributes to the identification of the underlying cause and is also useful for follow-up. The T1 and T2 signal intensities of hemorrhagic lesions vary with the age of the hemorrhage (Fig. 7.6). Perirenal hematoma must be differentiated from urinoma, which—just like urine in the bladder—has low signal intensity on T1w images and high signal intensity on T2w images. If injury of the pelvicaliceal system is suspected in the acute setting, a postcontrast T1w series will demonstrate leakage of contrast medium and thereby of urine (see Figs. 7.33 and 7.34).

Imaging modalities are typically used in patients with acute pyelonephritis to evaluate for complications such as pyonephrosis or abscess. Renal inflammation is seen on MRI as swelling of the kidney, altered signal intensities due to accompanying edema, blurring of renal contours, and loss of corticomedullary differentiation. There may be inflammatory stranding of perirenal fat and thickening of perirenal fasciae, best appreciated on unenhanced T1w images. Severe inflammation may be associated with focal perfusion defects on contrast-enhanced images. A fullblown abscess is a lesion with a central fluid collection of high T2 signal intensity surrounded by a wall that will enhance after contrast administration.15 Both solitary renal cysts and cysts in patients with polycstic kidney disease can become infected. An infected cyst resembles an abscess with a central fluid signal surrounded by a thickened wall showing intense enhancement on postcontrast T1w images (Fig. 7.7).16 Renal atrophy is usually due to chronic inflammation. MRI may be indicated for example if a tumor is suspected. Atrophied kidneys are only seen otherwise on MRI if the examination is performed for other purposes (Fig. 7.8).

152

7 The Kidneys and Upper Urinary Tract

a

b Fig. 7.6a–c Older perirenal hematoma following puncture of the left kidney. Axial breath-hold T1w GRE image (a) and breath-hold T2w TSE images in axial (b) and sagittal (c) planes. Intralesional area of high SI on T1w image and low-SI margin on T2w sequence indicate different stages of hematoma organization with marginal hemosiderin deposits (arrow).

c

a Fig. 7.7a–d Infected renal cyst in the right kidney in bilateral cystic kidney disease. a, b Axial (a) and coronal (b) breath-hold T2w TSE images. c Precontrast axial breath-hold T1w GRE image. d Fat-suppressed axial breath-hold T1w GRE image after IV contrast injection. The fluid in the infected cyst appears homogeneous on the unenhanced images and has a thickened wall, which enhances intensely on the postcontrast image.

b Fig. 7.7c, d

e

MRI Appearance of Pathologic Entities

c

153

d Fig. 7.7c, d

a

b

c Fig. 7.8a–d Pyelonephritic atrophy of both kidneys. a, b Breathhold axial (a) and (b) coronal T2w TSE images. c, d Breath-hold axial T1w GRE images before (c) and 2 min after (d) IV contrast injection. There is marked atrophy of both kidneys with only a little perfused residual parenchyma.

d

154

7 The Kidneys and Upper Urinary Tract

(Figs. 7.9 and 7.10). Chronic obstruction can be congenital or acquired. Causes of acquired obstruction are tumors of the urinary bladder, prostate, uterine cervix, or retroperitoneum, ureteral strictures, and benign prostatic hyperplasia. If the underlying cause cannot be identified by ultrasound, MRU performed as part of a renal soft tissue and vascular MRI protocol should be preferred to conventional intravenous urography or CT. The absence of radiation exposure is especially important in younger patients.

Renal Cysts Fig. 7.9 Small caliceal stone in the middle caliceal group of the left kidney. Respiratory-triggered, fat-suppressed T2w TSE image. The stone causes a signal void within the slightly dilated calix (arrow). Also present is a cyst in the lower pole of the right kidney.

Fig. 7.10 Urinary obstruction due to a stone in the right ureter. MIP image of excretory MRU performed with a coronal 3D GRE sequence ca. 10 min after IV contrast injection. Dilatation of the right collecting system and ureter down to the level of the stone at the iliac vessel crossing (arrow).

Urinary Obstruction Dilatation of the renal collecting systems and ureters occurs secondary to acute or chronic obstruction. The most common causes of acute obstruction are stones or blood clots, pregnancy, and ureteral edema developing after iatrogenic instrumentation, e. g., for stone extraction

Simple cysts (Fig. 7.11), the most common renal lesions, are found in ca. 50 % of adults after the age of 50. They are typically benign and detected incidentally. On MRI they are seen as sharply demarcated lesions of low T1 signal intensity and high T2 signal intensity which do not enhance on T1w images following intravenous contrast medium administration.1 Parenchymal, cortical, and parapelvic renal cysts are distinguished by location. Parapelvic cysts can mimic dilatation of the pelvicaliceal system on ultrasound and unenhanced T1w and T2w MR images. Unclear cases can be resolved by a contrast-enhanced T1w study including an excretory MR urogram (Fig. 7.12). Complicated cysts are due to inflammation or hemorrhage and contain proteinaceous material or blood. They typically have higher signal intensity on T1w images and lower signal intensity on T2w images; the signal intensity of hemorrhagic cysts varies with the age of the blood (Figs. 7.13 and 7.14). Septa, wall thickening, and calcifications are common. Differentiation of a complicated cyst from a neoplasm may occasionally pose a problem based merely on the MRI appearance. It has been reported that cancer is present in up to 30 % of kidneys with a hemorrhagic cyst. If imaging findings are inconclusive, surgical exposure of the kidney and histologic examination are needed for a definitive diagnosis.17 Autosomal dominant polycystic kidney disease (ADPKD) is the adult form of polycystic kidney disease with symptoms usually developing after the age of 40. There is an association with cysts in other organs, such as the liver, and cerebral artery aneurysms. Unlike contrast-enhanced CT, MRI can also be performed in patients with renal insufficiency.18 The typical MRI finding is that of massively enlarged kidneys containing multiple cysts of varying size and signal intensity (Fig. 7.14). MRI is typically performed to exclude superinfection (see Fig. 7.7) or renal cell carcinoma in patients presenting with fever or pain. However, differentiation between a complicated hemorrhagic cyst and inflammation or cancer may pose a considerable challenge, even with MRI. Hilpert and coworkers19 have shown that hemorrhagic cysts in patients with polycystic kidney disease can often be distinguished from cancer by the presence of visible fluid–hemosiderin levels.

MRI Appearance of Pathologic Entities

155

a

b Fig. 7.11a–c Simple renal cyst. a, b Axial images acquired with breath-hold T2w TSE sequence (a) and T1w GRE sequence (b). c Parasagittal reconstruction of a 3D GRE MRA sequence acquired 1 min after IV injection of Gd-based contrast medium. The cyst is seen as a smoothly demarcated lesion of uniform high SI on T2w image (a) and uniform low SI on T1w image (b). No enhancement of the lesion on postcontrast image (c).

c

Benign Renal Tumors Benign tumors of the kidney are usually small and asymptomatic and are often incidentally detected by ultrasound, CT, or MRI. They are of epithelial or mesenchymal origin (Table 7.3). Oncocytoma and renal cell adenoma are of epithelial origin. Benign mesenchymal tumors comprise angiomyolipomas and less common neoplasms such as hemangioma, lymphangioma, leiomyoma, lipoma, and juxtaglomerular tumors (reninoma). Oncocytoma is composed of large eosinophilic cells and is among the more common benign renal tumors. They are 1.6–2.5 times more common in men and occur at slightly higher ages than classic adenomas.20,21 Onco-

cytomas are rather large, measuring on average 5–8 cm, and have an excellent long-term prognosis. However, oncocytomas have been reported to contain focal malignant areas.22 Despite their size, about half of all renal oncocytomas cause no symptoms. In the other cases, hematuria is the most common symptom. The CT and MRI appearance of renal oncocytoma is characterized by a central stellate scar in an otherwise homogeneous lesion. The scar is present in ca. 50 % of all oncocytomas, but has also been reported in renal cell carcinoma.23 A spoke-wheel-like pattern of enhancement, known from conventional angiography, may be seen on early dynamic contrast-enhanced images24 (Figs. 7.15 and 7.16).

156

7 The Kidneys and Upper Urinary Tract

a

b

c

d Fig. 7.12a–d Bilateral parapelvic cysts. a, b Breath-hold axial T2w TSE image (a) and breath-hold sagittal T2w TSE image of the right side (b). c Breath-hold axial T1w GRE image. d MIP reconstruction of a 3D GRE MRA sequence acquired 10 min after IV injection of Gdbased contrast medium (MRU). Multiple large parapelvic cysts are

present, which appear as a septated mass within the dilated renal sinus, in particular on the T2w images, and may be misinterpreted as dilatation of the collecting system. MIP image depicts thin caliceal necks, and only the upper caliceal group on the right is slightly dilated because the necks are compressed by the cysts.

Fig. 7.14a–d Polycystic kidney disease. a Axial breath-hold T1w GRE image. b–d Axial (b), coronal (c), and sagittal (d) breath-hold T2w TSE w images. Both kidneys are massively enlarged and completely replaced by cysts. High SI of some cysts on T1w image (a) and fluid-hemosiderin levels indicate hemorrhage.

MRI Appearance of Pathologic Entities

157

a

b Fig. 7.13a, b Hemorrhagic cortical cyst next to an uncomplicated cyst. a Axial breath-hold T2w TSE image. b Axial breath-hold T1w GRE image. The hemorrhagic cyst (arrow) reverses the SI of the simple cyst with low T2 SI and moderately high T1 SI.

a

b

c

d

158

7 The Kidneys and Upper Urinary Tract

a

b

c

d

Fig. 7.15a–e Oncocytoma. a Axial breath-hold T2w image (TrueFISP). b–e Axial breath-hold T1w GRE images acquired before (b) and 15 s (c), 1 min (d), and 2 min (e) after IV contrast injection. The tumor shows typical early and nearly completely homogeneous enhancement, consistent with a hypervascular tumor such as oncocytoma. This MRI examination does not exclude RCC (images courtesy of Dr. B. Sander, Berlin).

e

Table 7.3 Overview of benign renal tumors Common

Rare

· Oncocytoma · Angiomyolipoma · Multilocular cystic nephroma

· · · · ·

Hemangioma Leiomyoma Lymphangioma Lipoma Juxtaglomerular tumor (reninoma)

MRI Appearance of Pathologic Entities

159

a

b

c

d

e

f Fig. 7.16a–f Oncocytoma. a Unenhanced axial T1w SE image. b Fatsuppressed axial T2w TSE image. c–f Dynamic breath-hold coronal GRE images before (c) and immediately (d), 1 min (e), and 3 min (f) after IV injection of contrast medium. Mass in the center of the right kidney (arrows). The lesion is not very conspicuous on T1w and T2w

images as it is nearly isointense to surrounding renal parenchyma. The dynamic contrast-enhanced images show a sharply marginated lesion with spoke-wheel enhancement, consistent with an oncocytoma.

160

7 The Kidneys and Upper Urinary Tract

a

b Fig. 7.17a–c Angiomyolipoma. a, b Unenhanced axial T1w GRE images acquired with IP (a) and OP (b) TEs (see Chapter 1 for details of IP/OP imaging). c Axial T2w HASTE image. Tumor in the right kidney (arrow in a) with high SI on IP image and loss of signal on OP image, resulting in a dark line around the tumor. The IP/OP signal characteristics suggest a large fatty component of the tumor, in keeping with an angiomyolipoma. The T2w features contribute no diagnostic information.

c

Renal angiomyolipoma is a hamartoma consisting of fat, smooth muscle, and abnormal blood vessels. Also common are intralesional hemorrhage and necrosis. Angiomyolipomas are rare, accounting for only 0.3 % of all renal tumors.25 About 50–80 % of patients with renal angiomyolipomas suffer from tuberous sclerosis, and these patients usually have smaller, multiple, and bilateral tumors.20,26 An angiomyolipoma may become as large as 20 cm. Symptomatic patients typically present with a palpable abdominal mass, flank pain, microhematuria, or retroperitoneal bleeding. Malignant behavior, including invasive local growth, regional lymphadenopathy, and inferior vena cava thrombosis, has occasionally been observed. However, renal angiomyolipomas are usually benign and definitive control is achieved by surgical resection.20 A confident diagnosis of an angiomyolipoma can usually be made by MRI. Fatty tumor components have high signal intensity on T1w images and can be confirmed by an additional fat-suppressed sequence, which clearly distinguishes fat from intralesional hemorrhage.27 On OP images, there may be a loss of signal intensity of intratumoral fatty areas, or the signal loss is confined to an interface between fat and normal renal tissue (Figs. 7.17 and 7.18). Angiomyolipoma must be differentiated from liposarcoma of the perirenal fatty tissue. Here, the multiplanar capability of MRI is helpful in showing the extrarenal

location of liposarcoma.28,29 In rare instances, an angiomyolipoma is predominantly composed of muscle tissue with only very little macroscopic fat; this tumor is difficult to distinguish from renal carcinoma or other tumors based on imaging features. On the whole, multilocular cystic nephroma is a rare tumor of the metanephric blastema, though it is one of the more common benign renal tumors. It appears to be a benign variant of Wilms tumor and has a biphasic age and sex distribution, affecting boys < 4 years of age and women aged 40–60 years.30,31 The tumor is usually large (ca. 10 cm on average) and consists of multiple noncommunicating cysts separated by thin septa with a thick capsule. The clinical behavior is usually benign, with a rare multilocular cystic nephroma undergoing malignant transformation to nephroblastoma in children or sarcoma in adults. There are no CT or MRI criteria to reliably differentiate multilocular cystic nephroma from malignant cystic tumors such as cystic Wilms tumor or cystic renal cell carcinoma. For this reason surgical excision is indicated in all cases.30 Rare benign tumors of the kidneys include hemangioma, lymphangioma, leiomyoma, and juxtaglomerular tumors (reninoma). Except for lipoma, the MRI appearance of these tumors is nonspecific and the diagnosis must be histologically confirmed.

MRI Appearance of Pathologic Entities

161

a

b

c

d Fig. 7.18a–d Small angiomyolipoma. a, b Unenhanced axial T1w GRE images acquired with IP (a) and OP (b) TEs (see Chapter 1 for details of IP/OP imaging). c Axial T2w HASTE image. d Axial fatsuppressed T2w TSE image. Small tumor in the parenchyma of the left kidney (arrow) with high SI on IP image and characteristic ring-

like loss of SI on OP image. The T2 SI is high without fat suppression and low with fat suppression. This constellation of SIs suggests a tumor with a large fatty component and is consistent with the diagnosis of angiomyolipoma.

In the past, the term renal adenoma was used to refer to solid tumors up to 2 cm in size, which are even difficult to differentiate histologically from malignant adenocarcinomas. Therefore, it has been proposed that what used to be referred to as renal adenoma is simply an early, nonmetastatic stage of malignant adenocarcinoma.20 A small vascularized solid tumor without features of an angiomyolipoma (fatty component) should be classified as a small renal cell carcinoma (RCC). Small solid tumors < 2 cm in diameter without pathologic, histologic, or imaging features of malignancy or aggressive behavior are found in 4–22 % of all autopsies.32 Since no definitive imaging criteria exist to distinguish between benign and malignant lesions, all lesions, including small ones, incidentally detected by MRI or another imaging modality should be regarded as potentially malignant and closely monitored or surgically removed. MR imaging in multiple planes provides detailed information on the site and extent of a renal tumor, which is especially important in patients scheduled for surgical enucleation or partial kidney resection rather than nephrectomy.

Malignant Renal Tumors Primary Renal Tumors An overview of primary and secondary renal malignancies is given in Table 7.4. Renal cell carcinoma (RCC) is the most common malignancy of the kidneys but accounts for only 2 % of all adult tumors. RCC is more common in men, with an age peak in the fifth and sixth decades. About 1 % of RCCs are bilateral, and multiple tumors in one kidney are present in ca. 5 % of cases. The signs and symptoms are unspecific and typically occur late. Only ca. 20 % of RCCs present with the classic triad of flank pain, hematuria, and a palpable mass.33 Exact pretherapeutic determination of the tumor extent is important because radical surgical resection is the only curative treatment option. RCC not extending beyond the renal capsule (stages T1 and T2) is treated by radical nephrectomy, supplemented by thrombectomy with venous graft repair as needed for tumors extending into the renal vein and inferior vena cava.

162

7 The Kidneys and Upper Urinary Tract

Table 7.4 Overview of primary and secondary malignant renal tumors Primary malignant renal tumors Common

· Renal cell carcinoma (hypernephroma, renal adenocarcinoma)

· Transitional cell carcinoma (TCC) · Wilms tumor (nephroblastoma) Rare

· · · · ·

Leiomyosarcoma Rhabdomyosarcoma Angiosarcoma Liposarcoma Fibrosarcoma

Secondary malignant renal tumors

· Metastases · Lymphoma

The Robson classification is widely accepted for staging RCC in the USA, whereas the TNM classification system is more widely used in Europe.34 The staging accuracy reported for multislice CT ranges between 70 % and 91 %, with that for MRI being comparable or slightly better.35–37 Both CT and MRI are highly accurate in staging RCC with tumor thrombus in the inferior vena cava.38 With regard to tumor characterization, MRI is slightly better than CT in correctly identifying cystic lesions.39 Like other solid renal tumors, RCC is typically isointense to surrounding tissue on unenhanced T1w images and can only be suspected if the tumor distorts the renal contour. The T2 appearance is variable. Liquid areas due to necrosis and intralesional hemorrhage are present in ca. 15 % of RCCs and have high signal intensity on T2w images. Both detection and characterization of RCC can be improved by administration of intravenous contrast medium.27,40 Even small tumors can be detected and characterized using the high-resolution sequences available today. Images obtained 1–2 min after administration of contrast medium are an integral part of the renal protocol (Figs. 7.19 and 7.20). Also important is the multiplanar capabil-

a

b

c

d Fig. 7.19a–d Small RCC in the right kidney, stage T1. a Fat-suppressed, respiratory-triggered axial T2w TSE image. b–d Dynamic contrast-enhanced, breath-hold axial T1w 2D GRE images acquired before (b) and 15 s (c) and 2 min (d) after IV injection of Gd-based contrast medium. Small parenchymal tumor in the anterior aspect

of the right kidney (arrow). The tumor has no cystic components on T2w image and shows a moderate SI increase on contrast-enhanced image. These features suggest a solid, vascularized tumor and are highly indicative of malignancy.

MRI Appearance of Pathologic Entities

163

a

b Fig. 7.20a–c Small RCC in the right kidney, stage T1, incidentally detected in an evaluation of a potential renal donor. a Breath-hold axial T2w TSE image. b Breath-hold axial T1w GRE image. c Coronal reconstruction of a 3D GRE MRA sequence acquired 1 min after IV contrast injection. Only focal contour distortion is seen on the unenhanced images. A malignant tumor is suggested by the inhomogeneous SI increase seen in this area on the postcontrast image (arrow in c).

c

ity of MRI, which affords accurate assessment of tumor extent with good delineation of the mass from surrounding organs on sagittal and coronal images (Figs. 7.21, 7.22, 7.23, 7.24). Venous invasion is present in ca. 20 % of patients with RCC at the time of diagnosis and is best detected by a venous phase 3D MRA sequence (MR venogram) (Fig. 7.23). At the same time, contrast-enhanced T1w images enable differentiation of a vascularized tumor thrombus from an appositional thrombus. The high susceptibility of MRI to the effects of contrast media, in conjunction with the high resolution afforded by state-of-the-art techniques, most notably 3D MRA sequences, allows good evaluation of delicate structures enhanced by contrast administration and facilitates the separation of complicated benign cysts and cystic RCC (Fig. 7.25). Above a size threshold of 1–1.5 cm, a lymph node is assumed to be metastatic; however, this criterion is unreliable and false positive results due to reactive lymph node enlargement and false negative results due to microscopic metastasis in normal-sized lymph nodes are not uncommon (Fig. 7.24).

Transitional cell carcinoma (TCC) of the renal pelvis accounts for 7 % of all renal neoplasms. Risk factors include urinary bladder cancer, analgesic abuse, and exposure to carcinogenic agents (aniline dyes, cyclophosphamide, tobacco smoke). TCC tends to be multicentric with ca. 8–40 % of TCC patients having a synchronous or metachronous tumor of the lower urogenital tract or on the contralateral side.41,42 The diagnostic role of MRI is limited and small tumors that do not dilate the pelvicaliceal system may be overlooked because they are typically isointense to surrounding tissues. However, TCC usually enhances early on postcontrast images.43 Imaging cannot differentiate advanced TCC invading the renal parenchyma from RCC; and vascular invasion, though less common than in RCC, has also been reported for TCC44 (Figs. 7.26 and 7.27). If a tumor cannot be reliably classified as RCC, the radiologist’s report should mention a TCC as a possible differential diagnosis. This is important because the surgical approach is different: nephrectomy for RCC vs nephrectoureterectomy for TCC.

164

7 The Kidneys and Upper Urinary Tract

a

b

c

d Fig. 7.21a–f RCC, stage T1. MRI to plan organ-sparing surgical resection. a, b Breath-hold axial (a) and sagittal (b) T2w TSE images. c, d Breath-hold T1w GRE images before (c) and 2 min after (d) IV injection of Gd-based contrast medium. e, f Contrast-enhanced 3D GRE MRA sequence reconstructed as an MIP image at 15 sec (e) and as a coronal MPR at 1 min (f). Tumor in the upper aspect of the right kidney causing contour distortion without invasion of perirenal

fat. Slices acquired in different planes and during different perfusion phases enable highly accurate definition of tumor extent, especially its relationship to the structures of the renal sinus (including the pelvicaliceal system). In the case shown, the MRA sequence of the renal protocol revealed three arteries supplying the right kidney. The tumor was laparoscopically removed and the kidney preserved. Fig. 7.21e, f

e

MRI Appearance of Pathologic Entities

165

e

f Fig. 7.21e, f

a

b Fig. 7.22a–e RCC of the right kidney, stage T3b. a–c Breath-hold axial (a), coronal (b), and sagittal (c) T2w TSE images. d, e Breathhold T1w GRE images before (d) and 2 min after (e) IV injection of Gd-based contrast medium. Large tumor in the upper half of the right kidney causes marked contour distortion while there are no

signs of extension to the Gerota fascia. The tumor invades the renal sinus and a small tumor thrombus extends through the renal vein into the inferior vena cava (arrow in e). Fig. 7.22c–e

e

166

7 The Kidneys and Upper Urinary Tract

c

d Fig. 7.22c–e

e

Fig. 7.23a–e RCC of the right kidney, stage T3c. a–c Breath-hold axial (a), coronal (b), and sagittal (c) T2w TSE images. d Coronal breath-hold 3D GRE MRA sequence acquired 1 min after IV contrast injection with reconstruction of coronal slice for evaluation of the tumor. e Coronal breath-hold thin-slice MIP image for evaluation of the tumor thrombus. Tumor in the upper third of the right kidney with contour distortion; no invasion of the perirenal fat but large tumor thrombus in the inferior vena cava with the tip extending into the right atrium (arrow in e). Fig. 7.23b–e

a

e

MRI Appearance of Pathologic Entities

b

167

c

d Fig. 7.23b–e

e

168

7 The Kidneys and Upper Urinary Tract

a Fig. 7.24a, b RCC of the right kidney, stage T4. a Breath-hold axial T1w GRE image. b 3D GRE MRA sequence acquired 1 min after IV contrast injection with reconstruction of coronal slice. The tumor appears small on the axial image, while extensive infiltration of the liver is seen on the coronal image. The axial image also reveals numerous enlarged retroperitoneal lymph nodes. Histology demonstrated these to be nonmetastatic.

b

a

c

d

b

Fig. 7.25a–e Cystic RCC of the right kidney, stage T1. a, b Breathhold axial (a) and coronal (b) T2w TSE images. c, d Breath-hold axial T1w GRE images obtained before (c) and 2 min after (d) IV injection of Gd-based contrast medium. e Coronal reconstruction of 3D GRE MRA sequence acquired 1 min after IV injection of contrast medium. Predominantly cystic tumor with some septa and small solid areas, which enhance after contrast administration. Fig. 7.25e e

MRI Appearance of Pathologic Entities

169

v Fig. 7.25e

a

b

c

d Fig. 7.26a–d TCC. Breath-hold dynamic coronary GRE images obtained before (a) and immediately (b), 1 min (c), and 3 min (d) after IV injection of Gd-based contrast medium. Small enhancing mass

(arrows) in the left ureteropelvic junction with obstruction and dilatation of the pelvicaliceal system.

170

7 The Kidneys and Upper Urinary Tract

a

b

c

d

e

f Fig. 7.27a–g TCC of the left kidney. a Precontrast axial T1w SE image. b Fat-suppressed axial fast T2w SE image. c–f Breath-hold coronal dynamic GRE images obtained before (c) and immediately (d), 1 min (e), and 3 min (f) after IV injection of Gd-based contrast medium. g Fat-suppressed contrast-enhanced axial T1w SE image. Mass in the left renal pelvis (arrow) of the same signal intensity as the renal parenchyma on the unenhanced T1w and T2w images (a, b). The tumor is revealed on the contrast-enhanced images because it enhances less intensely than the surrounding parenchyma (arrow in c–f, g). Small simple cyst in the right renal cortex (d).

g

MRI Appearance of Pathologic Entities

Nephroblastoma, or Wilms tumor, is a malignant embryonic tumor that occurs predominantly in children aged 2.5–3 years, affecting boys and girls equally. It is the most common abdominal malignancy of the first 8 years of life and is associated with urogenital anomalies, aniridia, and Beckwith–Wiedemann syndrome.20 The signal intensity is typically inhomogeneous due to the presence of necrosis and hemorrhage and does not allow differentiation from other renal tumors. Nevertheless, MRI may be useful for preoperative planning by differentiating tumor tissue from normal renal tissue and surrounding structures45,46 (Fig. 7.28). Leiomyosarcoma, rhabdomyosarcoma, liposarcoma, angiosarcoma, and fibrosarcoma are rare primary renal malignancies. Most of these tumors have nonspecific MRI findings and the final diagnosis is based on histology.

171

a

Secondary Renal Tumors Although renal metastases are detected in 2–20 % of all autopsies, they have only a small role in imaging as they are typically asymptomatic and are detected late or not at all.47 Renal metastases tend to be multifocal (Fig. 7.29). Despite the nonspecific MR findings, the diagnosis is obvious in patients with multiple solid lesions in both kidneys who have a known primary tumor or additional nonrenal metastases. However, large solitary renal metastases may occur in patients with cancer of the lung, breast, or colon, and these cannot be distinguished from RCC based on their MRI appearance alone. Primary renal lymphoma is very rare, but renal involvement through continuous extension from retroperitoneal lymph nodes or hematogenous spread in patients with systemic lymphoma is more common. Renal lymphoma is either diffuse or focal in the form of multiple or solitary lesions. Diffuse renal involvement is associated with enlargement of the kidney.48,49 Lymphoma nodules are usually hypovascular, making them conspicuous relative to the intensely enhancing normal renal parenchyma on early arterial phase contrast-enhanced T1w images (Figs. 7.30 and 7.31). MRI findings in diffuse renal lymphoma are loss of the corticomedullary differentiation, diffuse renal enlargement, and associated lymphadenopathy.24

Fig. 7.28a–c Nephroblastoma in an 18-month-old girl. a Axial T2w TSE image. b Axial contrast-enhanced T1w SE image. c Coronal contrast-enhanced T1w SE image. The tumor is of inhomogeneously high SI on the T2w image and poorly demarcated from the renal parenchyma (a). It enhances less intensely than surrounding normal parenchyma, and its extent can be accurately determined by evaluation in two planes (b, c).

b

c

172

7 The Kidneys and Upper Urinary Tract

a

b Fig. 7.29a–c Metastases from squamous cell carcinoma (ENT) in the left kidney. a Breath-hold axial T2w TSE image. b, c Breath-hold axial T1w GRE images obtained before (b) and 2 min after (c) IV injection of Gd-based contrast medium (image quality slightly degraded by ghosting artifacts). The metastases are markedly hyperintense on T2w image and hypointense on postcontrast image. Hypointensity indicates necrosis, which is typical of metastases from squamous cell carcinoma.

c

a

b Fig. 7.30a–d Lymphoma in the left kidney (solitary kidney). a, b Breath-hold axial (a) and coronal (b) T2w TSE images. Fig. 7.30c, d

e

MRI Appearance of Pathologic Entities

173

c Fig. 7.30c, d c Unenhanced breath-hold axial T1w GRE image. d Coronal reconstruction of a 3D GRE MRA sequence obtained 15 s after IV injection of Gd-based contrast medium. Tissue of nearly the same SI as renal tissue is present in the renal sinus and obstructs outflow from the calices (b). Reduced perfusion, predominantly in the upper third of the kidney (d). Patient had non-Hodgkin lymphoma. The imaging appearance is also consistent with TCC of the renal pelvis.

d

a

b

c

d Fig. 7.31a–d Lymphoma in both kidneys. Breath-hold dynamic coronal T1w GRE images before (a) and immediately (b), 1 min (c), and 3 min (d) after IV injection of Gd-based contrast medium. Both

kidneys are enlarged and contain multiple hypovascular, rounded parenchymal lesions.

174

7 The Kidneys and Upper Urinary Tract

Functional Renal Imaging With its high spatial resolution and good temporal resolution, MRI provides not only morphologic information on renal pathology but also an insight into renal function. Several conditions including glomerulonephritis, interstitial nephritis, renal involvement in diabetes mellitus, and chronic urinary obstruction are associated with global impairment of renal function, whereas extracorporeal shockwave lithotripsy can cause segmental functional impairment.50 MR contrast media, unlike the iodinated X-ray contrast media used in radiography and CT, are not nephrotoxic because only very small amounts are needed, and they can therefore also be used in patients with mildly impaired renal function without the risk of additional renal damage.27,51,52 For contrast medium administration in patients with severely compromised renal function, see the Contrast Media section earlier in this chapter. Gd-based contrast media are completely filtered in the renal glomeruli and then concentrated in the Henle loops and collecting tubes of the medulla by water reabsorption. Sequences acquired with a high temporal resolution, such as dynamic contrast-enhanced GRE or echo-planar sequences, are well suited to track the passage of the contrast medium as concentration-dependent changes in relaxation times. Patients with delayed elimination of the contrast medium due to impaired renal function show a less pronounced decrease in signal intensity than individuals with normal renal function. In patients with severely impaired renal function and a creatinine clearance < 30 mL/min, the typical contrast reversal in the medulla, which is due to the high concentration of Gd and the resulting predominance of the T2-shortening effect, may even be completely absent.24,44 Quantification of renal function by serial measurement of signal intensities after intravenous injection of contrast medium, using special pulse sequences such as inversion-prepared or saturationprepared fast sequences in conjunction with an appropriate analysis algorithm, has been proposed. However, the error rate in determining local contrast medium concentration is high, and the technique does not yet allow reliable determination of renal function parameters.53,54

Evaluation of Living Kidney Donors Kidneys from living donors are increasingly used to meet the demand of waiting recipients. Laparoscopic donor nephrectomy is minimally invasive and can promote living kidney donation.55 Careful evaluation of potential kidney donors and their kidneys contributes to the technical success of renal transplants and minimizes the periand postoperative risk for the donor. Renal vascular anatomy is important in selecting the side of donor nephrectomy. Ideally, the kidney to be donated should have a single artery and a single vein. Vascular renal anatomy is highly variable, with a 40 % rate of arterial and 20 % venous anomalies.56 This is why catheter-based angiography was

a

b Fig. 7.32a, b Evaluation of a potential living kidney donor. MIP reconstructions of a 3D GRE sequence acquired 15 s after IV injection of Gd-based contrast medium. a Left anterior oblique projection. b Right anterior oblique projection. There are two renal arteries on the left and two renal veins on the right (arrows).

used in the past to map vascular anatomy in potential kidney donors; however, these are healthy individuals and the risks of an invasive diagnostic test should be avoided. Several studies have shown that MRA is equivalent to conventional angiography in identifying relevant arterial anomalies and even superior in detecting venous variants (Fig. 7.32).57,58 Moreover, the different MRI techniques for renal imaging enable a more comprehensive morphologic evaluation and exclusion of anomalies of the kidneys and urinary tract in potential kidney donors. In very rare instances, kidney donor MRI may even detect an incidental malignant renal tumor (see Fig. 7.20).

MRI Appearance of Pathologic Entities

175

a

b Fig. 7.33a–c Urinoma in a renal transplant recipient. a, b Fast T1w GRE images before (a) and 10 min after (b) IV injection of Gd-based contrast medium. c Coronal reconstruction of a 3D GRE sequence performed ca. 5 min after injection of a test bolus for determination of circulation time and before the actual MRA sequence for vascular evaluation was acquired. Postcontrast images (b and c) show leakage of contrast medium from the ureter, which collects posterior to the kidney, consistent with urinoma (arrow in c).

c

Renal Transplants Several complications can occur in renal transplant recipients including postoperative hematoma, urinoma or lymphocele, urinary obstruction, infarction, acute tubular necrosis, toxicity of cyclosporine and other drugs, and acute or chronic rejection.59,60 MRI has gained an increasing role in evaluating complications in renal transplant recipients, not least because of technical refinements. MRI enables assessment of renal transplants in terms of morphology, function, and perfusion.59,60 The MRI protocol for kidney graft recipients includes fast T2w sequences (e. g., HASTE) in all three orthogonal planes and an axial T1w GRE sequence in conjunction with a contrast-enhanced 3D MRA sequence and a delayed T1w sequence. This protocol may be supplemented by a static or excretory MRU (see Tables 7.1 and 7.2). An excretory MRU will detect postoperative or postinterven-

tional complications such as ureteral leaks causing urinoma (Fig. 7.33). Vascular complications such as thrombosis of the transplant renal vein can be detected with a venous phase MRA sequence (for further details see Chapter 15). As with all operations, bleeding with hematoma formation may also occur after kidney transplant. Active bleeding is occasionally seen on contrast-enhanced images (Fig. 7.34). Rejection of a kidney transplant may be associated with loss of corticomedullary differentiation, which is best appreciated on T1w images. However, this is a nonspecific sign and may also occur in association with cyclosporine toxicity, congestion, and other complications.59,60 Renal infarction and cortical necrosis are characterized by increased signal intensity on T2w images and markedly reduced perfusion on dynamic contrast-enhanced sequences (Fig. 7.35).

176

7 The Kidneys and Upper Urinary Tract Fig. 7.34a–e Evaluation of a kidney graft recipient with bleeding from a parenchymal lesion. a Axial breath-hold T2w TSE image. b, c Axial breath-hold T1w GRE images obtained before (b) and 2 min after (c) IV injection of Gd-based contrast medium. d, e Coronal reconstruction (d) and MIP image (e) of a 3D MRA sequence acquired 15 s after contrast injection. Images depict a large hematoma lateral to and compressing the transplant kidney (arrows). In (c) and (d), active bleeding is indicated by leakage of contrast medium from the parenchyma.

a

b

c

d

e

MRI Appearance of Pathologic Entities

a

b

c

d

177

f Fig. 7.35a–f Vascular rejection in a transplanted kidney. a–d Breath-hold dynamic coronal T1w GRE images obtained before (a) and immediately (b), 1 min (c), and 3 min (d) after IV injection of Gd-based contrast medium. e MIP image of the arterial phase. f Explanted kidney. There is no perfusion of the renal cortex (arrows in a–d). The main branch of the transplant renal artery (arrow in e) and its segmental branches appear normal, while no peripheral perfusion is seen. The histologic diagnosis was cortical necrosis secondary to vascular rejection.

e

178

7 The Kidneys and Upper Urinary Tract

References 1. Israel GM, Bosniak MA. How I do it: evaluating renal masses. Radiology 2005;236(2):441–450 2. Helmberger T, Holzknecht N, Lackerbauer CA, et al. [Phasedarray superficial coil and breath holding technique in MRI of the liver. Comparison of conventional spin echo sequences with rapid fat suppressing gradient echo and turbo-spin sequences]. Radiologe 1995;35(12):919–924 3. Dobritz M, Radkow T, Nittka M, Bautz W, Fellner FA. [VIBE with parallel acquisition technique - a novel approach to dynamic contrast-enhanced MR imaging of the liver]. Rofo 2002;174(6): 738–741 4. Kim YK, Kim CS, Chung GH, Jeon SB, Lee JM. Feasibility of application of sensitivity encoding to the breath-hold T2weighted turbo spin-echo sequence for evaluation of focal hepatic tumors. AJR Am J Roentgenol 2005;184(2):497–504 5. Yoshioka H, Sato J, Takahashi N, et al. Dual double arterial phase dynamic MR imaging with sensitivity encoding (SENSE): which is better for diagnosing hypervascular hepatocellular carcinomas, in-phase or opposed-phase imaging? Magn Reson Imaging 2004;22(3):361–367 6. Vogt FM, Antoch G, Hunold P, et al. Parallel acquisition techniques for accelerated volumetric interpolated breath-hold examination magnetic resonance imaging of the upper abdomen: assessment of image quality and lesion conspicuity. J Magn Reson Imaging 2005;21(4):376–382 7. Asbach P, Klessen C, Kroencke TJ, et al. Magnetic resonance cholangiopancreatography using a free-breathing T2-weighted turbo spin-echo sequence with navigator-triggered prospective acquisition correction. Magn Reson Imaging 2005;23(9): 939–945 8. Nolte-Ernsting C, Staatz G, Wildberger J, Adam G. [MR-urography and CT-urography: principles, examination techniques, applications]. Rofo 2003;175(2):211–222 9. Memarsadeghi M, Riccabona M, Heinz-Peer G. [MR urography: principles, examination techniques, indications]. Radiologe 2005;45(10):915–923 10. O’Malley ME, Soto JA, Yucel EK, Hussain S. MR urography: evaluation of a three-dimensional fast spin-echo technique in patients with hydronephrosis. AJR Am J Roentgenol 1997;168(2): 387–392 11. Semelka RC, Kelekis N. Kidneys. In: Semelka RC, Ascher PM, Reinhold C, eds. MRI of the Abdomen and Pelvis. New York: Wiley-Liss; 1997: pp. 397–470 12. Pollack HM, Wein AJ. Imaging of renal trauma. Radiology 1989;172(2):297–308 13. Berrocal T, López-Pereira P, Arjonilla A, Gutiérrez J. Anomalies of the distal ureter, bladder, and urethra in children: embryologic, radiologic, and pathologic features. Radiographics 2002;22(5): 1139–1164 14. Staatz G, Rohrmann D, Nolte-Ernsting CC, et al. Magnetic resonance urography in children: evaluation of suspected ureteral ectopia in duplex systems. J Urol 2001;166(6):2346–2350 15. Aerts P, Van Hoe L, Bosmans H, Oyen R, Marchal G, Baert AL. Breath-hold MR urography using the HASTE technique. AJR Am J Roentgenol 1996;166(3):543–545 16. Chicoskie C, Chaoui A, Kuligowska E, Dember LM, Tello R. MRI isolation of infected renal cyst in autosomal dominant polycystic kidney disease. Clin Imaging 2001;25(2):114–117 17. Marotti M, Hricak H, Fritzsche P, Crooks LE, Hedgcock MW, Tanagho EA. Complex and simple renal cysts: comparative evaluation with MR imaging. Radiology 1987;162(3):679–684 18. Schuhmann-Giampieri G, Krestin G. Pharmacokinetics of GdDTPA in patients with chronic renal failure. Invest Radiol 1991;26(11):975–979 19. Hilpert PL, Friedman AC, Radecki PD, et al. MRI of hemorrhagic renal cysts in polycystic kidney disease. AJR Am J Roentgenol 1986;146(6):1167–1172 20. Leder LD, Richter HP. Pathology of renal and adrenal neoplasms. In: Lohr EL, Leder LD, eds. Renal and Adrenal Tumors. New York: Springer; 1987: pp.1–68 21. Romis L, Cindolo L, Patard JJ, et al. Frequency, clinical presentation and evolution of renal oncocytomas: multicentric experience from a European database. Eur Urol 2004;45(1):53–57, discussion 57

22. Tikkakoski T, Päivänsalo M, Alanen A, et al. Radiologic findings in renal oncocytoma. Acta Radiol 1991;32(5):363–367 23. Harmon WJ, King BF, Lieber MM. Renal oncocytoma: magnetic resonance imaging characteristics. J Urol 1996;155(3):863–867 24. Krestin GP. Morphologic and Functional MRI of the Kidneys and Adrenal Glands. Philadelphia: Field & Wood; 1990 25. Stanley RJ. Benign renal neoplasm. In: McClennan BL, ed. Syllabus: A Categorial Course in Genitourinary Radiology: RSNA, 1994 26. Hajdu SI, Foote FWJr. Angiomyolipoma of the kidney: report of 27 cases and review of the literature. J Urol 1969;102(4): 396–401 27. Semelka RC, Hricak H, Stevens SK, Finegold R, Tomei E, Carroll PR. Combined gadolinium-enhanced and fat-saturation MR imaging of renal masses. Radiology 1991;178(3):803–809 28. Friedman AC, Hartman DS, Sherman J, Lautin EM, Goldman M. Computed tomography of abdominal fatty masses. Radiology 1981;139(2):415–429 29. Vas W, Wolverson MK, Johnson F, Sundaram M, Salimi Z. MRI of an angioyolipoma. Magn Reson Imaging 1986;4:485–488 30. Madewell JE, Goldman SM, Davis CJJr, Hartman DS, Feigin DS, Lichtenstein JE. Multilocular cystic nephroma: a radiographicpathologic correlation of 58 patients. Radiology 1983;146(2): 309–321 31. Rha SE, Byun JY, Jung SE, et al. The renal sinus: pathologic spectrum and multimodality imaging approach. Radiographics 2004;24(Suppl 1):S117–S131 32. Harrison RB, Dyer R. Benign space-occupying conditions of the kidneys. Semin Roentgenol 1987;22(4):275–283 33. Riches EW, Griffiths IH, Thackray AC. New growths of the kidney and ureter. Br J Urol 1951;23(4):297–356 34. Wittekind Ch, Sobin LH. TNM Classification of Malignant Tumours. New York: Wiley-Liss; 2002 35. Ergen FB, Hussain HK, Caoili EM, et al. MRI for preoperative staging of renal cell carcinoma using the 1997 TNM classification: comparison with surgical and pathologic staging. AJR Am J Roentgenol 2004;182(1):217–225 36. Hallscheidt PJ, Bock M, Riedasch G, et al. Diagnostic accuracy of staging renal cell carcinomas using multidetector-row computed tomography and magnetic resonance imaging: a prospective study with histopathologic correlation. J Comput Assist Tomogr 2004;28(3):333–339 37. Kamel IR, Hochman MG, Keogan MT, et al. Accuracy of breathhold magnetic resonance imaging in preoperative staging of organ-confined renal cell carcinoma. J Comput Assist Tomogr 2004;28(3):327–332 38. Hallscheidt PJ, Fink C, Haferkamp A, et al. Preoperative staging of renal cell carcinoma with inferior vena cava thrombus using multidetector CT and MRI: prospective study with histopathological correlation. J Comput Assist Tomogr 2005;29(1):64–68 39. Kreft BP, Müller-Miny H, Sommer T, et al. Diagnostic value of MR imaging in comparison to CT in the detection and differential diagnosis of renal masses: ROC analysis. Eur Radiol 1997;7(4): 542–547 40. Eilenberg SS, Lee JK, Brown J, Mirowitz SA, Tartar VM. Renal masses: evaluation with gradient-echo Gd-DTPA-enhanced dynamic MR imaging. Radiology 1990;176(2):333–338 41. Browne RF, Meehan CP, Colville J, Power R, Torreggiani WC. Transitional cell carcinoma of the upper urinary tract: spectrum of imaging findings. Radiographics 2005;25(6):1609–1627 42. Yousem DM, Gatewood OM, Goldman SM, Marshall FF. Synchronous and metachronous transitional cell carcinoma of the urinary tract: prevalence, incidence, and radiographic detection. Radiology 1988;167(3):613–618 43. Barentsz JO, Ruijs SH, Strijk SP. The role of MR imaging in carcinoma of the urinary bladder. AJR Am J Roentgenol 1993;160(5):937–947 44. Krestin GP. Magnetic resonance imaging of the kidneys: current status. Magn Reson Q 1994;10(1):2–21 45. Belt TG, Cohen MD, Smith JA, Cory DA, McKenna S, Weetman R. MRI of Wilms’ tumor: promise as the primary imaging method. AJR Am J Roentgenol 1986;146(5):955–961 46. Müller MF, Krestin GP, Willi UV. [Abdominal tumors in children. A comparison between magnetic resonance tomography (MRT) and ultrasonography (US)]. Rofo 1993;158(1):9–14

MRI Appearance of Pathologic Entities 47. Choyke PL, White EM, Zeman RK, Jaffe MH, Clark LR. Renal metastases: clinicopathologic and radiologic correlation. Radiology 1987;162(2):359–363 48. Hauser M, Krestin GP, Hagspiel KD. Bilateral solid multifocal intrarenal and perirenal lesions: differentiation with ultrasonography, computed tomography and magnetic resonance imaging. Clin Radiol 1995;50(5):288–294 49 Reznek RH, Mootoosamy I, Webb JA, Richards MA. CT in renal and perirenal lymphoma: a further look. Clin Radiol 1990;42(4):233–238 50. Krestin G, Fischbach R, Vorreuther R, von Schlthess GK. Alterations in renal morphology and functionafter ESWL therapy: evaluation with dynamic contrast-enhanced MRI. Eur Radiol 1993;3:227–233 51. Haustein J, Niendorf HP, Krestin G, et al. Renal tolerance of gadolinium-DTPA/dimeglumine in patients with chronic renal failure. Invest Radiol 1992;27(2):153–156 52. Rofsky NM, Weinreb JC, Bosniak MA, Libes RB, Birnbaum BA. Renal lesion characterization with gadolinium-enhanced MR imaging: efficacy and safety in patients with renal insufficiency. Radiology 1991;180(1):85–89 53. Buonocore MH, Katzberg RW. Estimation of extraction fraction (EF) and glomerular filtration rate (GFR) using MRI: considerations derived from a new Gd-chelate biodistribution model simulation. IEEE Trans Med Imaging 2005;24(5):651–666

179

54. Levin YS, Chow LC, Pelc NJ, Sommer FG, Spielman DM. Estimation of renal extraction fraction based on postcontrast venous and arterial differential T1 values: an error analysis. Magn Reson Med 2005;54(2):309–316 55. Giessing M, Deger S, Schönberger B, Türk I, Loening SA. Laparoscopic living donor nephrectomy: from alternative to standard procedure. Transplant Proc 2003;35(6):2093–2095 56. Kadir P. Atlas of normal and variant angiographic anatomy. Philadelphia: WB Saunders Company; 1991 57. Giessing M, Kroencke TJ, Taupitz M, et al. Gadolinium-enhanced three-dimensional magnetic resonance angiography versus conventional digital subtraction angiography: which modality is superior in evaluating living kidney donors? Transplantation 2003;76(6):1000–1002 58. Israel GM, Lee VS, Edye M, et al. Comprehensive MR imaging in the preoperative evaluation of living donor candidates for laparoscopic nephrectomy: initial experience. Radiology 2002; 225(2):427–432 59. Hanna S, Helenon O, Legendre C, et al. MR imaging of renal transplant rejection. Acta Radiol 1991;32(1):42–46 60. Hélénon O, Attlan E, Legendre C, et al. Gd-DOTA-enhanced MR imaging and color Doppler US of renal allograft necrosis. Radiographics 1992;12(1):21–33

181

8

The Adrenal Glands M. Taupitz and G.P. Krestin

Introduction

Imaging Technique

Despite their small size and total weight of only 11 g, the adrenal glands are complex hormone-producing organs. As such, they are affected by dysregulation of the endocrine system, resulting in clinically manifest hormonal disorders and morphologic changes that can be detected by imaging. In addition, several nonfunctioning benign and malignant tumors can arise in the adrenal glands. Previously undetected adrenal adenomas are identified in 2–10 % of the population, and up to 26 % of patients with advanced extra-adrenal primary malignancies have adrenal metastases. Adrenal masses that can be demonstrated by imaging are present in ca. 9 % of individuals.1,2 Unsurprisingly, adrenal incidentalomas are common findings on abdominal imaging studies performed for other reasons. Characterization of such incidental adrenal masses is especially important in patients with a known extra-adrenal malignancy as it is likely to affect their further management.

See the respective sections in Chapter 1 for patient preparation, positioning, and coils. The usual axial sequences should be supplemented by a coronal sequence to define the relationship to the diaphragmatic crura. An additional sagittal sequence may be necessary to separate an adrenal lesion from the upper renal poles.

Indications The adrenal glands are visualized and evaluated in all MRI examinations of the abdomen or upper abdomen. Scrutinizing the adrenal glands for the presence of metastasis is especially important in patients who undergo MRI for a suspected or known malignancy. A dedicated MRI examination of the adrenal glands is indicated in the following situations: · Known extra-abdominal malignancy and indeterminate adrenal mass detected by other imaging modalities. · Characterization of renal incidentaloma. · Diagnostic work-up of patients in whom an adrenal mass is suspected on clinical grounds, e. g., hyperaldosteronism, Cushing syndrome, or hypertension.

Pulse Sequences If the adrenal glands are to be included in an MRI examination of an upper abdominal organ (e. g., kidneys, pancreas, liver), it is important to acquire both in-phase (IP) and opposed-phase (OP) T1w GRE images (Fig. 8.1). Otherwise, the pulse sequences used for MRI of the liver or the kidneys (see the respective sections in Chapters 1 and 7) are also suitable for adrenal MRI. The respiratorytriggered, fat-suppressed axial T2w TSE sequence suggested for upper abdominal MRI also yields images with excellent detail of the adrenal glands. Additional coronal and sagittal images should be acquired with a single-shot T2w TSE sequence (e. g., HASTE), which is characterized by short scan times and robustness to artifacts (Fig. 8.2, Table 8.1).

p

Fig. 8.1 Principle of opposed-phase (OP) and in-phase (IP) imaging. Because water and fat protons have slightly different precession frequencies, the amplitude of the echo in a GRE sequence is modulated (“beats”) in an environment consisting of a finely dispersed mixture of water and fat (as in a typical adrenal adenoma). Under OP conditions, the signal contributions of water and fat are subtracted in a volume element, resulting in low SI (left). Under IP conditions, the signal contributions are added together, resulting in high SI (right).

182

Slice distance, 10–20 % of slice thickness (distance factor, 0.1–0.2) The TEs given for acquiring IP and OP 2D GRE images are valid for 1.5 T; TEs for other field strengths are listed in Table 1.3. Note: The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time (for sequences with one signal average) but may come with a penalty in SNR.

Yes Yes No No 23 s 23 s 4–7 min 5–7 min 5 5 7 4 23 19 21 48 ∞ ∞ 5000 2500 T2 T2 T2 T2

HASTE HASE TSE TSE, respiratorytriggered

60–80 60–80 80–100 80

– – – –

Fixed Fixed 7–15 7–15

Yes/no Yes/no Yes Yes

116 × 256 192 × 256 128 × 256 168 × 320

300 (75 %) 300 (100 %) 300 (75 %) 300 (75 %)

1 1 3 2

Yes Yes Yes No 23 s 23 s 20–24 s 4–8 min 3–5 3–5 2.5 6 1 1 1 4 23 23 64 19 300 (75 %) 300 (75 %) 300 (75 %) 300 (75 %) ca. 170–200 ca. 170–200 5–7 500 2D GRE 2D GRE 3D GRE (VIBE) SE or TSE

Axial Axial Axial Axial (alternative) Axial Coronal Axial (optional) Axial (alternative) T1 IP T1 OP T1 FS T1

4.4–4.8 2.2–2.4 2.2–2.6 10–15

90 90 10 –

– – – –

No No Yes No

116 × 256 116 × 256 116 × 256 128 × 256

No. of slices TR (ms) Sequence type Plane Weighting

Table 8.1 Recommended pulse sequences and imaging parameters

TE (ms)

Flip (°)

ETL

FS

Matrix

FOV (mm)

No. of acquisitions

Slice thick- Scan ness (mm) time

Breathhold

8 The Adrenal Glands

To further characterize a very small suspicious adrenal mass, a set of T1w and T2w sequences with a slice thickness not exceeding 3–5 mm and a maximum interslice gap of 20 % should be acquired. The resolution in the phase-encoding and frequency-encoding directions should be less than 1.5 mm, which is typically achieved using a field of view (FOV) between 28 and 36 cm, according to patient habitus, and a 192 × 256 matrix. The acquisition of IP and OP images for identifying intracellular lipid is frequently referred to as chemical shift imaging (CSI) and must not be confused with spatially resolved MR spectroscopy, which is also designated as CSI. The latter is not used for MRI of the adrenal glands. IP and OP images can be acquired with two successive sequences or simultaneously using a double-echo sequence, which is available on newer scanners. Comparison of IP and OP signal intensities is especially important for evaluation of adrenal masses because it allows detection of finely dispersed intracellular lipid, which is highly characteristic of adrenal adenoma and rules out a malignant tumor with a high degree of likelihood.3–5

Contrast Media The precontrast sequences can be supplemented by a dynamic series performed after intravenous injection of nonspecific Gd-based contrast medium (e. g., Dotarem, Magnevist) at a standard dose of 0.1 mmol Gd per kg body weight. For details on nonspecific MR contrast media and the dynamic study protocol see the respective sections in Chapter 1. The postcontrast series is acquired using a conventional T1w GRE sequence with an IP echo time. Alternatively, the dynamic series can be performed using a 3D GRE sequence (e. g., VIBE). Images are acquired at the usual postcontrast time points for upper abdominal MRI: arterial phase at 15 s, portal venous phase at 55–60 s, and delayed phases at 2 and 5 min, optionally also at 10 min. If a double-echo T1w GRE sequence is available for IP/OP imaging, it can also be used for the dynamic series. Lipid-containing tumors such as adrenal adenoma may fail to enhance or even decrease in signal intensity on OP images acquired after contrast administration,6 while they increase in signal intensity on IP images. This paradoxical phenomenon is attributed to the fact that the contrast medium predominantly distributes in the watery compartment, resulting in a shift in signal contributions from fatty and watery compartments on OP images. No diagnostic benefit of this effect is known.

Image Analysis Determination of signal intensities on IP and OP images as well as on T2w images plays an important role in the MRI characterization of adrenal masses. Various quantitative methods have been proposed for region of interest (ROI) measurements in adrenal masses and, depending on the

Imaging Technique

183

a

b

c

d Fig. 8.2a–d Normal adrenal glands. a, b Breath-hold axial (a) and coronal (b) T2w HASTE images. c, d IP (c) and OP (d) images acquired simultaneously using a breath-hold axial T1w double-echo GRE sequence (1.5 T). Both adrenal glands are depicted as delicate structures.

method, in an adjacent reference tissue.4 Quantitative analysis is time-consuming and the results depend on the magnetic field strength and other technical parameters; published results and cut-off values for differentiating adenoma from malignant adrenal tumors may therefore not be directly applicable to every situation. Moreover, regressive changes and hemorrhage can alter the signal intensities of both benign and malignant tumors. Quantitative approaches were primarily pursued and favored in the early era of adrenal MRI, but are no longer considered to be superior to simple visual analysis. Some of the formulas proposed are briefly outlined below. The ROIs used for measuring signal intensity are placed in a homogeneous, solid area of an adrenal lesion and should be as large as possible while excluding partial volume effects and motion artifacts. Tsushima et al. proposed an index of signal intensities (SI) calculated from IP and OP images as follows7: Index = ([SIIP – SIOP]/SIIP) × 100 %. An index > 5 % is assumed to indicate a significant lipid component. A more straightforward way to quantify sig-

nal intensity is to calculate the percentage signal difference (SI%) between OP and IP images: SI% = 100 – (SIOP × 100/SIIP). An SI difference < 5 % calculated according to this equation is assumed to rule out a significant lipid component in an adrenal mass; a difference > 25 % is taken as positive evidence that a lesion contains lipid. Other SI ratios were investigated by Fujiyosji and coworkers and found to be of little diagnostic value.4 Other techniques use T2w signal intensities to separate benign and malignant adrenal lesions.8,9 One approach is to relate the signal intensity of an adrenal tumor on fatsuppressed T2w images (SIT) to that of a comparable area in the back muscle (SIM), the spleen (SIS), or the liver (SIL). Relative signal intensity can thus be calculated for interindividual comparison (e. g., SIrel = SIT/SIM). Because MR signal intensities of the liver and spleen are often altered by diffuse conditions (steatosis) or storage diseases (hemosiderosis), and because the fat signal is reduced to variable degrees when a fat suppression technique is used, SIrel should be calculated in relation to muscle. An

184

8 The Adrenal Glands

SIrel cut-off value of 3.5 was found to be highly sensitive and specific for the identification of adrenal adenomas. If dynamic contrast-enhanced images are used to characterize an adrenal mass, signal intensities are determined at different time points after contrast administration (SIt) and compared with the precontrast signal intensity (SIpre). Dynamic contrast ratios (SIdyn = SIt/SIpre) can thus be plotted over time. According to Krestin and coworkers,10 an early and persistent signal increase (SIdyn > 2) characterizes a malignant adrenal mass, whereas moderate enhancement (SIdyn < 2) with complete washout after 10 min indicates a benign lesion. The cut-off values for this method vary with the field strength, scanner, and sequence used. With some experience and training, every radiologist can acquire the skill to analyze images qualitatively and achieve the same results as with a quantitative analysis. The presumptive diagnosis of a benign tumor can be established if an adrenal mass or most of it appears markedly hypointense on OP images compared with the IP images. On T2w images, very high signal intensity is suggestive of a malignant adrenal tumor. On dynamic contrast-enhanced images, a moderate increase in signal intensity, which drops again within a few minutes, suggests a benign mass, whereas intense and persistent signal enhancement indicates a malignant process.

Anatomy and Normal MRI Appearance The adrenal glands are located lateral to the vertebral bodies at the level of the 11th to 12th rib and are enclosed by the upper portion of the Gerota fascia. They have the shape of an inverted Y or T. Embedded in perirenal fat, the adrenals are clearly demarcated from surrounding structures. The right adrenal gland lies above the upper renal pole, medial to the posterior contour of the right hepatic lobe, and lateral to the right crura of the diaphragm. Anteriorly, it borders on the inferior vena cava. The left adrenal gland comes to lie anteromedial to the upper pole of the left kidney, lateral to the left crura of the diaphragm, posterolateral to the aorta, and posteromedial to the splenic vessels and pancreas. On MRI, the adrenal glands are clearly demarcated from surrounding high-signal-intensity retroperitoneal fat on non-fat-suppressed T1w and T2w images. In most patients, the two glands can be easily identified on MRI. When little or no intra-abdominal fat is present, as in very slim or cachectic patients, identifying the adrenal glands may be difficult. They may also be difficult to detect in extreme hepatomegaly or when an adjacent mass such as a renal tumor is present. In patients with portal hypertension, a dilated left diaphragmatic vein may mimic thickening at the anterior border of the left adrenal. The true nature of vascular structures is revealed when flowrelated signal loss can be demonstrated, which is most obvious on single-shot TSE sequences.

On all conventional pulse sequences, a normal adrenal gland is of uniform low to intermediate signal intensity, slightly above that of paravertebral muscles and slightly below that of the liver and renal cortex. As the adrenal gland contains fat, it loses signal intensity on OP images compared with IP images.5 Following administration of nonspecific Gd-based contrast medium, the adrenal glands enhance early and washout within ca. 5 min.11

MRI Appearance of Pathologic Entities The first sign of a pathologic process of an adrenal gland is a pronounced contour distortion. Small adrenal lesions (< 1 cm) can be detected by MRI using surface coils and breath-hold acquisition to eliminate artifacts. Adrenal tumors can arise in the cortex or medulla and cause no overt clinical symptoms if they are nonfunctioning. Functioning masses become apparent due to clinical manifestations associated with hormonal hypersecretion or hyposecretion. Primary or secondary functioning adrenal masses can be diagnosed with clinical tests. MRI is not a first line modality in this setting but may occasionally provide useful findings for further treatment. Characterizing a nonfunctioning adrenal mass is a complex task. Such lesions are incidentally detected on abdominal CT scans in ca. 0.6–1.3 % of patients.12 At autopsies, incidentalomas are detected in 1.4–64.5 % of cases.13 They can range in size from 0.5 to 6.0 cm and are typically nonfunctioning adenomas. In patients with a known metastatic tumor such as lung cancer, it is important to establish the benign character of an incidentally detected adrenal mass and rule out a metastasis.

Benign Conditions Adrenocortical Insufficiency Adrenocortical insufficiency, or Addison disease, is either a primary condition or occurs secondary to an imbalance of the hypothalamic axis. It is now widely believed that the underlying mechanism is an idiopathic autoimmune process in which the adrenal glands are normal or almost normal in size.14 Destruction of the adrenal cortex can be due to inflammatory diseases such as tuberculosis or histoplasmosis, metastases, hemorrhagic infarction, and storage diseases.15 If the underlying cause is a mass (tuberculosis, metastasis), the adrenal gland is enlarged in the shape of that mass. Adrenal insufficiency becomes clinically apparent if at least 90 % of the adrenal tissue is lost.15

Inflammatory Conditions Before the era of tuberculostatic therapy, adrenal tuberculosis was a common cause of adrenal insufficiency. Inflammatory involvement of the adrenal glands has again

MRI Appearance of Pathologic Entities

185

become more common as there are now more immunocompromised individuals. The most common pathogens are Mycobacterium tuberculosis and Histoplasma capsulatum; adrenal infection with Blastomyces and Cryptococcus is seen occasionally. Concomitant involvement of other organs is common.16 Adrenal inflammation is usually symmetric and involves both glands. Inflammatory lesions typically appear inhomogeneous with a central area of liquefaction (abscess). Because they contain no intracellular lipid (identical signal intensities on IP and OP images), they cannot be distinguished from adrenal malignancies based on their MRI appearance. Solid components typically have low signal intensity on T2w images (SIrel < 3.5 compared with paravertebral muscle). On contrast-enhanced dynamic series, inflammatory lesions tend to enhance early and wash out rapidly.

a

Adrenocortical Hyperplasia and Hypersecretion Adrenocortical hyperplasia is difficult to diagnose on the basis of morphologic appearance, as no adrenal changes are apparent in most cases. The condition is suggested if there is diffuse enlargement with preservation of the normal adreniform shape (Fig. 8.3). Imaging is indicated only to confirm or rule out an adenoma; however, macronodular hyperplasia (which is primarily seen in hyperaldosteronism) cannot always be distinguished from a small adenoma. In such cases, functional scintigraphy or venous blood sampling is necessary. Endogenous hypercortisolism (Cushing syndrome) is caused by excessive ACTH secretion due to dysregulation of the hypothalamus–pituitary–adrenal axis in 80 % of cases and is associated with bilateral adrenal hyperplasia. Fewer than 20 % of patients have ectopic paraneoplastic ACTH secretion in the setting of an extra-adrenal primary tumor (lung cancer, breast cancer, melanoma). About 20 % of adult patients have the ACTH-independent form of Cushing syndrome, caused by a cortisol-producing tumor of the adrenal cortex, which is malignant in one third of cases (adrenal carcinoma). Primary hyperaldosteronism is caused by an aldosterone-producing adenoma of the adrenal cortex in 80 % of affected patients. The remaining 20 % of cases are accounted for by hyperplasia of the zona glomerulosa. Carcinoma is very rare (< 1 %). Adenoma is usually unilateral and more common in women and typically occurs between 30 and 50 years of age.15 Adrenocortical hyperplasia is bilateral and has no sex predilection. An autonomous adenoma is usually treated surgically, while hyperplasia is amenable to medical treatment. Excess androgen production typically occurs on the basis of congenital bilateral adrenocortical hyperplasia secondary to an enzyme defect of steroid synthesis. Twenty percent of patients have an autonomous adrenal tumor, which can occur at any age and affects men and women alike. The majority of adenomas are benign, although up to 25 % can be malignant.

b

c Fig. 8.3a–c Bilateral adrenal hyperplasia. a Breath-hold coronal T2w HASTE image. b, c IP (b) and OP (c) images acquired simultaneously using a breath-hold axial T1w double-echo GRE sequence (1.5 T). Normal adreniform shape is barely preserved and the adrenal limbs are thickened with additional nodular appearance of the left adrenal. Signal drop of both adrenal glands on the OP image (c) compared with the IP image (b).

186

8 The Adrenal Glands

Cysts Adrenal cysts are rare lesions that may occur at any age, with a female-to-male ratio of 3:1.17 They are of variable size and both adrenals are affected in 15 % of cases. Four histologic types are distinguished. The most common are endothelial cysts, accounting for 45 % of adrenal cysts. They are thought to be due to obstructed lymphatic drainage. About 39 % are pseudocysts, which represent residues of hemorrhage or degenerated adrenal neoplasms. Epithelial cysts (9 %) are rare and comprise cystic adenomas. The last group are parasitic cysts (7 %), which are caused by echinococcal infection. At MRI, adrenal cysts are characterized by low signal intensity on T1w images and high intensity on T2w images15 (Fig. 8.4), and they do not enhance on dynamic contrast-enhanced sequences. Unlike renal cysts, adrenal cysts often have thickened walls.

a

Adrenocortical Adenoma

b

c Fig. 8.4a–c Cyst of the left adrenal gland. a, b Breath-hold axial (a) and coronal (b) T2w HASTE images. c Breath-hold axial T1w GRE image (1.5 T). The adrenal cyst is depicted as a smoothly delineated mass anterosuperior to the left upper kidney pole. The cyst is of homogeneously high T2 SI and low T1 SI.

Adrenocortical adenomas are benign tumors that arise from adrenocortical cells. Histologically, they are predominantly composed of lipid-laden spongiocytes arranged in trabecular fashion or as solid clusters. They have a fibrous capsule and are clearly separated from the surrounding adrenal gland. MRI depicts adrenal adenomas as homogeneous masses ranging in size from < 1 cm to 8 cm. The majority of nonfunctioning adenomas are < 3 cm. While nonfunctioning adenomas have relaxation times comparable to those of the normal adrenal gland and liver,18,19 some hormonally active adenomas, in particular those producing aldosterone, have a moderately increased signal intensity on T2w images.14 However, there are also functioning tumors that are isointense to liver and nonfunctioning adenomas that are hyperintense to liver. A higher signal intensity is usually due to hemorrhagic infarction and necrosis.15 Hence, MRI does not enable reliable differentiation of functioning and nonfunctioning adrenal adenomas. Most adrenal adenomas can be differentiated from malignant tumors on T2w images. Regardless of whether an adenoma is functioning or nonfunctioning, SIrel relative to muscle is typically < 3.5. Using IP/OP imaging, adenomas can be identified with a high degree of accuracy. Because of their lipid content, adrenal adenomas display the expected decrease in signal intensity of > 25 % on OP images compared with IP images4,5,7,20 (Fig. 8.5). After contrast administration, adenomas enhance early and wash out rapidly.10 Larger adrenal adenomas may appear inhomogeneous on T2w images, OP images, and contrastenhanced images. Identification of an intralesional area clearly containing lipid on the basis of the OP signal intensity establishes the diagnosis of an adenoma (Fig. 8.6).

MRI Appearance of Pathologic Entities

187

a

b

c

d

e

f Fig. 8.5a–f Typical MRI appearance of a nonfunctioning adrenal adenoma on the left. a, b Breath-hold axial (a) and coronal (b) T2w HASTE images. c–f IP and OP images acquired simultaneously using a breath-hold axial T1w double-echo GRE sequence (1.5 T): IP (c) and OP (d) images acquired before contrast administration and IP (e) and OP (f) images acquired 1 min after IV injection of Gd-

based contrast medium. The markedly lower SI of the adrenal mass on unenhanced T1w OP image (d) compared with the IP image (c) is consistent with intracellular lipid and is typical of adrenal adenoma. The lack of high-SI areas on the T2w images indicates absence of regressive changes. The mass shows only little enhancement on postcontrast IP (e) and OP (f) images (see Fig. 8.10).

188

8 The Adrenal Glands

a

b

c

d Fig. 8.6a–d Atypical, nonfunctioning adrenal adenoma. a Breathhold axial T2w HASTE image. b, c IP (b) and OP (c) images acquired simultaneously using an axial T1w double-echo GRE sequence (1.5 T). d Fat-suppressed T1w GRE image obtained 2 min after IV injection of Gd-based contrast medium (1.5 T). T2w image shows inhomogeneous mass with areas of high and low SI. Nonenhanced T1w OP image (c) shows only small areas of signal loss compared

with IP image (b), consistent with a mass containing only little lipid (arrow in c). Contrast-enhanced image shows inhomogeneous, strong enhancement of the mass, sparing rounded intralesional areas (arrow in d). The histologic diagnosis was adrenal adenoma with numerous intralesional hemorrhagic foci. Images also show patchy hepatic steatosis and a liver cyst.

Myelolipoma

Hemangioma

Myelolipoma is a rare adrenal tumor that is composed of mature fatty tissue and hematopoietic tissue in variable proportions. It is usually detected incidentally. Large adrenal myelolipomas can cause abdominal signs and symptoms and may be complicated by spontaneous hemorrhage. The MRI appearance is determined by the large component of pure fatty tissue, resulting in very high signal intensity on non-fat-suppressed T1w and T2w sequences and low signal intensity on the corresponding fat-suppressed sequences. Because of their large fatty component, myelolipomas are of very high signal intensity on OP images. A myelolipoma containing connective tissue or a larger component of hematopoietic tissue appears inhomogeneous (Fig. 8.7).3

Adrenal hemangioma is rare and often an incidental finding. A cavernous hemangioma can cause clinical signs and symptoms based on size, which may be up to 25 cm.21 Adrenal hemangiomas are of uniform high signal intensity on T2w images and have the same signal intensity on IP and OP images. In a dynamic contrast-enhanced study, progressive centripetal enhancement, as in liver hemangioma, can occur, especially in larger hemangiomas. Confident characterization by MRI poses a challenge because adrenal hemangiomas lack two criteria of benign adrenal tumors—rather low T2 signal intensity and signal loss on OP images (Fig. 8.8).

MRI Appearance of Pathologic Entities

189

a

b

c

d Fig. 8.7a–e Myelolipoma of the right adrenal gland. a, b Breathhold axial (a) and coronal (b) T2w HASTE images. c, d IP (c) and OP (d) images acquired simultaneously using an axial T1w double-echo GRE sequence. e Breath-hold, fat-suppressed axial T1w GRE image (1.5 T). The mass is predominantly composed of fat, indicated by high SI on the non-fat-suppressed images (a–d). On the OP image, only the intralesional areas that still contain some residual nonfatty tissue (arrow in d) show a loss of SI compared with the IP image (c). The fat-suppressed image (e) shows nearly complete loss of signal of the lesion.

e

Malignant Tumors Primary Adrenocortical Carcinoma With an incidence of ca. 1 per 1 million population, adrenocortical carcinoma is a rare tumor.15 The mean age at presentation is 50 years, but the tumor can occur at all ages. Over 90 % of adrenocortical carcinomas reported had a size of > 6 cm. Involvement of the left adrenal gland is more common, and 10 % of adrenocortical carcinomas are bilateral. About half are functioning carcinomas (hypercorticoidism, virilization).22

The MRI appearance varies with size and vascularization. Small tumors (< 5 cm in diameter) tend to appear uniform, larger ones often have areas of necrosis and hemorrhagic infarction (Fig. 8.9). Infiltration of adjacent organs is common. Vascular invasion, especially of the vena cava, is best demonstrated on flow-sensitive GRE sequences or an MRA sequence (see Chapter 15). Adrenocortical carcinomas typically have high signal intensity on T2w images (SIrel relative to muscle > 3.5). Unlike adenoma, an adrenocortical carcinoma contains no lipid, and therefore the decrease in signal intensity on OP images vs IP images is less than 5 %. Because of their

190

8 The Adrenal Glands

a

b

c

d Fig. 8.8a–d Hemangioma of the left adrenal gland (arrow) in a patient with a large renal cell carcinoma of the right kidney (curved arrow). a Breath-hold axial T2w HASTE image. b, c IP (b) and OP (c) images acquired simultaneously using a breath-hold axial T1w double-echo GRE sequence. d T1w IP image acquired 2 min after IV injection of Gd-based contrast medium (1.5 T). The adrenal hemangioma is markedly hyperintense on the T2w image (a) and

markedly hypointense on the precontrast T1w images, without any difference between IP (b) and OP (c) intensities. Uniform high SI of the adrenal hemangioma on postcontrast image (d). Despite the typical MRI findings of an adrenal hemangioma, the left adrenal gland was removed because of the rarity of this lesion and suspected metastasis in this patient with a large renal cell carcinoma.

hypervascularity, adrenocortical carcinomas show intense enhancement within 1 min of intravenous injection of nonspecific contrast medium and persistent high signal intensity for > 10 min.23

Lymphoma

Metastases The adrenals are a common site of metastatic disease. Primary tumors that tend to metastasize to the adrenal glands are lung and breast cancer, melanoma, gastrointestinal tumors, medullary thyroid carcinoma, and pancreatic cancer.15 In an autopsy study, up to 27 % of patients with a malignant tumor were found to have adrenal metastases.16 Containing no intracellular lipid, metastases have similar signal intensities on IP and OP images (Fig. 8.10).

Adrenal involvement is more common in non-Hodgkin lymphoma compared with Hodgkin lymphoma. Primary adrenal lymphoma is rare.16 Most patients have concomitant retroperitoneal lymphomatosis. Lymphoma rarely causes adrenal insufficiency, although the condition typically involves both glands.15 MRI cannot distinguish adrenal lymphoma from other malignant tumors of the adrenal glands. The lesions have increased T2 signal intensity24 with marked and prolonged signal enhancement on postcontrast images.10

Pheochromocytoma Pheochromocytoma is a functional chromaffinoma of the adrenal medulla, and patients may present with signs and symptoms of excessive catecholamine secretion (hyper-

MRI Appearance of Pathologic Entities

a

191

b Fig. 8.9a–c Adrenocortical carcinoma. a Breath-hold axial T2w HASTE image. b, c Breath-hold axial T1w GRE images acquired before (b) and 2 min after (c) IV injection of Gd-based contrast medium (1.5 T). Giant mass in the left upper abdomen (arrows). Central necrotic areas of high SI on T2w image without enhancement on postcontrast image.

c

a

b Fig. 8.10a–c Metastasis in the right adrenal gland (from transitional cell carcinoma). a Breath-hold axial T2w HASTE image. b, c IP (b) and OP (c) images acquired simultaneously using a breath-hold axial T1w double-echo GRE sequence (1.5 T). Smoothly demarcated mass in the right adrenal gland of uniform SI on all pulse sequences and identical SI on IP and OP images indicate that the mass contains no lipid (see Fig. 8.5).

c

192

8 The Adrenal Glands

a

b

c

d Fig. 8.11a–d Pheochromocytoma in the right adrenal gland. a Breath-hold axial T2w HASTE image. b Respiratory-triggered, fatsuppressed axial T2w TSE image. c, d IP (c) and OP (d) images acquired simultaneously using a breath-hold axial T1w double-

echo GRE sequence (1.5 T). Smoothly demarcated mass in the right adrenal gland of uniform SI on all pulse sequences (typical high T2 SI and low T1 SI) and identical SI on IP and OP images indicate that the mass contains no lipid.

tension). The diagnosis can be established by the demonstration of catecholamines or their metabolites in urine or serum. Pheochromocytomas tend to occur in certain familial syndromes including multiple endocrine neoplasia (MEN) 2A and 2B, neurofibromatosis, and von Hippel–Lindau disease.25 About 65–100 % of patients with MEN 2A or 2B have pheochromocytomas, typically in both adrenal glands (80 %), and ca. 10 % of those with neurofibromatosis. Approximately 15 % of pheochromocytomas are located outside the adrenal gland and are termed paragangliomas. All pheochromocytomas occurring in the setting of an endocrine syndrome are intra-adrenal. Most pheochromocytomas are benign, only ca. 13 % are malignant.15 Malignant transformation is more common in extra-adrenal pheochromocytomas. A pheochromocytoma is typically depicted as a welldefined round or oval mass. The tumors have an average size of 5 cm and compress adjacent tissue (Fig. 8.11). Pheochromocytomas tend to be hypervascular and often hemorrhagic26 (Fig. 8.12). The MR signal intensity on T2w images is higher than that of muscle and the liver (Fig. 8.11). The appearance tends to be inhomogeneous due to necrosis with central

cystic areas (Fig. 8.12).15,19 Most pheochromocytomas lack fat and are therefore indistinct from other adrenal masses not containing cytoplasmic lipids on IP/OP images. Some pheochromocytomas with a fatty component have been reported in the literature and these have been misinterpreted as adenomas on fat-sensitive sequences.27 The contrast enhancement of pheochromocytomas resembles that of malignant adrenal tumors with an intense early signal increase and slow washout. Pheochromocytomas have been reported to enhance even more intensely than malignant adrenal tumors.23 MRI is comparable to CT and superior to MIBG scintigraphy in detecting adrenal pheochromocytoma (MRI has a higher sensitivity but lower specificity). MRI and MIBG scintigraphy are comparable regarding paragangliomas.

Neuroblastoma Neuroblastoma is the second most common tumor in childhood, accounting for 7–8 % of all neoplastic cases in this age group (see Chapter 17). It is derived from neural crest cells in the sympathetic ganglia26 and typically arises in the adrenal medulla or in the sympathetic chain.28 Most neuroblastomas secrete catecholamines or their less ac-

Rational Approach to the MRI Differentiation of Adrenal Lesions

193

a

b

c

d Fig. 8.12a–d Pheochromocytoma with intralesional hemorrhage in the left adrenal gland. a Breath-hold axial T2w HASTE image. b Respiratory-triggered, fat-suppressed axial T2w TSE image. c, d IP (c) and OP (d) images acquired simultaneously using a breath-

hold axial T1w double-echo GRE sequence (1.5 T). Mass in the left adrenal gland with liquid areas and sedimentation, consistent with intralesional hemorrhage. IP and OP signal intensities rule out a lipidcontaining mass.

tive precursors, which can be detected in urine. Neuroblastoma tends to metastasize extensively and early and often presents with symptoms of metastatic disease. Neuroblastomas are isointense or hyperintense to muscle on T1w images and typically hyperintense on T2w images. The tumors have an unchanged signal intensity on fat-saturated compared with non-fat-saturated sequences or on OP images compared with IP images because they contain no fat. After contrast injection, they enhance early and intensely. Approximately 40–50 % of neuroblastomas contain calcifications and these are difficult to detect by MRI. However, the multiplanar capability of MRI affords accurate definition of tumor extent and is helpful for staging.

with regard to their potential to provide information for characterizing adrenal masses.4,5,7,9,10,19,20,29–31 Published data on their sensitivities and specificities are variable. Acquisition of IP and OP images with evaluation of the signal change on OP images compared with IP images has shown to be highly sensitive and specific for the demonstration of cytoplasmic lipid and establishes the diagnosis of adrenal adenoma. IP and OP images will therefore also allow differentiation of benign from malignant adrenal masses in most cases. In the group of non-fat-containing tumors, only inflammatory lesions can be classified as benign on T2w images and dynamic contrast-enhanced series; for all other adrenal lesions, further differentiation by MRI is not possible (including metastases, primary adrenal carcinoma, lymphoma, pheochromocytoma). The following MRI approach is therefore recommended for the characterization of adrenal lesions (Fig. 8.13): Based on the reported 100 % sensitivity of IP/OP imaging, one may confidently establish the diagnosis of an adrenal adenoma if intracellular lipid is demonstrated using this technique, and the examination can end here. If no lipid is demonstrated in a lesion, a T2w sequence (SIrel in relation

Rational Approach to the MRI Differentiation of Adrenal Lesions The aim of MRI is to narrow the diagnosis of an adrenal lesion. Numerous pulse sequences have been investigated

194

8 The Adrenal Glands

Adrenal lesion

In-phase (IP) and opposedphase (OP) imaging Lipid+

Lipid−

Inflammation Pheochromocytoma Malignant tumor

Adenoma

T2w imaging Dynamic contrast-enhanced study T2+ CM+

+/–

T2– CM–

Malignant tumor Pheochromocytoma

Biopsy

Inflammation

Fig. 8.13 Flowchart for the MRI evaluation of adrenal lesions. If the fat-sensitive pulse sequences demonstrate intracellular lipid within an adrenal lesion, the diagnosis of adenoma is confirmed. If no fat can be demonstrated, the combination of T2w and dynamic contrast-enhanced sequences allows further differentiation of inflammatory lesions from malignant tumors and pheochromocytoma. CM = contrast medium.

to muscle < 3.5) and contrast-enhanced dynamic images (enhancement with rapid washout) can identify an inflammatory condition with a high degree of confidence. All other adrenal masses must be considered malignant tumors or pheochromocytomas until proven otherwise.

References 1. Dunnick NR, Korobkin M, Francis I. Adrenal radiology: distinguishing benign from malignant adrenal masses. AJR Am J Roentgenol 1996;167(4):861–867 2. Glazer HS, Weyman PJ, Sagel SS, Levitt RG, McClennan BL. Nonfunctioning adrenal masses: incidental discovery on computed tomography. AJR Am J Roentgenol 1982;139(1):81–85 3. Elsayes KM, Mukundan G, Narra VR, et al. Adrenal masses: mr imaging features with pathologic correlation. Radiographics 2004;24(Suppl 1):S73–S86 4. Fujiyoshi F, Nakajo M, Fukukura Y, Tsuchimochi S. Characterization of adrenal tumors by chemical shift fast low-angle shot MR imaging: comparison of four methods of quantitative evaluation. AJR Am J Roentgenol 2003;180(6):1649–1657 5. Mitchell DG, Crovello M, Matteucci T, Petersen RO, Miettinen MM. Benign adrenocortical masses: diagnosis with chemical shift MR imaging. Radiology 1992;185(2):345–351 6. Mitchell DG, Stolpen AH, Siegelman ES, Bolinger L, Outwater EK. Fatty tissue on opposed-phase MR images: paradoxical suppression of signal intensity by paramagnetic contrast agents. Radiology 1996;198(2):351–357 7. Tsushima Y, Ishizaka H, Matsumoto M. Adrenal masses: differentiation with chemical shift, fast low-angle shot MR imaging. Radiology 1993;186(3):705–709

8. Gruss LP, Newhouse JH. Eight echo T2 measurements of adrenal masses: limitations of differential diagnosis by relaxation time determination. J Comput Assist Tomogr 1996;20(5):792–797 9. Slapa RZ, Jakubowski W, Januszewicz A, et al. Discriminatory power of MRI for differentiation of adrenal non-adenomas vs adenomas evaluated by means of ROC analysis: can biopsy be obviated? Eur Radiol 2000;10(1):95–104 10. Krestin GP, Freidmann G, Fishbach R, Neufang KF, Allolio B. Evaluation of adrenal masses in oncologic patients: dynamic contrast-enhanced MR vs CT. J Comput Assist Tomogr 1991; 15(1):104–110 11. Krestin GP, Steinbrich W, Friedmann G. Adrenal masses: evaluation with fast gradient-echo MR imaging and Gd-DTPA-enhanced dynamic studies. Radiology 1989;171(3):675–680 12. Thompson NW, Cheung PS. Diagnosis and treatment of functioning and nonfunctioning adrenocortical neoplasms including incidentalomas. Surg Clin North Am 1987;67(2):423–436 13. Commons R, Callaway R. Adenomas of the adrenal cortex. Arch Intern Med 1948;81:37–41 14. Baker ME, Blinder R, Spritzer C, Leight GS, Herfkens RJ, Dunnick NR. MR evaluation of adrenal masses at 1.5 T. AJR Am J Roentgenol 1989;153(2):307–312 15. Dunnick NR. The adrenal gland. In: Taveras JM, et al., eds. Radiology: Diagnosis, Imaging, Intervention. Philadelphia: Lippincott; 1988 16. Moulton JS, Moulton JS. CT of the adrenal glands. Semin Roentgenol 1988;23(4):288–303 17. Cheema P, Cartagena R, Staubitz W. Adrenal cysts: diagnosis and treatment. J Urol 1981;126(3):396–399 18. Chang A, Glazer HS, Lee JK, Ling D, Heiken JP. Adrenal gland: MR imaging. Radiology 1987;163(1):123–128 19. Reinig JW, Stutley JE, Leonhardt CM, Spicer KM, Margolis M, Caldwell CB. Differentiation of adrenal masses with MR imaging: comparison of techniques. Radiology 1994;192(1):41–46 20. Zimmermann GG, Debatin JF, Krestin GP. The differentiation of adrenal gland tumors: an improvement in accuracy by a combination of fat-sensitive, T2-weighted and contrast-enhanced MR sequences. [Article in German] Rofo 1997;167:153–159 21. Yamada T, Ishibashi T, Saito H, et al. Two cases of adrenal hemangioma: CT and MRI findings with pathological correlations. Radiat Med 2002;20(1):51–56 22. Bodie B, Novick AC, Pontes JE, et al. The Cleveland Clinic experience with adrenal cortical carcinoma. J Urol 1989;141(2): 257–260 23. Ichikawa T, Ohtomo K, Uchiyama G, Fujimoto H, Nasu K. Contrast-enhanced dynamic MRI of adrenal masses: classification of characteristic enhancement patterns. Clin Radiol 1995;50(5): 295–300 24. Lee MJ, Mayo-Smith WW, Hahn PF, et al. State-of-the-art MR imaging of the adrenal gland. Radiographics 1994;14(5): 1015–1029, discussion 1029–1032 25. Mathieu E, Despres E, Delepine N, Taieb A. MR imaging of the adrenal gland in Sipple disease. J Comput Assist Tomogr 1987; 11(5):790–794 26. Leder LD, Richter HP. Pathology of renal and adrenal neoplasms. In: Lohr EL, Leder LD, eds. Renal and Adrenal Tumors. New York: Springer; 1987, pp.1–68 27. McLoughlin RF, Bilbey JH. Tumors of the adrenal gland: findings on CT and MR imaging. AJR Am J Roentgenol 1994;163(6): 1413–1418 28. Brady TM, Gross BH, Glazer GM, Williams DM. Adrenal pseudomasses due to varices: angiographic-CT-MRI-pathologic correlations. AJR Am J Roentgenol 1985;145(2):301–304 29. Haider MA, Ghai S, Jhaveri K, Lockwood G. Chemical shift MR imaging of hyperattenuating (> 10 HU) adrenal masses: does it still have a role? Radiology 2004;231(3):711–716 30. Korobkin M, Lombardi TJ, Aisen AM, et al. Characterization of adrenal masses with chemical shift and gadolinium-enhanced MR imaging. Radiology 1995;197(2):411–418 31. Schwartz LH, Panicek DM, Koutcher JA, Heelan RT, Bains MS, Burt M. Echoplanar MR imaging for characterization of adrenal masses in patients with malignant neoplasms: preliminary evaluation of calculated T2 relaxation values. AJR Am J Roentgenol 1995;164(4):911–915

195

9

The Retroperitoneum G. Krupski-Berdien, C.R. Habermann, and V. Nicolas

Introduction In recent years MRI has evolved into a powerful imaging tool. The technical innovations that are most relevant in the setting of retroperitoneal disease include parallel imaging techniques and improved gradient echo sequences for 3D volume acquisition (e. g., VIBE).1 These advances have been made possible by the advent of high-performance gradient systems (30–40 mT/m), ultrafast sequences that can be acquired during breath-holding, and better techniques for magnetization preparation. Of increasing importance are optimized body coils and combinations of several surface coils for imaging of the entire retroperitoneum in a single acquisition (e. g., total imaging matrix— TIM, Siemens). Taken together, these new techniques enable full retroperitoneal evaluation in a short time while virtually eliminating artifacts. MRI is now the modality of choice for imaging benign soft-tissue tumors throughout the body because it enables evaluation of a set of malignancy criteria and identification of several specific tissues (e. g., well-differentiated liposarcoma). Furthermore, MRI allows excellent definition of local tumor extent, providing crucial information for biopsy guidance and subsequent resection of tumors thought to be malignant on the basis of clinical behavior (e. g., growth) and imaging appearance. This information is important for predicting the prognosis of patients with soft-tissue tumors, which is mainly determined by the tumor-free resection margin and the histologic subtype.2 Precise evaluation of tumor extent in the retroperitoneum by MRI requires imaging in two planes, which should show the long and short axis of the tumor and take into account known patterns of spread of retroperitoneal soft-tissue tumors—these tend to grow in natural spaces such as muscle compartments and extend along anatomic pathways such as vessels or nerve sheaths. Use of standardized imaging protocols is especially important in the retroperitoneum to ensure adequate interpretation of follow-up findings obtained on the same or a different MR scanner. In addition, MR images should show anatomic landmarks such as typical bone structures, joints, and unchanging soft-tissue structures (e. g., vessel bifurcations, muscles), which can help identify the site of a disease process on serial MR studies, especially after surgical resection.

This chapter discusses MRI of those retroperitoneal tumors and diseases that are not addressed in the organspecific chapters.

Indications MRI is indicated in patients with a retroperitoneal mass suggested by clinical findings, palpation, ultrasound, or CT or in patients with suspected diffuse disease such as retroperitoneal fibrosis. It is estimated that more than half of all retroperitoneal MRI examinations are performed for softtissue tumors. CT tends to be the preferred method in patients with inflammatory disease and abscess formation because it can also be used for interventional procedures. However, MRI is preferred for the evaluation of retroperitoneal fistulas, which may occur as a rare complication of Crohn disease.

Imaging Technique All retroperitoneal MRI examinations are performed with the patient in the supine position. Intravenous access is established before imaging so that a contrast-enhanced dynamic study can be performed and an antispasmodic agent (butylscopolamine or glucagon) can be injected if necessary. In general, the retroperitoneum is imaged using a combination of spinal and abdominal phasedarray surface coils. Phased-array multicoils can be combined with fast pulse sequences and parallel imaging techniques and are preferable to the whole-body resonator.3,4 Infants are often examined using a multichannel head coil, knee coil, or flexible surface coil.5

Imaging Planes The examination usually begins with a coronal sequence to estimate the craniocaudal extent of a disease process and its topographic relationship to the iliopsoas muscle, the kidneys, and the adrenals. Axial images are useful for comparison with CT scans. The volume to be imaged and slice thickness are tailored to the specific clinical question.

Note: Alternatively, T1w images can be acquired using a TSE sequence and adjusting parameters accordingly. The suggested parameters are only examples and have to be adjusted for use on different brands of scanners.

– – – – – 4.14 min 2.11 min 3.45 min 34 s 3.10 min 6–8 8 6–8 2 6 1 1 1 1 3 21 23 19 36 21 350 360 360 360 360 384 × 512 280 × 512 300 × 512 182 × 256 330 × 512 – – + + – 15 14 13 1.81 99 TIRM SE SE 3D FLASH (VIBE) TSE Axial Axial Coronal Axial Axial/coronal

4860 601 840 5.12 5540

+ + + + 12 s 40 s 23 s 22 s 6–8 6–8 6–8 6–8 1 1 1 1 19 19 19 19 450 350 350 350 128 × 256 180 × 256 196 × 256 128 × 256 – – – + 55 101 4.76 2.47 653 3200 183 208 HASTE TSE 2D FLASH 2D FLASH Axial/coronal Axial/coronal Axial Axial

T2 T2 T1 ± IV contrast Fat suppression: T1 post IV contrast Optional T2 (see text) or optional T1 (see text) T1 with fat suppression T1 T2

Sequence type Plane

TR (ms)

TE (ms) FS

Matrix

FOV (mm)

No. of slices

No. of acquisitions

Slice thickness (mm)

Scan time

Breath-hold

9 The Retroperitoneum

Weighting

Table 9.1 Recommended pulse sequences and imaging parameters for MRI of the retroperitoneum (parameters for imaging at 1.5 T, gradient strength of 30 mT/m, phased-array coils, parallel imaging)

196

As a rule of thumb, a slice thickness of 4–10 mm (with no interslice gap) is chosen, depending on tumor size.

Pulse Sequences A T2w HASTE sequence acquired during breath-hold allows rapid documentation of tumor extent in the retroperitoneum. Breath-hold imaging minimizes artifacts and will improve conformity with the prescribed slice locations. However, fast pulse sequences accentuate the fat signal compared with conventional sequences, making it more difficult to differentiate between fat and fluid. This problem can be overcome by using a fat suppression technique, which enables good separation of retroperitoneal fat and fluid (e. g., inversion recovery sequence or T2w sequence with spectral fat saturation). Provided that the necessary equipment is available, T1w imaging before and after contrast administration should be performed with a gradient echo (GRE) sequence (e. g., 2D FLASH), which allows acquisition of larger slabs during a breath-hold of less than 30 s. GRE sequences are not only faster than spin echo (SE) sequences but also allow good identification of heterotopic ossification or hemorrhage as sources of artifacts distorting field homogeneity. Most indications for retroperitoneal MRI require intravenous contrast administration for improved delineation of tumor borders and internal architecture. For comparability, postcontrast imaging should be performed with the same parameters as the precontrast T1w series (Table 9.1). If a 2D FLASH sequence is used, it can be repeated with fat suppression after contrast administration, typically with acquisition of two phases. Alternatively, a 3D GRE sequence (VIBE) can be used, which combines fast acquisition with high resolution, but, owing to inherent fat saturation, provides no information on the presence of fat. If T1w imaging during breath-hold is precluded for technical reasons, an SE or GRE sequence can be used, whereas T2w images should always be acquired with a fast SE (FSE or TSE) sequence. A T2*w sequence may be useful in patients with a rapidly growing tumor and suspected intralesional hemorrhage or to confirm a suspected hematoma. These pulse sequences can also be used for imaging below the iliac crest, where artifacts due to respiratory motion are negligible. Details of the imaging parameters are summarized in Table 9.1.

Contrast Media An MRI examination for suspected retroperitoneal pathology is not complete without intravenous administration of contrast medium (e. g., Magnevist, Dotarem) at a standard dose of 0.1 mmol Gd per kg body weight. Newer contrast agents, in particular ultra-small superparamagnetic iron oxide particles (USPIO), have not yet proven beneficial in this setting, except for lymph node imaging.

Classification of Soft-Tissue Tumors

Using the techniques described above the radiologist can fully exploit the advantages of MRI, i. e., its exquisite soft-tissue contrast and multiplanar capability, and provide important information to improve planning of preoperative chemotherapy and/or radiotherapy in patients with retroperitoneal disease.

Anatomy The retroperitoneum is the connective tissue space behind the peritoneal cavity. Posteromedially, the lumbar spine and the psoas muscles on either side protrude into this space. Medially within the retroperitoneal space lie the abdominal aorta, the inferior vena cava, the ascending lumbar vein, portions of the sympathetic trunk, networks of vegetative nerves, and the chyle cistern. Lateral to the paired psoas muscle, the retroperitoneum widens to enclose the renal compartments. Through the posterolateral border of the retroperitoneum pass the lumbar plexus nerve, the subcostal nerve, and the segmental blood vessels and lymphatics. The retroperitoneum is bounded posterosuperiorly by the fascial sheath of the diaphragm and inferiorly by the peritoneal reflection onto the rectum and inguinal ligament. At its largest extent in the upper abdomen, the retroperitoneal space is subdivided into three compartments: the anterior pararenal space between the peritoneum and the anterior renal fascia (Gerota fascia), the perinephric space between the anterior renal fascia and the posterior renal fascia (Zuckerkandl fascia), and the posterior pararenal space between the posterior renal fascia and the fascial sheath of the autochthonous back muscles arising from the transversalis fascia. The fascia are very accurate anatomic boundaries but are not consistently visualized by MRI in healthy individuals, while they are often clearly delineated in the presence of inflammation.

Classification of Soft-Tissue Tumors Soft-tissue tumors are neoplasms of mesenchymal origin that arise in the soft tissues of the extremities or in any other organ. Since connective tissue cells retain their ability to divide, these tumors can differentiate into various types of mesenchymal tissue elements (e. g., connective tissue, fat, muscle, cartilage, bone, nerve tissue, synovial tissue). The most widely used classification of soft-tissue tumors is the one by Enzinger and Weiss,6 which is based on the different mesenchymal tissue types imitated by these tumors. This classification distinguishes benign, malignant, and indeterminate soft-tissue tumors on the basis of their biological behavior. A fourth class consists of tumorlike lesions, defined as nonneoplastic disease processes that mimic tumors. Like tumors, they can recur, but they have not been observed to metastasize. An example

197

is deep fibromatosis. The very detailed classification presented in Table 9.2 provides an overview of the large group of soft-tissue tumors. The staging classification of the American Joint Committee on Cancer, which is based on the TNM system, is most widely used in the clinical setting. The TNM staging system for soft-tissue sarcomas is presented in Table 9.3. The second important factor for therapeutic decisionmaking is the histologic tumor grade, which describes the degree of differentiation. Well-differentiated G1 tumors do not require adjuvant postoperative treatment if R0 resection has been achieved, but nearly all patients with dedifferentiated G3 tumors, which have a high recurrence rate, are candidates for postoperative radiotherapy and/or adjuvant chemotherapy. Soft-tissue tumors are rare, and it is almost impossible to give exact frequency data. Especially for benign tumors, the incidence can only be estimated: a figure of ca. 300 per 100 000 individuals has ben proposed.6 Statistical data on malignant neoplasms are more reliable, since one may safely assume that all malignant soft-tissue tumors will come to clinical attention sooner or later and are then biopsied. For the USA, it has been estimated that a total of 486 000 malignant tumors occurred in1990, among them 5700 soft-tissue sarcomas with ca. 3100 sarcoma-related deaths. Based on these estimates, soft-tissue sarcomas constitute 1 % of all malignant tumors. The incidence reported in the literature is 1.35 to 1.4 per 100 000 population,6 and it is assumed that the incidence of soft-tissue sarcomas of the extremities is higher in old age (up to 8 per 100 000 in those over 80).7 Important epidemiologic data are summarized in Table 9.4. Some soft-tissue tumors have a predilection for specific sites—leiomyosarcomas constitute nearly half of all retroperitoneal, intra-abdominal, and visceral tumors. Next in frequency in the retroperitoneum are liposarcomas and schwannomas.4 Common tumors in the extremities are liposarcoma (26 %) and malignant fibrous histiocytoma (or the “newer” entities derived from these) (24 %). The classic histologic categorization (leiomyosarcoma, liposarcoma, malignant fibrous histiocytoma, etc.) is increasingly being abandoned in favor of a more refined classification based on immunohistochemical and molecular biological properties.8,9 At the same time these properties have the potential to serve as targets for new therapies: gastrointestinal stromal tumor (GIST), a variant of sarcoma, was established as a separate tumor entity quite recently. On the molecular level, it is characterized by partially excessive c-kit expression, making the tumor amenable to chemotherapeutic treatment with imatinab (Glivec/Gleevec), a tyrosine kinase inhibitor.

198

9 The Retroperitoneum

Table 9.2 Classification of soft-tissue tumors (according to Enzinger and Weiss, Salzer and Kuntschik) Tissue of origin

Tumorlike

Benign

Intermediate

Malignant

Fat

· Lipomatosis and variants · Lipoblastomatosis

· Lipoma and variants

· Atypical lipoma

· Liposarcoma

· · Fibroblastic tissue

· Fibromatosis

· Fibroma with subtypes · Atypical fibroxanthoma · Nasopharyngeal · Dermatofibrosarcoma

·

· Giant cell fibro-

· · · · · · · · · · · · ·

Smooth muscle

– spindle cell lipoma – pleomorphic lipoma – angiolipoma Lipoblastoma Hibernoma

– fascial – musculoaponeurotic Infantile myofibromatosis – solitary – multicentric Infantile and juvenile fibromatosis with subtypes Myofibromatosis – congenital/acquired Nodular fasciitis Proliferative fasciitis Proliferative myositis Elastofibroma Keloid Fibrous hamartoma of infancy Calcifying aponeurotic fibroma Fibromyxoid inflammatory pseudotumor Xanthoma Generalized eruptive histiocytoma Reticulohistiocytoma – localized – multicentric

· Leiomyomatosis – intravenous – peritoneal

angiofibroma blastoma

· Intranodal myofibro· ·

protuberans

· Fibrosarcoma with subtypes · Congenital and infantile fibrosarcoma

tumor (PFT)

– – – – – –

blastoma Fibrous histiocytoma Juvenile xanthogranuloma

· Leiomyosarcoma with

· ·

· Epitheloid leiomyosarcoma · Myxoid leiomyosarcoma

subtypes Angiomyoma Epitheloid leiomyoma with subtypes

subtypes

· Rhabdomyosarcoma

– fetal type – adult type – genital type

· Hemangiomatosis · Pyogenic granuloma · Papillary endothelial hyperplasia

Lymph vessels

· Lymphangiomatosis · Lymphangiomyomatosis

storiform pleomorphic myxoid giant cell inflammatory angiomatoid

· Leiomyoma with

· Blood vessels

well-differentiated lipomalike sclerosing inflammatory myxoid round cell pleomorphic dedifferentiated

· Bednar tumor · Malignant fibrous histio· Plexiform fibrohistiocytic cytoma

· Rhabdomyoma

Striated muscle

– – – – – – – –

· Hemangioma with · · · ·

subtypes Epithelioid hemangioma Glomus tumor Hemangiopericytoma Infantile hemangiopericytoma

· Lymphangioma · Lymphangiomyoma

– embryonal – botryoid – alveolar – pleomorphic – mixed Ectomesenchymoma – triton tumor

· Hemangioendothelioma · Angiosarcoma with – epitheloid – spindle cell – papillary angioendothelioma

subtypes

· Kaposi sarcoma · Malignant glomus tumor · Malignant hemangiopericytoma

· Angiosarcoma

Classification of Soft-Tissue Tumors

199

Table 9.2 (continued) Tissue of origin

Tumorlike

Synovial tissue

· Tenosynovial giant cell

Benign

Intermediate

Malignant

· Malignant synovialoma (synovial

tumor – localized – diffuse

· Bone and cartilage

· Ossifying pseudotumors ·

– panniculitis ossificans – myositis ossificans Progressive ossifying fibrodysplasia

· Extraskeletal osteoma · Extraskeletal chondro-

sarcoma) – biphasic – monophasic – fibrous – epithelial – calcifying – undifferentiated Malignant giant cell tumor of the tendon sheath

· Extraskeletal osteosarcoma · Extraskeletal chondrosarcoma · Extraskeletal myxoid chondro-

ma

sarcoma

· Extraskeletal mesenchymal chondrosarcoma

Peripheral nerve tissue

· Traumatic neuroma · Schwannoma with subtypes · Von Recklinghausen neurofibromatosis with subtypes · Neurofibroma with

· Malignant schwannoma with subtypes

subtypes

· Granular cell tumor · Pigmented neuroecto· · · · · · Uncertain

dermal tumor of infancy Ectopic meningioma Nasal glioma Neurothecoma Ganglioneuroma Melanocytic schwannoma Paraganglion with subtypes

· ·

neuroectodermal tumor

schwannoma

· Malignant paraganglioma · Extraspinal ependymoma · · · · ·

cell tumor Myxoma with subtypes Parachordoma Aggressive angiomyxoma

Table 9.3 TNM classification system for soft-tissue sarcomas (according to American Joint Committee on Cancer, 1992) T Primary tumor T1: Tumor ≤ 5.0 cm in greatest dimension T2: Tumor > 5.0 cm in greatest dimension N Regional lymph nodes N0: No regional lymph node metastasis N1: Regional lymph node metastasis M Distant metastasis M0: No distant metastasis M1: Distant metastasis G Histologic grade of malignancy G1: Well differentiated (low malignancy grade) G2: Moderately differentiated (intermediate malignancy grade) G3: Poorly differentiated (high malignancy grade) G4: Undifferentiated (anaplastic)

ectodermal tumor

· Malignant primitive peripheral · Neuroblastoma · Ganglioneuroblastom · Malignant melanocytic

· Congenital granular ·

· Malingnant triton tumor · Malignant granular cell tumor · Malignant pigmented neuro-

Alveolar soft-tissue sarcoma Epitheloid sarcoma Clear cell sarcoma Extraskeletal Ewing sarcoma Extrarenal malignant rhabdoid tumor

Table 9.4 Epidemiology of soft-tissue sarcomas (Memorial SloanKettering Cancer Center, New York, 1982–1992 [n = 2044]) Male-to-female ratio

1.2:1

Age distribution

< 30 years

18 %

31–50 years

33 %

51–70 years

26 %

> 70 years

12 %

200

9 The Retroperitoneum

MRI in Initial Diagnostic Work-up Tumors Although they can occur throughout the body, soft-tissue tumors have a fairly consistent clinical presentation, the single most notable feature shared by nearly all of them being painless swelling. However, patients tend to delay seeing a doctor, which is why the “swelling” is often detected during a routine examination. The role of imaging in this situation is to define the site and extent of the tumor and to identify a suitable access for mandatory open biopsy. Several criteria have been proposed to improve diagnostic confidence in characterizing soft-tissue lesions on the basis of their MRI appearance. The most important are: · homogeneity/heterogeneity of the lesion · margination · growth behavior · enhancement patterns after intravenous contrast administration.10 Whether a lesion is homogeneous or heterogeneous is an important criterion for distinguishing benign from malignant soft-tissue tumors. A consistently uniform appearance on all pulse sequences favors a benign process. Histologically, rapidly growing tumors are of mixed composition, containing viable tissue and necrotic areas with scars and hemorrhage. However, there are benign tumors with a heterogeneous MR appearance. In a hemangioma, for instance, areas of low signal intensity due to old blood may be present side by side with fresh thrombi of high signal intensity as well as hyperintense, slow-flowing blood and focal signal voids due to saturation phenomena or fast-flowing blood. Central necrosis is considered a fairly reliable indicator of a malignant tumor in conjunction with strong marginal enhancement indicating viable tumor. Central necrosis alone is not a reliable distinguishing feature. Tumor margination is another very important criterion. Although smooth margins generally indicate a benign lesion, this assumption can be misleading when dealing with sarcoma. Soft-tissue sarcomas typically have a pseudocapsule resulting from compression of surrounding tissue and therefore appear as well-defined lesions on T2w images. However, they may appear ill-defined if peritumoral edema is present. Also confounding are inflammatory or hemorrhagic changes, which invariably make a tumor appear blurred. In a retrospective analysis of the MRI findings in 95 benign and malignant soft-tissue tumors carried out in 1990,11 Berquist and coworkers found that 40 % of the benign lesions also had irregular margins. Another feature that might contribute to the differentiation of benign and malignant soft-tissue tumors is growth behavior. While high-grade sarcomas tend to invade surrounding tissue at an early stage, a benign process is usually confined to the compartment in which it arises

and displaces surrounding structures while respecting fascial boundaries until fairly late in the disease process. One must be aware, however, that some soft-tissue structures such as periosseous spaces, neurovascular tracts, subcutaneous fatty tissue, skin, and also the fascia are in effect unbounded structures. Thus, tumor extension beyond any of these structures is not a reliable criterion either. Enhancement patterns are also of limited value. The distribution models established for X-ray contrast media also apply to nonspecific Gd-based MR contrast media. Both types of contrast media distribute in the extracellular (interstitial) space after an initial intravascular phase. Malignant tumors are highly vascularized lesions with an enlarged interstitial space and therefore tend to enhance more intensely than normal tissue. The enhancement can be quantified by measuring MR signal intensities. However, it must be borne in mind that there are hypovascular soft-tissue malignancies and mixed benign tumors composed of different tissues, which tend to enhance more slowly after intravenous contrast administration (e. g., hamartoma, lipomatous hemangioma). Dynamic contrast-enhanced MRI may also contribute to lesion characterization; but again, there is overlap between benign and malignant tumors.12 Petterson and coworkers13 calculated T1 and T2 relaxation times of benign and malignant musculoskeletal tumors to investigate their potential for tissue characterization, or at least for separating benign from malignant masses. A basic problem noted by the authors was that the relaxation times determined in different regions of interest (ROIs) in the same tumor already varied between 20 % and 60 %. Therefore, the authors failed to establish a relaxation-based malignancy criterion, concluding that the heterogeneous composition of soft-tissue tumors precludes direct correlations between tumor histology and relaxation times, or MR appearance in general. However, another group found serial intra-individual measurement of relaxation times to be useful for monitoring response to chemotherapy in patients with Ewing sarcoma.14 Figure 9.1 presents two cases of liposarcoma to illustrate the malignancy criteria just outlined and to show differences in the MR appearance of high-grade and low-grade tumors. All of the above criteria have to be taken into account in characterizing a soft-tissue tumor, since none of them alone is reliable. When all imaging features are assessed in conjunction with the clinical presentation (rapidly growing mass), correct differentiation was achieved in up to 95 % of cases.11 The currently accepted procedure to establish a diagnosis in patients with soft-tissue tumors is to surgically resect small lesions or to perform open biopsy in patients with extensive disease that cannot be resected without excessive collateral damage to neighboring structures. Malignant soft-tissue tumors constitute a fairly uniform group of neoplasms in terms of their MRI appearance. Nearly all soft-tissue sarcomas, regardless of their histologic composition, have long T1 and T2 relaxation

Classification of Soft-Tissue Tumors

201

a, b

c Fig. 9.1a–d Illustration of malignancy criteria using two cases of liposarcoma as examples. a Coronal T2w SE image. b, c Coronal T1w images before (b) and after (c) IV administration of contrast medium. Images show a G3 dedifferentiated liposarcoma consisting of solid and fatty components in the left retroperitoneal space. The tumor contains necrotic areas and nodules with strong enhancement (arrow in b and c). d Axial T1w GRE image from another patient shows extensive mass in the right retroperitoneum containing areas isointense to fat (*) but also some thick septa (arrow) in the periphery.The tumor was classified as G1 liposarcoma. Important malignancy criteria are septa, cysts, necrosis, solid tumor components with contrast enhancement, and ill-defined margins.

d

times. As a result, highly malignant liposarcoma, malignant fibrous histiocytoma, fibrosarcoma, and rhabdomyosarcoma are characterized by low signal intensity on T1w images and high signal intensity on T2w images.12 Welldifferentiated tumors may occasionally exhibit the same MR signal intensities as the tissues of which they are composed. On T1w images, malignant soft-tissue tumors cannot be distinguished from muscle tissue because of their signal intensities are identical. After intravenous administration of Gd-based contrast medium, all sarcomas enhance intensely and tumor architecture can be assessed. It must be noted, however, that the extent of a mixed tumor containing nonviable tissue may be underestimated on postcontrast MRI. Infiltrative growth is best appreciated on T2w sequences, which invariably depict all tumors with high signal intensity. High signal intensity on T1w sequences in conjunction with fairly low T2 signal intensity is indicative of a tumor containing paramagnetic substances (e. g., blood products). These mainly include hematoma, hemangioma, hemorrhagic tumors, and some cysts (see possible differ-

Table 9.5 Narrowing of differential diagnosis on the basis of MR signal intensities MR appearance

Possible diagnosis

Bright on T1w sequences Lipoma / differentiated liposarcoma Fresh hematoma / hemangioma Melanoma Dark on T2w sequences Pigmented villonodular synovitis Giant cell tumor Fibrous tumors Old hematoma Deposits (hemosiderin, amyloid) Melanoma Dark on T1w and bright on T2w sequences

All other tumors Cysts (may be confused with myxoid liposarcoma)

ential diagnoses based on MR signal intensities in Table 9.5). Overall, the ability of MRI to provide a specific diagnosis of a soft-tissue tumor is limited even if all parameters discussed above are taken into account. This is not

202

9 The Retroperitoneum

b

a

c Fig. 9.2a–c Lipoma. a Sagittal T1w SE image. b Axial T2w fast SE image. c Axial T1w SE image after IV administration of Gd-based contrast medium. Large, dumbbell-shaped tumor in the retroperitoneum extending to the minor trochanter. The tumor is of

very homogeneous SI and isointense to fat with unchanged SI after contrast administration compared with precontrast T1w image. All pulse sequences acquired allow delineation of the tumor from surrounding tissue.

surprising, since the WHO classification distinguishes > 140 soft-tissue tumors.15 It would therefore make little sense to comprehensively describe the whole range of MR findings of soft-tissue tumors; instead, the following discussion will focus on the most relevant tumors, either because they are common or because they have specific MR features worth pointing out.

cystic, and dedifferentiated. They are of very high signal intensity on T2w images and appear inhomogeneous. Another characteristic feature is infiltrative growth, and even bone destruction is not uncommon. Published data on the MR signal pattern of well-differentiated liposarcomas are inconsistent. In our experience, this subtype invariably contains tissue areas with T1 signal intensity isointense to muscle and high signal intensity on T2w images (Fig. 9.1). However, such suspicious tissue components in the form of septa or marginal compartments may be small and appear insignificant at first sight. Liposarcoma can only be ruled out with a high degree of confidence if a lesion is exclusively composed of fat and there is no septation or contrast enhancement on a fat-suppressed T1w sequence.16–19 Conversely, it is also impossible to diagnose a liposarcoma if MRI demonstrates fatty portions in a malignantappearing lesion because fat can also be present in other sarcomas.

Lipoma Lipomas are very rare in the first two decades of life and usually present between 40 and 60 years of age when fat starts to accumulate in the body as a result of decreasing physical activity.5 They are often stable in size after initial growth. Unsurprisingly, lipomas display MR signal intensities paralleling those of “normal” fat (Fig. 9.2), i. e., high signal intensity on both T1w and T2w images. In addition, fat suppression techniques allow confident tissue characterization. Lipomas are thus among the soft-tissue tumors with distinctive MR features and many of them show a capsule on MRI. Liposarcoma Liposarcomas are the most common sarcomas throughout the body. Several histologic subtypes are distinguished: well-differentiated, myxoid, round-cell, pleomorphic,

Leiomyosarcoma Leiomyosarcomas are the most common retroperitoneal sarcomas but account for only 5–10 % of all sarcomas. About 50 % of leiomyosarcomas originate within the retroperitoneum. Other common sites are the omentum, mesentery, and stomach in the peritoneal cavity. Over

Classification of Soft-Tissue Tumors

203

b

a Fig. 9.3a–c Leiomyosarcoma. a Coronal T1w SE image. b Axial T2w fast SE image. c Coronal T1 SE image after IV administration of contrast medium. Unenhanced T1w image (a) shows poorly demarcated retroperitoneal tumor (arrow) extending from the renal hilum to the liver hilum. The mass has much higher SI on the unenhanced T2w image (b), enabling good delineation from surrounding structures. The tumor shows very mild enhancement after contrast administration, which is discernible only in direct comparison with the unenhanced image (arrows in a and c).

c

60 % of all retroperitoneal leiomyosarsomas occur in women, making this tumor an exception among softtissue sarcomas. The median age at diagnosis is 60 years. Most retroperitoneal leiomyosarcomas are moderately differentiated. Histologically, leiomyosarcomas are the most heterogeneous group of soft-tissue sarcomas, which is why only three clearly defined subtypes are distinguished: · epitheloid leiomyosarcoma with a predilection for the extremities · myxoid leiomyosarcoma, which is most common in the uterus · granular cell leiomyosarcoma, which has no apparent predilection for a particular localization. Retroperitoneal leiomyosarcomas are very aggressive and are typically very large at presentation, precluding complete en bloc resection. The 5-year survival rates reported in the literature range between 0 % and 29 %.2 Leiomyosarcomas are also isointense to muscle on T1w images and hyperintense on T2w images. They enhance intensely after intravenous contrast administration, often have central areas of necrosis, and show infiltrative growth in 50 % of cases (Fig. 9.3).8

Fibrosarcoma Fibrosarcomas may develop at any age with a peak incidence between 30 and 55 years. These tumors can arise anywhere in the body where connective tissue is found but are most common in the lower extremities (45 %), followed by the upper extremities (28 %) and the trunk (17 %). Fibrosarcomas of the head and neck region are rare and involve the nasal cavity, paranasal sinuses, or the nasopharynx. Histologically, well-differentiated fibrosarcoma with sclerosing epitheloid fibrosarcoma as a subset is separated from undifferentiated fibrosarcoma. It is histologically difficult to differentiate fibrosarcoma from other spindle cell tumors such as malignant fibrous mesothelioma, dermatofibrosarcoma protuberans, and spindle cell rhabdomyosarcoma. Fibrosarcomas are isointense to muscle on T1w images and hyperintense on T2w images (Fig. 9.4), thus representing another subgroup of soft-tissue sarcomas without specific MRI features. Over 60 % of fibrosarcomas show exophytic growth, depicted in most cases on T2w and contrast-enhanced T1w images as infiltration of surrounding tissue.17

204

9 The Retroperitoneum

a

b Fig. 9.4a–c Fibrosarcoma. a Axial T1w SE image. b Axial T2w fast SE image. c Axial T1w SE image after IV contrast administration. Precontrast T1w image shows left-sided mass (arrow in a) of muscle SI which is contiguous with but does not infiltrate the iliopsoas muscle anteriorly. On T2w image (b), the tumor is of high SI and clearly delineated from the iliopsoas muscle but delineation from anterior structures is more difficult. Postcontrast image (c) shows slight enhancement but provides no definitive proof of necrosis.

c

Malignant Fibrous Histiocytoma Malignant fibrous histiocytoma (MFH) is most commonly diagnosed between 50 and 70 years and very rarely presents before age 20.6 Four histologic subtypes exist: · storiform-pleomorphic · myxoid · giant cell · inflammatory. MFH is most common in the extremities, apart from inflammatory MFH, which has a predilection for the retroperitoneum.6 The inflammatory subtype has a local recurrence rate of 50 % in the retroperitoneum, and half of the patients have metastases.20 MFH is more common in men and in white people. Patients with inflammatory MFH leukocytosis often have neutrophilia, eosinophilia, and fever. Again, the MR features are the same as for the aforementioned entities. Edema of surrounding tissue can be seen in about half of cases17 (Fig. 9.5). MFH grows rapidly, which is why central necrosis can be demonstrated on contrast-enhanced T1w sequences in nearly 50 % of cases.17 Rare Tumors The tumors discussed so far constitute well over 90 % of retroperitoneal tumors. Less common tumors in the retroperitoneum include neurogenic tumors such as schwannoma (Fig. 9.6) and peripheral nerve sheath tumors as well as extremely rare entities such as primitive neuroectodermal tumors (PNET), extraosseous Ewing sarcoma, clear-cell sarcoma (the soft-tissue counterpart of melanoma), and many others. Here as well, the role of MRI is

to assess the tumor extent and contribute to tissue characterization.

Diffuse (Nontumorous) Diseases Retroperitoneal Fibrosis (RPF) Idiopathic retroperitoneal fibrosis, or Ormond disease, is twice as common in men than in women and typically presents between 40 and 60 years. This entity is said to be closely related to inflammatory aortic disease (without aortic dilatation) and constitutes 90 % of cases of RPF. The remaining 10 % (nonidiopathic RPF) are considered to be caused by certain medications such as methysergide or adrenergic blockers. Other secondary causes include radiation treatment and prior trauma or infection. Both forms have the same imaging features. In most patients, a tissue plate up to 2 cm thick is demonstrated in the lumbosacral retroperitoneum. The defining feature of RPF is encasement of retroperitoneal nerves and vessels without invasion. At MRI, retroperitoneal fibrotic tissues correspond in signal intensity to connective tissue, which is intermediate between that of muscle and fat. After contrast administration, the fibrotic tissue shows mild to intense enhancement (Fig. 9.7), best demonstrated on a fat-suppressed T1w sequence. Retroperitoneal Xanthofibrogranulomatosis Retroperitoneal xanthofibrogranulomatosis is a distinct histologic entity that differs from idiopathic RPF by the presence of giant cells. The MR signal intensities of both entities are the same, but, unlike RPF, xanthofibrogranu-

Classification of Soft-Tissue Tumors

205

a

c Fig. 9.5a–c Malignant fibrous histiocytoma (G2) of the right psoas muscle with infiltration of the second lumbar vertebra (arrow). a, b Axial T1w 2D FLASH images obtained before (a) and after (b) contrast administration. c Coronal contrast-enhanced T1w GRE image acquired with spectral fat saturation. Tumor appears to have a pseudocapsule due to compression of surrounding normal tissue. There is marked inflammation around the lesion, which is otherwise well circumscribed (arrow).

b

a Fig. 9.6a, b Neurofibromas in a patient with neurofibromatosis type II. a Axial T2w SE image shows typical, hourglass-shaped tumor extending into the spinal canal and through the posterior transversalis fascia. b Coronal fat-suppressed, contrast-enhanced T1w FLASH image demonstrates typical homogeneous enhancement of the tumor.

b

206

9 The Retroperitoneum

a

b Fig. 9.7a–c Retroperitoneal fibrosis. a Axial T1w SE image. b Axial T2w fast SE image. c Axial T1w SE image after IV contrast administration. Images demonstrate a retroperitoneal mass (arrow) encasing major vessels. There is no evidence of vascular invasion and adjacent vertebral bodies appear intact. The mass is of high SI on T2w image (b) and enhances slowly after contrast administration.

c

lomatosis shows infiltrative growth, which may involve all retroperitoneal structures.

MRI in Tumor Recurrence MRI has an important role in postoperative follow-up. Soft-tissue sarcomas tend to recur with a reported rate of 7–38 %;6 and 80 % of locally recurrent tumors present during the first 2 years after surgical resection. In a study of 54 patients, Simon and Enneking observed all local recurrences within 30 months of surgery.21 When MRI is used in the early phase after surgical resection or radiotherapy, the radiologist must distinguish recurrent tumor from therapy-related changes in signal intensities caused by lymphedema, seroma, abscess, or infected lymphocele. Contrast-enhanced MRI will distinguish these entities in most cases. However, a baseline MRI examination should be done immediately after radical resection and/or radiotherapy to serve as a basis of comparison for early identification of a recurrent tumor. It has been shown that scar tissue takes up contrast medium for up to 6 months postoperatively but tends to have low signal intensity on T2w SE images. Recurrent tumor is also characterized by enhancement, even when still small.17–19 The postoperative MRI follow-up scheme practiced by the authors comprises a baseline examination immediately after surgery or at the end of radiotherapy to identify recurrent tumors while they are still small. Further MRI

examinations are performed at 3, 6, and 12 months in the first year, at 6-month intervals in the second year, and once yearly thereafter. The close initial follow-up intervals are dictated by the fact that most tumors recur within 2 years of treatment.6 MRI has become the standard imaging tool for evaluating soft-tissue tumors. It is more flexible than the other cross-sectional imaging modalities and affords better sensitivity and specificity.17,19,21–24 Another major advantage of MRI is that it provides the information on tumor localization needed for optimal planning of surgery and radiotherapy. However, MRI is limited in identifying or ruling out metastatic disease, in particular metastatic spread to the lungs. Evaluation for metastatic disease is the domain of other modalities—conventional radiography, ultrasound, scintigraphy, or CT with or without PET; the most suitable method is determined by the organ system involved.

References 1. Bammer R, Schoenberg SO. Current concepts and advances in clinical parallel magnetic resonance imaging. Top Magn Reson Imaging 2004;15(3):129–158 2. Brennan MF. Soft-tissue sarcoma: advances in understanding and management. Surgeon 2005;3(3):216–223 3. Campeau NG, Johnson CD, Felmlee JP, et al. MR imaging of the abdomen with a phased-array multicoil: prospective clinical evaluation. Radiology 1995;195(3):769–776

Classification of Soft-Tissue Tumors 4. Erzen D, Sencar M, Novak J. Retroperitoneal sarcoma: 25 years of experience with aggressive surgical treatment at the Institute of Oncology, Ljubljana. J Surg Oncol 2005;91(1):1–9 5. Chung T, Muthupillai R. Application of SENSE in clinical pediatric body MR imaging. Top Magn Reson Imaging 2004;15(3): 187–196 6. Enzinger FM, Weiss SW. Soft-tissue Tumors. 3 rd ed. St. Louis: Mosby; 1995 7. Rydholm A, Berg NO, Gullberg B. Epidemiology of soft-tissue sarcomas in the locomotor system: a retrospective populationbased study of the interrelationship between clinical and morphological variables. Acta Pathol Microbiol Scand 1984;92: 363–379 8. Coindre JM. Immunohistochemistry in the diagnosis of softtissue tumours. Histopathology 2003;43(1):1–16 9. Czerniak B. Pathologic and molecular aspects of soft-tissue sarcomas. Surg Oncol Clin N Am 2003;12(2):263–303, v 10. Totty WG, Murphy WA, Lee JK. Soft-tissue tumors: MR imaging. Radiology 1986;160(1):135–141 11. Berquist TH, Ehman RL, King BF, Hodgman CG, Ilstrup DM. Value of MR imaging in differentiating benign from malignant softtissue masses: study of 95 lesions. AJR Am J Roentgenol 1990; 155(6):1251–1255 12. Erlemann R, Vassallo P, Bongartz G, et al. Musculoskeletal neoplasms: fast low-angle shot MR imaging with and without GdDTPA. Radiology 1990;176(2):489–495 13. Pettersson H, Slone RM, Spanier S, Gillespy TIII, Fitzsimmons JR, Scott KN. Musculoskeletal tumors: T1 and T2 relaxation times. Radiology 1988;167(3):783–785 14. Shapeero LG, Vanel D. Imaging evaluation of the response of high-grade osteosarcoma and Ewing sarcoma to chemotherapy

15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

207

with emphasis on dynamic contrast-enhanced magnetic resonance imaging. Semin Musculoskelet Radiol 2000;4(1):137–146 Katenkamp D. [Morphology and molecular biology of malignant soft-tissue sarcomas]. Praxis (Bern 1994) 1998;87(34): 1043–1049 Einarsdottir H, Söderlund V, Larson O, Jenner G, Bauer HC. MR imaging of lipoma and liposarcoma. Acta Radiol 1999;40(1): 64–68 Habermann CR, Nicolas V, Beese M. MR imaging in soft-tissue sarcomas. Eur Radiol 1997;135(Suppl. 7):670 Jelinek JS, Kransdorf MJ, Shmookler BM, Aboulafia AJ, Malawer MM. Liposarcoma of the extremities: MR and CT findings in the histologic subtypes. Radiology 1993;186(2):455–459 Nicolas V, Beese M, Habermann CR, Maas R, Schwarz R, Zornig C. Primär- und Rezidivdiagnostik bei Weichteilsarkomen der Extremitäten in der MRT. Zentralbl Radiol. 1993;147:900–905 Fukuda T, Tsuneyoshi M, Enjoji M. Malignant fibrous histiocytoma of soft parts: an ultrastructural quantitative study. Ultrastruct Pathol 1988;12(1):117–129 Simon MA, Enneking WF. The management of soft-tissue sarcomas of the extremities. J Bone Joint Surg Am 1976;58(3): 317–327 Hyslop WB, Balci NC, Semelka RC. Future horizons in MR imaging. Magn Reson Imaging Clin N Am 2005;13(2):211–224 Petasnick JP, Turner DA, Charters JR, Gitelis S, Zacharias CE. Softtissue masses of the locomotor system: comparison of MR imaging with CT. Radiology 1986;160(1):125–133 Reuther G, Mutschler W. Detection of local recurrent disease in musculoskeletal tumors: magnetic resonance imaging versus computed tomography. Skeletal Radiol 1990;19(2):85–90

209

10

The Urinary Bladder V. Nicolas and D. Beyersdorff

Introduction Ultrasound, cystography, voiding cystourethrography, and intravenous urography are the primary imaging modalities for examining the urinary bladder. If these diagnostic tests are inconclusive, CT or MRI can be employed to obtain important supplementary information. As elsewhere in the body, MRI affords excellent soft-tissue contrast and enables multiplanar imaging. However, CT is gaining on MRI now that multislice spiral CT scanners provide thin-slice reconstruction with the option of additional multiplanar reconstruction.

Indications MRI of the bladder is mainly used to stage known bladder cancer, detect tumor recurrence, and evaluate for secondary invasion of the bladder by gynecologic, prostatic, or rectal neoplasms. In addition, the multiplanar capabilities of MRI may provide important information in selected patients with congenital anomalies or inflammatory disease.

Imaging Technique Phased-array coil systems have become the standard in abdominal MRI. When imaging the urinary bladder and perivesical fat, a phased-array coil (usually consisting of four to eight elements) markedly improves the signal-tonoise ratio (SNR) compared with the integrated wholebody resonator and thus enables acquisition of higherresolution images. Optimal bladder filling is important for a complete evaluation and the detection of small lesions. This is achieved by asking the patient to void a couple of hours before the examination and then not again until completion of the test. On the other hand, overdistention must also be avoided as it may obscure small lesions and can lead to motion artifacts because the patient is uncomfortable and more likely to move. To

minimize such artifacts, it is also important to ensure comfortable positioning, e. g., by elevating the patient’s knees with a small pad. Respiratory artifacts can be reduced by placing a compression band across the abdomen. The same effect is achieved by exerting slight pressure above the pubic bone with the phased-array coil. Artifacts caused by bowel motion are minimized by intravenous/intramuscular injection of an antispasmodic agent.

Imaging Planes The bladder protocol should include an imaging plane perpendicular to the interface between a suspected lesion and the bladder wall, i. e., an axial or coronal plane for evaluating a lesion of the lateral wall and a sagittal or coronal plane for lesions involving the bladder dome or bladder outlet. Occasionally, a double oblique imaging plane is necessary to acquire images perpendicular to the affected wall.1

Pulse Sequences The urinary bladder is imaged with a combination of unenhanced T2w and T1w sequences in at least two planes and a T1w sequence after intravenous contrast administration (Table 10.1). The slice thickness should not exceed 4–5 mm. We use the following protocol for MRI of the bladder: · a multiplanar localizer scan · standard fast T1w and T2w sequences in axial, coronal, and sagittal planes for tumor localization · an axial T2w TSE sequence for acquisition of highresolution images (oblique plane perpendicular to the tumor–bladder wall interface if necessary) · a T1w sequence oriented perpendicular to the tumorbladder wall interface before and after administration of nonspecific Gd-based contrast medium. Alternatively, thin-slice T1w sequences or 3D sequences can be used.

210

10 The Urinary Bladder

Table 10.1 Recommended pulse sequences and imaging parameters for MRI of the urinary bladder Weight- Plane ing T2 T1

Sequence type

Axial and as speci- TSE fied below* As specified TSE below*

TR (ms)

TE (ms)

ETL

FS

Matrix

FOV (mm)

No. of slices

No. of acquisitions

Slice thickness (mm)

3500

100–150

5–15

–

270 × 512

300

15–19

1

3–5

823

10

3

–

270 × 512

300

15–19

3

3–5

*An oblique plane is acquired perpendicular to the lesion/involved bladder wall (which may require a double oblique plane) using an unenhanced T2w sequence followed by a T1w sequence before and after IV contrast administration. Note: The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time (for sequences with one signal average) but may come with a penalty in SNR.

Contrast Media Contrast-enhanced bladder MRI is performed using one of the commercially available Gd-based contrast media (e. g., Magnevist, Dotarem, Prohance) injected at a standard dose of 0.1 mmol Gd per kg body weight. When images are acquired with a conventional T1w TSE sequence, it is important to start acquisition immediately after intravenous bolus injection in order not to miss the early phase of maximum contrast between urine, mucosa, and bladder wall. Bladder neoplasms typically enhance early in a dynamic study. In patients with suspected tumor recurrence, quantification of the signal increase on dynamic contrastenhanced images may occasionally help distinguish scar tissue from recurrent tumor. A tumor may be obscured on postcontrast MRI if images are acquired after distribution of the contrast medium in the urine. Since the signal intensity of urine varies with the Gd concentration, inhomogeneous distribution of the contrast medium may occasionally lead to layering within the bladder. This is seen as low signal intensity in the dependent portion of the bladder, where the Gd concentration is highest; as high signal intensity of a middle layer, reflecting urine containing less contrast medium; and as low signal intensity of a third layer anteriorly, representing unenhanced urine (see Fig. 10.3). Direct intravesical administration of iron-oxidebased contrast medium causes uniform signal loss of urine on T2-weighted images but offers no diagnostic advantage because delineation of the mucosa is not improved.2

MRI Appearance of Normal Anatomy The standard protocol for MRI of the urinary bladder comprises unenhanced T1w and T2w sequences and a T1w sequence after intravenous contrast administration. With their rather long T1 times, urine and the bladder wall have low signal intensity on T1w images. On heavily T1-weighted images, the distinction between urine and the wall may become difficult if the bladder is very full. On T2w images, the bladder wall is seen as a low-signalintensity band between the high-signal-intensity urine

and the rather high signal intensity of the perivesical fat. One must be aware that chemical shift artifacts which occur at the bladder/perivesical fat interface can obscure the bladder wall or mimic pathology. The true nature of such artifacts will be revealed by changing the phaseencoding direction. Following intravenous contrast administration, the mucosa enhances earlier than the outer muscular layer.3 On later images, the entire bladder wall becomes enhanced but the SI remains below that of the mucosa. Contrast-enhanced urine entering the bladder can result in inhomogeneous enhancement around the ureteral orifices on early postcontrast images with subsequent layering before the urine finally shows relatively uniform enhancement.

MRI Appearance of Pathologic Entities Anomalies The most common urinary bladder anomalies are agenesis, hypoplasia, duplication, exstrophy, and diverticula. Since ultrasound and cystography enable reliable diagnosis in most cases, MRI is very rarely used in patients with suspected bladder anomalies. Urinary bladder diverticula are often detected incidentally in patients undergoing pelvic MRI for other reasons. The majority are acquired (pseudo-)diverticula that develop as a result of increased intravesical pressure due to conditions causing bladder outlet obstruction such as benign prostatic hyperplasia (BPH), prostate cancer, or urethral stricture. MRI will reveal a bladder diverticulum as an outpouching of the bladder wall. The wall is thinner than the normal bladder wall and has low signal intensity on T2w images. The diverticular neck is typically narrow, but some diverticula have a wide neck (Fig. 10.1). Most bladder diverticula have no clinical relevance. However, stagnant urine in the pouch can lead to inflammation and stone formation, and bladder neoplasms more often arise in a diverticulum (see below).

MRI Appearance of Pathologic Entities

211

a

b Fig. 10.1a, b Urethral diverticulum. a Axial T2w TSE image. b Coronal T2w TSE image. Typical horseshoe configuration of the diverticulum, which wraps around the urethra (arrow).

Inflammation The diagnosis of cystitis is made by clinical and bacteriologic examinations in most cases, occasionally supplemented by cystoscopy with biopsy. Bacteria are the most common etiologic agents. Older women often have concomitant urethritis and vaginitis. Other causes are bladder voiding dysfunction, surgery, and radiotherapy, e. g., for gynecologic tumors. Inflammation causes focal or diffuse thickening of the bladder wall with an intermediate signal intensity or heterogeneous appearance on T1w images. On T2w images, inflammatory areas are of higher signal intensity than uninvolved wall segments. According to Hricak4 acute radiation cystitis has a characteristic MR appearance with the wall showing uniform signal intensity on unenhanced T1w images, while T2w images reveal a thin inner layer of lower signal intensity compared with the thickened outer layer. Following intravenous contrast administration, the inflamed bladder wall will show inhomogeneous enhancement. The high signal intensity on T2w images and the strong enhancement are due to hyperemia.5 Chronic cystitis is associated with marked bladder shrinkage and wall thickening. Postcontrast imaging often reveals areas of inhomogeneous mucosal enhancement, which cannot be distinguished from small superficial tumors.

Benign Tumors Benign tumors of the urinary bladder are rare. They include leiomyoma, neurofibroma, pheochromocytoma, polyps, hemangioma, hamartoma, and fibroma. While some of these tumors have characteristic clinical presentations, their MRI appearance and signal intensities may not allow reliable differentiation from malignant bladder neoplasms.

Leiomyomas typically occur in young women and usually arise near the bladder trigone, less commonly from the posterolateral wall. They are well-circumscribed tumors with signal intensity similar to that of the bladder wall, i. e., intermediate on T1w images and low on T2w images. As with leiomyomas of the uterus, regressive changes such as hemorrhage or calcifications may alter the signal intensity.6 Pheochromocytomas of the urinary bladder constitute ca. 1 % of the extra-adrenal manifestations of this tumor and are very rare, accounting for less than 1 in 1000 of all bladder tumors. In the bladder, they involve the trigone, ureteral orifices, dome, and rarely the lateral wall. The majority are benign extra-adrenal pheochromocytomas, but 30–40 % are malignant. Pheochromocytomas are highly vascularized tumors and typically contain areas of hemorrhage and necrosis. Most patients present with hypertension, intermittent hematuria, and episodes of headache and sweating triggered by bladder distention or voiding. On T1w images, pheochromocytoma is similar in signal intensity to muscle and enhances intensely after intravenous contrast injection. T2w images reveal pheochromocytoma as a high-signal-intensity lesion. Regressive changes may alter the SI in the tumor center.7 Few data are available on the MRI appearance of hemangiomas of the urinary bladder.8 They resemble hepatic hemangiomas with low signal intensity on T1w images and high signal intensity on T2w images.

Malignant Tumors Over 90 % of primary urinary bladder tumors are of epithelial origin and the majority are malignant. Transitional cell carcinoma (TCC), also known as urothelial carcinoma, is by far the most common malignant tumor of the urinary bladder, accounting for 90 % of cases. TCC tends to be

212

10 The Urinary Bladder

Table 10.2 TNM staging system for urinary bladder cancer Ta Tis T1

Noninvasive papillary tumor Carcinoma in situ Tumor invades subepithelial connective tissue

T2

Tumor invades muscle of bladder wall

T2a

Tumor invades superficial muscle (inner half)

T2b

Tumor invades deep muscle (outer half)

T3

Tumor invades perivesical fat

T3a

Tumor invades perivesical fat (microscopic tumor)

T3b

Tumor invades perivesical fat (macroscopic tumor

T4

Tumor invades one of the following structures: prostate, uterus, vagina, pelvic sidewall, abdominal wall

T4a

Tumor invades prostate, uterus, vagina

T4b

Tumor invades pelvic sidewall, abdominal wall

Subepithelial connective tissue

Perivesical fat

Epithelium

Muscle

T2a Tis

T1

T2b

Ta

T3 T4b

T4a

T= pT Fig. 10.2 Stages of urinary bladder cancer.

multifocal and typically involves the lateral and posterior bladder walls, the area around the ureteral orifices, and the trigone. There is a high recurrence rate of 30–85 % after transurethral resection (TUR). Squamous cell carcinomas constitute ca. 2–5 % of bladder neoplasms and typically occur in patients with chronic cystitis and schistosomiasis. Another 2–3 % are adenocarcinomas. Based on their growth pattern, papillary vs nonpapillary and superficial vs infiltrative tumors are distinguished. The prognosis and therapeutic approach depend on the depth of bladder wall invasion, degree of differentiation, and the presence of nodal or distant metastases at the time of diagnosis.

Staging Bladder carcinomas are staged according to the TNM classification system (Table 10.2, Fig. 10.2)9. Initial diagnosis and staging predominantly rely on cystoscopy. Stages up to T1G3 are classified as superficial bladder carcinomas. Up to 70 % of patients have superficial tumors at presentation. In > 80 % of these cases, local tumor control is achieved by TUR, which may be supplemented by intravesical chemotherapy or immunotherapy. Imaging is indicated to assess the depth of bladder wall invasion in patients with large tumors at presentation or in whom muscle invasion is suggested by histologic workup after TUR. MRI is usually performed after cystoscopy to determine the depth of muscle invasion. Imaging planes are prescribed perpendicular to the bladder wall bearing the tumor on the basis of the cystoscopy findings. Polypoid or broad-based tumors are of the same signal intensity as the normal bladder wall on unenhanced T1w images. On T2w images, they typically have higher signal intensity than the unaffected wall but may occasionally be isointense or even hypointense (Fig. 10.3 and Fig. 10.5). Since the individual layers of the bladder wall cannot be distinguished and signal differences between a tumor and the wall are usually small, MRI allows only a two-way distinction between tumors confined to the bladder wall and those extending through the wall. After intravenous administration of contrast medium, bladder cancer shows early significant selective enhancement, which improves differentiation from the normal bladder wall. Contrast-enhanced MRI may occasionally help differentiate a superficial tumor (T1) (Figs. 10.3 and 10.4) from muscle invasion (T2). The initial sign of a muscle-invasive tumor is a plateaulike wall configuration. Tumor extension through the bladder wall (T3) is indicated by decreased signal intensity of the adjacent perivesical fatty tissue (Figs. 10.5 and 10.6). Spread to adjacent organs can be demonstrated by T2w sequences or contrast-enhanced imaging. T2w images have the advantage of visualizing the zonal anatomy of the prostate and uterus and the grapelike configuration of the seminal vesicles. As with CT, obliteration of the urethrovesical or seminal vesicle angle is the first sign of seminal vesicle involvement with a loss of signal on T2w images indicating infiltration (Figs. 10.7 and 10.8). MRI is more accurate in staging bladder cancer than CT. On one hand, the multiplanar capabilities of MRI improve the evaluation of the trigone and dome compared with axial CT. However, this advantage is lost when multislice spiral CT with the option of multiplanar reconstruction is used. On the other hand, CT cannot distinguish tumor tissue from the normal bladder wall because they have the same attenuation. CT and unenhanced MRI have nearly the same accuracies, 73–89 %, in differentiating tumors confined to the bladder wall (≤T2b) from tumors extending through the wall (≥T3a) . Contrast administration can improve the differentiation between T2 and T3a tumors and in particular that between T1 and T2a tumors,

MRI Appearance of Pathologic Entities

213

a, b

a

c, d

r Fig. 10.3a–d Papillary bladder carcinoma, stage T1. a, b Axial T1w SE (a) and axial T2w TSE (b) images obtained before contrast administration. c, d Axial T1w SE (c) and sagittal T2w TSE (d) images obtained after IV bolus injection of nonspecific Gd-based contrast medium. On T1w image (a) the lesion is similar in SI to muscle and indistinguishable from the bladder wall. On T2w image (b) the tumor is well delineated from both the high-SI urine and the lowSI bladder wall. Contrast between the tumor and bladder wall is poorer on delayed postcontrast T1w image (c) compared with postcontrast T2w image (d). Postcontrast images show the typical layering effect in the bladder with reversal of signal intensities on T2w image.

b

c Fig. 10.4 Papillary bladder carcinoma, stage T1. Axial T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Early postcontrast image shows selective enhancement of the tumor without enhancement of the underlying bladder wall, indicating superficial tumor without invasion of the bladder wall (arrowheads).

Fig. 10.5a–c Bladder carcinoma, stage T3. a Axial T2w SE image. b Axial T1w SE image before contrast administration. c Axial T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. T2w image shows low-SI tumor surrounded by high-SI urine; image is degraded by motion artifacts (a). Depth of bladder wall invasion cannot be determined on unenhanced T1w image (b). Postcontrast image (c) shows enhancement of the tumor, which improves delineation from the bladder wall and reveals extensive infiltration of perivesical fat (arrowheads in a and c).

214

10 The Urinary Bladder

a

b Fig. 10.6a, b Bladder carcinoma, stage T3. a Coronal T1w SE image before contrast administration. b Coronal T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Large tumor with central postbiopsy defect of the left superolateral blad-

der wall. Improved delineation of the tumor from the normal bladder wall on contrast-enhanced image (b). Invasion of perivesical fat (arrows).

Fig. 10.7a, b Bladder carcinoma, stage T4. a Sagittal T1w SE image before contrast administration. b Sagittal T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. The polypoid mass (●) at the bladder outlet is contiguous with the prostate

(P). Intense enhancement of the tumor on postcontrast image enables good delineation from unaffected bladder wall (arrowhead) and shows prostate invasion (arrows). Central zone of the prostate (*), seminal vesicle (S).

resulting in an improved overall accuracy of MRI of 83–96 %.5,10–16

mor, shows little or no enhancement on postcontrast images (Figs. 10.9 and 10.10). Occasionally, quantification of the signal intensity increase on dynamic postcontrast images may help make the distinction.10

a

b

Tumor Recurrence Patients who have undergone TUR or partial bladder resection are usually followed up by cystoscopy. MRI is used to differentiate focal wall thickening, scar tissue, and recurrent tumor. Older scar tissue (> 6 months) is of low signal intensity on T2w images and, unlike recurrent tu-

Rare Malignant Tumors Rare malignant tumors of the urinary bladder are nonepithelial in origin (rhabdomyosarcomas and leiomyosarcomas) and secondary tumors such as metastases.

MRI Appearance of Pathologic Entities

215

a

Fig. 10.8 Urethral carcinoma. Coronal T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Large mass of high SI with extensive invasion of the pelvic floor (arrows).

b

Fig. 10.10 Recurrent bladder carcinoma after repeat TUR. Coronal T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Two enhancing bladder lesions (4 mm in size) are revealed (arrows).

The most common bladder sarcoma is rhabdomyosarcoma of childhood. This tumor typically presents as a large mass and often shows invasive growth at the time of diagnosis. Sarcomas have the same MR signal characteristics as carcinomas. Central necrosis is seen especially in large tumors (Fig. 10.11). MRI is mainly used to assess tumor extent and monitor response to chemotherapy.

Fig. 10.9a, b Recurrent bladder carcinoma in the presence of trabecular hyperplasia (secondary to BPH). a Axial T1w SE image before contrast administration. b Axial T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Diffuse thickening of the bladder wall secondary to BPH with a maximum wall thickness of 2 cm. Recurrent tumor located in the right anterior wall, where a biopsy was taken (rectangular defect). Absence of perivesical fat indicates extravesical tumor extension (arrowhead in a, b). Intense enhancement of the tumor on postcontrast image (b) enables good differentiation from the thickened bladder wall (arrows).

Primary lymphoma of the urinary bladder is rare. Typically, there is bladder involvement in the setting of malignant lymphoma. Lymphomas have the same MRI features as other malignant bladder tumors, but they seem to enhance less intensely than transitional cell carcinoma (Fig. 10.12).

216

10 The Urinary Bladder

a

b Fig. 10.11a–c Botryoid rhabdomyosarcoma in a 2-year-old girl. a Axial T1w SE image before contrast administration. b, c Axial (b) and coronal (c) T1w SE images after IV bolus injection of nonspecific Gd-based contrast medium. Large grape-shaped tumor that does not extend through the bladder wall. Tumor extension beyond the wall can only be ruled out after contrast administration.

c

a

b Fig. 10.12a, b Primary lymphoma of the urinary bladder in a 54-yearold woman. a Axial T2w TSE image. b Axial T1w SE image after IV bolus injection of nonspecific Gd-based contrast medium. Large broad-based tumor of the anterolateral bladder wall on the right.

Precontrast T2w image depicts an irregular tumor consisting of two tissue components. Signal increase on postcontrast image (b) is less pronounced compared with TCC.

MRI Appearance of Pathologic Entities

References 1. Narumi Y, Kadota T, Inoue E, et al. Bladder tumors: staging with gadolinium-enhanced oblique MR imaging. Radiology 1993; 187(1):145–150 2. Beyersdorff D, Taupitz M, Giessing M, et al. [The staging of bladder tumors in MRT: the value of the intravesical application of an iron oxide-containing contrast medium in combination with high-resolution T2-weighted imaging]. Rofo 2000;172(6): 504–508 3. Tanimoto A, Yuasa Y, Imai Y, et al. Bladder tumor staging: comparison of conventional and gadolinium-enhanced dynamic MR imaging and CT. Radiology 1992;185(3):741–747 4. Hricak H. The bladder and female urethra. In: Hricak H, Carrington B, eds. MRI of the Pelvis. Norwalk: Appleton & Lange; 1991 5. Hawnaur JM, Johnson RJ, Read G, Isherwood I. Magnetic resonance imaging with Gadolinium-DTPA for assessment of bladder carcinoma and its response to treatment. Clin Radiol 1993;47(5):302–310 6. Maya MM, Slywotzky C. Urinary bladder leiomyoma: magnetic resonance imaging findings. Urol Radiol 1992;14(3):197–199 7. Langkowski JH, Nicolas V. [Pheochromocytoma of the bladder]. Rofo 1990;153(4):479–480 8. Amano T, Kunimi K, Hisazumi H, Kadoya M, Matsui O. Magnetic resonance imaging of bladder hemangioma. Abdom Imaging 1993;18(1):97–99 9. Sobin LH, Wittekind C. In: Sobin LH, Wittekind C, eds. International Union against Cancer (UICC): TNM Classification of Malignant Tumors: New York: Wiley; 2002

217

10. Barentsz JO, Jager GJ, van Vierzen PBJ, et al. Staging urinary bladder cancer after transurethral biopsy: value of fast dynamic contrast-enhanced MR imaging. Radiology 1996;201(1): 185–193 11. Beyersdorff D, Zhang J, Schöder H, Bochner B, Hricak H. Bladder cancer: can imaging change patient management? Curr Opin Urol 2008;18(1):98–104 12. Kim B, Semelka RC, Ascher SM, Chalpin DB, Carroll PR, Hricak H. Bladder tumor staging: comparison of contrast-enhanced CT, T1- and T2-weighted MR imaging, dynamic gadolinium-enhanced imaging, and late gadolinium-enhanced imaging. Radiology 1994;193(1):239–245 13. Neuerburg JM, Bohndorf K, Sohn M, Teufl F, Guenther RW, Daus HJ. Urinary bladder neoplasms: evaluation with contrast-enhanced MR imaging. Radiology 1989;172(3):739–743 14. Nicolas V, Spielmann R, Maas R, et al. [The diagnostic value of MR tomography following gadolinium-DTPA compared to computed tomography in bladder tumors]. Rofo 1990;153(2): 197–203 15. Persad R, Kabala J, Gillatt D, Penry B, Gingell JC, Smith PJ. Magnetic resonance imaging in the staging of bladder cancer. Br J Urol 1993;71(5):566–573 16. Sparenberg A, Hamm B, Hammerer P, Samberger V, Wolf KJ. [The diagnosis of bladder carcinomas by NMR tomography: an improvement with Gd-DTPA?]. Rofo 1991;155(2):117–122

219

11

The Prostate and Seminal Vesicles V. Nicolas, D. Beyersdorff, U.G. Mueller-Lisse, W. Pennekamp, and C.M. Heyer

Introduction In recent years MRI of the pelvic organs has gained increasing importance. Technical refinements and the advent of magnetic resonance spectroscopy (MRS) have markedly improved the sensitivity and specificity of MRI in imaging the prostate. But only through the combination of an endorectal coil with a body or torso phased-array coil has it become possible to obtain detailed images of the zonal anatomy of the prostate, which are the basis for evaluating pathology.

Indications The indications for prostate MRI derive directly from the findings of the basic clinical examinations, digital rectal examination (DRE), and transrectal ultrasound (TRUS). In the setting of prostate cancer, MRI is used for the staging of known cancer or as a problem-solving tool, for example in patients with elevated prostate-specific antigen (PSA) but inconclusive palpation findings or normal TRUS, or in patients with increasing PSA levels and repeated negative biopsies. In addition, MRI is used to detect recurrent cancer after radical prostatectomy and to determine the extent of prostate sarcoma. The sensitivity of prostate MRI can be improved by additional MR spectroscopy and a dynamic contrast-enhanced study, especially in terms of differentiating malignancies from inflammatory conditions, fibrosis, and prostatic intraepithelial neoplasia (PIN). It is also expected that MRI can help lower the false-negative rate of prostate biopsy by enabling better biopsy guidance on the basis of the combined findings of T2w sequences, dynamic contrast-enhanced imaging, and MR spectroscopy.

Fig. 11.1 Endorectal coil.

ever, the signal-to-noise ratio (SNR) was inadequate for high-resolution prostate imaging. Only thanks to the introduction of endorectal coils (Fig. 11.1), which are now state of the art in prostate imaging, used in combination with a body or torso phased-array coil, has it become possible to obtain detailed images of the prostate and surrounding structures1,2. The patient is examined in the supine position. After a DRE, the endorectal coil is inserted and kept in place at the level of the prostate by inflating the surrounding balloon with a total volume of 60–100 mL until the patient reports mild pressure. To secure the coil firmly in place, it may be additionally fixed to the thigh. Next, the body phasedarray coil is placed. Making the patient comfortable, e. g., with the knees supported, can help reduce motion artifacts. Respiratory artifacts are minimized by using an abdominal belt, which is fitted above the symphysis pubis with slight pressure. Artifacts caused by bowel movement are minimized by intravascular/intramuscular injection of butylscopolamine or glucagon (after excluding contraindictions).

Imaging Planes

Imaging Technique When MRI was first introduced, only the whole-body resonator was available for imaging of the prostate. The resulting images provided good overviews of the pelvic organs, lymphatic drainage, and bony structures; how-

Accurate slice prescription for axial imaging, e. g., parallel to the acetabular roof, is important for symmetrical visualization of the peripheral zone and facilitates image interpretation. Coronal images should be acquired parallel to the longitudinal axis of the seminal vesicles, which ensures optimal evaluation of the base of the seminal

– – – – ca. 5 min ca. 5 min ca. 5 min ca. 5 min 3 3 3 5 3 3 3 3 16 16 16 16 162 162 162 120 230 × 256 230 × 256 230 × 256 512 × 360 TSE TSE TSE TSE T1 T2 T2 PD

Axial Axial Coronal Axial

617 4660 4000 1300

13 100 100 10

3 13 13 3

– – – –

FOV (mm) Matrix FS ETL TE (ms) TR (ms) Sequence type Plane Weighting

Table 11.1 Recommended pulse sequences and imaging parameters for MRI of the prostate

No. of slices

No. of acquisitions

Slice thickness (mm)

Breathhold

11 The Prostate and Seminal Vesicles

Scan time

220

vesicles, the vasa deferentia, and seminal vesicle involvement in patients with peripheral zone cancer. Only slight angulation is required in most cases because the rectum is expanded by the coil. If the axial and coronal images are inconclusive (e. g., bladder neck), additional sagittal images may be acquired to supplement the examination.

Pulse Sequences T2w TSE images in at least two planes are acquired to evaluate prostate zonal anatomy and localize pathologic processes (Table 11.1). Slices covering the prostate should not be > 3 mm thick. GRE sequences are inferior for delineation of prostate cancer and its zonal localization. Additional T1w images, acquired with fat suppression in selected cases, are needed to differentiate fatty tissue, demonstrate postbiopsy changes, and evaluate locoregional lymph nodes. The following sequence protocol is recommended for MRI of the prostate: · multiplanar scout view (localizer) · axial T2w TSE sequence · coronal T2w TSE sequence parallel to the longitudinal axis of the seminal vesicles · axial T1w TSE sequence · axial PD TSE sequence up to the aortic bifurcation.

Contrast Media No contrast medium is needed for a standard MRI examination of the prostate; however, a contrast-enhanced scan may occasionally provide useful information on inflammatory conditions and malignant tumors (prostate cancer with suspected rectal or bladder involvement, sarcoma). Contrast-enhanced MRI of the prostate is performed using a nonspecific extracellular Gd-based contrast medium at a dose of 0.1 mmol Gd per kg body weight. A dynamic MRI study consists of a series of T1w images of the prostate acquired at short intervals after contrast administration. Quantitative analysis of the enhancement slope and peak enhancement may improve the differentiation of normal and abnormal tissue in individual cases. Dynamic prostate MRI can be performed with an inversion recovery TurboFLASH sequence (TR/TE, 1300 ms/4.2 ms; TI, 654 ms; flip angle, 13°; field of view [FOV], 360 mm; 128 × 128 matrix). Acquisition of 10 4-mm slices results in a scan time of 13 s for each phase.2–5

MRI Appearance of Normal Anatomy The normal adult prostate is roughly the size and shape of a chestnut and weighs ca. 15–20 g. The prostate gland surrounds the urethra between the neck of the urinary bladder and the urogenital diaphragm. In shape it is like

MRI Appearance of Normal Anatomy

 %

 %

 &"

 "$%$&

56

0'

221

13%4 &%

0'

5 6

5 6

 %% #'6

 "$% $&

#'6

Fig. 11.2 Coronal, sagittal, and axial diagrams of prostate anatomy.

an inverted pyramid with its base directed upward and the apex pointing down. Laterally, the prostate borders on the middle portions of the levator ani muscles. Each of the paired seminal vesicles is approximately 5–6 cm long and 1 cm wide and extends from the top of the prostate in a superolateral direction lateral to the vasa deferentia. The prostate is surrounded by the visceral leaf of the abdominal fascias. It is separated from the rectum by the Denonvilliers fascia, which consists of a layer that is contiguous with the external longitudinal rectal muscles and a thicker fibroelastic membrane that covers the entire posterior surface of the prostate and extends upward to above the seminal vesicles and the rectovesical pouch, the lowermost extension of the peritoneal cavity. Anterolaterally, there is an extensive venous complex. At the prostate base, the neurovascular pathways divide into an upper bundle (the neurovascular bundle), whose nerve fibers penetrate the prostate capsule on the posterolateral aspect, and a lower bundle, which courses to the apex and the corpora cavernosa through the Denonvilliers fascia. The widely accepted zonal subdivision of the glandular parts of the prostate goes back to the detailed anatomic and histologic studies presented by McNeal et al. in 1972.6 The anatomic landmark is the urethra, dividing the prostate into an anterior fibromuscular part and a posterolateral glandular part. The prostatic urethra turns 35° at its midpoint in the gland (the urethral angle). The ejaculatory ducts run parallel to the urethra and open into it in the area of the seminal colliculus. The glandular portion of the prostate is subdivided into three zones: · The wedge-shaped central zone surrounds the ejaculatory ducts and extends from the seminal colliculus to behind the neck of the bladder. · The peripheral zone comprises up to 75 % of the glandular volume of the normal prostate. It is contiguous with the central zone at the base, while its distal portion extends from below the seminal colliculus to the apex, surrounding the urethra.

· The transitional zone is the portion at the level of the proximal urethra above the seminal colliculus and consists of two lobes. The anterior third of the prostate consists of fibromuscular tissue (anterior fibromuscular band) that extends from the bladder neck to the apex (Fig. 11.2). On T1w images, the normal prostate is of homogeneous intermediate signal intensity, slightly higher than that of muscle. The zonal anatomy is best appreciated on T2w images. The peripheral zone is seen as a sickle-shaped area of homogeneous high signal intensity on axial images. The central gland comprising the central and transitional zones contains areas of high and low signal intensity. The transitional zone cannot be reliably differentiated, for two reasons: on the one hand, it constitutes only a small portion of the entire prostate (5–10 %) and, on the other hand, it is identical to the peripheral zone in composition because both zones have the same embryonic origin. The signal difference between the peripheral zone and the central glandular tissue is due to their different histologic makeup. While the stromal tissue of the central zone contains acini surrounded by long, densely packed smooth muscle fibers, the acini in the peripheral zone are small and thin-walled and surrounded by only a few muscle fibers. The low signal intensity of the central zone is thus attributable to the high content of smooth muscle.7 At the same time, differences in histologic composition suggest functional differences in secretion or fluid content. The anterior fibromuscular band is isointense to muscle on T2w images. When an endorectal coil is used, the prostate capsule is seen as a line of low signal intensity around the high-signal-intensity peripheral zone in the lower and mid portions of the prostate. On T2w TSE images, the capsule is also delineated from the high-signal-intensity fat. The neurovascular bundle is seen as a triangle, in which the nerves and vessels may be discernible as punctiform structures (Fig. 11.3).

222

11 The Prostate and Seminal Vesicles

Table 11.2 summarizes the signal intensities of the prostate and adjacent tissues on T2w and T1w images.

(1H) MR Spectroscopy

Fig. 11.3 Prostate cancer, stage T2a. Axial T2w TSE image. Circumscribed hypointense tumor focus in the right peripheral zone (large arrow) and heterogeneous central gland with areas of high and low SI, consistent with BPH. The prostate capsule and neurovascular bundle (arrowhead) appear normal. Also seen is the pseudocapsule (small arrows). Table 11.2 MRI signal intensities of the prostate

Peripheral zone Central gland Prostate capsule Seminal vesicles Fatty tissue Muscle Benign prostatic hyperplasia Tumor Postbiopsy hemorrhage

T2w

T1w

++ +/− – ++ + – +/− – +

0 0 Not seen 0 ++ 0 0 Not seen ++

The seminal vesicles have the same signal intensity as the prostate on T1w images. Their lobulated glandular structure is only seen on T2w images. Intraluminal signal intensity is higher than that of muscle in adult men and decreases slightly with age. The borders of the tubules and the capsule have low signal intensity.

Three-dimensional MR spectroscopy (3D MRS), also known as chemical shift imaging (CSI) and MR spectroscopic imaging (MRSI), is predominantly performed using spin echo sequences (SE; PRESS—point-resolved spectroscopy)8,9 or STEAM sequences (stimulated echo acquisition mode)10 for excitation and readout as well as spatial encoding. In general, STEAM sequences are used when TEs < 30 ms are desired, while PRESS and SE sequences are preferred for examinations with longer TEs.11 In 3D MRS, the number of adjacent voxels in the different spatial dimensions is determined by the number of phase-encoding steps in each direction (Fig. 11.4). The SNR improves with each phase-encoding step, but the duration of the examination also increases in direct proportion to the number of steps. In MR spectroscopy, the received signal is decomposed into its frequencies and displayed in the form of a frequency spectrum (Fig. 11.5). The resonance frequency of characteristic hydrogen bonds of a metabolic product is given using the chemical shift of trimethylsilyl propionate (TSP) as reference at a given field strength and is expressed in parts per million (ppm). Using the ppm scale, it is easier to identify a metabolite because it is independent of magnetic flow density and resonance frequencies are always at the same site in each 1H-MR spectrum. In this way it is possible to identify a biochemical substance in vivo if its magnetic resonance frequency is known from in-vitro studies. If the strong signals of water and fat are suppressed by applying frequency-selective saturation pulses, one can obtain frequency spectra that represent the peaks of metabolic products of which only small concentrations are present. There are three main metabolites that are of interest in the prostate: · citrate, which is an intermediate of the Krebs cycle and is accumulated in the prostate and secreted into the glandular ducts

Fig. 11.4a, b MR spectroscopy and MR image with superimposed MR spectral array. a The white arrows indicate the directions of the phase-encoding steps in the plane of the MR image. b Example of an MR spectrum showing the peaks of citrate (Cit), creatinine (Cr), and choline (Cho).

a

b

223

MRI Appearance of Pathologic Entities

a

b

Fig. 11.5a, b The MR signal (a) serves to generate the frequency spectrum (b) of MR spectroscopy by means of Fourier transform. Invivo MR spectroscopy of the prostate does not allow direct quanti-

fication. The ratio of the area integrals of choline (black) plus creatine (hatched) to citrate (dotted) differentiates normal tissue from prostate cancer in the peripheral zone.

· choline, which arises in association with the composi-

ing the space where citrate can be found. The higher cellular turnover of tumor tissue increases the concentration of free, choline-rich molecules in the cytosol and interstitial space,24 which are important components of cell membranes. 1H MR spectroscopy of the prostate differentiates normal prostate tissue with high citrate peaks and low choline peaks from cancer tissue with low citrate peaks and elevated choline.12–14,18,25

·

tion/decomposition of cell membranes creatine, a component of high-energy phosphates.12–18

In-vivo detection of these metabolic products requires not only suppression of fat and water signals but also shimming by additional small magnetic fields to eliminate inhomogeneities of the specimen or patient. The SNR is improved by multiple acquisitions or phase-encoding steps. Saturation bands placed around the target volume will reduce signal contamination from outside the volume.8,19–21 The signal acquired in MRS in the form of FID (free induction decay) needs to be postprocessed, typically in several steps: zero filling to improve digital frequency resolution, apodization to enhance the spectral signal versus electrostatic noise by means of Gauss or Lorentz filtering, Fourier transform into a frequency spectrum, and correction of the baseline and phase of the frequency spectrum.11

MR Spectroscopy of the Prostate: Biochemical Basis Normal epithelial cells of the human prostate accumulate citrate from the Krebs cycle. The citrate is stored in the epithelial cells and glandular ducts in the presence of zinc ions.13,22,23 For this reason the citrate concentration is very high in the peripheral zone and very low in the central zone. In the transitional zone, which is the site of benign prostatic hyperplasia (BPH), the citrate concentration is the same as in the peripheral zone if the glandular component of BPH is large but markedly lower if there is a large stromal component. With decreasing cellular differentiation, prostate cancer loses the enzymes that are necessary for storing and secreting citrate. The higher the cell turnover of cancer, the more citrate is introduced into energy metabolism.13,22 Moreover, prostate cancer compresses or infiltrates the glandular ducts, thereby reduc-

MRI Appearance of Pathologic Entities Anomalies Congenital prostate anomalies such as agenesis and hypoplasia are frequently associated with other anomalies of the urogenital tract. There may be complete absence of the prostate or the proportion of glandular tissue is reduced. Congenital prostate cysts are the result of anomalous development of remnants of the müllerian or wolffian ducts. The most common cysts are retention cysts that develop in the prostatic acini. Other causes are prostatitis and trauma.26 MRI permits accurate determination of the site and size of prostatic cysts.27 Müllerian duct cysts, which are attached to the verumontanum by a connective tissue stalk, and utricle cysts, which may communicate with the urethra, are midline lesions. The signal intensity of the cyst fluid can vary, for example, when there is superinfection or intralesional hemorrhage. Uncomplicated cysts containing serous fluid are of the same signal intensity as urine on T1w and T2w images (Fig. 11.6). In agenesis of the seminal vesicles, one or both seminal vesicles may be absent. The condition is typically associated with renal agenesis, absence of the vas deferens, or a congenital vasoureteral communication. Less common are ectopic seminal vesicles and ectopic openings of the ureters and seminal vesicle ducts.

224

11 The Prostate and Seminal Vesicles

a

b Fig. 11.6a–c Hemorrhagic prostatic cyst. a Axial T1w SE image. b Axial T2w TSE image. c Coronal T2w TSE image. MRI was performed because of a sonographically indeterminate prostatic lesion. T1w image demonstrates a well-defined lesion of high SI. On T2w images, the lesion has low SI compared with the surrounding peripheral zone (●). Decreased SI of the vas deferens (small arrow) and left seminal vesicle (large arrow) secondary to postbiopsy hemorrhage.

c

Inflammatory Conditions Bacterial prostatitis (typically caused by E. coli, gonococci, staphylococci, or streptococci) is most commonly due to an ascending infection through the urethra or a descending infection via the urinary bladder or the vas deferens, or occurs after prior surgery. Hematogenous spread is less common. A prostatic abscess usually develops as a complication of acute prostatitis. The diagnosis is established clinically and sonographically. MRI does not provide any clinically relevant diagnostic information, and the changes seen on MRI are nonspecific without knowledge of the patient’s clinical signs and symptoms. Use of an endorectal coil is contraindicated in acute prostatitis because placement may lead to dissemination of pathogens in the bloodstream.

Acute prostatitis is seen as diffuse enlargement of the prostate on T1w images. On T2w images, inflamed tissue has high signal intensity and may be difficult to distinguish from the high-signal-intensity peripheral zone. Prostatic abscess is depicted as a focal lesion of low signal intensity on T1w images and high signal intensity on T2w images.28 Chronic prostatitis leads to connective tissue proliferation with scar formation in the prostate, seen as low-signal-intensity areas on T2w images. Such scars cannot be differentiated from cancer on the basis of their imaging appearance and signal intensity when they are located in the peripheral zone. Inflammation of the seminal vesicles typically occurs secondary to prostatitis. The MR appearance is variable. In the acute stage, there is enlargement of the seminal vesicle with normal T1 signal intensity. Hemospermia

MRI Appearance of Pathologic Entities

225

a

b Fig. 11.7a, b Two cases of stage T3a prostate cancer with BPH. Axial T2w TSE images. a Large prostate carcinoma in the posterior aspect of the right peripheral zone (T) with extracapsular extension (lower arrow) along the neurovascular bundle (arrowhead). Mixed BPH. Upper arrow indicates the prostate capsule. b Image from second patient shows localized carcinoma in the posterior aspect

of the left peripheral zone (arrow) with extracapsular extension along the neurovascular bundle. Compression of the right peripheral zone (●) by extensive, predominantly cystic, BPH, which is separated from the peripheral zone by a pseudocapsule (open arrow).

Fig. 11.8a, b BPH. Axial (a) and coronal (b) T2w TSE images. There is diffuse enlargement of the prostate with cystic components, resulting in heterogeneous appearance of the central gland. BPH

is separated from the compressed peripheral zone (PZ in b) by a pseudocapsule (arrowheads in a, arrows in b). Arrows in a indicate the outer capsule of the prostate.

leads to an increase in signal intensity. T2 signal intensity is increased in acute inflammation, while chronic seminal vesicle inflammation results in a loss of signal intensity on both T1w and T2w images.

BPH has a characteristic MR appearance, but definitive tissue characterization is not possible. The signal intensity varies considerably, depending on the ratio of glandular to stromal tissue. T2w images show multiple high-signalintensity nodules, representing hyperplastic glands with retained secretions. The nodules are surrounded by a lowsignal-intensity pseudocapsule (Figs. 11.7 and 11.8). More areas of low signal intensity are depicted in BPH with a larger fibrous and muscular component. In extensive BPH, there may be marked compression of the peripheral zone, leaving only a thin band of high signal intensity on MRI. The MRI signal intensity and appearance alone do not always allow reliable differentiation of BPH from a malignant process.

a

b

Benign Prostatic Hyperplasia BPH begins as a localized proliferation in the periurethral glands and the transitional zone. As the condition progresses, there will be compression of adjacent normal glandular tissue and the urethra with subsequent urinary obstruction.

226

11 The Prostate and Seminal Vesicles

Malignant Tumors Prostate Cancer Prostate cancer is the most common malignant tumor in men and the second most common cause of cancer death. The increasing incidence observed in recent years is mainly attributable to the introduction of the PSA test in the mid-1980s and the wider acceptance of clinical screening. It is estimated that up to 50 % of men have prostate cancer at age 50 and ca. 90 % at age 80. The cumulative risk of being diagnosed with prostate cancer is 24 % until age 85. The survival rate of prostate cancer patients has improved from 67 % to 93 % over the past 20 years: 72 % of all patients survive for over 10 years, 53 % for over 15 years.29 Various factors have contributed to this development including earlier detection of prostate cancer, better radiotherapeutic approaches, improved surgical techniques, and antiandrogen treatment. Knowledge of the site and local extent of prostate cancer as well as presence of lymphadenopathy and/or distant metastases is crucial to devising the most suitable therapeutic strategy. The basic urologic examinations comprise DRE, TRUS, and PSA determination. If cancer is suspected, a systematic biopsy will be performed.30 Depending on patient age and concomitant diseases, radical prostatectomy is the first-line treatment for local cancer confined within the prostate, while more advanced cancer is treated by radiotherapy, possibly in combination with androgen blockade. The role of imaging is to define local tumor extent and rule out secondary tumor foci. The vast majority of malignant prostate tumors are adenocarcinomas, accounting for up to 97 % of cases. The prognosis depends on the tumor volume,31 the TNM stage, the Gleason score,32 and the PSA level.33 A general criterion of a malignant process is neovascularization, which is seen when there is transition from hyperplasia to neoplasia.21,34 Table 11.3 TNM classification of prostate cancer T1

T1c

Clinically inapparent tumor not palpable or visible by imaging Tumor incidental histologic finding in ≤ 5 % of tissue resected Tumor incidental histologic finding in > 5 % of tissue resected Tumor identified by needle biopsy

T2 T2a T2b T2c

Tumor confined within prostate Tumor involves ≤ 50 % of one lobe Tumor involves > 50 % of one lobe Tumor involves both lobes

T3 T3a T3b

Tumor extends through the prostate capsule Extracapsular extension (unilateral or bilateral) Tumor invades seminal vesicle(s)

T4

Tumor is fixed or invades adjacent structures other than seminal vesicles

T1a T1b

Up to 70 % of all prostate carcinomas arise in the peripheral zone, most commonly in the posterolateral aspect, 20 % in the transitional zone, and 10 % in the central zone. Staging of Prostate Cancer Prostate cancer is staged according to the TNM staging classification, last updated in 2003 (Table 11.3). Stage T1 tumors are so small that they are not detectable by clinical methods or imaging. Most stage 1 cancers are incidentally detected by histologic examination, e. g., in patients who undergo transurethral resection of the prostate (TURP) for BPH. If there is tumor in ≤ 5 % of the specimen resected, the stage is T1a, and if tumor tissue accounts for > 5 % of the specimen it is T1b. Stage T1c is defined as tumor identified by needle biopsy, e. g., because of elevated PSA. Stage T2 tumors are confined to the prostate. Stage T2a involves half of one lobe or less and stage T2b more than half of one lobe; stage T2c is defined as tumor involving both lobes. Stage T3 tumors extend through the prostate capsule on one or both sides with invasion of periprostatic tissue (T3a) and/or invasion of one or both seminal vesicles (T3b). Stage T4 tumors are fixed to the pelvic sidewall or invade adjacent structures other than the seminal vesicles. These include the bladder neck, external sphincter, rectum/sigmoid colon, and levator muscles. Endorectal Coil MRI of Prostate Cancer The signal intensity and detectability of prostate cancer by MRI depend on its localization within the gland and the pulse sequence used. T1 signal intensity is similar to or slightly lower than that of normal glandular tissue. Prostate cancer in the high-signal-intensity peripheral zone is revealed as a low-signal-intensity lesion on T2w images. The low signal intensity is due to the compact arrangement of cellular elements with only small amounts of mucin or fluid. Cancer extent within the prostate can be difficult to define if there is concomitant BPH with a predominant stromal component because both entities may have similar signal intensities. Additional cancer foci in the central gland may therefore be difficult to detect in this setting. Carcinomas arising in the transitional zone are relatively large at the time of diagnosis because they may escape detection by DRE. As with cancer in the peripheral zone, they typically have uniform low signal intensity on T2w images (Figs. 11.9 and 11.10). The lack of a characteristic signal intensity precludes the use of MRI as a screening test for prostate cancer. The MRI diagnosis is solely based on the site of the cancer within the prostate, i. e., the presence of a low-signalintensity lesion in the peripheral zone. In patients with clinically suspected prostate cancer and inconclusive DRE and negative ultrasound findings, MRI can help determine the most suitable site for biopsy.

MRI Appearance of Pathologic Entities

Fig. 11.9 Prostate cancer in the transitional zone after TURP. Axial T2w TSE image. Large tumor (T) with homogeneous SI in the center. Contiguous with the prostatic urethra (●). Invasion of periprostatic tissue on the left (arrow).

227

Fig. 11.10 Multifocal prostate cancer in the transitional zone. Coronal T2w TSE images. Tumor of homogeneous signal intensity (T) in the central gland with displacement of the prostatic urethra. Multiple additional tumor foci in the left peripheral zone (dashes).

Fig. 11.11 Extensive bilateral hemorrhage (arrows) after biopsy. Axial T1w SE image.

Hemorrhage after prostate biopsy will alter the MR signal, which must be borne in mind when interpreting MR images.4,35 Postbiopsy hemorrhage is of low signal intensity on T2w images and may suggest prostate cancer. If the findings are unclear, the extent of hemorrhage can be assessed on T1w images (Figs. 11.11 and 11.12). The same holds true for hemorrhage in the seminal vesicles. Since such postbiopsy changes limit the diagnostic yield of MRI, it is advisable to perform MRI before prostate biopsy or wait 6 weeks after biopsy. The criteria employed to assess local tumor extent are similar to those used for other organ cancers and include

appearance of the prostate contour, intactness of the capsule, signal intensity changes of peritumoral tissues, and delineation from surrounding structures.36–39 Evaluation of the prostatic capsule is important to differentiate cancer confined within the gland from extracapsular infiltration. Errors in distinguishing stage 2 from stage 3 on conventional MRI are due to the poor differentiation of cancer from the prostatic capsule. Differentiation is markedly improved when an endorectal coil is used. Outwater et al.36 proposed criteria for capsular penetration. These include indirect signs such as retraction, capsular thickening, and prostatic bulge and direct signs of extracapsular

228

11 The Prostate and Seminal Vesicles

a

b Fig. 11.12a, b Fresh postbiopsy hemorrhage in the left seminal vesicle (S), vas deferens (D), and paravesical tissue (B). Axial (a) and coronal (b) T2w TSE images.

Fig. 11.13a–d Prostate cancer, stage T3a. Axial T2w TSE images. Multifocal prostate cancer (arrows). Images show various signs of capsular penetration: bulging (a), stranding (b), extracapsular tumor (c and d), and retraction (d).

a

b

c

d

tumor such as stranding and tissue of tumor signal intensity outside the normal confines of the prostate (Figs. 11.13, 11.14, 11.15). Cancer spread beyond the prostate occurs along the nerve fibers penetrating the capsule including the neurovascular bundle.40 Small solitary tumors confined within the prostate can be operated on by nervesparing prostatectomy. In these cases, MRI enables accurate evaluation of the unaffected side (normal peripheral zone and intact capsule) including the neurovascular bundle to be preserved. MRI can identify early seminal vesicle invasion and is therefore superior in this respect to other imaging modalities such as CT and TRUS. Involvement is revealed as diminished T2 signal intensity of the affected seminal vesicle with loss of the normal lobulated appearance before changes in configuration or size become apparent.

There are two routes of spread of prostate cancer to the seminal vesicles, either by continuous growth along the seminal colliculus or by gross extracapsular extension along the seminal vesicles. Reliable evaluation for seminal vesicle invasion requires imaging in two planes (Figs. 11.14, 11.15, 11.16, 11.17). An imaging plane parallel to the long axis of the seminal vesicle enables simultaneous evaluation of cancer in the peripheral zone and possible seminal vesicle invasion. Stage T4 is a fixed tumor with invasion of adjacent organs such as the urinary bladder or rectum. T2w images enable adequate evaluation of the local tumor and extension to adjacent structures. Bladder invasion is seen on T2w images as a disruption of the low-signal-intensity wall, which is clearly delineated from high-signal-intensity urine. Similarly, rectal invasion is seen as disruption of

MRI Appearance of Pathologic Entities

229

b

a

Fig. 11.14a, b Prostate cancer, stage T3b. Axial T2w TSE images acquired with combined endorectal/phased-array coils. a Low SI (T) of most of the left peripheral zone and invasion of the periprostatic fat along the neurovascular bundle. b Invasion of both seminal vesicles (T).

v Fig. 11.15a–d Prostate cancer, stage T3b. a–c Axial T2w TSE images through the prostate apex (a), midgland (b), and base (c). Large cancer (arrows) in the left peripheral zone with transcapsular extension into the periprostatic fat. d Two routes of cancer spread to the left seminal vesicle: along the vas deferens (white arrow) and extension of extracapsular cancer into the seminal vesicle (black arrow). a

b

c

d

Fig. 11.16 Prostate cancer, stage T3b. Coronal T2w TSE image. Diffuse bilateral prostate carcinoma (T) with complete replacement of the right seminal vesicle (arrow) and infiltration of the proximal third of the left seminal vesicle with distal congestion (arrow). Arrowheads indicate vasa deferentia.

230

11 The Prostate and Seminal Vesicles

a

b

c

d Fig. 11.17a–d Prostate cancer, stage T3b, with bone metastasis. a–c Axial T2w images. d Coronal T2w image. Large bilateral prostate carcinoma (upper arrows in a, b) with invasion of periprostatic fatty

tissue (lower arrows in a, b) and both seminal vesicles (arrows in c and arrowheads in d). Black arrow (d) indicates vas deferens.

the Denonvilliers fascia and high-signal-intensity interruption of the low-signal-intensity rectal muscle layer. As reported for bladder cancer, contrast-enhanced T1w imaging can occasionally be helpful in identifying organ invasion by improving delineation of tumor from the bladder or rectal wall (Figs. 11.18 and 11.19). MR images of the prostate cannot be interpreted without taking into account the patient’s clinical findings, most recent PSA and course of PSA levels, age, and prior treatment (hormone therapy, TUR, radiotherapy, biopsy).41–44 However, in a study by Dhingsa et al. (2004), knowledge of clinical data did not improve sensitivity but rather tended to increase the rate of false-positive findings.45 Another important factor influencing results is the radiologist’s experience with the interpretation of prostate MR images acquired with use of an endorectal coil.46–48 Promising results regarding the detection of prostate cancer have been reported for use of the endo-

rectal coil alone in patients with increased PSA and prior negative core biopsy. In this subset of patients, MRI was found to detect prostate cancer with an 85 % sensitivity, while the absence of suspicious findings on MRI had a 94.4 % negative predictive value for absence of cancer in a subsequent biopsy.49,50 MRI has an accuracy ranging between 55 % and 91 % for evaluating local tumor extent (confined within prostate vs capsular penetration). This wide variation is due to the fact that the signal intensity of prostate cancer and criteria for extracapsular extension are nonspecific and that the patient populations investigated differed in terms of prostate cancer size and capsular penetration. The sensitivity of MRI for identifying seminal vesicle invasion is 90–95 % with a specificity of up to 90 %.38,46,51–53 Although whether contrast-enhanced MRI can improve tissue characterization and local tumor staging is still an open question, particularly with regard to distinguishing

MRI Appearance of Pathologic Entities

231

a

b Fig. 11.18a, b Prostate cancer, stage T4a. a Axial T2w SE image shows large carcinoma (T) with extensive invasion of the urinary bladder (arrows). b Sagittal T1w SE image after IV administration of

a

nonspecific Gd-based contrast medium (body coil) additionally reveals infiltration of the rectum (arrowhead) and sigmoid colon (large arrows). Small arrows indicate bladder invasion.

b Fig. 11.19a–c Prostate cancer, stage T4; hematuria. a Axial T2w TSE image shows low-SI tumor with extensive contact to the posterior bladder wall. Obliteration of the vesicourethral angle/seminal vesicle angle (arrowhead). Complete invasion of the left and partial invasion of the right seminal vesicle (arrow). b Axial T1w SE image with fat suppression after contrast administration shows tumor of high SI relative to the uninvolved bladder wall. c Axial TrueFISP image after contrast administration illustrates poor delineation of cancer from the normal bladder wall and loss of differentiation of the internal structure of the preserved portion of the right seminal vesicle.

c

232

11 The Prostate and Seminal Vesicles

of criteria such as washout, determination of relative peak enhancement, and the use of postprocessing algorithms.3,17

a

b Fig. 11.20a, b Biopsy-proven prostate cancer in a 62-year-old man with a PSA of 7 ng/ml. a Axial T2w TSE image reveals a circumscribed low-SI area in the posterolateral aspect of the left peripheral zone. b Comparison with the contralateral side on dynamic MRI following IV bolus injection of contrast medium reveals a steeper increase of the SI curve and higher peak (from ref. 41).

between T2 and T3 tumors, promising results have been achieved with dynamic contrast-enhanced MR studies in delineating prostate cancer from normal glandular tissue.54–56 Prostate cancer tends to show a steeper increase in signal intensity and higher peak enhancement compared with normal glandular tissue in the peripheral zone (Fig. 11.20). In one study, the sensitivity for demonstrating prostate cancer in the peripheral zone was found to increase from 76 % to 83 % when the time course of signal enhancement and peak signal intensity were analyzed; however, local tumor staging was not significantly improved compared with nonenhanced T2w images.57 Other approaches that have been proposed to improve MR tissue characterization in the prostate are evaluation

Role of (1H) MR Spectroscopy in Prostate Cancer Clinical MR spectroscopy (MRS) for the differentiation of normal prostate tissue from cancer usually relies on the ratio of choline plus creatine to citrate ([Cho+Cr]/Cit ratio) since it is not possible to directly quantify the areas under the peaks of individual metabolic products.12,13,58 Analysis of MR spectra from the peripheral zone (where ca. 70–75 % of prostate cancers are located) has revealed that the [Cho+Cr]/Cit ratio is more than three standard deviations above the average ratio of normal prostate tissue in up to 96 % of patients with prostate cancer. In a study of 53 patients, it has been shown that additional 3D MRS significantly improves the accuracy of sextant biopsy in detecting or excluding prostate cancer by 10 %, irrespective of the radiologist’s experience with prostate MRI. If both MRI and 3D MRS suggest cancer, the positive predictive value is 89–92 %. If neither of the two tests is indicative of cancer, the negative predictive value is 74–82 %15 (Figs. 11.21 and 11.22). A study of 47 patients showed that combined MRI/3D MRS improves sextant localization of prostate cancer compared with core biopsy, especially in the apex of the prostate.59 In 49 patients with postbiopsy hemorrhage, combination of MRI and MRS increased specificity from 26 % to 66 % and accuracy from 52 % to 75 %.60 MRI has a specificity of up to 95 % for excluding extracapsular tumor extension, while its sensitivity is only 17–54 %. At MRS, the likelihood of extracapsular tumor increases with the number of suspicious spectra in the peripheral zone, significantly improving the detection of extracapsular prostate cancer by 8–13 %.61 Combined MRI/MRS is also superior to MRI alone in estimating cancer volume if the volume is at least 0.5 cm3, as demonstrated in a study of 37 patients.62 The [Cho+Cr]/Cit ratios vary widely in the abnormal range three standard deviations above the average of normal prostate tissue.12,14,16 Initial results show that choline levels in prostate cancer increase with decreasing tumor cell differentiation,63 suggesting that MRS may have the potential to assess tumor aggressiveness. Recent studies indicate that 3D MRS improves coverage of the prostate if the target volume is imaged with angulation in two dimensions, and also that the technique can be used on 3-T whole-body MR scanners. Being noninvasive, MRS is a candidate for following up metabolic changes occurring in the prostate in response to hormone treatment, radiotherapy, or cryotherapy. Dihydrotestosterone controls not only cell proliferation and apoptosis and PSA expression but also the production, accumulation, and secretion of citrate in the prostate. Hormone deprivation therapies that either suppress testosterone production (castration, LHRH agonists) or competitively block dihydrotestosterone-binding sites in the prostate (antiandrogens) reduce prostatic citrate levels.16,23 Citrate appears to decrease earlier or more

MRI Appearance of Pathologic Entities

233

Fig. 11.21a, b 1H MR spectroscopy in a 72-year-old man with a PSA increase from 3.0 to 4.5 ng/mL over a period of ca. 2 years; no biopsy; negative DRE and negative TRUS of the prostate. a Axial T2w image with superimposed spectral array. b Example of spectrum suggesting prostate cancer. The magnified spectrum was obtained in the right posterior aspect of the prostate (indicated by arrowhead in a).

Fig. 11.22a, b 1H MR spectroscopy in a 72-year-old man with a PSA increase from 3.0 to 4.5 ng/mL over a period of ca. 2 years; no biopsy; negative DRE and negative TRUS of the prostate (same patient as Fig. 11.21). a Axial T2w image with superimposed grid for MR spectroscopy. b Example of spectrum suggesting normal prostate tissue.

markedly than choline in patients on hormone therapy, leading to increases in [Cho+Cr]/Cit ratios in both normal prostate tissue and cancer in the peripheral zone. Nevertheless, the ability to separate normal prostate tissue and cancer on the basis of [Cho+Cr]/Cit ratios is preserved as long as these metabolic products are still detectable. If citrate levels drop below the detection limit of 1H MRS under hormone therapy, demonstration of choline suggests cancer tissue, as shown in a study of 95 patients.16 An investigation of 64 patients who underwent radical prostatectomy after MRI and 3D MRS, among them 16 patients receiving neoadjuvant hormone therapy, has demonstrated that the accuracy of combined MRI and MRS is maintained for at least the first 4 months of hormone treatment despite therapy-related changes in [Cho+Cr]/Cit ratios.17 In conclusion, these results suggest that MRS is able to differentiate normal prostate tissue from cancer as long as metabolic products are detectable and can therefore be used to monitor the response to hormone therapy in combination with MRI. In the treatment of localized prostate cancer, the results of radiotherapy are now comparable to those of radical prostatectomy. The outcome of radiotherapy crucially relies on applying a maximum radiation dose to the target organ while minimizing the dose to adjacent structures. In planning radiotherapy aimed at dose escalation in cancer, superimposing combined MRI/MRS data on

ultrasound and CT scans has become a straightforward task.64,65 This method is most useful when individual tumor foci are identified by MRI/MRS, while it does not improve tumor control probability (TCP) or reduce radiation-induced morbidity in patients with diffuse prostate cancer.64 There is typically no citrate detectable in prostate tissue following irradiation. Nevertheless, a multivariate analysis of in-vitro MR spectra obtained after radiotherapy was found to clearly differentiate core biopsies with and without histopathologic proof of prostate cancer in a study of 116 biopsy specimens obtained in 35 patients 16–36 months after radiotherapy and examined by MRS with histopathologic correlation.66 These results suggest that combined MRI/MRS has the potential to identify persistent or recurrent cancer foci in patients in whom PSA levels increase again after radiotherapy (PSA failure). In patients who have undergone focal cryotherapy of the prostate, combined MRI/MRS can differentiate necrotic tissue in the treated area, residual normal tissue, and residual or recurrent cancer. In patients with PSA failure after cryotherapy, combined MRI/MRS has been shown to identify cancer foci even if core biopsy is negative.67,68 Three-dimensional 1H MR spectroscopy of the prostate is able to detect and localize prostate cancer with high sensitivity and specificity, especially when combined with conventional MRI techniques. Differentiation of normal prostate tissue from cancer is also possible in patients

234

11 The Prostate and Seminal Vesicles

who have undergone hormone treatment, radiotherapy, or cryotherapy. Combined MRI/MRS of the prostate is noninvasive and does not involve radiation exposure, which is why it can be used repeatedly and would be suitable for planning prostate biopsies and treatment and also for monitoring treatment response. MR spectroscopy is available only at specialized centers. To encourage its wider clinical use, handling of the equipment and data analysis need to be simplified. MR-Guided Prostate Biopsy Systematic tissue sampling with an 18 G biopsy needle, which is the routine procedure in the clinical setting, was shown to yield a different Gleason score than prostatectomy specimens in up to 65 % of cases, especially in patients with better-differentiated tumors.69 In a site-by-site analysis of prostate biopsy results and histology of prostatectomy specimens in comparison with the findings of conventional MRI combined with MRS, the combined MR technique was found to have higher sensitivity but lower specificity than prostate biopsy. Tissue sampling will therefore continue to be necessary.59 Being more sensitive

a

b Fig. 11.23a, b Targeted MR-guided transrectal prostate biopsy. a Diagram of biopsy device consisting of a needle sheath filled with contrast medium for visualization by MRI and a stand for height adjustment, forward and backward movement, and rotation. b The needle guide is aimed at the target area in the prostate using a T2w HASTE sequence and then fixed in the desired position. Images show the needle guide aimed at the apex (left) and base (right) of the prostate (from ref. 41).

than DRE and TRUS, combined MRI/MRS, possibly supplemented by a dynamic contrast-enhanced study, might help reduce the number of false-negative core biopsy specimens by identifying representative tumor areas for subsequent targeted biopsy. This will enable a reliable estimate of the patient’s prognosis. Technically, this can be accomplished by transferring the MR data to TRUS for TRUS-guided biopsy or by making direct use of the MRI datasets for MR-guided prostate biopsy. Ideally, MR-guided biopsy should be performed immediately after the diagnostic MR examination. MR-guided biopsies using a transperineal or transgluteal access are usually obtained in an open MR scanner, which typically operates at a lower field strength and has a less homogeneous magnetic field.70,71 Because of the poorer image quality of open MRI, the diagnostic examination has to be performed on a different scanner. A device for performing transrectal MR-guided prostate biopsy in a closed 1.5-T MR scanner using a torso or body phased-array coil was developed by Beyersdorff et al. in cooperation with an industrial partner (MRI-Devices/Daum GmbH, Germany).72 A prototype of the device is available for clinical use and comprises a needle guide that is visualized by MRI and can be fixed in different positions by means of a flexible arm, a base plate, and a cushion allowing the patient to lie comfortably in the prone position (Fig. 11.23). Following patient positioning, the MRvisible needle guide is inserted into the rectum and then attached to the arm, which allows rotation of the needle guide, forward and backward movement, and height adjustment. Rotational and horizontal movement for targeting a specific region within the prostate under MR guidance is accomplished from outside by means of a telescope rod. The needle is then fixed in the desired position and tissue sampled with an MR-compatible needle. Tissue specimens can be removed from the device without the need for repositioning the patient. With this device, MRguided prostate biopsy and the diagnostic MR examination can be performed in the same scanner. A T2w TSE sequence acquired with the same parameters as the preceding diagnostic MR images yields reproducible results and enables spatial matching of the intraprocedural image information with that of the preceding diagnostic MR examination. In a study of 12 patients in whom the biopsy device was used for sampling prostate tissue, cancer was histologically confirmed in five patients.41 The method may be improved by integrating coils into the biopsy device, which has been investigated in an animal experiment.73 Instead of sampling tissue under MR guidance, MR datasets can be fused with TRUS images. Provided that adequate fusion is achieved, the advantages of superior tumor detection by MRI could be combined with TRUSguided prostate biopsy, which is faster and less expensive than MRI-guided biopsy. Experience with this method is still limited to a study of patients with suspected recurrence of prostate cancer after radiotherapy.74

MRI Appearance of Pathologic Entities

Diagnosis of Recurrent Prostate Cancer The risk of cancer recurrence after radical prostatectomy strongly correlates with the preoperative PSA level, histopathologic tumor stage, Gleason score, and surgical resection margin.75,76 Up to 50 % of patients with a preoperative PSA of > 10 ng/mL or a Gleason score > 7 develop tumor recurrence within 7 years of surgery; patients with extracapsular cancer extension (stage T3) have a recurrence rate of ca. 35 %.75 Patients with a positive resection margin will develop recurrence within 5 years in 25 % of cases regardless of the tumor stage.77 PSA level determination is the most important test in the follow-up of patients after surgery, radiotherapy, or androgen deprivation therapy for prostate cancer. Following radical prostatectomy, the PSA is typically < 0.4 ng/mL or below the limit of detection. Demonstration of PSA in patients with undetectable PSA in earlier tests or a PSA elevation in patients with constantly low levels previously indicates residual prostate tissue, recurrent tumor, or metastasis. Prostate cancer after radical prostatectomy recurs close to the resection site (positive margins), particularly in the area of the apex and the vesicourethral anastomosis. DRE plays only a minor role in diagnosing recurrence, for two reasons: a minimum amount of volume must be present to be amenable to palpation, and the preceding operation alters the anatomic situation in the true pelvis to such an extent that it is difficult in most cases to differentiate between postoperative scar tissue and recurrent cancer.78,79 Following radiotherapy, shrinkage of the prostate in conjunction with fibrotic replacement precludes differentiation between malignant and benign tissue. When performing MRI to evaluate patients for recurrent prostate cancer, the use of an endorectal coil or a high-resolution surface coil is necessary to also detect smaller lesions. In addition to detailed evaluation of the prostate compartment, including the area of the former seminal vesicles, thin slices should be obtained at least of the vesicourethral anastomosis, which should be imaged in several planes if necessary (Figs. 11.24, 11.25, 11.26, 11.27). Recurrent prostate cancer is of higher signal intensity than muscle, which has low signal intensity on T2w images, and can often be identified as an asymmetry when the appearance is compared with the contralateral side. In patients with negative DRE and normal TRUS findings, MRI may help define the most suitable site for tissue sampling. Another advantage of MRI over CT and US is that it allows detection of asymmetrically arranged groups of lymph nodes, even when they are small, and evaluation of nodal regions that are more difficult to assess by other imaging modalities (Fig 11.28). MRI cannot differentiate between active, suppressed, and necrotic tissue after androgen deprivation therapy or radiotherapy because fibrosis, residual glandular tissue, and viable tumor tissue have overlapping signal intensities. Differentiation of these tissues might be improved by the use of MR spectroscopy or contrast-enhanced dynamic MRI.80

235

Fig. 11.24 Recurrent tumor (T) in the apex of the prostate. Axial T2w TSE image.

a

b Fig. 11.25a, b Recurrent tumor (arrow) in the prostate compartment. a Axial T2w TSE image. b Sagittal T2w TSE image.

236

11 The Prostate and Seminal Vesicles

Fig. 11.26 Patient after TURP. Axial T2w TSE image showing widened prostatic urethra and bladder neck after TURP (arrows).

Fig. 11.28 Prostate cancer with histologically proven metastases in the obturator lymph nodes. Axial T1w SE images show multiple lymph nodes up to 5 mm in size in the primary lymphatic drainage area of the prostate.

Other Malignant Prostate Tumors

a

b

Numerous variants of prostate carcinoma have been described in the last two decades. These atypical carcinomas account for < 5 % of all malignant prostate tumors (Table 11.4) and may arise from epithelial or stromal prostate tissue or ectopic cells in the prostate. While these variants do not differ from typical adenocarcinoma in terms of clinical presentation, they have a different etiology and differ in their response to therapy. Few data are available on the imaging appearance of these tumors, especially their MRI features. Their MRI appearance is the same as that of classic adenocarcinoma of the peripheral zone in most cases. Mucinous carcinoma of the prostate has no characteristic echo pattern on ultrasound. They differ from the typical low-signal-intensity carcinoma of the peripheral zone in that they usually have high signal intensity on T2w images because they secrete and store mucin.76,81

v Fig. 11.27a, b Recurrent prostate cancer (arrows) after radical prostatectomy. a Axial T2w image. b Sagittal T2w image. Sagittal image shows the recurrent tumor (T) to be located at the resection margin near the apex.

MRI Appearance of Pathologic Entities

237

Table 11.4 Atypical malignant prostate tumors Epithelial tumors

Variants of adenocarcinoma Endometrioid type adenocarcinoma Comedocarcinoma Mucinous carcinoma Adenoid cystic carcinoma Signet ring carcinoma Adenosquamous carcinoma

· · · · · ·

Other epithelial carcinomas Squamous cell carcinoma Transitional cell carcinoma Neuroendocrine tumors Carcinoid and small-cell carcinoma

· · · · Nonepithelial tumors

Tumors with muscular differentiation

· Rhabdomyosarcoma · Leiomyosarcoma

a

Other rare sarcomas · Fibrosarcoma · Malignant fibrous histiocytoma · Osteosarcoma Angiosarcoma

· Chondrosarcoma · Malignant nerve sheath tumor Mixed tumors

· Carcinosarcoma · Malignant phylloid tumor Hematolymphatic malignancies

· Malignant lymphoma · Leukemic involvement of the prostate Metastases

b

Transitional cell carcinoma of the prostate is either a primary neoplasm arising from the glandular ducts/acini or a synchronous or metachronous manifestation of bladder or urethral cancer. Both primary and secondary tumors have been observed to invade the stroma. The primary form accounts for 2–4 % of all prostate carcinomas. Secondary invasion of the transitional zone occurs in 7–43 % of patients with bladder or urethral cancer. Most transitional cell carcinomas are advanced at presentation and have uniform low MR signal intensity. Prostate sarcoma is a rare malignant tumor of the prostate. At diagnosis the tumor is typically locally advanced and presents with symptoms related to local growth such as disturbed voiding and defecation, hematuria, and obstructive nephropathy. CT and MRI are predominantly used to identify possible invasion of adjacent organs. Histologically, rhabdomyosarcoma is typically found in boys and young men, while leiomyosarcoma predominates in adult men (Figs. 11.29 and 11.30). Especially when a sarcoma is large, not even MRI may allow accurate determination of the site of origin because the tumor destroys the normal glandular architecture of the prostate. De-

c Fig. 11.29a–c Rhabdomyosarcoma. a Axial PD SE image. b Coronal PD SE image. c Axial T1w SE image after IV injection of nonspecific Gd-based contrast medium (body coil). Large tumor (T) invading the levator ani muscle, obturator muscle (arrow in a), and seminal vesicles. Central necrosis is revealed on the contrast-enhanced image (c).

238

11 The Prostate and Seminal Vesicles

References

Fig. 11.30 Leiomyosarcoma. Sagittal T1w SE image after IV injection of nonspecific Gd-based contrast medium (body coil). Giant tumor (T) with large central necroses and invasion of the posterior urinary bladder wall (arrows).

pending on the pulse sequence used, MRI will reveal the different tissue components of the tumor. While the inhomogeneous makeup is only suggested on T1w images, PD and T2w sequences will show the characteristic appearance of a rapidly growing tumor, namely a septated mass in the true pelvis with a high-signal-intensity center, which may be similar in intensity to fluid on T2w images and represents central necrosis. A sarcoma often has a capsule, which may be helpful in the differential diagnosis, especially when the peripheral gland is still intact. A characteristic set of MRI features on different pulse sequences obtained before and after contrast administration that would allow diagnosis of a specific histologic tumor type is not available for any of the aforementioned prostate tumors. However, in contrast to prostate carcinoma, administration of Gd-based contrast medium may be helpful in prostate sarcoma to distinguish perfused and necrotic tumor components and to evaluate for involvement of the urinary bladder and rectum in those cases where cystoscopy is precluded. MRI is also used to monitor the response to chemotherapy. It will reveal the tumor size and contrastenhanced images allow precise evaluation of viable residual tumor.

1. Beyersdorff D, Darsow U, Stephan C, Schnorr D, Loening S, Taupitz M. [MRI of prostate cancer using three different coil systems: image quality, tumor detection, and staging]. Rofo 2003;175(6):799–805 2. Tempany CM, Zhou X, Zerhouni EA, et al. Staging of prostate cancer: results of Radiology Diagnostic Oncology Group project comparison of three MR imaging techniques. Radiology 1994;192(1):47–54 3. Engelbrecht MR, Huisman HJ, Laheij RJ, et al. Discrimination of prostate cancer from normal peripheral zone and central gland tissue by using dynamic contrast-enhanced MR imaging. Radiology 2003;229(1):248–254 4. Preziosi P, Orlacchio A, Di Giambattista G, et al. Enhancement patterns of prostate cancer in dynamic MRI. Eur Radiol 2003;13(5):925–930 5. Schlemmer HP, Merkle J, Grobholz R, et al. Can pre-operative contrast-enhanced dynamic MR imaging for prostate cancer predict microvessel density in prostatectomy specimens? Eur Radiol 2004;14(2):309–317 6. McNeal JE. The prostate and prostatic urethra: a morphologic synthesis. J Urol 1972;107(6):1008–1016 7. Sommer FG, McNeal JE, Carrol CL. MR depiction of zonal anatomy of the prostate at 1.5 T. J Comput Assist Tomogr 1986; 10(6):983–989 8. Bottomley PA, Edelstein WA, Foster TH, Adams WA. In vivo solvent-suppressed localized hydrogen nuclear magnetic resonance spectroscopy: a window to metabolism? Proc Natl Acad Sci U S A 1985;82(7):2148–2152 9. Ordridge RJ, Bendall MR, Gordon RE, Conelly A. Volume selection for in-vivo biological spectroscopy. In: Govil G, Khetrapal CL, Saran A, eds. Magnetic Resonance in Biology and Medicine. New Dehli: Tata McGraw-Hill; 1985: pp. 387 10. Moonen CTW, von Kienlin M, van Zijl PC, et al. Comparison of single-shot localization methods (STEAM and PRESS) for in vivo proton NMR spectroscopy. NMR Biomed 1989;2(5-6):201–208 11. Cousins JP. Clinical MR spectroscopy: fundamentals, current applications, and future potential. AJR Am J Roentgenol 1995; 164(6):1337–1347 12. Kurhanewicz J, Vigneron DB, Hricak H, Narayan P, Carroll P, Nelson SJ. Three-dimensional H-1 MR spectroscopic imaging of the in situ human prostate with high (0.24-0.7-cm3) spatial resolution. Radiology 1996a;198(3):795–805 13. Costello LC, Franklin RB, Narayan P. Citrate in the diagnosis of prostate cancer. Prostate 1999;38(3):237–245 14. Heerschap A, Jager GJ, van der Graaf M, Barentsz JO, Ruijs SH. Proton MR spectroscopy of the normal human prostate with an endorectal coil and a double spin-echo pulse sequence. Magn Reson Med 1997;37(2):204–213 15. Scheidler J, Hricak H, Vigneron DB, et al. Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study. Radiology 1999;213(2):473–480 16. Mueller-Lisse UG, Swanson MG, Vigneron DB, et al. Time-dependent effects of hormone-deprivation therapy on prostate metabolism as detected by combined magnetic resonance imaging and 3D magnetic resonance spectroscopic imaging. Magn Reson Med 2001a;46(1):49–57 17. Mueller-Lisse UG, Vigneron DB, Hricak H, et al. Localized prostate cancer: effect of hormone deprivation therapy measured by using combined three-dimensional 1 H MR spectroscopy and MR imaging: clinicopathologic case-controlled study. Radiology 2001b;221(2):380–390 18. Mueller-Lisse UG, Scherr M. [1 H magnetic resonance spectroscopy of the prostate]. Radiologe 2003;43(6):481–488 19. Star-Lack J, Nelson SJ, Kurhanewicz J, Huang LR, Vigneron DB. Improved water and lipid suppression for 3D PRESS CSI using RF band selective inversion with gradient dephasing (BASING). Magn Reson Med 1997;38(2):311–321 20. Males RG, Vigneron DB, Star-Lack J, et al. Clinical application of BASING and spectral/spatial water and lipid suppression pulses for prostate cancer staging and localization by in vivo 3D 1 H magnetic resonance spectroscopic imaging. Magn Reson Med 2000;43(1):17–22

MRI Appearance of Pathologic Entities 21. Schricker AA, Pauly JM, Kurhanewicz J, Swanson MG, Vigneron DB. Dualband spectral-spatial RF pulses for prostate MR spectroscopic imaging. Magn Reson Med 2001;46(6):1079–1087 22. Costello LC, Franklin RB. Concepts of citrate production and secretion by prostate. 1. Metabolic relationships. Prostate 1991a;18(1):25–46 23. Costello LC, Franklin RB. Concepts of citrate production and secretion by prostate: 2. Hormonal relationships in normal and neoplastic prostate. Prostate 1991b;19(3):181–205 24. Daly PF, Lyon RC, Faustino PJ, Cohen JS. Phospholipid metabolism in cancer cells monitored by 31P NMR spectroscopy. J Biol Chem 1987;262(31):14875–14878 25. García-Segura JM, Sánchez-Chapado M, Ibarburen C, et al. In vivo proton magnetic resonance spectroscopy of diseased prostate: spectroscopic features of malignant versus benign pathology. Magn Reson Imaging 1999;17(5):755–765 26. Hendry WF, Pryor JP. Müllerian duct (prostatic utricle) cyst: diagnosis and treatment in subfertile males. Br J Urol 1992; 69(1):79–82 27. Gevenois PA, Van Sinoy ML, Sintzoff SAJr, et al. Cysts of the prostate and seminal vesicles: MR imaging findings in 11 cases. AJR Am J Roentgenol 1990;155(5):1021–1024 28. Papanicolaou N, Pfister RC, Stafford SA, Parkhurst EC. Prostatic abscess: imaging with transrectal sonography and MR. AJR Am J Roentgenol 1987;149(5):981–982 29. American Cancer Society. Cancer Facts and Figures 2002. Atlanta: ACS; 2002 30. Babaian RJ, Toi A, Kamoi K, et al. A comparative analysis of sextant and an extended 11-core multisite directed biopsy strategy. J Urol 2000;163(1):152–157 31. McNeal JE, Bostwick DG, Kindrachuk RA, Redwine EA, Freiha FS, Stamey TA. Patterns of progression in prostate cancer. Lancet 1986;1(8472):60–63 32. Gleason DF, Mellinger GT. Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol 1974;111(1):58–64 33. Berner A, Waere H, Nesland JM, Paus E, Danielsen HE, Fosså SD. DNA ploidy, serum prostate specific antigen, histological grade and immunohistochemistry as predictive parameters of lymph node metastases in T1-T3/M0 prostatic adenocarcinoma. Br J Urol 1995;75(1):26–32 34. Folkman J, Watson K, Ingber D, Hanahan D. Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 1989;339(6219):58–61 35. White S, Hricak H, Forstner R, et al. Prostate cancer: effect of postbiopsy hemorrhage on interpretation of MR images. Radiology 1995;195(2):385–390 36. Outwater EK, Petersen RO, Siegelman ES, Gomella LG, Chernesky CE, Mitchell DG. Prostate carcinoma: assessment of diagnostic criteria for capsular penetration on endorectal coil MR images. Radiology 1994;193(2):333–339 37. Huch Böni RA, Boner JA, Lütolf UM, Trinkler F, Pestalozzi DM, Krestin GP. Contrast-enhanced endorectal coil MRI in local staging of prostate carcinoma. J Comput Assist Tomogr 1995;19(2): 232–237 38. Tempany CM, Rahmouni AD, Epstein JI, Walsh PC, Zerhouni EA. Invasion of the neurovascular bundle by prostate cancer: evaluation with MR imaging. Radiology 1991;181(1):107–112 39. Ogura K, Maekawa S, Okubo K, et al. Dynamic endorectal magnetic resonance imaging for local staging and detection of neurovascular bundle involvement of prostate cancer: correlation with histopathologic results. Urology 2001;57(4):721–726 40. Villers A, McNeal JE, Redwine EA, Freiha FS, Stamey TA. The role of perineural space invasion in the local spread of prostatic adenocarcinoma. J Urol 1989;142(3):763–768 41. Beyersdorff D, Hamm B. [MRI for troubleshooting detection of prostate cancer]. Rofo 2005;177(6):788–795 42. Mueller-Lisse UG, Mueller-Lisse UL, Haller S, et al. Likelihood of prostate cancer based on prostate-specific antigen density by MRI: retrospective analysis. J Comput Assist Tomogr 2002; 26(3):432–437 43. Padhani AR, MacVicar AD, Gapinski CJ, et al. Effects of androgen deprivation on prostatic morphology and vascular permeability evaluated with mr imaging. Radiology 2001;218(2):365–374 44. Qayyum A, Coakley FV, Lu Y, et al. Organ-confined prostate cancer: effect of prior transrectal biopsy on endorectal MRI

45.

46.

47. 48.

49.

50.

51.

52. 53. 54. 55.

56.

57.

58.

59.

60.

61.

62. 63.

64.

239

and MR spectroscopic imaging. AJR Am J Roentgenol 2004; 183(4):1079–1083 Dhingsa R, Qayyum A, Coakley FV, et al. Prostate cancer localization with endorectal MR imaging and MR spectroscopic imaging: effect of clinical data on reader accuracy. Radiology 2004;230(1):215–220 Mullerad M, Hricak H, Wang L, Chen HN, Kattan MW, Scardino PT. Prostate cancer: detection of extracapsular extension by genitourinary and general body radiologists at MR imaging. Radiology 2004;232(1):140–146 Harris RD, Schned AR, Heaney JA. Staging of prostate cancer with endorectal MR imaging: lessons from a learning curve. Radiographics 1995;15(4):813–829, discussion 829–832 Mueller-Lisse UG, Mueller-Lisse U, Scheidler J, Klein G, Reiser M. Reproducibility of image interpretation in MRI of the prostate: application of the sextant framework by two different radiologists. Eur Radiol 2005;15(9):1826–1833 Perrotti M, Han KR, Epstein RE, et al. Prospective evaluation of endorectal magnetic resonance imaging to detect tumor foci in men with prior negative prostastic biopsy: a pilot study. J Urol 1999;162(4):1314–1317 Beyersdorff D, Taupitz M, Winkelmann B et al. Patients with a History of Elevated Prostate-Specific Antigen Levels and Negative Transrectal US-guided Quadrant for Sextant Biopsy Results: Value of MR Imaging. Radiology 2002;224:701–706 Engelbrecht MR, Jager GJ, Laheij RJ, Verbeek AL, van Lier HJ, Barentsz JO. Local staging of prostate cancer using magnetic resonance imaging: a meta-analysis. Eur Radiol 2002;12(9): 2294–2302 Langlotz CP, Schnall MD, Pollack H. Staging of prostatic cancer: accuracy of MR imaging. Radiology 1995;194(3):645–646, discussion 647–648 Nicolas V, Beese M, Keulers A, Bressel M, Kastendieck H, Huland H. [MR tomography in prostatic carcinoma: comparison of conventional and endorectal MRT]. Rofo 1994;161(4):319–326 Huisman HJ, Engelbrecht MR, Barentsz JO. Accurate estimation of pharmacokinetic contrast-enhanced dynamic MRI parameters of the prostate. J Magn Reson Imaging 2001;13(4):607–614 Ito H, Kamoi K, Yokoyama K, Yamada K, Nishimura T. Visualization of prostate cancer using dynamic contrast-enhanced MRI: comparison with transrectal power Doppler ultrasound. Br J Radiol 2003;76(909):617–624 van Dorsten FA, van der Graaf M, Engelbrecht MR, et al. Combined quantitative dynamic contrast-enhanced MR imaging and (1)H MR spectroscopic imaging of human prostate cancer. J Magn Reson Imaging 2004;20(2):279–287 Jager GJ, Ruijter ET, van de Kaa CA, et al. Dynamic TurboFLASH subtraction technique for contrast-enhanced MR imaging of the prostate: correlation with histopathologic results. Radiology 1997;203(3):645–652 Fütterer JJ, Scheenen TW, Huisman HJ, et al. Initial experience of 3 tesla endorectal coil magnetic resonance imaging and 1 Hspectroscopic imaging of the prostate. Invest Radiol 2004; 39(11):671–680 Wefer AE, Hricak H, Vigneron DB, et al. Sextant localization of prostate cancer: comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectroscopic imaging with step section histology. J Urol 2000;164(2):400–404 Kaji Y, Kurhanewicz J, Hricak H, et al. Localizing prostate cancer in the presence of postbiopsy changes on MR images: role of proton MR spectroscopic imaging. Radiology 1998;206(3): 785–790 Yu KK, Scheidler J, Hricak H, et al. Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and threedimensional proton MR spectroscopic imaging. Radiology 1999; 213(2):481–488 Coakley FV, Kurhanewicz J, Lu Y, et al. Prostate cancer tumor volume: measurement with endorectal MR and MR spectroscopic imaging. Radiology 2002;223(1):91–97 Vigneron DB, Males R, Noworolski P, et al. 3D MRSI of prostate cancer: correlation with histologic grade. Proceedings of the International Society for Magnetic Resonance in Medicine, Sixth Scientific Meeting and Exhibition, Sydney, Australia, April 18–24, 1998, Volume 2: S488 Zaider M, Zelefsky MJ, Lee EK, et al. Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging. Int J Radiat Oncol Biol Phys 2000;47(4):1085–1096

240

11 The Prostate and Seminal Vesicles

65. DiBiase SJ, Hosseinzadeh K, Gullapalli RP, et al. Magnetic resonance spectroscopic imaging-guided brachytherapy for localized prostate cancer. Int J Radiat Oncol Biol Phys 2002; 52(2):429–438 66. Menard C, Smith ICP, Somorjai RL, et al. Magnetic resonance spectroscopy of the malignant prostate gland after radiotherapy: a histopathologic study of diagnostic validity. Int J Radiat Oncol Biol Phys 2001;50(2):317–323 67. Kurhanewicz J, Vigneron DB, Hricak H, et al. Prostate cancer: metabolic response to cryosurgery as detected with 3D H-1 MR spectroscopic imaging. Radiology 1996b;200(2):489–496 68. Parivar F, Hricak H, Shinohara K, et al. Detection of locally recurrent prostate cancer after cryosurgery: evaluation by transrectal ultrasound, magnetic resonance imaging, and three-dimensional proton magnetic resonance spectroscopy. Urology 1996;48(4):594–599 69. Bostwick DG. Gleason grading of prostatic needle biopsies. Correlation with grade in 316 matched prostatectomies. Am J Surg Pathol 1994;18(8):796–803 70. D’Amico AV, Tempany CM, Cormack R, et al. Transperineal magnetic resonance image guided prostate biopsy. J Urol 2000; 164(2):385–387 71. Cormack RA, D’Amico AV, Hata N, Silverman S, Weinstein M, Tempany CM. Feasibility of transperineal prostate biopsy under interventional magnetic resonance guidance. Urology 2000; 56(4):663–664 72. Beyersdorff D, Winkel A, Hamm B, Lenk S, Loening SA, Taupitz M. MR imaging-guided prostate biopsy with a closed MR unit at 1.5 T: initial results. Radiology 2005;234(2):576–581

73. Susil RC, Krieger A, Derbyshire JA, et al. System for MR imageguided prostate interventions: canine study. Radiology 2003;228(3):886–894 74. Kaplan I, Oldenburg NE, Meskell P, Blake M, Church P, Holupka EJ. Real time MRI-ultrasound image guided stereotactic prostate biopsy. Magn Reson Imaging 2002;20(3):295–299 75. Catalona WJ, Smith DS. Cancer recurrence and survival rates after anatomic radical retropubic prostatectomy for prostate cancer: intermediate-term results. J Urol 1998;160(6 Pt 2): 2428–2434 76. Grossfeld GD, Stier DM, Flanders SC, et al. Use of second treatment following definitive local therapy for prostate cancer: data from the caPSURE database. J Urol 1998;160(4):1398–1404 77. Morton RA, Steiner MS, Walsh PC. Cancer control following anatomical radical prostatectomy: an interim report. J Urol 1991;145(6):1197–1200 78. Goldenberg P, Carter M, Dashefsky P, et al. Sonographic characteristics of the urethrovesical anastomosis in the early postradical prostatectomy patient. J Urol 1992;147:1307 79. Lightner DJ, Lange PH, Reddy PK, Moore L. Prostate specific antigen and local recurrence after radical prostatectomy. J Urol 1990;144(4):921–926 80. Rouvière O, Valette O, Grivolat S, et al. Recurrent prostate cancer after external beam radiotherapy: value of contrast-enhanced dynamic MRI in localizing intraprostatic tumor—correlation with biopsy findings. Urology 2004;63(5):922–927 81. Schiebler ML, Schnall MD, Outwater E. MR imaging of mucinous adenocarcinoma of the prostate. J Comput Assist Tomogr 1992;16(3):493–494

241

12

The Uterus and Vagina C. Kluener and B. Hamm

Introduction

Imaging Technique

As the imaging modality with the highest sensitivity and specificity for many benign and malignant diseases of the female pelvis, MRI has an increasingly important role in women’s imaging. Evaluation of the uterus benefits from the excellent multiplanar capability of MRI (e. g., in the sagittal plane), while the high intrinsic soft-tissue contrast is unsurpassed in depicting uterine zonal anatomy. New coils and fast pulse sequences ensure high image quality and have further contributed to the evolution of MRI into a powerful diagnostic tool for imaging the female pelvis. Although ultrasound secondary to a pelvic examination continues to be the most widely used imaging tool, it has several well-known limitations such as being examiner dependent and providing poor tissue characterization. The most notable drawback in the setting of female pelvic malignancy is the fact that it does not allow adequate tumor staging. CT offers inadequate soft-tissue contrast for reliable local tumor staging and should only be used in women with advanced pelvic cancer. MRI, on the other hand, visualizes the female pelvis with exquisite morphologic detail and is the most accurate imaging modality for tumor staging. Recent studies have also shown MRI to be a cost-effective test for the pretherapeutic staging of cervical and endometrial cancer. Finally, MRI is increasingly being used for diagnostic work-up and therapeutic decision-making in women with benign uterine disease.

A woman undergoing a pelvic MRI examination should be asked not only the standard questions regarding known contraindications to MRI (pacemaker, etc.) but also a set of questions pertaining to her gynecologic history and hormonal status. This information is important for tailoring the pelvic MRI study to the individual situation and will help the radiologist to adequately interpret the results since the appearance of the pelvic organs varies through the menstrual cycle and with the patient’s hormonal status. Also important is information on the presence of foreign bodies, such as an intrauterine device (IUD), and previous surgery and radiotherapy. We use a special form for the patient’s gynecologic data, so the information is available at a glance when the radiologist interprets the MR images (Table 12.1). Earlier fears that IUDs might lead to overheating or become displaced have proved unfounded. An IUD is no longer considered a contraindication to MRI, which will depict the device as a low-signal-intensity structure in the uterine cavity. It is unnecessary for patients to fast before undergoing pelvic MRI. The bladder should be empty or only moderately distended at the beginning of the examination to minimize motion artifacts and avoid the need to interrupt the examination if the patient has an urge to void. Placement of a tampon for improved delineation of the vaginal wall is not necessary and may even cause susceptibility artifacts on GRE sequences. Pelvic MRI can be performed with the patient in the prone position (which will reduce respiratory artifacts) or in the preferred and more comfortable supine position. Various measures are available to reduce ghosting and blurring resulting from respiratory motion, intestinal peristalsis, and vessel pulsation. Respiratory artifacts from the abdominal wall can be minimized by instructing the patient to breathe quietly during data acquisition. If MRI is performed with the patient supine, it is helpful to place a presaturation band over the anterior abdominal wall to reduce the signal from abdominal fat (Figs. 12.1 and 12.2). Unless contraindicated, intestinal peristalsis can be effectively suppressed by intravenous split-dose administration of a spasmolytic agent (40 mg of butylscopolamine or 2 mg of glucagon). The venous line can also

Indications MRI is the diagnostic test of choice for evaluating malignant tumors of the uterus (cervical and endometrial cancer) and carcinoma of the vagina. It is increasingly being used for planning radiotherapy and monitoring the response to treatment. As an adjunct modality, MRI can help resolve inconclusive ultrasound findings, determine the site of origin of a pelvic mass, and provide useful information for lesion characterization.

242

12 The Uterus and Vagina

Table 12.1 Useful clinical information to obtain before an MRI examination of the female pelvis 1. What is the patient’s menstrual status? Premenopausal

Postmenopausal

Date of last period: Is pregnancy possible? No

Yes

2. Does the patient take hormones? Estrogens: Oral contraceptives:

No No

Yes Yes

Multipara

Cesarean

3. Parity? Nullipara

4. Presence of foreign bodies (e. g., tampon, IUD, cervical pessary)? No If yes, specify:

Yes

5. Did the patient have prior gynecologic surgery (including D&C) or radiotherapy? Surgery:

No

Yes

If yes, type and date of operation: Radiotherapy:

No

Yes

If yes, date of completion: 6. Does the patient have contraindications to Contrast media? Butylscopolamine? Glucagon?

be used for later injection of contrast medium. Presaturation pulses applied superior and inferior to the field of view (FOV) reduce arterial and venous inflow artifacts (Fig. 12.1). An anteroposterior phase-encoding direction should be chosen for acquisition of axial (and sagittal) images to move vascular pulsation artifacts away from the target anatomy in the center of the pelvis. Although pelvic MRI can be performed with the inbuilt body coil, use of a body or torso phased-array coil is now preferred because it markedly improves signal-to-noise ratio (SNR) and spatial resolution (Fig 12.3). Paired coils in Helmholtz arrangement and endoluminal coils are no longer used for pelvic imaging. An endorectal coil images nearby structures with high resolution (e. g., cervical cancer) but does not improve overall staging accuracy and is not well accepted by patients.

No No No

Yes Yes Yes

Imaging Planes Proper selection of imaging planes is critical in planning a pelvic MR examination. Axial images enable adequate evaluation of the uterus and cervix, uterosacral ligaments, and presacral space and are optimal for parametrial and lymph node assessment. Sagittal images cover the entire length of the uterus, which is an advantage over axial images, and they are more suitable for evaluating the vesicouterine ligament and the anatomic relationship of the uterus and vagina to the urinary bladder and rectum. Sequence planes angled to the uterus or cervix to obtain true long-axis and short-axis views of these organs are helpful in assessing uterine anomalies (long-axis view) and depth of myometrial invasion in endometrial cancer (short-axis view of the uterine corpus). Otherwise, additional coronal images are rarely acquired in uterine and vaginal MRI.

Pulse Sequences

Fig. 12.1 Sagittal T1w GRE image showing placement of presaturation bands on the high-SI abdominal wall for reduction of motion artifacts and of presaturation bands superior and inferior to the imaging volume for elimination of pulsation artifacts.

The protocol for MRI of the uterus and vagina comprises T1w and T2w sequences. T2w images best depict the zonal anatomy of the uterus and are most suitable to evaluate the integrity of muscular structures (e. g., bladder and rectal wall, pelvic sidewall). T1w sequences provide optimal contrast between pelvic organs and surrounding fat, making them well suited for lymph node assessment. Imaging parameters for conventional T1w and T2w SE sequences as well as their fast counterparts (FSE or TSE) are summarized in Table 12.2. Part of the gain in imaging speed resulting from the use of TSE or FSE sequences should be used to increase the imaging matrix and number of signal averages. Combining an FSE or TSE sequence with the use of a phased-array body coil yields images with an excellent spatial resolution while also shortening acquisition time. Fat-suppressed SE sequences will help differentiate between hemorrhagic and fatty lesions (e. g., endometriosis cyst vs dermoid) and improve detection of

Imaging Technique

243

a

b Fig. 12.2a, b Improved image quality resulting from use of presaturation pulses. T1w SE images obtained without and with abdominal presaturation. a Without presaturation, respiratory artifacts from the high-SI abdominal wall degrade the image through the

true pelvis. b Presaturation of the abdominal wall eliminates respiratory motion artifacts. Normal appearance of the ovaries in the ovarian fossae. Sharp visualization of the intestine due to administration of spasmolytic agent.

a Fig. 12.3a, b Improved SNR resulting from use of a phased-array body coil. T2w TSE images obtained with identical sequence parameters. Images through the endometrial cavity showing intramural fibroids. a Body coil. b Phased-array body coil.

small, hyperintense endometrial implants, which may escape detection on conventional images when surrounded by hyperintense fat. A fat suppression technique also improves the differentiation of fat from enhancing tissue on postcontrast images.

b

Because of their poorer spatial resolution, gradient echo (GRE) sequences have not established themselves as an alternative MR technique for imaging of the female pelvis. They are occasionally used for dynamic contrastenhanced MR studies.

244

5 5 5 5 8 8 5 5 5 5 2 2 3 2 3 4 4 4 4 3

No No No No No No No No No No

After intravenous injection of a nonspecific, Gd-based contrast medium (e. g., Magnevist, Dotarem), the uterine zonal anatomy is also seen on T1w images (see Figs. 12.13 and 12.19), but less clearly so than on T2w images. A contrast-enhanced study is generally performed using a high-resolution T1w SE sequence, but is only required for specific indications (e. g., endometrial cancer, differentiation of viable tumor and necrosis). Routine administration of an oral contrast agent is not necessary for a dedicated MRI examination of the uterus or vagina.

6.2 11 5.6 11 5 8 4 8 8 5.6

Slice thickness Scan (mm) time (min)

General Imaging Strategy

Note: Use of a body or torso phased-array coil is recommended for acquisition of TSE/FSE sequences with high spatial resolution. All sequences are acquired with an interslice gap of 20 % of the slice thickness (distance factor, 0.2).

19 19 19 19 23 19 15 15 13 19 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 320 (75 %) 220 × 512 128 × 256 240 × 512 128 × 256 228 × 512 192 × 256 270 × 512 192 × 256 192 × 256 240 × 512 No No No No No No No No Yes Yes 5 – 15 – 3 – 3 – – 15 – – – – – – – – – – 20/120 90 112 90 10 15 10 15 15 112 TSE SE TSE SE TSE SE TSE SE SE TSE Sagittal Sagittal (alternative) Axial Axial (alternative) Axial Axial (alternative) Sagittal Sagittal (alternative) Axial Axial T2 T2 T2 T2 PD/T1 T1 T1 T1 T1 T2

4200 2500 7000 2500 1150 500 650 500 500 7000

No. of slices FOV (mm) Matrix FS ETL (example) Flip (°) TE (ms) TR (ms) Sequence type Weighting Plane

Table 12.2 Recommended pulse sequences and imaging parameters for MRI of the uterus and vagina

Contrast Media

No. of acquisitions

Breathhold

12 The Uterus and Vagina

The basic MRI protocol for the uterus comprises a sagittal T2w sequence and axial T2w and T1w sequences. This is the standard combination of pulse sequences recommended for evaluating uterine anomalies as well as benign and malignant tumors (e. g., fibroids, endometrial and cervical cancer). Sagittal images are acquired to detect endometrial cancer and define its longitudinal extent as well as to identify possible invasion of the urinary bladder and rectum. Axial T2w images are most useful to evaluate for invasion of the parametrium, sacrouterine ligaments, and pelvic sidewall. Axial images also provide supplementary information for assessment of the urinary bladder and rectal wall. As a rule, T1w images (from the aortic bifurcation to the pelvic floor) are only acquired in the axial plane; they are most useful for assessing the lymph node status. When evaluating women for uterine anomalies, the standard orthogonal planes should be replaced with oblique images oriented parallel to the endometrial cavity to obtain true long-axis views of the uterus (Fig. 12.4). In patients with endometrial cancer, additional contrast-enhanced T1w images in sagittal and oblique axial planes (short-axis view of uterus) will increase diagnostic confidence. An axial T2w sequence is the basic sequence for MRI of the vagina, with axial T1w images providing useful additional information (see Fig. 12.21). After contrast administration, zonal anatomy is also seen on T1w images (see Fig. 12.21c). An additional sagittal T2w sequence is useful to determine the longitudinal extent of a vaginal tumor and demonstrate invasion of the urinary bladder and rectal wall. A coronal sequence (T2w) is indicated if invasion of the levator ani muscle is suspected.

MRI Appearance of Normal Anatomy

245

a

b Fig. 12.4a, b Diagrams showing standard axial vs angled views.

Fig. 12.5 Diagram of the uterus showing the corpus, isthmus, and cervix (according to Martius).

MRI Appearance of Normal Anatomy The normal MRI appearance of the uterus and vagina varies with the patient’s age and hormonal status.1–6 The uterus can be divided into three distinct segments—the corpus, the isthmus (or lower uterine segment), and the cervix (Fig. 12.5). The wall of the uterine corpus (unlike that of the cervix) is predominantly composed of smooth muscle. The slitlike uterine cavity is lined by endometrium. In the

nonpregnant uterus, the isthmus, together with the internal os, constitutes the junction of the uterine corpus and the cervix. The isthmus is only ca. 0.5 cm long but is quite distinct from the cervix in that it enlarges considerably during pregnancy to become the lower part of the uterine cavity. The uterine cervix consists of supravaginal and vaginal parts. The rounded vaginal part (ectocervix or portio vaginalis cervix) projects into the vagina. The wall of the cervix is mostly composed of firm connective tissue and ca. 10 % smooth muscle fibers arranged in a ringlike fashion. The cervical canal is lined by columnar epithelium

246

12 The Uterus and Vagina

The uterus is covered by peritoneum (Fig. 12.6). Anteriorly, the peritoneum is reflected from the uterus onto the superior surface of the urinary bladder, creating the vesicouterine pouch. Posteriorly, the peritoneum extends down to the level of the posterior vaginal fornix, from where it is reflected onto the anterior rectal wall, forming the rectouterine pouch (pouch of Douglas, cul-de-sac). The uterine vessels pass through the parametrium and reach the uterus at the level of the uterine isthmus (i. e., at the internal os of the cervical canal). The ureters course through the parametrial tissue on either side ca. 2 cm lateral to the cervix. The parametrium is mainly composed of connective tissue with a larger proportion of fatty tissue toward the pelvic sidewalls. Below, the parametrium is separated from the paravaginal connective tissue (the paracolpium) by the cardinal ligaments (Mackenrodt ligaments), which extend from the cervix to the pelvic sidewalls. The uterus is supported and suspended by a total of eight ligaments. The uterovesical and uterosacral ligaments are important landmarks, for example, when evaluating the extent of cervical carcinoma (Fig. 12.7). The uterovesical ligament extends from the cervix to the base and posterior surface of the urinary bladder, while the uterosacral ligament extends from the cervix to the sacrum, surrounding the rectum. In women of reproductive age, the entire uterus is 6–9 cm long. It has uniform low signal intensity on T1w images (Fig. 12.8), whereas three distinct zones—the en-

Fig. 12.6 Sagittal T2w TSE image of the uterus showing the parts of the uterus covered by peritoneum (white line). The peritoneal cavity in the true pelvis appears black.

and the ectocervix by stratified squamous epithelium. The transition between these two types of epithelium is at the level of the external os.

Uterosacral ligament

Rectum Suspensory ligament of ovary

Uterine tube Cardinal ligament

Ovarian ligament Uterovesical ligament

Round ligament of uterus

Urinary bladder

Fundus of uterus

Fig. 12.7 Ligaments of the internal female genital organs (according to Martius).

MRI Appearance of Normal Anatomy

Fig. 12.8 Sagittal T1w SE image of the uterus.

dometrium, the junctional zone, and the myometrium— can be recognized on T2w images (Fig. 12.9). The endometrium is of high T2 signal intensity. Its thickness varies during the menstrual cycle from 1–3 mm at the beginning of the proliferative phase (Fig. 12.10) to 5–7 mm in the mid-secretory phase (Figs. 12.9 and 12.11). A blood clot may occasionally be present in the endometrial cavity at the time of menstruation and must not be mistaken for a foreign body or pathology (Fig. 12.12). The outer myometrium is of intermediate signal intensity on T2w images and increases in signal intensity during the secretory phase due to a higher water content (compare Figs. 12.9 and 12.11 with Fig. 12.10). Myometrial vessels are also prominent during this menstrual phase. The junctional zone represents the innermost myometrium and is of low signal intensity because it has a lower water content and higher nucleus-to-cytoplasm ratio than the outer myometrium.7,8 A thickness of the junctional zone of up to 5 mm is normal, while a thickness > 12 mm is considered to be diagnostic of adenomyosis. Following intravenous injection of a paramagnetic, Gdbased contrast medium, uterine zonal anatomy is also visualized on T1w images (Fig. 12.13). There is marked enhancement of the endometrium and myometrium, whereas the junctional zone remains low in signal intensity, which has been attributed to its more compact structure and smaller extracellular space for distribution of the contrast agent.8,9

247

Fig. 12.9 Normal zonal anatomy of the uterus in mid-cycle in a 30-year-old woman. T2w SE image shows high-SI endometrium surrounded by low-SI junctional zone. The outer myometrium is of moderately high SI.

Fig. 12.10 Uterus in the first half of the menstrual cycle (proliferative phase) in a 32-year-old woman. T2w SE image shows thin endometrial layer and rather low SI of myometrium.

On MRI of the neonatal uterus, the myometrium and endometrium can be differentiated because of the prenatal influence of maternal hormones. Before menarche, the endometrium is indistinct or appears as a thin stripe of high signal intensity, while the junctional zone cannot be distinguished from the low-signal-intensity myometrium. At this age, the uterine corpus is small and shorter than the cervix (Fig. 12.14).

248

12 The Uterus and Vagina

Fig. 12.11 Uterus in the secretory phase in an 18-year-old woman. T2w TSE image shows thick endometrium and increased SI of the myometrium. Some ascites is present in the rectouterine pouch with good delineation of the peritoneal fold as it is reflected from the uterus onto the rectum.

Fig. 12.12 Small blood clot (arrow) in the endometrial cavity on T2w TSE image.

Fig. 12.13 Normal appearance of the uterus after IV administration of Gd-based contrast medium. Same patient as Fig. 12.12. T1w SE image shows prominent enhancement of the endometrium and marked enhancement of the vaginal mucosa. Low SI of the vaginal muscle.

Fig. 12.14 Normal uterus in a sexually mature 14-year-old girl on T2w TSE image. Note the length of the uterine cervix relative to the length of the corpus.

In postmenopausal women, T2w images will demonstrate a thin stripe of high-signal-intensity endometrium, while the junctional zone and myometrium are indistinct because both are of low signal intensity (Fig. 12.15). The endometrial thickness is ca. 3–5 mm. In women on hormone replacement therapy, the endometrium may be up to 10 mm thick and the premenopausal signal characteristics and zonal anatomy are preserved. Ultrasonography also distinguishes three uterine layers, but these do not

exactly match the zones seen on MRI, which explains the divergent thickness data.10 Hormone treatment also affects the MRI appearance of the uterus and vagina. The myometrium has higher signal intensity in women who take oral contraceptives (Fig. 12.16). T2w images of the adult cervix show an innermost area of high signal intensity surrounded by a rather thick layer of low signal intensity and an outer layer of intermediate

MRI Appearance of Normal Anatomy

signal intensity (Figs. 12.11 and 12.17). The high-signalintensity area corresponds to intracervical mucus and the endocervical glands. High-resolution MRI will occasionally depict an additional thin stripe of moderate hyperintensity (Fig. 12.18), which most likely represents the cervical mucosa. Also seen on high-resolution images are the palmate folds. The low-signal-intensity layer represents the cervical stroma, which is rich in connective tissue. In contrast, the outer layer, which is similar in signal intensity to the uterine myometrium, has a less compact structure.11 Nabothian cysts are common benign cervical lesions that develop when mucous glands become enlarged or obstructed. They are depicted on T2w images as welldefined, round or oval lesions of high signal intensity (Figs. 12.6 and 12.42).12 After intravenous contrast administration, the cervical zonal anatomy is also seen on T1w images with the cervical mucosa, the ectocervix, and the outer cervical stroma showing stronger enhancement (Fig. 12.19). The MRI appearance of the cervical zonal anatomy does not change much after menopause, nor is it much different in women taking exogenous hormones or oral contraceptives. The parametrium has low signal intensity on T1w images, due to its large connective tissue component and the presence of blood and lymphatic vessels, and cannot be reliably differentiated from the uterine cervix (Fig. 12.20a). On T2w images, the higher signal intensity of the parametrial tissue allows good differentiation from the inner layer of low-signal-intensity cervical stroma (Fig. 12.20b) but may be indistinct from the outer cervical stroma, which is of similar T2 signal intensity. In advanced endometrial cancer with parametrial invasion, tumor in the fatty tissue near the pelvic sidewalls is best appreciated on T1w images. The vagina is of intermediate signal intensity on T1w images, similar to that of the urethra and rectum (Fig. 12.21a). Differentiation of these structures is much better on T2w and contrast-enhanced T1w images (Fig. 12.21b and c). Insertion of a tampon into the vagina is not necessary and may even distort anatomy. Sagittal images are most suitable to differentiate the posterior vaginal fornix from the cervix and anterior rectal wall (Figs. 12.18 and 12.19), whereas the smaller anterior fornix is less well delineated from the cervix. The paravaginal tissues (the paracolpium) contain many veins and therefore have high T2 signal intensity (Figs. 12.12 and 12.21b). The MRI appearance of the vagina, like that of the uterus, responds to hormones. In the early proliferative phase, the vagina is of low signal intensity on T2w images with a central hyperintense stripe corresponding to mucus and epithelium. During the secretory phase, the central stripe is somewhat thicker, and, in ca. 70 % of women, the vaginal wall increases in signal intensity. Thus, the contrast between the wall and mucosa/mucus is reduced

249

Fig. 12.15 Postmenopausal uterus on T2w SE image.

Fig. 12.16 Normal uterus in a woman taking oral contraceptives. T2w SE image shows thin endometrium and increased myometrial SI.

compared with the proliferative phase.13 In postmenopausal women not taking exogenous hormones, T2w images show a low-signal-intensity vagina with a very thin central stripe of high signal intensity. In postmenopausal women on hormone replacement therapy, the vagina has the same MR appearance as during the proliferative phase in premenopausal women.

250

12 The Uterus and Vagina

a

b Fig. 12.17a–c Normal appearance of the uterine cervix. a Sagittal T2w TSE image. The hypointense junctional zone of the uterine corpus blends into the hypointense cervical stroma. Some ascites is present in the rectouterine pouch. b Axial T2w TSE image. Arrows indicate the low-SI rim of the inner cervical stroma surrounding the narrow high-SI cervical canal. c Sagittal T1w SE image after IV contrast injection. There is marked enhancement of the endometrium and cervical mucosa.

c

Fig. 12.18 Normal appearance of the uterine cervix on sagittal T2w TSE image. Four zones can be distinguished: the innermost high-SI mucus, the palmate folds, the low-SI inner cervical stroma, and the intermediate-SI outer stroma. The ectocervix protrudes into the vagina. Good delineation of the posterior vaginal fornix (arrow) from the high-SI mucus.

Fig. 12.19 Normal appearance of the uterine cervix on sagittal T1w SE image after IV contrast administration. There is enhancement of the cervical endometrium and mucosa and of the vagina (moderate enhancement of the outer myometrium), but no enhancement of the mucus in the cervical canal and vagina. Good delineation of the posterior and anterior vaginal fornices. Very high SI of the urine in the bladder following IV contrast injection.

MRI Appearance of Normal Anatomy

251

a

b Fig. 12.20a, b Cervix and parametria on axial images. a On the T1w image, it is nearly impossible to differentiate between the cervix and the parametria. High SI of the parametria near the pelvic sidewalls is

a

due to fatty tissue. b On the T2w image, the cervix is clearly identified by the low-SI rim of normal stroma. The outer cervical stroma blends into the parametrial tissue.

b Fig. 12.21a–c Normal axial view of the vagina. a Poor delineation of the vagina from the paracolpium, urethra, and rectum on T1w SE image. b Good differentiation of the vagina from the high-SI paravaginal tissues, urethra, rectum, and levator ani muscle on T2w image. c Axial T1w image of the vagina after IV contrast administration.

c

252

12 The Uterus and Vagina

MRI Appearance of Pathologic Entities Congenital Uterine Anomalies The lower part of the vagina develops from the urogenital sinus, while the uterine tubes, uterus, cervix, and upper part of the vagina derive from the paired Müllerian ducts of the embryo. Müllerian duct anomalies result from agenesis, absent or incomplete fusion of the two ducts, or failure of septal resorption (Fig. 12.22). Such anomalies are present in 2–3 % of all women14,15 and are diagnosed in 9–25 % of women presenting with infertility or a history of miscarriage.16,17 Clinical signs and symptoms of uterine anomalies comprise pain due to primary amenorrhea,

recurrent miscarriage, complications during delivery, and symptoms of endometriosis. It is important to note that 20–40 % of women have concomitant renal anomalies (e. g., agenesis, ectopia, malrotation). Besides diagnostic laparoscopy, there are several noninvasive tests such as hysterosalpingography, endovaginal ultrasound, and MRI for diagnosing uterine anomalies. MRI allows comprehensive evaluation of pelvic anatomy and also permits concomitant assessment of the entire urinary tract18,19 and is therefore considered the modality of choice, at least in women who are candidates for surgery.20–22 MRI should be performed with the highest possible spatial resolution. The use of an FSE or TSE sequence with multiple signal averages23–25 and a high matrix in

Uterus didelphys

Uterus bicornis bicollis

Septate uterus

Fig. 12.22 Uterine anomalies.

Uterus bicornis unicollis

Subseptate uterus

Unicornuate uterus

MRI Appearance of Pathologic Entities

conjunction with a phased-array body coil is recommended. There are no special requirements regarding patient preparation and positioning. No intravenous contrast medium is needed. The pulse sequences to be acquired (Table 12.3) are coronal T1w images with a large FOV for evaluation of the kidneys and urinary tract, followed by T2w sequences of the true pelvis in sagittal and axial planes with a slice thickness of 4–5 mm. These are the most important pulse sequences for evaluation of the uterine contour and zonal anatomy as well as of the ovaries. The axial images can be used to prescribe an oblique sequence angled to the long axis of the uterus (Fig. 12.4), which is important for differentiating a septate from a bicornuate uterus. The shape of the endometrial cavity is best appreciated when the endometrium is thick during the secretory phase. Axial T1w images additionally allow characterization of ovarian cysts, fibromas, and hemorrhage. If endometriosis is suspected, a fat-saturated T2w sequence should be obtained to identify small endometriotic deposits.

253

Table 12.3 Congenital anomalies—recommended pulse sequences Sequence

Scan area/FOV (plane)

Comment

T1w SE

Entire urinary tract (coronal) Uterus and vagina including ovaries (axial and sagittal) Uterus (long-axis view) Same as T2w TSE (axial or sagittal) True pelvis (axial)

Renal and urinary tract anomalies Uterine anomalies, vaginal septum, ovaries Uterine anomalies Ovaries: characterization, hemorrhage Ectopic endometrium

T2w TSE

T2w TSE T1w SE (optional) T2w with FS (optional)

Complete and Partial Agenesis MRI allows good evaluation for agenesis of the different parts of the urogenital tract, except for the ovarian tubes (Fig. 12.23). The most common form of urogenital agenesis is Mayer–Rokitansky–Küster–Hauser syndrome, which is characterized by the absence of the uterus, cervix, and upper part of the vagina, while the tubes and ovaries are normal in most cases. It is the second most common cause of primary amenorrhea after gonadal dysgenesis.26 Isolated agenesis or hypoplasia of the uterus is very rare. Absence of the uterus is best seen on sagittal T2w images, while absence of the cervix and proximal vagina is easier to identify on axial images. In Mayer–Rokitansky– Küster–Hauser syndrome, there is prominent absence of the uterus and cervix with a blind-ending lower vagina. A hypoplastic uterus with absence of the upper vagina is rare.27

a

Uterus Didelphys (Double Uterus) A uterus didelphys results from complete failure of müllerian duct fusion and is characterized by the presence of two fully developed uterine bodies and cervices. This anomaly may be associated with a longitudinal or transverse septum dividing the upper vagina. Obstruction of one of the uterine cavities by the septum results in hematometrocolpos with ipsilateral hematosalpinx.25 Although the two uterine bodies are quite far apart, the cervical canals lie side by side but are completely separate (Fig. 12.24). Whenever two cervical canals are demonstrated by imaging, special attention should be paid to the presence of a vaginal septum or a double vagina (Fig. 12.24). Since the vaginal septum is not always seen

b Fig. 12.23a, b Absence of the uterus and vagina on sagittal (a) and axial (b) T2w SE images.

254

12 The Uterus and Vagina

a

b Fig. 12.24a, b Uterus didelphys. Axial T2w SE images. a Good visualization of the two parallel cervical canals. b Demonstration of two vaginas (arrows).

Hematometra

Fig. 12.25 Unicornuate uterus with noncommunicating left horn. Endometrium from the noncommunicating horn can enter the peritoneal cavity through the uterine tube and cause endometriosis. Fig. 12.26 Uterus bicornis unicollis. T2w TSE image.

on MRI, the clinical findings must be taken into account to differentiate a uterus didelphys from a uterus bicornis bicollis.25,26

Unicornuate Uterus A unicornuate uterus is the result of incomplete or absent development of one of the müllerian ducts. The former is characterized by the presence of a rudimentary horn with or without functioning endometrium. Clinically, a unicornuate uterus is associated with a very high rate of miscarriage and complications during delivery. A noncommunicating horn containing functioning endometrial tissue can cause endometriosis (Fig. 12.25). A unicornuate uterus appears banana-shaped and tends to be deviated

to one side. The typical triangular appearance of the uterine body is absent, but the zonal anatomy is preserved. If a rudimentary horn is present, it is similar in signal intensity to myometrium and may be distended by blood if it has an endometrial cavity.27,28 Women with this anomaly should be evaluated for the presence of ectopic endometrial tissue outside the uterus.

Bicornuate Uterus A bicornuate uterus results from partial nonfusion of the müllerian ducts. The septum separating the two horns is composed of myometrium and may extend to the internal os (bicornis unicollis) or the external os (bicornis bicollis) (Figs. 12.26, 12.27).

MRI Appearance of Pathologic Entities

255

a

b Fig. 12.27a, b Uterus bicornis bicollis. Axial T2w TSE images. a The typical zonal anatomy (endometrium, junctional zone, myometrium) allows good identification of the two uterine horns (arrows). b Parallel arrangement of the two cervical canals.

MRI will demonstrate two uterine cavities lined by normal endometrium. The best image quality is again achieved during the secretory phase. The fundus (or roof of the uterus) is concave (with a depression of > 1 cm) and the intercornual distance is increased (> 4 cm).28,29 Fundal concavity is best appreciated on oblique coronal T2w images angled to the long axis. The septum has the same signal intensity as endometrium on all pulse sequences. The lower portion of the septum may be composed of connective tissue, reflected by lower T1w and T2w signal intensities. A cervical septum, if present, also has connective tissue signal intensity. Arcuate uterus is the mildest form of bicornuate uterus. It is characterized by a flattened fundus and a short septum more than 1 cm long.

Septate Uterus A septate uterus results from incomplete resorption of the fibrous septum between the two horns. The septum may be short or extend down to the external os. Differentiation of a septate uterus from a bicornuate uterus is important for treatment planning because a septum can be removed endoscopically, while correction of a bicornuate uterus requires open surgery. These two anomalies are best distinguished on coronal T2w images angled to the uterine axis. Septate uterus differs from bicornuate uterus in that the distance between the two horns is smaller (< 4 cm) and the fundus is convex or slightly flattened (Fig. 12.28).29

Fig. 12.28 Septate uterus. T2w SE image.

Acquired Benign Uterine Disorders Leiomyomas Leiomyomas or fibroids are the most common tumors of the uterus, occurring in 30–40 % of women during their reproductive years. They are composed of smooth muscle tissue with variable amounts of connective tissue and can occur as single or multiple tumors. Most leiomyomas arise in the body of the uterus (90 %) and 5 % in the cervix. An occasional fibroid is found outside the uterus. They may be submucosal, intramural, or subserosal. The majority of

256

12 The Uterus and Vagina

Table 12.4 Leiomyomas and adenomyosis—recommended pulse sequences Sequence

Scan area/FOV (plane)

Comment

T2w SE

Uterus (axial and sagittal)

Determination of number, size, and site of fibroids; differential diagnosis: fibroid versus adenomyosis; characterization

Fig. 12.29 Intramural leiomyomas. Sagittal T2w TSE image reveals two low-SI tumors in the myometrium. The larger tumor (arrow) in the posterior uterine wall shows degenerative changes. The smaller tumor is located in the anterior wall.

Fig. 12.30 Small intramural lyomyoma in the area of the left tubal ostium on coronal T2w SE image.

leiomyomas are intramural and are typically detected incidentally. However, depending on their size and location in the uterus, fibroids can block a uterine tube and cause infertility or obstruct delivery. Another 5–10 % are submucosal or protrude into the endometrial cavity. They most often cause clinical symptoms (e. g., hypermenorrhea). Subserosal leiomyomas occasionally mimic a solid ovarian mass. Very rarely, torsion of a pedunculated fibroid can cause an acute abdomen. Other common clinical symptoms are enlargement of the uterus, menorrhagia, infertility, frequent miscarriage, and delivery complications. Appropriate treatment depends on the size and location of the fibroids as well as the severity of symptoms. In women who wish to have children, small submucosal leiomyomas can be removed by hysteroscopy, while intramural tumors are an indication for myomectomy. A more recent therapeutic option is uterine artery embolization (UAE), which is suitable for treating larger and multiple leiomyomas and can spare women from hysterectomy.30,31 Fibroids can be diagnosed by pelvic examination or by ultrasound. The diagnosis rarely poses a problem. MRI is mainly used to examine women scheduled for uterussparing surgery or interventional treatment (UAE) and occasionally for lesion characterization.32,33 The high spatial resolution afforded by MRI allows detection of very small tumors (e. g., around the tubal ostia), reliable localization (e. g., subserosal leiomyomas), and detailed morphologic characterization. If uterine leiomyomas are found incidentally on MRI, the radiologist must reliably establish their benign character to obviate the need for further diagnostic testing. MRI Appearance T2w pulse sequences in two planes are the basis for fibroid localization and characterization (Table 12.4). T1w images are not required but may be useful to demonstrate intralesional hemorrhage. Uterine leiomyomas have a characteristic MR appearance. They are depicted as round, well-defined lesions and are clearly delineated from the higher signal intensity of the myometrium. Their low signal intensity on T2w images distinguishes them from malignant tumors (Figs. 12.29, 12.30, 12.31, 12.32).34,35 Most fibroids have a pseudocapsule due to compression of surrounding tissue. Additionally, a thin hyperintense rim of dilated lymphatic clefts, dilated veins, and mild edema may be seen on T2w images (Figs. 12.32 and 12.38) in about one third of cases36. T1w images only show an enlarged uterus, while the fibroids are isointense to the uterus. MRI allows good localization of leiomyomas. A submucosal leiomyoma elevates the endometrium or is seen within the endometrial cavity when pedunculated. Their low T2 signal intensity clearly differentiates fibroids from endometrial polyps and endometrial carcinoma.37 Intramural leiomyomas (Fig. 12.29) are well demarcated from the higher-signal-intensity myomtrium. A large subserosal leiomyoma (Fig. 12.33) may initially appear as a pel-

MRI Appearance of Pathologic Entities

257

Fig. 12.32 Small intramural leiomyoma with a thin high-SI rim in the posterior uterine wall. Sagittal T2w SE image.

v Fig. 12.31 Myomatous uterus with multiple intramural (arrow), submucosal, and subserosal (curved arrow) leiomyomas on sagittal T2w TSE image.

Fig. 12.33 Giant leiomyomas of the uterus. Sagittal T2w TSE image. The low SI suggests benign lesions. The leiomyoma located in the pelvis displaces the uterus and bladder anteriorly, while the second tumor mass makes the uterus appear elongated. The arrow indicates the connection of the tumor with the uterus (arrow), which is diagnostic of leiomyoma and excludes ovarian fibroma.

vic mass; however, the lower signal intensity should alert the radiologist to the possibility of a uterine fibroid (differential diagnosis: ovarian fibroma). Demonstration of a pedicle connecting the mass to the uterus (on T2w image) is diagnostic of a fibroid (Figs. 12.33 and 12.34). Uterine fibroids without degenerative changes have nearly homogeneous low signal intensity on T2w images. Areas of high signal intensity within a lesion indicate degenerative changes and are seen in up to 60 % of leiomyomas, especially large ones (Figs. 12.34 and 12.35). Hyaline degeneration is the most common form (high T2 signal, low T1 signal). Intralesional hemorrhage (red degeneration) is far less common and is identified as high signal intensity on T1w images and variable T2 signal intensity, depending on the age of the hemorrhage.38 Calcifications are another typical feature but are difficult to detect by MRI. They are not relevant because the aforementioned features enable a reliable diagnosis to be made by MRI. Intravenous contrast administration does not provide useful additional information39 because fibroid enhancement varies with the degree of vascularization (Fig. 12.35). However, it has been suggested that a dynamic contrast-enhanced study using fast GRE sequences might improve the differentiation of leiomyoma and leiomyosarcoma.40 Moreover, contrast-enhanced MRI is useful for preinterventional planning and follow-up of UAE (Figs. 12.36 and 12.37).30,31

258

12 The Uterus and Vagina

a

b Fig. 12.34a, b Large subserosal leiomyoma of the posterior uterine wall. a Axial T2w TSE image. b Sagittal T2w TSE image. Inhomogeneous areas of high SI (arrows) within the low-SI mass indicate degeneration.

a

b Fig. 12.35a, b Large myomatous uterus with a nondegenerative and a degenerative leiomyoma. a Sagittal T2w SE image shows a uniformly hypointense leiomyoma without degenerative changes above a predominantly hypointense mass with inhomogeneous

hyperintensities, consistent with degenerative changes. Blood clots, indicated by low SI, are present in the endometrial cavity following D&C. b T1w SE image obtained immediately after contrast administration shows only mild enhancement of the two leiomyomas.

Adenomyosis

ectopic tissue is functional). Uterine adenomyosis is demonstrated in up to 25 % of all hysterectomy specimens and is thus more common than clinically apparent.42 One in four women with adenomyosis has concomitant fibroids. Adenomyosis typically becomes manifest in the fourth or fifth decade of life and is more common in multiparous

Adenomyosis is the invasion of the myometrium by endometrial tissue.41 The ectopic endometrium mostly consists of tissue of the stratum basale, which changes only little during the menstrual cycle, and typically does not contain blood (in contradistinction to endometriosis, in which the

MRI Appearance of Pathologic Entities

259

a

b Fig. 12.36a, b Myomatous uterus. a Sagittal T2w TSE image. b Coronal MIP image from MR angiography shows increased tortuosity of both uterine arteries (arrows).

a

b Fig. 12.37a–c Intramural leiomyoma of the posterior uterine wall. a Sagittal T2w TSE image. b Postcontrast T1w SE image obtained before UAE shows nearly homogeneous enhancement of the leiomyoma. c Postcontrast T1w SE image after bilateral UAE shows complete absence of perfusion within the leiomyoma and normal myometrial perfusion.

c

260

12 The Uterus and Vagina

Fig. 12.38 Adenomyosis and leiomyomas of the uterus. Sagittal T2w TSE image shows thickening of the low-SI junctional zone with foci of high SI and intramural leiomyomas of the posterior and anterior uterine wall.

Fig. 12.39 Focal adenomyosis of the uterus. Sagittal T2w TSE image shows focal thickening of the low-SI junctional zone (arrows).

UAE has evolved into a valid alternative for fibroid treatment and has also been used in adenomyosis, but experience is still limited. Hormone treatment can shrink fibroids but has no effect on adenomyosis. MRI permits reliable differentiation of these two entities and is clearly superior to transabdominal or endovaginal ultrasound, especially in patients with adenomyosis and concomitant fibroids.43–45

Fig. 12.40 Diffuse uterine adenomyosis. Sagittal T2w SE image reveals generalized thickening of the low-SI junctional zone.

women. Symptoms may resemble those of fibroids and include hypermenorrhea, dysmenorrhea, and an enlarged uterus. On pelvic examination, the uterus is enlarged but is of softer consistency than in women with fibroids. Reliable differentiation of adenomyosis from uterine leiomyomas is critical because the treatment options are different: uterus-sparing myomectomy is an option in women with fibroids, while hysterectomy is the traditional treatment of choice for symptomatic adenomyosis.

MRI Appearance Adenomyosis is diagnosed by T2w imaging (Table 12.4). The crucial diagnostic criterion for adenomyosis is abnormal thickening of the low-signal-intensity junctional zone to > 12 mm (Figs. 12.38, 12.39, 12.40).46–49 Less pronounced thickening (5–10 mm) may occasionally be seen in a normal uterus, e. g., due to transitional contraction, and should not be interpreted to indicate adenomyosis.50 The thickened junctional zone often contains punctate foci of high signal intensity (Figs. 12.38 and 12.39), which most likely correspond to interspersed glandular tissue if only present on T2w images.51 If there is both T1 and T2 hyperintensity, the foci are most likely caused by hemorrhage. Adenomyosis may be diffuse or focal. In the diffuse form, changes are present throughout the uterus (Fig. 12.40), which can be considerably enlarged. Focal adenomyosis is characterized by localized thickening of the junctional zone with poor demarcation from adjacent myometrium (Fig. 12.39) and is occasionally difficult to distinguish from a fibroid because both entities have low signal intensity on T2w images. A leiomyoma is more rounded and a better-defined tumor in comparison with focal adenomyosis, which appears as an irregular and poorly defined lesion without mass effect. In most cases,

MRI Appearance of Pathologic Entities

261

focal adenomyosis arises in the junctional zone. A localized form in submucosal or subserosal location without a direct connection to the junctional zone and with mass effect, called adenomyoma, is very rarely found48. Administration of contrast medium adds no information for diagnosing adenomyosis or differentiating the focal form from leiomyoma. Contrast-enhanced imaging is typically indicated only for preinterventional planning and followup of UAE.52,53

Gestational Trophoblastic Disease: Hydatidiform Mole and Choriocarcinoma Gestational trophoblastic disease (GTD) is a cytogenetically and clinically heterogeneous spectrum of tumors characterized by the proliferation of trophoblastic tissue. The tumors range in biological behavior from benign to highly malignant (hydatiform mole—invasive mole—choriocarcinoma). GTD is the most common neoplasm occurring in pregnant women. Clinical signs and symptoms comprise bleeding from the uterus, uterine enlargement, and failure of the uterus to return to its normal size after miscarriage. Human chorionic gonadotropin (β-hCG) is an important diagnostic marker; its serum levels closely correlate with tumor behavior and allow accurate monitoring of response to treatment (progression or regression). The diagnosis being primarily based on β-hCG, the role of MRI is limited to defining the site and extent of the process and follow-up during treatment. MRI Appearance GTD distorts or obliterates the uterine zones. The bulk of the process is found in the myometrium with rare endometrial infiltration. The MRI appearance is usually inhomogeneous due to necrosis and hemorrhage. The mole is hypervascular and MRI demonstrates dilated, tortuous vessels. Arteriovenous shunting within the mole results in dilatation of the uterine and internal iliac vessels. Normal uterine zonal anatomy should return 6–9 months after initiation of chemotherapy.54

Endometrial Polyps Endometrial polyps are seen in ca. 10 % of women; they are a frequent finding during menopause. Most polyps arise from the fundus, typically in the area of the tubal ostia. They are often associated with uterine fibroids. The risk of malignant transformation is less than 1 %.55 Endometrial polyps present with irregular or persistent bleeding. Isolated endometrial polyps are demonstrated in > 20 % of women with postmenopausal bleeding. Many endometrial polyps are asymptomatic. The diagnosis of endometrial polyps may pose a challenge. Even very careful dilatation and curettage (D&C) may not remove them all. Hysteroscopy is the most reliable diagnostic tool; endovaginal ultrasound has limita-

Fig. 12.41 Endometrial polyp. Sagittal T2w SE image. The endometrial polyp has slightly lower SI than the endometrium. The junctional zone is intact and has normal low SI. Polyps cannot be differentiated from polypoid endometrial carcinoma on the basis of their MR appearance.

tions. Since women with suspected endometrial polyps will invariably undergo D&C, MRI is rarely indicated. Nevertheless, a radiologist should be familiar with the MRI appearance of endometrial polyps as MRI is used in the diagnostic work-up of women presenting with postmenopausal bleeding. MRI Appearance The diagnosis mostly relies on sagittal and axial T2w images, on which endometrial polyps are isointense or slightly hypointense to endometrium (Fig. 12.41). An occasional large polyp may distend the endometrial cavity. On T1w images, most endometrial polyps are not revealed. Following contrast administration, polyps enhance intensely, similar to endometrium.39,56 The MR signal intensity allows differentiation from submucosal fibroids, which are of lower signal intensity on T2w images. Contrast-enhanced images can be helpful in unclear cases. Fibroids enhance moderately at most.

Uterine Cervical Stenosis Stenosis of the uterine cervix can develop secondary to infection, surgery, or radiotherapy. It is typically seen around the external os and may cause serometra, hematometra, or pyometra. A cervical stenosis is suggested on T2w MR images by high-signal-intensity fluid retained in the endometrial cavity in conjunction with a typically short narrowing of the otherwise hyperintense cervical canal (on T2w image). The normal low signal intensity of the stroma is preserved (in contrast to tumor-related stricture).

262

12 The Uterus and Vagina

Cervical Cysts (Nabothian Cysts)

Malignant Tumors of the Uterus

Nabothian cysts (Fig. 12.42) are retention cysts that cause no symptoms and do not require treatment. They are typically detected incidentally and their MR appearance clearly distinguishes them from other cervical lesions such as cancer.12 Nabothian cysts usually have a diameter of a few millimeters but may occasionally grow to 2–4 cm.

Endometrial Carcinoma

a

b Fig. 12.42a, b Nabothian cysts (arrows). a Axial T2w TSE image. b Sagittal T2w TSE image.

Endometrial carcinoma is the most common malignant tumor of the female genital tract (annual incidence of 28 cases per 100 000 women). It is a disease of postmenopausal women with only 2–5 % of cases occurring before age 40. The cancer typically arises in the functional layer of the endometrium, and it is assumed that its growth is stimulated by estrogen. Characteristically, endometrial cancer arises when the functional endometrium is no longer transformed and discharged in the course of the menstrual cycle. Risk factors for developing endometrial cancer are long-term unopposed estrogen replacement therapy after menopause and estrogen-producing tumors. Many women with endometrial carcinoma share a set of factors comprising overweight, hypertension, and reduced glucose tolerance or diabetes mellitus, which are occasionally referred to as the “endometrial carcinoma syndrome.” Endometrial cancer is characterized by invasive growth into the myometrium but rarely extends through the serosa into the abdominal cavity. The cervix is involved in ca. 10 % of cases. Lymphatic and distant metastases occur late compared with cervical cancer. Depth of myometrial invasion is an important prognostic factor and closely correlates with the presence of lymph node metastases.55,57 Only 3 % of women with superficial myometrial invasion (stage IB) have positive lymph nodes, as opposed to 40 % of those with deep myometrial invasion (stage IC). Retrograde spread to the vagina is common (5–10 %) and must be taken into account when planning the MRI examination. Special attention must be paid to the upper third of the vagina and to the anterior vaginal wall and the urethral crest in the lower third of the vagina. Organ metastases are rare and occur late in the course of disease. Postmenopausal bleeding is the most common symptom of endometrial cancer and often the only one. The most important diagnostic procedure for confirming suspected endometrial carcinoma is fractional curettage (to obtain an endocervical and an endometrial sample to confirm or rule out stage II tumor). Transvaginal ultrasound allows reliable evaluation of the depth of myometrial invasion in stage I cancer but is unsuitable for general tumor staging (e. g., cervical involvement).58 MRI enables excellent evaluation of myometrial invasion and good overall tumor staging. It has evolved into the modality of choice for staging endometrial cancer and provides important information for treatment planning (e. g., hysterectomy with/without lymphadenectomy, chemotherapy, gestagen therapy).

MRI Appearance of Pathologic Entities

263

Table 12.5 Endometrial carcinoma—recommended pulse sequences Sequence

Scan area/FOV (plane)

Comment

T2w TSE

Uterus and vagina (axial and sagittal) (3–5 mm slice thickness) Entire pelvis (axial) Uterus and vagina (optimal angled views as for T2w images) Same as T1w SE

Depth of myometrial invasion, invasion of cervix, vaginal metastasis Lymph node staging, pelvic sidewall Baseline SI for evaluation of signal enhancement on postcontrast images Tumor extent, depth of myometrial invasion; differential diagnosis: tumor versus necrosis or fluid

T1w SE or PD TSE T1w SE T1w SE after contrast administration

MRI Appearance and Tumor Staging Endometrial carcinoma has no specific MR features. Small endometrial carcinomas without myometrial invasion have the same MR signal intensity as normal endometrium and cannot be distinguished from endometrial hyperplasia or a polyp. The task of MRI, therefore, is to stage endometrial carcinoma after the diagnosis has been established. The MRI protocol begins with T2w sequences of the uterus and vagina in two planes (Table 12.5) to determine the depth of myometrial invasion and identify possible involvement of the cervix and vagina. The entire pelvis should be imaged with an axial T1w or PD sequence for evaluation of pelvic lymph nodes and the pelvic sidewalls before performing a contrast-enhanced MR study. Unenhanced T1w images usually do not contribute any information for staging cancer confined to the uterus but provide the baseline signal intensity for evaluating enhancement on the subsequent postcontrast images. Postcontrast T1w images (using an SE or TSE sequence) acquired immediately after bolus administration of contrast medium (in the plane deemed most promising on the basis of the preceding T2w sequence) will improve delineation of endometrial cancer from the surrounding myometrium. T2w images depict endometrial cancer as a mass of high signal intensity, typically associated with abnormally increased thickness or lobulation of the endometrium; an occasional tumor may contain heterogeneous areas of lower signal intensity (Figs. 12.43 and 12.44). These changes are not specific and may also occur in endometrial hyperplasia or polyps or when a blood clot is present. The final diagnosis is therefore always based on histologic work-up. MRI depicts myometrial invasion as high-signalintensity interruption of the junctional zone and therefore allows determination of the depth of invasion (staging of endometrial carcinoma, Table 12.6). Imaging in two planes is essential for accurately assessing the depth of myometrial invasion (Fig. 12.45). Even T2-weighted images may not always allow reliable differentiation of an endometrial tumor from the surrounding myometrium (Fig. 12.46a). Large polypoid endometrial carcinomas may be overstaged by MRI (they are associated with thinning of the myometrium but do not invade it). Staging errors may also occur in patients with concomitant large uterine fibroids, congenital anomalies, or a small atrophic uterus, or if there is complete loss of the zonal antomy.59

Fig. 12.43 Endometrial carcinoma, stage IA. Sagittal T2w TSE image. Abnormal thickening of the endometrium, which contains areas of lower SI. The thin low-SI junctional zone is intact. Also present are uterine fibroids. The image is degraded by motion artifacts from the hyperintense urinary bladder.

Staging of endometrial cancer is incomplete without postcontrast imaging.60–62 On postcontrast images, endometrial cancer becomes more conspicuous because it enhances less intensely than the myometrium and endometrium (Figs. 12.46 and 12.47). Contrast-enhanced MRI also improves the differentiation of viable tumor from necrosis and fluid (e. g., hematometra or pyometra). MRI staging of endometrial cancer is based on the FIGO classification (Table 12.6). Stage I tumor is confined to the corpus. Three subtypes with different prognosis based on myometrial invasion are distinguished: IA: No myometrial invasion IB: Invasion of < 50 % of the myometrium IC: Invasion of > 50 % of the myometrium.

264

12 The Uterus and Vagina

a

b Fig. 12.44a, b Endometrial carcinoma, stage IB. a Sagittal T2w TSE image. Abnormal thickening of the endometrium. There appears to be invasion of the myometrium (at least of the inner half) of the

anterior uterine wall (arrow). b Contrast-enhanced T1w SE image improves tumor conspicuity and confirms invasion of < 50 % of the myometrium.

a

b Fig. 12.45a–c Endometrial carcinoma, stage IC. a Sagittal T2w SE image. Invasion of the anterior and posterior uterine wall (arrows) by endometrial cancer is already apparent without contrast administration. b, c Sagittal and axial T1w SE images acquired immediately after contrast administration improve contrast between the hypovascular endometrial carcinoma and the enhancing myometrium. The images confirm deep myometrial invasion of the anterior uterine wall and superficial invasion of the posterior wall.

c

MRI Appearance of Pathologic Entities

a

265

b Fig. 12.46a–c Endometrial carcinoma, stage II. a The histologically proven endometrial carcinoma is not seen on the sagittal T2w image because it is isointense to myometrium. The endometrial cavity has low SI due to the presence of blood clots. Atypical appearance of the uterine cervix after conization. b, c T1w SE images before (b) and immediately after (c) contrast administration. Contrast administration markedly improves conspicuity of the tumor, which invades > 50 % of of the myometrium and the cervix.

c Fig. 12.47 Endometrial carcinoma, stage IV. Sagittal T1w SE image. The uterus is enlarged and has a slightly irregular contour. Ascites is present and there is tumor invasion of the major omentum (arrows).

266

12 The Uterus and Vagina

Table 12.6 Staging of endometrial carcinoma FIGO* stage

Criteria

MR findings on T2w and contrast-enhanced T1w images

0 I IA

Carcinoma in situ Tumor confined to uterine corpus No myometrial invasion

No tumor identifiable

IB

Invasion of < 50 % of the myometrium

IC

Invasion of > 50 % of the myometrium

II IIA IIB III IIIA

Cervical invasion, otherwise confined to uterus No invasion of cervical stroma Invasion of cervical stroma Tumor beyond uterus but within true pelvis Invasion of serosa and/or adnexa or positive peritoneal cytology Vaginal invasion Lymph node metastasis Tumor outside true pelvis, or invasion of bladder or rectal mucosa

IIIB IIIC IV

Intact junctional zone; if juntional zone is not visible, endometriummyometrium interface should be smooth and sharp Circumscribed disruption of junctional zone or irregular endometriummyometrium interface; all changes seen involve < 50 % of myometrium Deep invasion of myometrium (> 50 %); thin outer stripe of normal myometrium indicates that serosa is still spared Altered signal intensity extends into endocervix or cervical stroma Normal low-SI cervical stroma (T2w image) High-SI invasion of the cervical stroma (T2w image) Complete myometrial invasion (seen in 2 planes); peritoneal carcinosis is identified by abnormal enhancement on postcontrast T1w images High-SI mass in the vagina Enlarged pelvic lymph nodes (1 cm) High-SI disruption of the low-SI muscle of the bladder or rectum

*International Federation of Gynecology and Obstetrics.

An intact low-signal-intensity junctional zone indicates stage IA endometrial cancer (Fig. 12.43). If the uterus is atrophic and the junctional zone is not clearly visualized, stage IA is indicated by a sharp tumor–myometrium interface. Stage IB is characterized by disruption of the junctional zone or a blurred/irregular tumor–myometrium interface; the tumor signal intensity extends into < 50 % of the myometrium. The defining feature of stage IB is disruption of the junctional zone, which, however, is not well defined in postmenopausal women. Contrastenhanced T1w images are helpful to establish myometrial invasion. Deep myometrial invasion, i. e., involvement of > 50 % of the myometrium, defines stage IC endometrial cancer (Fig. 12.45). Staging is difficult when there is symmetrical rarefaction of the myometrium (e. g., large polypoid endocavitary tumor or hematometra). The former FIGO criterion of length of the endometrial cavity would be easy to determine by MRI; however, it does not correlate with the tumor stage and depends primarily on the patient’s hormonal status. The depth of myometrial invasion is a more important prognostic factor. Stage II is characterized by tumor invasion of the cervix or endocervix (Fig. 12.46c). Cervical invasion is suggested on T2w images obtained in two planes but should be confirmed by a postcontrast study. Only contrast-enhanced images will avoid false positive findings (caused for instance by a blood clot in the cervical canal or hematometra) and reveal the true extent of cervical involvement (endocervix or cervical stroma). MRI also identifies stage III and IV endometrial carcinoma with tumor invasion beyond the uterus (tumor confined to the true pelvis in stage III versus invasion of the mucosa of the urinary bladder or rectum in stage IV).

Spread to the bladder or rectal wall is indicated by circumscribed high-signal-intensity disruption of the normal low-signal-intensity muscle layer on T2w images. An irregular appearance of the uterine contour indicates extrauterine invasion if ascites is present (Fig. 12.47). Lymph node staging by MRI relies on the same size criterion as CT (short-axis diameter). Nevertheless, MRI is somewhat superior because the far better T-staging provides some indirect clues as to the differentiation of reactive versus metastatic nodal enlargement. The new iron oxide particle contrast media hold promise for distinguishing benign and malignant lymph nodes for the first time (see Chapter 16). Role of MRI in Staging Endometrial Carcinoma A multicenter study including 83 patients with endometrial carcinoma performed in 199163 found an overall accuracy of 85 % for general tumor staging, while the accuracy in determining the depth of myometrial invasion (stages IA–IC) was only 74 %. A similar accuracy of only 78 % was reported for the prognostically important differentiation of stage IA/B from stage IC tumors in a later study.55 Intravenous contrast administration (e. g., Magnevist, Dotarem) was shown to markedly improve the accuracy of MRI in assessing myometrial invasion.60,61,64,65 A more recent study of contrast-enhanced MRI reported a 91 % accuracy for the preoperative differentiation of stages IA/B from IC.66 MRI is now the most accurate imaging modality for staging endometrial cancer. Its advantages over endovaginal ultrasound are obvious for stage II or greater disease.33,67,68 While nonenhanced T2w MRI is comparable to transvaginal ultrasound in evaluating stage I endome-

MRI Appearance of Pathologic Entities

trial cancer, contrast-enhanced MRI is clearly superior.33,58,69,70 Although expensive, MRI has been shown to be cost-effective because it can help avoid unnecessary lymph node dissection.71,72

Table 12.7 Cervical carcinoma—recommended pulse sequences Sequence

Scan area/FOV (plane)

Comment

T2w TSE

Uterus and vagina (axial and sagittal); may be supplemented by true short-axis views of the cervix

T1w SE or PD TSE

Entire pelvis (axial)

Depth of cervical stroma invasion, parametrial invasion, tumor extension to vagina and uterine corpus Tumor invasion of pelvic sidewall, lymph node staging

Cervical Carcinoma The annual incidence of cervical carcinoma is 16 per 100 000 women, and there are two peaks in incidence— between 35 and 45 years and between 65 and 75 years. It is the most common malignant neoplasm in women under 50. Cervical cancer used to be the most common cancer in women of reproductive age but is becoming a tumor of elderly women. This shift is primarily due to the success of cervical cancer screening programs. If women under 35 are diagnosed with invasive cervical cancer, they typically have a very aggressive form. It is now generally assumed that cervical carcinoma develops in several steps—from dysplasia, to carcinoma in situ, to microinvasive carcinoma. Dysplasia and carcinoma in situ are now collectively referred to as cervical intraepithelial neoplasia. Up to 10 % of patients with microinvasive carcinoma (stage IA2: invasive component < 5 mm and horizontal spread < 7 mm) already have metastatic lymph nodes.55,73 As the tumor progresses, there is continuous local extension to the vaginal wall, parametria, and uterine corpus, and lymphatic metastatic spread. The order of lymphatic involvement is usually as follows: lymph nodes in the obturator fossa, along the internal iliac vessels, external iliac vessels, and finally the para-aortic lymph nodes. Distant metastases primarily involve the lungs and bones but are rare and occur late. The initial symptoms are vaginal bleeding outside the menstrual cycle and discharge. Screening programs improve detection of early stages as no other gynecologic carcinoma is more easily accessible to physical examination. Pregnancy does not exclude cervical cancer. MRI Appearance and Tumor Staging The role of MRI is to stage known cervical cancer. As with endometrial carcinoma, T2w images are most useful (Table 12.7), depicting cervical cancer with high signal intensity and clearly delineating it from the surrounding low-signal-intensity cervical stroma (Figs. 12.48, 12.49, 12.50). On T1w images, cervical cancer has the same signal intensity as the cervical stroma, myometrium, and vagina. Nevertheless, a T1w sequence is helpful for identifying tumor spread to the parametria and for evaluating the lymph nodes in advanced cervical cancer. While intravenous contrast administration improves staging of endometrial cancer, it is not helpful in staging cervical cancer.74,75 The latter shows highly variable enhancement, and the increase in signal intensity even reduces lesion conspicuity in the parametrian fat. The improved differentiation of viable and necrotic tumor areas on contrastenhanced images rarely has any clinical implications (Fig. 12.51). Cervical cancer has been reported to enhance

267

a

b Fig. 12.48a, b Cervical carcinoma, stage IB. T2w TSE images. a Sagittal image shows small cancer in the posterior cervix. The cervix and tumor protrude into the vagina. The adjacent vaginal wall is of normal low SI. b Axial image shows normal hypointense cervical stroma anteriorly and the cervical tumor at the 3–9 o’clock position.

268

12 The Uterus and Vagina

Fig. 12.49 Cervical carcinoma, stage IIA. Axial T2w TSE image. Most of the cervical stroma is replaced by moderately hyperintense tumor, and only little low-SI cervical stroma is left (compare Fig. 12.17b). No parametrial invasion.

early in dynamic contrast-enhanced GRE studies compared with the surrounding cervical stroma,76 but again, tumor conspicuity is not improved compared with nonenhanced T2w images. However, it has been proposed that tumors showing intense early enhancement will respond well to radiotherapy.21,77 Another possible indication for contrast-enhanced MRI is the assessment of advanced disease with suspected invasion of the urinary bladder or rectum. SE images obtained immediately after the contrast bolus occasionally improve the detection of tumor in the muscular wall layers of these organs compared with T2w images (Fig. 12.52). MRI staging of cervical cancer is based on the FIGO classification (Table 12.8). Preclinical cervical cancer (stage IA) can only be demonstrated microscopically and is not seen on MRI. Stage IB is cancer confined to the cervix (with or without invasion of the uterine corpus) and is indicated on T2w images by an intact low-signalintensity rim of normal cervical stroma surrounding the tumor (Figs. 12.53 and 12.54). Diagnostic confidence is improved if this criterion is evaluated in two planes. T1w images will show a normal interface between the cervix

b

c

a Fig. 12.50a–d Cervical carcinoma, stage IIB. T2w SE images. a Sagittal image. High-SI cervical tumor anteriorly and posteriorly and protrusion into the vagina. b–d Axial images through the cervix at different levels. The most proximal image shows a normal cervix (b), while the tumor has nearly completely replaced the cervical stroma at the level of the ectocervix (d). Parametrial invasion on the left with complete replacement of the cervical stroma (arrow).

d

MRI Appearance of Pathologic Entities

269

a

b Fig. 12.51a, b Cervical carcinoma, stage IIIA. Large tumor invading uterine corpus and lower third of vagina. a Nonenhanced T2w SE image. b Contrast-enhanced T1w SE image. Differentiation of viable and necrotic tumor portions is improved after contrast administration.

a

b Fig. 12.52a, b Cervical carcinoma, stage IV. Advanced cervical carcinoma extending to lower third of vagina and urinary bladder. a Nonenhanced T2w TSE image. The air inclusion (arrow) in the urinary bladder indicates a tumor-related fistula. b Contrastenhanced T1w SE image improves evaluation of the extent of tumor

necrosis and invasion of the urinary bladder wall (compare SI of the posterior and anterior wall). Fluid is retained in the endometrial cavity due to obstruction of the cervical canal by the tumor. Normal appearance of the anterior rectal wall without signs of infiltration.

270

12 The Uterus and Vagina

Table 12.8 Staging of cervical carcinoma FIGO* stage

Criteria

0 I

Carcinoma in situ No tumor identifiable Cervical carcinoma confined to uterus (extension to corpus should be disregarded) Preclinical invasive carcinoma, diagnosed by No tumor identifiable microscopy only Minimal microscopic stromal invasion (< 2 mm) Tumor with invasive component < 5 mm in depth and < 7 mm in horizontal spread Tumor larger than stage IA2 High SI of tumor partially disrupts the low-SI cervical stroma Tumor invades beyond uterus but without invasion of pelvic sidewall or lower third of vagina Without parametrial invasion Tumor invasion of upper two thirds of vagina; high SI of tumor disrupts the low-SI vaginal wall (tumor protrusion into the vagina should not be mistaken for vaginal wall invasion) With parametrial invasion Demonstration of tumor mass invading the parametrial tissue; SI of tumor extending through both inner and outer cervical stroma as an indirect sign of early parametrial invasion Tumor extends to pelvic sidewall and/or involves lower third of vagina and/or causes hydronephrosis Tumor invades lower third of vagina; no extension to High SI of tumor disrupts the wall of the lower third of the vagina pelvic sidewall Tumor invades pelvic sidewall and/or causes hydroHigh-SI tumor invasion of pelvic sidewall (most commonly of nephrosis obturator internus, piriform, or levator ani muscle); dilated ureter Tumor extends outside true pelvis or involves mucosa of bladder or rectum Tumor invades mucosa of bladder or rectum T2w images demonstrate direct tumor extension seen as disruption of the low-SI muscle of the bladder or rectal wall; postcontrast images demonstrate tumor invasion by abnormal enhancement Distant metastasis Distant metastasis

IA IA1 IA2 IB II IIA

IIB

III IIIA IIIB IV IVA

IVB

MR findings on T2w images

*International Federation of Gynecology and Obstetrics.

a

b Fig. 12.53a, b Cervical carcinoma, stage IB. a T2w SE image shows the low-SI lesion to be completely surrounded by an intact rim of cervical stroma. b T1w SE image does not enable differentiation of

the tumor, cervical stroma, and paracervical tissue. Normal appearance of the fatty tissue in the lateral portions of the parametria on both sides.

MRI Appearance of Pathologic Entities

271

a

b Fig. 12.54a, b Cervical carcinoma, stage IB. T2w TSE images. a Sagittal image. The small cervical carcinoma is seen as a high-SI lesion predominantly within the cervix (arrows). b Axial image. Both images show that the cancer is surrounded by intact low-SI cervical

stroma. The vaginal wall appears normal and shows no signs of tumor invasion. The patient also has uterine prolapse and fibroids in the anterior uterine wall.

a

b Fig. 12.55a, b Cervical carcinoma, stage IIB. T2w TSE images. Cervical carcinoma with complete infiltration of the anterior cervical stroma. a Sagittal image shows tumor invasion in the area of the anterior vaginal fornix, while the posterior vaginal wall is of normal

low SI. b Axial image shows complete tumorous replacement of the cervical stroma at the 9–12 o’clock position, while a thin intact stripe of normal cervical stroma is seen on the left.

and surrounding fatty tissue (Fig. 12.53b). Stage IIA cervical cancer involves the upper two thirds of the vagina (and is best detected clinically). Vaginal wall involvement is seen as a high-signal-intensity disruption of the normal low-signal-intensity muscle stripe on T2w images (Figs. 12.55a and 12.58). If the tumor only protrudes into the vagina, T2w images will show an intact low-signal-intensity wall of the vaginal fornices.

Stage IIB cervical cancer is defined as parametrial invasion. Extensive parametrian invasion may be seen as an irregular parametrial mass on T1w and T2w images. Early parametrial tumor invasion is not directly seen on MRI (Fig. 12.55b). The most important criterion of possible parametrial tumor extension is complete (hyperintense) replacement of the normal hypointense cervical stroma on T2w images78,79 (Figs. 12.55 and 12.56); however, this

272

12 The Uterus and Vagina

a

b Fig. 12.56a, b Cervical carcinoma, stage IIB. a T2w SE image shows complete invasion of the cervical stroma on the right. b Corresponding T1w SE image does not allow differentiation of the tumor from

the cervix. Image shows normal appearance of the fatty tissue of the lateral parametria.

a

b Fig. 12.57a, b Cervical carcinoma, stage IIIB. a Axial T2w TSE image. Bilateral parametrial infiltration; on the right side, the tumor extends to the ureter (arrow) and invades the uterosacral ligament

(curved arrow). b Sagittal T2w TSE image shows dilatation of the right ureter (as a sign of urinary obstruction caused by the tumor).

indirect criterion of parametrial invasion is less reliable (higher false-positive rate80) than an intact rim of cervical stroma around the tumor as an indicator of absent parametrian invasion. In stage IIIA cervical cancer, high-signal-intensity tumor extends to the lower third of the vagina (which has different lymphatic drainage and is the reason why the inguinal lymph nodes should be included in the MRI evaluation). Stage IIIB tumor invades the pelvic sidewall or causes obstruction of one or both ureters (Fig. 12.57).

From a surgeon’s perspective, complete obliteration of the lateral fat plane due to tumor already constitutes invasion of the pelvic sidewall. A direct sign of tumor spread to the pelvic wall muscle layer is high-signal-intensity invasion on T2w images. Invasion of the bladder and/or rectum establishes stage IVA disease and is seen as high-signal-intensity disruption of the low-signal-intensity muscle of the bladder or rectal wall on T2w images (Figs. 12.58 and 12.59). If the findings are indeterminate, contrast-enhanced images may be

MRI Appearance of Pathologic Entities

273

helpful for identifying wall invasion by improving conspicuity of the enhancing tumor (Fig. 12.52b) or of a tumorassociated fistula. The FIGO classification is based on the endoscopic criterion of tumor infiltration of the mucosa of the bladder or rectal wall, while MRI will already show invasion of the muscle layer. Size remains the only imaging criterion for identifying metastatic lymph nodes (Fig. 12.59b). Lymph node conspicuity is improved by high-resolution imaging using a phased-array body coil.81 Nonspecific extracellular paramagnetic contrast media do not improve the differentiation of malignant from benign nodes,82 but the situation may improve in the future with the use of superparamagnetic iron oxide particles. Role of MRI in Staging Cervical Carcinoma Although the pelvic examination reliably diagnoses cervical cancer, tumor staging by bimanual examination has a high error rate when compared with intraoperative staging, ranging from 20–30 % for stage IB cervical cancer to 50–60 % for stages IIA–IIIB.83,84 Overstaging is common. The results of preoperative staging are not much improved by CT because of the poor soft-tissue contrast of this modality (Table 12.9).46 MRI is now considered the most accurate preoperative staging modality for cervical cancer84–86 (Table 12.9). Preoperative assessment of cervical cancer by MRI has also been shown to be cost-effective because it can replace various additional diagnostic tests.87 Moreover, MRI has also become the method of choice for identifying patients who are candidates for surgery and those who should undergo radiotherapy. Finally, MRI has an important role in planning and following up radiotherapy because it offers unique evaluation of tumor extent in all three orthogonal planes and will show early therapeutic effects.21,77,88

Fig. 12.58 Cervical carcinom, stage IV. Sagittal T2w TSE image. Large cervical tumor invading the anterior vaginal fornix and muscle layer of the posterior bladder wall (arrow).

Table 12.9 Reported diagnostic accuracies of clinical examination, CT, and MRI for staging cervical carcinoma Method

Kim et al. 1990

Parametrial Clinical invasion examination CT MRI

78 %

Staging

70 %

Clinical examination CT MRI

70 % 92 %

63 % 83 %

Subak et al. 1995

Sironi et al. 1991

76 % 94 %

62 % 81 %

65 % 90 %

a

b Fig. 12.59a, b Cervical carcinoma, stage IV. a T2w TSE image shows advanced cervical carcinoma invading the posterior vaginal wall, rectouterine pouch, and rectum. b PD TSE image reveals suspiciously enlarged para-aortic lymph nodes (arrows).

274

12 The Uterus and Vagina

Vaginal Pathology The main indications for vaginal MRI are congenital anomalies and vaginal tumors. These entities are rare, and experience with vaginal MRI is still limited.89 MRI is the imaging modality with the highest diagnostic yield. The vagina is best imaged with axial and sagittal T2w sequences using a maximum slice thickness of 4 mm.90

Congenital Anomalies Vaginal anomalies include agenesis, duplication (Fig. 12.60), and septation. Vaginal septa are typically associated with other müllerian duct anomalies such as uterus didelphys. Fibrous septa at the level of the introitus

can occur in isolation and may cause hematocolpos in newborn girls (Fig. 12.61). Vaginal agenesis may be partial or complete (Fig. 12.24) and often occurs in association with unilateral renal agenesis or ectopia, which is why an MRI examination should always include the kidneys in these cases.91 Gartner duct cysts are developmental remnants of the wolffian duct. They are retention cysts and are typically found in the anterolateral vaginal wall (Fig. 12.62). The cysts are usually small and asymptomatic, but they can become large and compress the urethra. On MRI, Gartner duct cysts are diagnosed on the basis of the characteristic cystic appearance and their location. The cysts commonly contain proteinaceous fluid and therefore have high signal intensity on T1w MR images.

Bartholin Cysts Bartholin cysts are retention cysts arising as a result of obstruction of the greater vestibular glands, typically due to an inflammatory process or trauma. Most Bartholin cysts are asymptomatic incidental findings on MR images. Their most common location is posterolaterally in the lower vagina. The signal intensity of the cyst fluid is high on T2w images and moderate to high on T1w images, depending on the protein content.

Vaginal Tumors

Fig. 12.60 Double vagina (arrows) on axial T2w SE image.

Most vaginal tumors are malignant and these include primary vaginal carcinoma and metastases from other malignancies. Primary vaginal cancer is rare, accounting for < 2 % of all gynecologic malignant neoplasms. Vaginal

a

b Fig. 12.61a, b Hematocolpos in a 17-day-old girl. a T2w HASTE image. b T1w GRE image. Hematocolpos and hematometra (asterisk) with sedimentation. Arrow indicates urinary bladder.

MRI Appearance of Pathologic Entities

275

b

a Fig. 12.62a–c Gartner duct cysts. a, b Sagittal (a) and axial (b) T2w TSE images show two high-SI lesions in the vagina. c T1w TSE image with increased SI of the cyst fluid (arrow).

c

Table 12.10 Staging of vaginal carcinoma FIGO* stage

Criteria

MR findings on T2w and contrast-enhanced T1w images

0 I

Carcinoma in situ Tumor confined to vagina

II

Tumor invades paravaginal tissues but does not extend to pelvic sidewall Tumor extends to or invades pelvic sidewall

No tumor identifiable High-SI tumor without/with disruption of the low-SI vaginal wall and normal appearance of paravaginal tissues High SI of tumor extends to paravaginal tissues

III IVA

Tumor involves mucosa of bladder and/or rectum and/or extends outside true pelvis

IVB

Distant metastasis

High-SI tumor invasion of pelvic sidewall (most commonly of levator ani muscle) T2w images demonstrate tumor extension as disruption of the low-SI muscle of the bladder or rectal wall; postcontrast images demonstrate tumor invasion by abnormal enhancement Distant metastasis

*International Federation of Gynecology and Obstetrics

metastases are more common, most frequently from endometrial and cervical cancer, followed by colon carcinoma, renal cell carcinoma, and malignant melanoma. The MRI appearance does not allow differentiation of primary and secondary vaginal malignancies and both may on occasion be difficult to differentiate from a focal inflammatory process. Contrast-enhanced images are not helpful either because concomitant inflammation with intense enhancement is common in the setting of vaginal

malignancy. Difficulties may also arise in differentiating cervical cancer invading the vagina from primary cancer of the proximal vagina. If the bulk of the tumor is in the vagina, primary vaginal carcinoma is more likely. The MRI criteria are nonspecific, and a biopsy is needed for a definitive diagnosis. The main role of MRI is to stage known vaginal cancer (Fig. 12.63; see FIGO staging classification in Table 12.10). The MR pulse sequences recommended for assessing vag-

276

12 The Uterus and Vagina

Table 12.11 Vaginal carcinoma—recommended pulse sequences Sequence

Scan area/FOV (plane)

Comment

T2w TSE

Uterus to pelvic floor (axial and sagittal)

T1w SE or PD TSE

Aortic bifurcation to pelvic floor including inguinal lymph nodes (axial) Uterus to pelvic floor (axial)

Evaluation of craniocaudal tumor extent; depth of infiltration; infiltration of bladder, urethra, and rectum; infiltration of pelvic sidewall muscles (most commonly of levator ani muscle) Lymph node staging, fatty tissue infiltration

T1w SE after contrast

Tumor staging, same criteria as for T2w TSE; fistulas

inal carcinoma are summarized in Table 12.11. In patients with stage I disease, T2w images may already demonstrate higher-signal-intensity tumor within the low-signal-intensity vaginal wall. In stage II cancer, there is extension of low-signal-intensity tumor to paravaginal fatty tissue on T1w images; however, this is an unreliable criterion because there is only a thin fat plane around the vagina. On T2w images, tumor extension beyond the vagina may be seen as a mass within the high-signal-intensity venous network of the paravaginal tissues. T2w sequences in at least two planes are necessary for reliable identification of high-signal-intensity invasion of the bladder and rectal wall or of the muscle of the pelvic sidewall or pelvic floor (obturator, piriform, and levator ani muscles) in advanced stage III and IV disease. Small vesicovaginal fistulas are best identified by intense enhancement on postcontrast images.

MRI Appearance after Dilatation and Curettage Uterine changes seen on MRI after uncomplicated dilatation and curettage (D&C) are confined to the endometrium. T2w images show curvilinear areas of low signal intensity in the endometrial cavity, which represent blood clots. The junctional zone and myometrium were found to be of normal width and signal intensity 2–7 days after uncomplicated D&C.24

a

MRI Appearance after Radiotherapy

b Fig. 12.63a, b Vaginal carcinoma, stage IVA. a Sagittal T2w TSE image. Large tumor of the vagina with invasion of the urinary bladder base and anterior cervix. b Axial T2w TSE image obtained using an endorectal coil. Demonstration of tumor in the anterior vaginal wall and posterior bladder wall (arrow) with bilateral urinary obstruction (images courtesy of R. Kubik-Huch, Zurich).

Ultrasound and CT play only a minor role in diagnosing recurrent tumor and differentiating it from scars. The better soft-tissue contrast of MRI enables identification of radiation-induced changes of all pelvic organs and tissues. The MR signal changes (T2w images) are proportional to the total radiation dose applied and increase significantly above a threshold of 45 Gy.92 The observed changes occur invariably after percutaneous irradiation and brachytherapy.92 MRI demonstrates recurrent tumor and differentiates radiation fibrosis from residual tumor tissue. Uterus. Various myometrial and endometrial changes have been reported in women of reproductive age.23 The T2 signal intensity of the myometrium decreases, resulting in a loss of zonal anatomy (Fig. 12.64). The endome-

MRI Appearance of Pathologic Entities

277

trial stripe is thinner and has decreased signal intensity on T2w images. These changes are best seen 6 months or more after radiation therapy; during the first 6 months they may be obscured by edema. The uterus may atrophy as a direct result of radiation or because of indirect effects due to loss of hormonal stimulation resulting from radiation-induced ovarian dysfunction. Postmenopausal women show no radiation-induced uterine changes (unchanged low signal intensity and indistinct junctional zone on T2w images). Vagina. Acute and subacute radiation-induced edema and inflammation are reflected by increased vaginal signal intensity on T2w images and more intense enhancement after contrast administration. Six months or more after completion of radiotherapy, all sequences will show decreased vaginal signal intensity due to fibrosis (Fig. 12.65). In some patients, radiation may cause vaginal stenosis with serometra or hematometra (Fig. 12.66). Vaginal fistulas are best identified by enhancement of the inflamed fistulous wall on contrast-enhanced T1w images and, if present, low-signal-intensity fluid in the lumen.

a

b

c Fig. 12.64 Normal MR appearance of the uterus in a 40-year-old woman 18 months after radiotherapy. Sagittal T2w TSE image. Small uterus of low SI. A small amount of fluid is retained in the upper third of the vagina secondary to radiation-induced vaginal stenosis.

Fig. 12.65a–c Cervical carcinoma. Follow-up after radiotherapy. Sagittal T2w SE images. a High-SI cervical carcinoma before radiotherapy. b Normal appearance of the cervix 4 weeks after completion of radiotherapy. c Marked hypointensity of the cervical stroma 6 months after radiotherapy. Findings suggest cervical stenosis with fluid retention in the endometrial cavity.

278

12 The Uterus and Vagina

a

b Fig. 12.66a, b Vaginal stenosis causing hematometra and hematosalpinx. Patient underwent radiotherapy 1 year earlier. Hemorrhagic fluid retention in the endometrial cavity and upper third of the

Residual Tumor and Recurrence after Radiotherapy The diagnosis of residual or recurrent tumor after radiotherapy is primarily relevant for cervical cancer. MRI will yield reliable results only if the examination is performed at least 6 months after the end of radiotherapy.93 In the early phase after irradiation, residual tumor cannot be differentiated from reactive edema or inflammation on the basis of MR signal intensities (Fig. 12.67). Even fibrosis has higher signal intensity during the first 6 months following radiotherapy.94 Six months after treatment, the cervical stroma and myometrium have low signal intensity, and a recurrent tumor is best seen as a high-signalintensity mass on T2w images (Figs. 12.68, 12.69, 12.70). The MRI criteria are the same as in the initial evaluation.

Tumor Recurrence after Surgery Postoperative tumor recurrence and fibrosis have different MR signal intensities, making MRI an ideal modality for differentiating these two entities.78 Ultrasound and CT cannot make the distinction because they rely on lesion configuration as the only criterion. The extensive postoperative changes also limit the value of the clinical examination, since suspicious palpation findings may indicate a recurrent tumor or postoperative scar formation.

vagina. a Sagittal T2w SE image. b Axial T1w SE image. Bilateral hematosalpinx (arrows).

A preferred site of recurrent cervical carcinoma is the vaginal stump (Fig. 12.71). Recurrent tumor tends to be inhomogeneous and typically has increased T2 signal intensity,95 whereas postoperative fibrosis is of low T2 signal intensity. However, a scar may also display increased signal intensity due to inflammation or neovascularization in the first weeks or months after surgery, which is why it is again important to wait at least 6 months before performing an MRI examination. Even after this interval, there may be overlap of the signal intensities of recurrent tumor and fibrosis on T2w images. Inconclusive findings may be resolved by a dynamic contrast-enhanced study with a fast GRE sequence (Fig. 12.72), which will identify a recurrent tumor by its significantly earlier and more intense enhancement compared with scar tissue.96 Contrast-enhanced images will also reveal invasion of the bladder or rectal wall, seen as circumscribed enhancement (Fig. 12.72). Pelvic sidewall invasion by recurrent tumor is diagnosed using the same criteria as for the primary tumor and is again best seen on T2w images. Seventy percent of all cases of recurrent endometrial cancer are observed during the first 3 years after treatment. Recurrent endometrial cancer tends to involve the pelvic sidewall, the parametria, and the vaginal stump. The MRI criteria for identifying recurrent endometrial cancer are the same as for cervical cancer.

MRI Appearance of Pathologic Entities

279

Fig. 12.67a–c Cervical carcinoma. Follow-up after radiotherapy. Sagittal T2w SE images. a Barrel-shaped cervical carcinoma before radiotherapy. Fluid retention in the endometrial cavity due to obstruction of the cervical canal by the tumor. Multiple uterine fibroids of low SI. b Image obtained immediately upon completion of radiotherapy. Normal cervical structures can be identified, while cervical SI is still abnormally increased (differential diagnosis: residual tumor, radiation-induced edema). c Image obtained 6 months after completion of radiotherapy. Normal appearance of the cervix without signs of residual tumor. Unchanged demonstration of multiple fibroids.

a

b

c

280

12 The Uterus and Vagina

Fig. 12.68 Recurrent endometrial carcinoma after radiotherapy. Sagittal T2w SE image. High-SI mass within the low-SI myometrium.

Fig. 12.69 Recurrent cervical carcinoma after radiotherapy. Sagittal T2w SE image shows a recurrent tumor of high SI (arrow) above a low-SI scar. Radiation-induced thickening of the posterior bladder wall (image courtesy of M. Lorenzen, Hamburg).

Fig. 12.70 Recurrent cervical carcinoma after radiotherapy. Axial T2w TSE image. Suspicious hyperintensity of the cervical stroma at the 12–5 o’clock position. The vaginal wall is depicted with normal low SI on both sides (image courtesy of M. Lorenzen, Hamburg).

Fig. 12.71 Normal postoperative appearance after hysterectomy for cervical carcinoma. Sagittal T2w TSE image. The vaginal stump and interface with the peritoneum appear normal.

MRI Appearance of Pathologic Entities

281

b

a

c

d Fig. 12.72a–d Recurrent cervical carcinoma after hysterectomy. a, b Sagittal (a) and axial (b) T2w TSE images. Circumscribed lowSI mass at the proximal end of the vaginal stump with extension to the right lateral urinary bladder wall. The morphologic appearance suggests a recurrent tumor, while the low SI of the lesion is more in

keeping with a scar. c, d Axial T1w SE image before (c) and T1w GRE image 1 min after (d) IV bolus injection of contrast medium. The mass is contiguous with a sigmoid loop and extends to the right pelvic sidewall. The intense enhancement of the mass is consistent with a recurrent tumor (which was histologically confirmed).

References

7. McCarthy S, Scott G, Majumdar S, et al. Uterine junctional zone: MR study of water content and relaxation properties. Radiology 1989;171(1):241–243 8. Scoutt LM, Flynn SD, Luthringer DJ, McCauley TR, McCarthy SM. Junctional zone of the uterus: correlation of MR imaging and histologic examination of hysterectomy specimens. Radiology 1991;179(2):403–407 9. Brown HK, Stoll BS, Nicosia SV, et al. Uterine junctional zone: correlation between histologic findings and MR imaging. Radiology 1991;179(2):409–413 10. Mitchell DG, Schonholz L, Hilpert PL, Pennell RG, Blum L, Rifkin MD. Zones of the uterus: discrepancy between US and MR images. Radiology 1990;174(3 Pt 1):827–831 11. deSouza NM, Hawley IC, Schwieso JE, Gilderdale DJ, Soutter WP. The uterine cervix on in vitro and in vivo MR images: a study of zonal anatomy and vascularity using an enveloping cervical coil. AJR Am J Roentgenol 1994;163(3):607–612 12. Li H, Sugimura K, Okizuka H, et al. Markedly high signal intensity lesions in the uterine cervix on T2-weighted imaging: differentiation between mucin-producing carcinomas and nabothian cysts. Radiat Med 1999;17(2):137–143

1. De Graef M, Karam R, Juhan V, Daclin PY, Maubon AJ, Rouanet JP. High signals in the uterine cervix on T2-weighted MRI sequences. Eur Radiol 2003;13(1):118–126 2. Haynor DR, Mack LA, Soules MR, Shuman WP, Montana MA, Moss AA. Changing appearance of the normal uterus during the menstrual cycle: MR studies. Radiology 1986;161(2):459–462 3. Hötzinger H, Spätling L. MRI in der Gynäkologie und Geburtshilfe. Berlin: Springer; 1994 4. Lee JKT, Gersell DJ, Balfe DM, Worthington JL, Picus D, Gapp G. The uterus: in vitro MR-anatomic correlation of normal and abnormal specimens. Radiology 1985;157(1):175–179 5. Masui T, Katayama M, Kobayashi S, et al. Changes in myometrial and junctional zone thickness and signal intensity: demonstration with kinematic T2-weighted MR imaging. Radiology 2001;221(1):75–85 6. Togashi K, Nakai A, Sugimura K. Anatomy and physiology of the female pelvis: MR imaging revisited. J Magn Reson Imaging 2001;13(6):842–849

282

12 The Uterus and Vagina

13. Hricak H, Chang YCF, Thurnher S. Vagina: evaluation with MR imaging. Part I. Normal anatomy and congenital anomalies. Radiology 1988;169(1):169–174 14. Golan A, Langer R, Bukovsky I, Caspi E. Congenital anomalies of the müllerian system. Fertil Steril 1989;51(5):747–755 15. Stampe Sørensen S. Estimated prevalence of müllerian anomalies. Acta Obstet Gynecol Scand 1988;67(5):441–445 16. Ansbacher R. Uterine anomalies and future pregnancies. Clin Perinatol 1983;10(2):295–304 17. Harger JH, Archer DF, Marchese SG, Muracca-Clemens M, Garver KL. Etiology of recurrent pregnancy losses and outcome of subsequent pregnancies. Obstet Gynecol 1983;62(5):574–581 18. Console D, Tamburrini S, Barresi D, Notarangelo L, Bertucci B, Tamburrini O. [The value of the MR imaging in the evaluation of Müllerian duct anomalies]. Radiol Med (Torino) 2001;102(4): 226–232 19. Saleem SN. MR imaging diagnosis of uterovaginal anomalies: current state of the art. Radiographics 2003;23(5):e13 20. Lang IM, Babyn P, Oliver GD. MR imaging of paediatric uterovaginal anomalies. Pediatr Radiol 1999;29(3):163–170 21. Loncaster JA, Carrington BM, Sykes JR, et al. Prediction of radiotherapy outcome using dynamic contrast enhanced MRI of carcinoma of the cervix. Int J Radiat Oncol Biol Phys 2002;54(3): 759–767 22. Marten K, Vosshenrich R, Funke M, Obenauer S, Baum F, Grabbe E. MRI in the evaluation of müllerian duct anomalies. Clin Imaging 2003;27(5):346–350 23. Arrivé L, Chang YCF, Hricak H, Brescia RJ, Auffermann W, Quivey JM. Radiation-induced uterine changes: MR imaging. Radiology 1989;170(1 Pt 1):55–58 24. Ascher SM, Scoutt LM, McCarthy SM, Lange RC, DeCherney AH. Uterine changes after dilation and curettage: MR imaging findings. Radiology 1991;180(2):433–435 25. Ballesio L, Andreoli C, De Cicco ML, et al. Hematocolpos in double vagina associated with uterus didelphus: US and MR findings. Eur J Radiol 2003;45:150–153 26. Wentz AC. Congenital anomalies and intersexuality. In: Jones HW, Wentz AC, Burnett LP. NovakTextbook of Gynecology. 11th ed. Baltimore: Williams & Wilkins; 1988, pp.140–186 27. Reinhold C, Hricak H, Forstner R, et al. Primary amenorrhea: evaluation with MR imaging. Radiology 1997;203(2):383–390 28. Carrington BM, Hricak H, Nuruddin RN, Secaf E, Laros RKJr, Hill EC. Müllerian duct anomalies: MR imaging evaluation. Radiology 1990;176(3):715–720 29. Pellerito JS, McCarthy SM, Doyle MB, Glickman MG, DeCherney AH. Diagnosis of uterine anomalies: relative accuracy of MR imaging, endovaginal sonography, and hysterosalpingography. Radiology 1992;183(3):795–800 30. Katsumori T, Nakajima K, Tokuhiro M. Gadolinium-enhanced MR imaging in the evaluation of uterine fibroids treated with uterine artery embolization. AJR Am J Roentgenol 2001;177(2): 303–307 31. Kröncke TJ, Gauruder-Burmester A, Hamm B. [Uterine fibroid embolization - a new therapeutic option for symptomatic leiomyomata of the uterus]. Rofo 2002;174(10):1227–1235 32. Bazot M, Salem C, Frey I, Daraï E. [Imaging of myomas: is preoperative MRI usefull?]. Gynecol Obstet Fertil 2002;30(9): 711–716 33. Dueholm M, Lundorf E, Hansen ES, Ledertoug S, Olesen F. Accuracy of magnetic resonance imaging and transvaginal ultrasonography in the diagnosis, mapping, and measurement of uterine myomas. Am J Obstet Gynecol 2002;186(3):409–415 34. Hricak H, Tscholakoff D, Heinrichs L, et al. Uterine leiomyomas: correlation of MR, histopathologic findings, and symptoms. Radiology 1986;158(2):385–391 35. Schnall MD. Magnetic resonance evaluation of acquired benign uterine disorders. Semin Ultrasound CT MR 1994;15(1):18–26 36. Mittl RLJr, Yeh IT, Kressel HY. High-signal-intensity rim surrounding uterine leiomyomas on MR images: pathologic correlation. Radiology 1991;180(1):81–83 37. Murase E, Siegelman ES, Outwater EK, Perez-Jaffe LA, Tureck RW. Uterine leiomyomas: histopathologic features, MR imaging findings, differential diagnosis, and treatment. Radiographics 1999;19(5):1179–1197 38. Kawakami S, Togashi K, Konishi I, et al. Red degeneration of uterine leiomyoma: MR appearance. J Comput Assist Tomogr 1994;18(6):925–928

39. Hricak H, Finck S, Honda G, Göranson H. MR imaging in the evaluation of benign uterine masses: value of gadopentetate dimeglumine-enhanced T1-weighted images. AJR Am J Roentgenol 1992;158(5):1043–1050 40. Goto A, Takeuchi S, Sugimura K, Maruo T. Usefulness of Gd-DTPA contrast-enhanced dynamic MRI and serum determination of LDH and its isozymes in the differential diagnosis of leiomyosarcoma from degenerated leiomyoma of the uterus. Int J Gynecol Cancer 2002;12(4):354–361 41. Parrott E, Butterworth M, Green A, White IN, Greaves P. Adenomyosis—a result of disordered stromal differentiation. Am J Pathol 2001;159(2):623–630 42. Reinhold C, Atri M, Mehio A, Zakarian R, Aldis AE, Bret PM. Diffuse uterine adenomyosis: morphologic criteria and diagnostic accuracy of endovaginal sonography. Radiology 1995; 197(3):609–614 43. Bazot M, Cortez A, Darai E, et al. Ultrasonography compared with magnetic resonance imaging for the diagnosis of adenomyosis: correlation with histopathology. Hum Reprod 2001;16(11):2427–2433 44. Dueholm M, Lundorf E, Hansen ES, Sørensen JS, Ledertoug S, Olesen F. Magnetic resonance imaging and transvaginal ultrasonography for the diagnosis of adenomyosis. Fertil Steril 2001;76(3):588–594 45. Kinkel K, Vincent B, Balleyguier C, Hélénon O, Moreau J. [Value of MR imaging in the diagnosis of benign uterine conditions]. J Radiol 2000;81(7):773–779 46. Boss EA, Barentsz JO, Massuger LF, Boonstra H. The role of MR imaging in invasive cervical carcinoma. Eur Radiol 2000;10(2): 256–270 47. Togashi K, Nishimura K, Itoh K, et al. Adenomyosis: diagnosis with MR imaging. Radiology 1988;166(1 Pt 1):111–114 48. Togashi K, Ozasa H, Konishi I, et al. Enlarged uterus: differentiation between adenomyosis and leiomyoma with MR imaging. Radiology 1989;171(2):531–534 49. Byun JY, Kim SE, Choi BG, Ko GY, Jung SE, Choi KH. Diffuse and focal adenomyosis: MR imaging findings. Radiographics 1999; 19(Spec No):S161–S170 50. Kang S, Turner DA, Foster GS, Rapoport MI, Spencer SA, Wang JZ. Adenomyosis: specificity of 5 mm as the maximum normal uterine junctional zone thickness in MR images. AJR Am J Roentgenol 1996;166(5):1145–1150 51. Reinhold C, Tafazoli F, Mehio A, et al. Uterine adenomyosis: endovaginal US and MR imaging features with histopathologic correlation. Radiographics 1999;19(Spec No):S147–S160 52. Jha RC, Takahama J, Imaoka I, et al. Adenomyosis: MRI of the uterus treated with uterine artery embolization. AJR Am J Roentgenol 2003;181(3):851–856 53. Siskin GP, Tublin ME, Stainken BF, Dowling K, Dolen EG. Uterine artery embolization for the treatment of adenomyosis: clinical response and evaluation with MR imaging. AJR Am J Roentgenol 2001;177(2):297–302 54. Hricak H, Demas BE, Braga CA, Fisher MR, Winkler ML. Gestational trophoblastic neoplasm of the uterus: MR assessment. Radiology 1986;161(1):11–16 55. Martius G, Breckwoldt M, Pfleiderer A. Lehrbuch der Gynäkologie und Geburtshilfe. Stuttgart: Thieme: 1996 56. Grasel RP, Outwater EK, Siegelman ES, Capuzzi D, Parker L, Hussain SM. Endometrial polyps: MR imaging features and distinction from endometrial carcinoma. Radiology 2000;214(1): 47–52 57. Chen SS, Lee L. Retroperitoneal lymph node metastases in Stage I carcinoma of the endometrium: correlation with risk factors. Gynecol Oncol 1983;16(3):319–325 58. Yamashita Y, Mizutani H, Torashima M, et al. Assessment of myometrial invasion by endometrial carcinoma: transvaginal sonography vs contrast-enhanced MR imaging. AJR Am J Roentgenol 1993;161(3):595–599 59. Scoutt LM, McCarthy SM, Flynn SD, et al. Clinical stage I endometrial carcinoma: pitfalls in preoperative assessment with MR imaging. Work in progress. Radiology 1995;194(2):567–572 60. Hricak H, Hamm B, Semelka RC, et al. Carcinoma of the uterus: use of gadopentetate dimeglumine in MR imaging. Radiology 1991;181(1):95–106 61. Yamashita Y, Harada M, Sawada T, Takahashi M, Miyazaki K, Okamura H. Normal uterus and FIGO stage I endometrial carci-

MRI Appearance of Pathologic Entities

62. 63. 64. 65.

66.

67. 68. 69.

70. 71. 72.

73. 74. 75. 76.

77.

78. 79. 80.

noma: dynamic gadolinium-enhanced MR imaging. Radiology 1993;186(2):495–501 Frei KA, Kinkel K. Staging endometrial cancer: role of magnetic resonance imaging. J Magn Reson Imaging 2001;13(6):850–855 Hricak H, Rubinstein LV, Gherman GM, Karstaedt N. MR imaging evaluation of endometrial carcinoma: results of an NCI cooperative study. Radiology 1991;179(3):829–832 Ito K, Matsumoto T, Nakada T, Nakanishi T, Fujita N, Yamashita H. Assessing myometrial invasion by endometrial carcinoma with dynamic MRI. J Comput Assist Tomogr 1994;18(1):77–86 Savci G, Ozyaman T, Tutar M, Bilgin T, Erol O, Tuncel E. Assessment of depth of myometrial invasion by endometrial carcinoma: comparison of T2-weighted SE and contrast-enhanced dynamic GRE MR imaging. Eur Radiol 1998;8(2):218–223 Cunha TM, Félix A, Cabral I. Preoperative assessment of deep myometrial and cervical invasion in endometrial carcinoma: comparison of magnetic resonance imaging and gross visual inspection. Int J Gynecol Cancer 2001;11(2):130–136 Smith RC, McCarthy S. Magnetic resonance staging of neoplasms of the uterus. Radiol Clin North Am 1994;32(1):109–131 Taïeb S, Ceugnart L, Leblanc E, Chevalier A, Cabaret V, Querleu D. [MR imaging of endometrial carcinoma: role and limits]. Bull Cancer 2002;89(11):963–968 Kim SH, Kim HD, Song YS, Kang SB, Lee HP. Detection of deep myometrial invasion in endometrial carcinoma: comparison of transvaginal ultrasound, CT, and MRI. J Comput Assist Tomogr 1995;19(5):766–772 Kinkel K, Kaji Y, Yu KK, et al. Radiologic staging in patients with endometrial cancer: a meta-analysis. Radiology 1999;212(3): 711–718 Hardesty LA, Sumkin JH, Nath ME, et al. Use of preoperative MR imaging in the management of endometrial carcinoma: cost analysis. Radiology 2000;215(1):45–49 Todo Y, Sakuragi N, Nishida R, et al. Combined use of magnetic resonance imaging, CA 125 assay, histologic type, and histologic grade in the prediction of lymph node metastasis in endometrial carcinoma. Am J Obstet Gynecol 2003;188(5):1265–1272 Fleischer AC, Javitt MC, Jeffrey RB, Jones HW. Clinical Gynecologic Imaging. Philadelphia: Lippincott-Raven; 1997 Hawighorst H, Knapstein PG, Weikel W, et al. Cervical carcinoma: comparison of standard and pharmacokinetic MR imaging. Radiology 1996;201(2):531–539 Sironi S, De Cobelli F, Scarfone G, et al. Carcinoma of the cervix: value of plain and gadolinium-enhanced MR imaging in assessing degree of invasiveness. Radiology 1993;188(3):797–801 Yamashita Y, Baba T, Baba Y, et al. Dynamic contrast-enhanced MR imaging of uterine cervical cancer: pharmacokinetic analysis with histopathologic correlation and its importance in predicting the outcome of radiation therapy. Radiology 2000; 216(3):803–809 Gong QY, Brunt JN, Romaniuk CS, et al. Contrast enhanced dynamic MRI of cervical carcinoma during radiotherapy: early prediction of tumour regression rate. Br J Radiol 1999; 72(864):1177–1184 Lorenzen M, Nicolas V, Kopp A. [MRT diagnosis of recurrence of gynecologic tumors]. Rofo 1994;161(6):526–530 Sironi S, Belloni C, Taccagni GL, DelMaschio A. Carcinoma of the cervix: value of MR imaging in detecting parametrial involvement. AJR Am J Roentgenol 1991;156(4):753–756 Lien HH, Blomlie V, Iversen T, Tropé C, Sundfør K, Abeler VM. Clinical stage I carcinoma of the cervix. Value of MR imaging in determining invasion into the parametrium. Acta Radiol 1993; 34(2):130–132

283

81. Kim SH, Kim SC, Choi BI, Han MC. Uterine cervical carcinoma: evaluation of pelvic lymph node metastasis with MR imaging. Radiology 1994;190(3):807–811 82. Delgado G, Bundy B, Zaino R, Sevin BU, Creasman WT, Major F. Prospective surgical-pathological study of disease-free interval in patients with stage IB squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oncol 1990; 38(3):352–357 83. Lagasse LD, Creasman WT, Shingleton HM, Ford JH, Blessing JA. Results and complications of operative staging in cervical cancer: experience of the Gynecologic Oncology Group. Gynecol Oncol 1980;9(1):90–98 84. Sheu MH, Chang CY, Wang JH, Yen MS. Preoperative staging of cervical carcinoma with MR imaging: a reappraisal of diagnostic accuracy and pitfalls. Eur Radiol 2001;11(9):1828–1833 85. Sironi S, Zanello A, Rodighiero MG, et al. [Invasive carcinoma of the cervix uteri (Stage IB-IIB). Comparison of CT and MR for the assessment of the parametrium]. Radiol Med (Torino) 1991;81(5):671–677 86. Subak LL, Hricak H, Powell CB, Azizi L, Stern JL. Cervical carcinoma: computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 1995;86(1):43–50 87. Hricak H, Powell CB, Yu KK, et al. Invasive cervical carcinoma: role of MR imaging in pretreatment work-up—cost minimization and diagnostic efficacy analysis. Radiology 1996;198(2): 403–409 88. Ohara K, Tanaka Y, Tsunoda H, Nishida M, Sugahara S, Itai Y. Assessment of cervical cancer radioresponse by serum squamous cell carcinoma antigen and magnetic resonance imaging. Obstet Gynecol 2002;100(4):781–787 89. Rouanet JP, De Graef M, Teissier JM, Daclin PY, Kassem Z, Maubon A. [Imaging of the cervix and the vagina]. J Radiol 2001;82(12 Pt 2):1845–1853 90. McCarthy S, Hricak H. The uterus and vagina. In: Higgins CB, Hricak H, Helms CA. Magnetic Resonance Imaging of the Body. Philadelphia: Lippincott-Raven; 1997, pp. 761–814 91. Griffin JE, Edwards C, Madden JD, Harrod MJ, Wilson JD. Congenital absence of the vagina. The Mayer-Rokitansky-KusterHauser syndrome. Ann Intern Med 1976;85(2):224–236 92. Sugimura K, Carrington BM, Quivey JM, Hricak H. Postirradiation changes in the pelvis: assessment with MR imaging. Radiology 1990;175(3):805–813 93. Hricak H, Swift PS, Campos Z, Quivey JM, Gildengorin V, Göranson H. Irradiation of the cervix uteri: value of unenhanced and contrast-enhanced MR imaging. Radiology 1993;189(2): 381–388 94. Flückiger F, Ebner F, Poschauko H, Arian-Schad K, Einspieler R, Hausegger K. [Value of magnetic resonance tomography after primary irradiation of carcinoma of the cervix uteri: evaluation of therapeutic success and follow-up]. Strahlenther Onkol 1991;167(3):152–157 95. Jeong YY, Kang HK, Chung TW, Seo JJ, Park JG. Uterine cervical carcinoma after therapy: CT and MR imaging findings. Radiographics 2003;23(4):969–981, discussion 981 96. Hawighorst H, Knapstein PG, Schaeffer U, et al. Pelvic lesions in patients with treated cervical carcinoma: efficacy of pharmacokinetic analysis of dynamic MR images in distinguishing recurrent tumors from benign conditions. AJR Am J Roentgenol 1996;166(2):401–408

285

13

The Adnexa M. Reuter and M. Lorenzen

Introduction

Imaging Technique

Ovarian cancer accounts for 25 % of all malignancies of the female genital tract. In terms of the absolute number of deaths, it ranks second among gynecologic cancers following breast cancer.1 Early ovarian cancer is often asymptomatic, and most women have advanced disease at presentation.1 A reliable diagnostic test is needed, especially to detect ovarian cancer at an early stage. Transvaginal ultrasound (TVUS) is now the primary imaging modality;2,3 MRI is typically used as a supplementary tool for further characterization of a known ovarian mass.

A woman undergoing pelvic MRI is asked about her hormonal status (day of menstrual cycle) as the ovaries and functional cysts vary in size in relation to the menstrual cycle. Information about prior gynecologic operations is also important for image interpretation. The adnexa can be imaged at intermediate field strength; however, a 1.5-T scanner is preferred if available. Bladder emptying before scanning will reduce the likelihood of motion artifact. The patient is placed in a comfortable supine position, with a foam pad to elevate the knees if desired. Unless contraindicated, an antispasmodic agent is injected intravenously (butylscopolamine or glucagon) to reduce peristalsis. A high-resolution phased-array surface coil (with at least four elements) is used for read-out.8 The coil is applied with slight pressure above the pubic symphysis to restrict abdominal wall and respiratory motion. Respiratory artifacts are further reduced by asking the patient to breathe quietly during acquisitions.

Indications The vast majority of ovarian tumors operated on because of suspected malignancy turn out to be benign cysts. Gynecologists routinely use TVUS to supplement the pelvic examination. Because of the direct access, TVUS can be performed at high frequency and yields high-resolution images.3 MRI is a useful secondary imaging tool in women with complex or solid ovarian masses and indeterminate ultrasound findings.4,5 Tissue characterization based on the analysis of MR signal intensities on different pulse sequences enables reliable differentiation of serous, mucinous, hemorrhagic, fatty, and other solid pelvic masses. Moreover, the known high soft-tissue contrast and multiplanar capability of MRI allow better definition of the spatial relationship of a pelvic mass to other structures in the narrow confines of the true pelvis and to the pelvic sidewall. By thus improving preoperative characterization of ovarian masses, MRI can help reduce overtreatment. Ovarian tumors that are very likely benign based on their MRI appearance can be managed by pelvic laparoscopy without compromising therapeutic outcome, reserving laparotomy for women with suspicious ovarian masses.6 Although expensive, MRI can therefore help contain health-care expenditure when its use is carefully limited to those women in whom reliable pretherapeutic characterization of ovarian tumors can obviate the need for radical surgery.7

Imaging Planes The standard adnexal protocol comprises axial and coronal sequences. Axial images allow reliable evaluation of the relationship of a pelvic mass to the iliac vessels, the pelvic sidewalls, the bladder, and the rectum. Coronal images facilitate identification of the site of origin of a mass in the ovarian fossa. This standard protocol is occasionally supplemented by sagittal images to determine the relationship of a mass to the uterus, rectum, and rectouterine pouch.

Pulse Sequences Both T1w and T2w sequences are needed to characterize an ovarian mass. The MRI protocol (Tables 13.1 and 13.2) should ideally begin with an axial fat-suppressed inversion recovery (IR) sequence (which requires an inversion delay of ca. 150 ms at 1.5 T). Although spatial resolution is limited, this pulse sequence will sensitively demonstrate cystic components, thereby providing important information for tissue characterization in conjunction with the

* TI = 150 ms Note: Alternatively, T1w images can be acquired using a TSE sequence and adjusting parameters accordingly. The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time but may come with a penalty in SNR.

No No No No No 4.33 5.25 2.5 3.5 3.5 6 6 6 6 6 19 19 19 13 13 7 11 – – – Axial Coronal Axial Axial Coronal T2 T2 T1 T1 + Gd T1 + Gd

IR (e. g., TIRM*) TSE SE SE SE

5035 3882 456 637 637

15 99 14 20 20

180 180 90 90 90

Yes No No Yes Yes

210 × 256 220 × 256 205 × 256 205 × 256 205 × 256

260 (100 %) 250 (100 %) 260 (100 %) 260 (100 %) 280 (100 %)

1 2 1 1 1

Breathhold Slice thickness (mm) No. of acquisitions No. of slices FOV (mm) Matrix FS ETL Flip (°) TE (ms) TR (ms) Sequence type Plane Weighting

Table 13.2 Recommended pulse sequences and imaging parameters for MRI of the adnexa

13 The Adnexa

Scan time (min)

286

Table 13.1 Standard pulse sequences for MRI of the adnexa Sequence

Plane

Comment

T2w IR

Axial

T2w TSE

Coronal

T1w SE or TSE T1w SE or TSE, fat-suppressed after Gd-based contrast T1w SE or TSE, fat-suppressed after Gd-based contrast

Axial Axial

Sensitive tumor detection, tissue characterization, distance from pelvic sidewall, involvement of bowel and posterior bladder wall, presence of peritoneal implants, lymph node metastasis, ascites Tumor architecture, tissue characterization, delineation from uterus, vagina, vessels; presence of ascites Tissue characterization Tumor architecture, tissue characterization

Coronal

Tumor architecture, tissue characterization

nonenhanced T1w SE sequence to be performed next. Coronal T2w TSE images provide excellent resolution and high contrast for assessing the architecture of an adnexal mass. Contrast-enhanced series are acquired in two planes (axial and coronal) using a fat-suppressed T1w SE or TSE sequence. GRE sequences are inadequate for characterizing an adnexal mass due to poor contrast. A dynamic contrast-enhanced study does not add diagnostic information since most adnexal tumors are cystic or mixed cystic and solid. Single-shot TSE or HASTE sequences afford high contrast and eliminate artifacts but their spatial resolution is markedly lower than that of a T2w TSE sequence. Their use should be limited to those rare cases where T2w TSE images are suboptimal.9

Contrast Media Negative intestinal contrast medium is helpful in reliably differentiating cystic tumor components from fluid-containing bowel loops.6 To effectively suppress signal from the bowel lumen in the true pelvis, patients should start drinking the contrast solution at least 2 h before MRI is performed. An effective negative contrast solution can be prepared by dissolving one tablet of an iron gluconate formulation approved for treating iron deficiency in 1 L of water, of which the patient is asked to drink about one quarter every 30 min starting 2 h before the examination. Features relevant for differentiating benign from malignant ovarian masses such as subtle wall irregularities, septa, and nodules often only become apparent on images acquired after administration of nonspecific Gd-based contrast medium. This is why the adnexal protocol comprises contrast-enhanced T1w SE or TSE sequences in two planes perpendicular to each other, unless the precontrast images reliably identify a hemorrhagic cyst (endometriotic or chocolate cyst) or teratoma.

MRI Appearance of Normal Anatomy

MRI Appearance of Normal Anatomy Basic Anatomy The uterine adnexa consist of the ovaries, uterine tubes, and broad ligament. The broad ligament of the uterus is a double layer of peritoneum extending from the sides of the uterus to the lateral walls and floor of the pelvis. Its largest part forms the mesentery of the uterus (mesometrium). The superior portion of the broad ligament encasing the uterine tube on either side is the mesosalpinx. Between the layers of the broad ligament and attached to the uterus just below the uterotubal junction lie the ligament of the ovary posterosuperiorly and the round ligament of the uterus anteroinferiorly. The ovarian ligament attaches the ovary to the uterus, and the round ligament extends into the inguinal canal. The connective tissue of the broad ligament conveys the ovarian artery and vein, nerves, and lymphatic vessels. A uterine tube (fallopian tube or salpinx) is 8–20 cm long and is supported by the mesosalpinx. It is divisible into four parts: the infundibulum—the funnel-shaped distal end that is not directly attached to the ovary but opens into the peritoneal cavity through the abdominal ostium; the ampulla—the widest and longest part; the isthmus— the thick-walled part that enters the uterine horn; and the uterine part—the short intramural segment that passes through the wall of the uterus. The abdominal ostium is surrounded by fringed processes, the fimbriae, the longest of which is attached to the ovary. The ovaries are almond-shaped organs, each ca. 2.5– 5 cm long and 0.5–1 cm thick. They have a connective tissue capsule that adheres firmly to the outer layer of modified epithelium of the peritoneal fold by which it is enveloped. The capsule encloses the ovarian cortex, the site of follicular maturation. The inner region, the ovarian medulla, has an irregular interface with the cortex and contains highly tortuous blood vessels, lymphatics, and nerves, which enter the ovary at the hilum. During follicular maturation, primordial germ cells migrate into the ovary and develop into oogonia, which become surrounded by follicular epithelium and continue to develop into secondary follicles. Some of the secondary follicles become graafian follicles. The cells surrounding the follicle form the matrix for the follicular theca, which produces estrogen after ovulation. In the first year of life, the ovaries and uterine tubes migrate from the thoracolumbar junction into the ovarian fossae, depressions in the pelvic sidewall bounded by the obliterated umbilical arteries anteriorly and the ureters posteriorly. The ovaries are usually lodged in the ovarian fossae in nulliparous women, but they may be displaced posteriorly into the rectouterine pouch (pouch of Douglas, cul-de-sac) or anteriorly toward the abdominal wall during pregnancy, where they may persist after delivery. The ovaries derive their blood supply from the ovarian arteries and ovarian branches of the uterine arteries. The

287

paired ovarian arteries arise from the abdominal aorta and pass through the retroperitoneum and the ovarian suspensory ligament to give off branches which pierce the ovaries. The uterine arteries arise from the internal iliac arteries and give off ovarian and tubal branches, which course along the attachment of the broad ligament at the lateral edge of the uterus toward the uterotubal junction and also anastomose with the ovarian arteries. Veins leaving the hilum of the ovary form a vinelike network of vessels, the pampiniform plexus. The ovarian vein formed by this plexus empties into the renal vein on the left side and directly into the inferior vena cava on the right. Blood from the uterine veins drains into the internal iliac veins. Lymphatic drainage from the ovaries and uterine tubes is to the aortic lymph nodes in the lumbar region. Lymphatics from the region of the uterotubal junction pass through the suspensory ligament toward the superficial inguinal lymph nodes in the groin.

MRI Appearance The normal uterine tubes are not routinely seen on MRI. In rare instances, they are depicted as horizontally oriented, tubular structures of intermediate signal intensity on T2w images. During the reproductive years, the ovaries are regularly seen on T2w images. The ovarian cortex contains the characteristic rounded, hyperintense follicles (Fig. 13.1). The size of these cystlike structures ranges from a few millimeters to 2.5 cm and varies during the menstrual cycle. The largest follicle is also known as the dominant follicle. Persistence of the dominant follicle can give rise to a follicular cyst. The latter is a functional cyst and can be

Fig. 13.1 Normal appearance of both ovaries in a 27-year-old woman. Axial T2w TSE image depicts multiple high-SI follicles.

288

13 The Adnexa

a

b Fig. 13.2a, b MR appearance at different phases of the menstrual cycle in a 19-year-old woman. Axial T2w TSE images before (a) and after ovulation (b) show the change in size of the dominant follicle during the menstrual cycle. Note some fluid in the rectouterine pouch.

differentiated from normal ovarian follicles on serial images from different phases of the menstrual cycle. Normal ovarian follicles are characterized by regular changes in size and signal intensity (Fig. 13.2). Serial examinations should be performed using TVUS. On MRI, ovarian follicles are difficult to identify on T1w images because they are isointense to muscle. Ovarian zonal anatomy, i. e., the depiction of the cortex and medulla as two distinct regions, is best appreciated on T2w images obtained before ovulation. The cortex is typically isointense to muscle, while the medulla is slightly hyperintense (intermediate in signal intensity between muscle and fat). Zonal anatomy varies during the menstrual cycle. Outwater et al. were able to identify ovarian zonal anatomy in 85 % of premenopausal women and 28 % of postmenopausal women.10 Postmenopausal ovaries are less conspicuous. They have lower T2 signal intensity, which is due to atrophic fibrosis and a decrease in the number and size of follicles.

MRI Appearance of Pathologic Entities Pelvic Inflammatory Disease (PID) Most cases of pelvic inflammatory disease in women result from infection with Chlamydia trachomatis or Neisseria gonorrhoeae, which ascend through the cervix and uterus, from where they can spread to the uterine tubes and ovaries. Tubal blockage secondary to salpingitis is usually not detected with routine MRI techniques but can be ascertained using MR hysterosalpingography (Fig. 13.3). With

this technique, tubal patency is assessed by acquiring a 3 D MR angiography sequence following instillation of diluted Gd-based contrast medium solution (for intravenous injection) into the endometrial cavity.11 Early salpingitis is suggested by fluid retention in the uterine tubes, which would not be identified otherwise. Coronal T2w images are well suited for this purpose because they depict the tubes as high-signal-intensity structures extending nearly horizontally from the uterus to the ovaries. The MR morphology of pyosalpinx is indistinct from that of uncomplicated tubal dilatation. Tubo-ovarian abscess is a complication of inflammation of the serosal coat of the ovaries secondary to spread of salpingitis into the rectouterine pouch. An abscess is depicted by MRI as edema around the ovary, which is usually not enlarged.

Ectopic Pregnancy Ectopic pregnancy is diagnosed using TVUS, which can detect 62 % of all ectopic pregnancies as early as 7 weeks of gestation. MRI is rarely indicated in early pregnancy, which is why ectopic pregnancy is usually an incidental finding at MRI. The MR appearance is that of a cystlike mass of predominantly high signal intensity. Solid or necrotic areas of intermediate signal intensity may be present, depending on whether or not the product of conception is intact. If an ectopic pregnancy is suspected at MRI, a sagittal sequence should be acquired for better delineation of the chorionic cavity. Confident differentiation from cystic adnexal tumors is not always possible.

MRI Appearance of Pathologic Entities

289

a

b Fig. 13.3a–c Dynamic MR hysterosalpingography. MIP images during early (a, b) and late (c) filling. Images show normal appearance of the endometrial cavity (a) and uterine tubes (b). Tubal patency is indicated by the presence of contrast medium in the peritoneal cavity on both sides (c). (Images courtesy of Priv.-Doz. Dr. W. Wiesner, Klinik Stephanshorn, St. Gallen, Switzerland.)

c

Ovarian Torsion Torsion typically occurs in women with a benign mass of the affected ovary; however, a normal ovary may also undergo torsion, especially in girls. MRI will reveal follicular cysts with thickened walls and massive edema of the ovarian stroma, resulting in a distancing of the follicular cysts. The presence of blood suggests hemorrhagic infarction.12 Enhancement of the follicle walls on postcontrast images indicates a viable ovary.13

Benign Tumors MRI can help characterize the tissue composition of ovarian masses, mainly by differentiating between cystic and solid components.

Predominantly Cystic Benign Tumors Functional Ovarian Cysts. Ovarian cysts are the most common benign adnexal tumors in premenopausal women. Two types of functional cysts are distinguished: follicular cysts, which develop from persistent follicles, and corpus luteum cysts, which are products of failed ovulation. Func-

tional cysts are usually asymptomatic and detected incidentally; however, they can cause ovarian torsion and present with acute peritoneal symptoms. The MR appearance does not allow differentiation of follicular and corpus luteum cysts. They are therefore collectively reported as simple (serous) cysts (Fig. 13.4) or, if bleeding is present, as hemorrhagic cysts (Fig. 13.5). On MRI, these cysts appear as well-circumscribed, thinwalled unilocular masses without septa or wall irregularities. Simple cysts containing serous fluid have uniform low signal intensity on T1w images and high T2 signal intensity.14–17 They are up to 4 cm in size. Hemorrhage into ovarian cysts is occasionally seen; these cysts are known as chocolate cysts because of the brown color of old blood. The MR signal pattern of hemorrhagic cysts depends on the age and amount of blood present. Typical hemorrhagic cysts are of high T1 and T2 signal intensity.14,17,18 Polycystic Ovary Syndrome (PCOS). Polycystic ovary syndrome, or Stein–Leventhal syndrome, is characterized by bilaterally enlarged ovaries and the presence of multiple immature follicles of similar size. Clinical features of PCOS are abnormal menses (oligomenorrhea, amenorrhea), obesity, and infertility. Associated imaging findings include a hypoplastic uterus with normal zonal anatomy

290

13 The Adnexa

a Fig. 13.4a, b Simple cyst. a Axial T2w TSE image. b Coronal T1w SE image. Parovarian serous cyst with the typical features of a simple cyst: isointensity to fluid on T2w and T1w sequences, unilocularity, and thin wall. No solid components. The cyst is unusually large. b

pending on the cyst contents: intracystic hemorrhage or mucinous fluid shortens T1 relaxation time, resulting in a bright signal on both T1w and T2w images, whereas serous cystadenomas have the same signal intensity as simple cysts.15,20,21 Dilated Uterine Tube. A dilated uterine tube may be difficult to differentiate from benign cystic adnexal tumors.15 Depiction of a C- or S-shaped tubular fluid-filled structure with smooth walls has been reported as a characteristic MRI finding (Fig. 13.7).15,18 Nevertheless, a dilated uterine tube is frequently confused with simple/hemorrhagic cysts or, because of the multicystic appearance, with cystadenoma. The MR signal pattern of hematosalpinx is similar to that of hemorrhagic cysts. Fig. 13.5 Hemorrhagic cyst. Axial T1w SE image. Small left ovarian cyst of predominantly high SI.

on T2w images. The MRI finding of multiple peripheral ovarian cysts is not specific for polycystic ovaries.19 Cystadenoma. Cystadenomas are the most common benign ovarian neoplasms after menopause. They can be divided into serous and mucinous tumors (Fig. 13.6). Serous cystadenomas are unilocular, bilocular, or multilocular, whereas mucinous cystadenomas are always multilocular. On MRI, cystadenomas have thin walls and may contain a few thin septa. Vegetations are present in rare instances. Most cystadenomas are larger than functional ovarian cysts. There are no reliable criteria for distinguishing serous cystadenomas from other benign cystic ovarian tumors. The MR signal characteristics are variable, de-

Parovarian Cyst. Parovarian cysts are the second most frequent benign adnexal cystic lesion of nonovarian origin. On MRI, parovarian cysts are indistinct from simple ovarian cysts (see Fig. 13.4). Endometriosis. Endometriosis is characterized by foci of ectopic endometrial tissue that respond to hormonal stimulation and may cause the same cycle-dependent symptoms as the normal uterine endometrium. Endometrial tissue can enter the peritoneal cavity through the uterine tubes. The most common manifestations of endometriosis are bilateral cystic blood collections on the surface of the ovaries. Other common sites of endometrial implants are the pelvic peritoneum and uterine ligaments. Endometriotic Cyst. Endometrial implants can enlarge to form endometriotic cysts, or endometriomas. These are typically multilocular cystic structures enclosed by a thick,

MRI Appearance of Pathologic Entities

291

a Fig. 13.6a, b Mucinous cystadenoma. a Coronal T2w TSE image. b Sagittal T1w SE image after contrast administration. Unusually large, multilocular mass with mostly thin walls and septa and only a few, very subtle irregularities. The T2 SI of the cyst contents is isointense to fluid (a) with two cranial locules having slightly higher T1 SI (b).

b

a Fig. 13.7a, b Dilated uterine tube. a Axial T2w TSE image. b Sagittal T2w TSE image. The S-shape of the tubular structure (b) suggests a tubal origin, while the multicystic appearance on the axial image (a) is more in keeping with a cystadenoma.

b

292

13 The Adnexa

a

b Fig. 13.8a, b Endometriotic cyst. a Axial T2w TSE image. b Axial T1w SE image. Unilocular cyst with characteristic capsulelike wall and characteristic shortening of both T1 and T2 relaxation time.

blood present in the cyst. Low T2 signal intensity results if there is marked T2 shortening due to the presence of large amounts of hemosiderin. A distinguishing MR feature of endometriotic cysts is the shading sign, a fluid–fluid level with hypointensity of the dependent portion (Fig. 13.9). A combination of unenhanced T1w and T2w sequences is necessary for reliable diagnosis of endometriosis using MRI.15,17,18,22,23

Fig. 13.9 Endometriotic cyst. Axial T2w TSE image shows characteristic shading with low-SI fluid in the dependent portion and higher-SI fluid in the nondependent portion of the cyst. A thrombus is present in the posterior aspect of the cyst.

capsulelike wall of low signal intensity. The wall helps differentiate endometriotic cysts from thin-walled hemorrhagic cysts (Fig. 13.8). Most endometriomas have illdefined margins. The presence of paramagnetic blood products—methemoglobin and hemosiderin—causes shortening of both T1 and T2 relaxation times. As a result, the cysts have inherently high T1 signal intensity, which is why intravenous contrast administration is not necessary for characterization. T2 signal intensity is high to intermediate, depending on the age and degree of absorption of the

Dermoid Cyst. Dermoid cysts, or mature cystic teratomas, are the most common benign neoplasms during puberty, accounting for 20 % of these tumors. They consist predominantly of ectodermal tissue derived from all three germ layers and therefore contain fat interspersed with hair, teeth, or bone. The primordial tooth or bone may be enclosed in a dermoid plug (Rokitansky protuberance). Dermoid cysts are congenital, and bilateral occurrence is common. They are usually asymptomatic and detected incidentally, unless they cause ovarian torsion. There is a potential risk of malignant transformation, which is up to 20 % for serous dermoid cysts and requires histologic work-up in all cases. Malignant transformation of dermoid cysts has been described as transmural extension of solid tumor components and invasive growth into adjacent organs.24 Identification of the fat component is helpful for the MR diagnosis. Dermoid cysts are characterized by a signal intensity isointense to fat on all pulse sequences14–18,25,26 (Fig. 13.10). A chemical shift artifact will be seen at the interface between intracystic fat and surrounding bowel or intracystic fluid. Teratomas frequently contain fluid, causing characteristic fat–fluid levels with fat of intermediate signal intensity and fluid of high signal intensity on T2w images. Dermoid plugs may be detectable as outgrowths from the inner surface of the cyst wall.

MRI Appearance of Pathologic Entities

293

Apart from endometriotic cysts and teratomas, MRI often fails to provide a specific diagnosis. A classification of ovarian masses based on MRI features alone would therefore be less refined than histologic classification. In all cases where a specific diagnosis cannot be established, the radiologist should report definitively benign cystic adnexal masses as simple or hemorrhagic cysts or, if multilocular with septa, as complex cysts.18,27 From a clinical perspective, confident characterization of an ovarian mass as benign or malignant is more relevant than a specific diagnosis.

Predominantly Solid Benign Tumors Ovarian Fibroma. Ovarian fibromas are of mesenchymal origin and account for ca. 4 % of all ovarian tumors. Most ovarian fibromas are diagnosed in the sixth decade of life. They are hormonally inactive and typically occur unilaterally. MRI depicts them as well-defined, solid tumors of low signal intensity on T1w images. On T2w images, they also have predominantly low signal28 but typically appear heterogeneous due to hyperintense cystic components (Fig. 13.11). Ovarian fibromas can be confused with pedunculated uterine leiomyomas, especially when they are contiguous with the uterine wall. Malignant transformation has been reported and MRI alone does not allow reliable differentiation from benign fibroma (Fig. 13.12). One percent of cases are associated with Meigs syndrome, a benign condition characterized by ovarian fibroma, ascites, and pleural effusion. Functioning Adnexal Tumors. Hormonally active adnexal tumors produce either estrogen or androgen. The most common estrogen-producing tumors include granulosa cell tumors and thecomas. They predominantly occur in postmenopausal women and are usually unilateral. Estrogen production by these tumors leads to endometrial hyperplasia, vaginal bleeding, and cystic mastopathy. Granulosa cell tumors in childhood induce precocious pseudopuberty. The most important androgen-producing tumors of the ovary are Sertoli–Leydig cell tumors. The testosterone secreted by these tumors causes sterility, excessive growth of body hair, and enlargement of the clitoris. Malignant transformation can occur. Typical MRI findings include a solid mass with variable cystic components29,30 and moderate enhancement on postcontrast images (Fig. 13.13).

Fig. 13.10a–c Mature teratoma. a Axial TIRM image. b Axial T1w SE image. c Sagittal T2w TSE image. The lesion is isointense to fat on all sequences with a cystic component in the posterior aspect.

a

b

c

294

13 The Adnexa

a

Fig. 13.11 Ovarian fibroma. Sagittal T2w TSE image. Heterogeneous mass consisting of low-SI solid tissue and multiple cystic areas of high SI.

b Fig. 13.12a, b Moderately differentiated adenocarcinoma that has developed from serous cystadenofibroma. a Coronal T2w TSE image. b Fat-suppressed coronal T1w SE image obtained after IV contrast administration. Predominantly solid mass with some necrotic areas. The MR appearance does not allow reliable differentiation from the benign ovarian fibroma shown in Fig. 13.11.

Fig. 13.13 Hemorrhagic infarcted thecoma. Axial T1w SE image after contrast administration. Mixed cystic and solid mass.

Malignant Tumors Ovarian cancer is the sixth most common cancer and is the fifth leading cause of cancer death in women. Relative to the number of cases, ovarian cancer is the most common cause of death due to gynecologic cancers.31 The risk of developing ovarian cancer increases with age. It is rare before the age of 40, and almost half of all ovarian cancer patients are over 60. Ovarian cancer accounts for less than

5 % of adnexal tumors diagnosed before menopause. Specific early signs and symptoms are lacking, which is why most ovarian cancers are detected late, often during a routine pelvic examination. Rapid progression further contributes to a poor prognosis. Women diagnosed with early ovarian cancer confined to the true pelvis (FIGO stage I or II) have a much better prognosis. Ovarian cancer is an extremely heterogeneous group of malignancies. There are several major histologic classes,

MRI Appearance of Pathologic Entities

295

often with further subclasses. Epithelial ovarian cancers constitute by far the largest group, accounting for ca. 60 % of ovarian neoplasms. The most common epithelial subtype is cystadenocarcinoma.32 There is a continuum from definitely benign ovarian masses (cystadenoma, adenofibroma) to definitely malignant tumors.33 Borderline ovarian tumors are indeterminate and constitute a separate class in the histologic WHO classification of ovarian cancer. About 17 % of all epithelial ovarian tumors are borderline tumors. They are indeterminate between benign and malignant in terms of clinical behavior and morphologic features (Figs. 13.14 and 13.15). Histologically, borderline tumors are characterized by epithelial proliferation without destructive growth. The macroscopic criteria of malignancy used by pathologists32 have been adapted for the characterization of ovarian masses on MRI. This set of criteria is widely accepted and used by most investigators in the radiologic literature.5,15,17,18,20,27,34–36 An ovarian mass is classified as malignant if at least one of the following criteria is fulfilled (Figs. 13.15, 13.16, 13.17): · a solid or predominantly solid mass with or without necrosis · a predominantly cystic mass with a wall or septa > 3 mm in thickness · a cystic mass with peripheral nodules or papillary projections · marked multilocularity · invasion of adjacent organs or of the pelvic sidewall · peritoneal, mesenteric, or omental tumor implants. Note, however, that mural nodules and solid components have occasionally been demonstrated by MRI in cystadenomas (see Fig. 13.6)20,27,34 and cystadenofibromas.35 Therefore, tissue characterization using macroscopic MRI criteria alone has the same limitations as gross pathologic evaluation.33 Nevertheless, as long as other, more specific criteria are lacking, the radiologist’s main task in interpreting the MRI data remains to carefully search for evidence of solid tumor elements. Because of the limited specificity of the malignancy criteria, up to 20 % of diagnoses are false positive.6 Conversely, by excluding solid components in those cases where TVUS demonstrates complex cystic adnexal tumors and thus ruling out malignancy with a high degree of confidence, MRI can help identify women who can be treated with less radical surgery.6 Although peritoneal, mesenteric, or omental spread as well as ascites and lymphoma can be detected by MRI,37,38 staging laparotomy is the standard approach to define the disease extent in clinical practice.39 Thorough and systematic intraoperative exploration serves to identify and sample any tissue that appears suspicious on gross inspection or palpation. However, staging laparotomy is less reliable than MRI or CT in identifying retroperitoneal lymphoma and subdiaphragmatic implants. Like other gynecologic malignancies, ovarian cancer is staged using the FIGO or UICC classification system (Table 13.3).

a

b Fig. 13.14a, b Mucinous cystadenoma in transition to borderline tumor. a Axial TIRM image. b Fat-suppressed axial T1w SE image after IV contrast injection. Multilocular mass with complex architecture in the posterior aspect. Thick walls and thick septa are revealed in this portion of the mass on the postcontrast image (b).

Ovarian metastases are rare. They affect one or both ovaries and typically cannot be distinguished from primary cancer, although multilocularity favors the diagnosis of primary ovarian malignancy.40 Tumors metastasizing to the ovaries include breast cancer, malignant melanoma, and gastrointestinal cancer. Drop metastasis to both ovaries from signet-ring gastric cancer is known as Krukenberg tumor (Fig. 13.18). Depending on the primary tumor, ovarian metastases are solid or cystic and can markedly enlarge or diffusely replace the affected ovary. Ascites may be demonstrated in patients with peritoneal carcinosis. The MRI appearance is extremely heterogeneous. Enhancement after intravenous contrast administration is moderate to intense, depending on the degree of vascularization of the primary tumor.

296

13 The Adnexa

a

b Fig. 13.15a–c Very highly differentiated adenocarcinoma with features suggestive of borderline tumor. a Axial T2w TSE image. b Axial T1w SE image. c Axial fat-suppressed T1w SE image after IV contrast administration. Large multilocular mass of inhomogeneous SI on precontrast T1w image (b). The higher SI of some locules indicates mucinous or hemorrhagic fluid. Postcontrast image (c) reveals suspicious projections from the walls and septa in the anterior aspect of the mass.

c

a

b Fig. 13.16a, b Poorly differentiated serous papillary ovarian cancer, FIGO stage IC. a Axial T2w TSE image. b Axial T1w SE image after IV contrast injection. Multilocular, predominantly cystic mass with

nodular projections and uniformly thickened wall. Free fluid in the peritoneal cavity.

MRI Appearance of Pathologic Entities

297

a

b Fig. 13.17a, b Bilateral ovarian cystadenocarcinoma. a Axial T2w TSE image. b Axial T1w SE image after IV contrast injection. Large cystic and solid mass consisting predominantly of confluent solid portions.

a

b Fig. 13.18a, b Bilateral ovarian metastases from gastric cancer. a Axial T2w TSE image. b Coronal fat-suppressed T1w SE image after IV contrast injection. Solid ovarian masses with inhomogeneous enhancement.

Table 13.3 FIGO and UICC stages of ovarian cancer FIGO UICC

Tumor extent

I IA IB IC

T1 T1A T1B T1C

II

T2

IIA IIB III

T2A T2B T3

IV

T4

Cancer confined to the ovaries Unilateral tumor, no ascites Bilateral tumor, no ascites Unilateral or bilateral tumor, tumor cells in ascites Extension beyond ovaries; contained within the true pelvis Tumor involves uterine tube or uterus Tumor involves other pelvic structures Peritoneal spread outside pelvis, or nodal spread of tumor Distant spread of cancer

Tubal Cancer Cancer of the uterine tubes is very rare, more common in nulliparous women, and bilateral in 10–15 % of cases. The age peak is between 50 and 60 years. Histologically, tubal cancers are adenocarcinomas resembling serous cystadenocarcinomas. The tumor initially grows in the tubal lumen and is characterized by early lymphatic and distant metastatic spread to para-aortic, iliac, and inguinal lymph nodes and into the peritoneum. On MRI, tubal cancer cannot be differentiated from cystic ovarian cancer.

Dysgerminoma Dysgerminoma is the ovarian counterpart of testicular seminoma. These rare embryonal tumors develop in the second or third decade of life in 75 % of cases. Most dysgerminomas are solid and occur unilaterally. One third

298

13 The Adnexa

show aggressive growth with destructive invasion of surrounding tissue and early lymphatic and hematogenous metastatic spread. They are of intermediate to high signal intensity on T2w images.

Ovarian Lymphoma

Fig. 13.19 Ovarian involvement in disseminated non-Hodgkin lymphoma. Axial T2w TSE image. Both ovaries are enlarged and of intermediate SI.

Primary lymphoma of the ovaries without bone marrow involvement or infiltration of a lymph node station is rare. In women with disseminated malignant lymphoma, autopsy reveals ovarian lymphoma in up to 30 % of cases, typically in the form of diffuse infiltration and symmetrical ovarian enlargement. The MRI appearance of ovarian lymphoma41 is that of a purely solid mass of intermediate signal intensity on T1w and T2w images, with the T2 signal being slightly above that of muscle (Fig. 13.19). No follicles will be seen in the diffusely infiltrated ovaries, which may therefore be confused with enlarged lymph nodes. There is moderate enhancement after intravenous contrast administration but this does not contribute to the diagnosis.

Recurrent Tumor

Fig. 13.20 Recurrent tumor at the vaginal vault after removal of left ovary and supracervical hysterectomy, chemotherapy, and radiation. Axial fat-suppressed T1w SE image after IV contrast injection. Unilocular, predominantly cystic mass with suspicious nodular projections on the inner cyst wall.

Recurrent or residual tumor is not uncommon, as many patients with ovarian cancer present with advanced disease. Recurrence after hysterectomy typically involves the vaginal vault or rectouterine pouch (Figs. 13.20 and 13.21). Recurrent ovarian cancer is similar in morphologic appearance and signal characteristics to the primary cancer. Lymphatic metastases primarily involve pelvic sidewall and para-aortic nodes. Patients with ascites need to be carefully evaluated for interintestinal tumor cell clusters, implants on the liver surface, and hepatic metastases. Follow-up imaging is performed only if recurrent disease is suspected on the basis of clinical or laboratory findings (tumor marker CA-125).42

Fig. 13.21 MRI to evaluate for suspected repeated recurrence of borderline tumor after removal of right ovary and resection of recurrent tumor of left ovary. Axial TIRM image. Irregular cystic mass in the pelvis; the cyst appears to be folded onto itself in the left lateral aspect. Histologically, the mass was a peritoneal cyst.

MRI Appearance of Pathologic Entities

References 1. Yancik R. Ovarian cancer. Age contrasts in incidence, histology, disease stage at diagnosis, and mortality. Cancer 1993; 71(2, Suppl)517–523 2. Cohen CJ, Jennings TS. Screening for ovarian cancer: the role of noninvasive imaging techniques. Am J Obstet Gynecol 1994;170(4):1088–1094 3. Granberg S, Norström A, Wikland M. Tumors in the lower pelvis as imaged by vaginal sonography. Gynecol Oncol 1990;37(2): 224–229 4. Reuter M, Steffens JC, Schüppler U, et al. Critical evaluation of the specificity of MRI and TVUS for differentiation of malignant from benign adnexal lesions. Eur Radiol 1998;8(1):39–44 5. Yamashita Y, Hatanaka Y, Torashima M, Takahashi M, Miyazaki K, Okamura H. Characterization of sonographically indeterminate ovarian tumors with MR imaging. A logistic regression analysis. Acta Radiol 1997;38(4 Pt 1):572–577 6. Reuter M, Steffens JC, Schüppler U, et al. [Preoperative differential diagnosis of cystic adnexal tumors: double-contrast MRT]. Rofo 1996;164(5):394–400 7. Schwartz LB, Panageas E, Lange R, Rizzo J, Comite F, McCarthy S. Female pelvis: impact of MR imaging on treatment decisions and net cost analysis. Radiology 1994;192(1):55–60 8. Smith RC, Reinhold C, McCauley TR, et al. Multicoil high-resolution fast spin-echo MR imaging of the female pelvis. Radiology 1992;184(3):671–675 9. Yamashita Y, Tang Y, Abe Y, Mitsuzaki K, Takahashi M. Comparison of ultrafast half-Fourier single-shot turbo spin-echo sequence with turbo spin-echo sequences for T2-weighted imaging of the female pelvis. J Magn Reson Imaging 1998;8(6): 1207–1212 10. Outwater EK, Talerman A, Dunton C. Normal adnexa uteri specimens: anatomic basis of MR imaging features. Radiology 1996; 201(3):751–755 11. Wiesner W, Ruehm SG, Bongartz G, Kaim A, Reese E, De Geyter C. Three-dimensional dynamic MR hysterosalpingography: a preliminary report. Eur Radiol 2001;11(8):1439–1444 12. Rhap E, Byun JY, Jung PE, et al. CT and MR imaging features of adnexal torsion. Radiographics 2002;22:283–294 13. Haque TL, Togashi K, Kobayashi H, Fujii S, Konishi J. Adnexal torsion: MR imaging findings of viable ovary. Eur Radiol 2000;10(12):1954–1957 14. Dooms GC, Hricak H, Tscholakoff D. Adnexal structures: MR imaging. Radiology 1986;158(3):639–646 15. Kombächer P, Hamm B, Becker R, Hese S, Weitzel HK, Wolf KJ. [Tumors of the adnexa—a comparison of magnetic resonance tomography, endosonography and the histological findings]. Rofo 1992;156(4):303–308 16. Mitchell DG, Mintz MC, Spritzer CE, et al. Adnexal masses: MR imaging observations at 1.5 T, with US and CT correlation. Radiology 1987;162(2):319–324 17. Thurnher SA. MR imaging of pelvic masses in women: contrastenhanced vs unenhanced images. AJR Am J Roentgenol 1992;159(6):1243–1250 18. Jain KA, Friedman DL, Pettinger TW, Alagappan R, Jeffrey RBJr, Sommer FG. Adnexal masses: comparison of specificity of endovaginal US and pelvic MR imaging. Radiology 1993;186(3): 697–704 19. Kimura I, Togashi K, Kawakami S, et al. Polycystic ovaries: implications of diagnosis with MR imaging. Radiology 1996; 201(2):549–552 20. Ghossain MA, Buy JN, Lignères C, et al. Epithelial tumors of the ovary: comparison of MR and CT findings. Radiology 1991; 181(3):863–870 21. Mawhinney RR, Powell MC, Worthington BS, Symonds EM. Magnetic resonance imaging of benign ovarian masses. Br J Radiol 1988;61(723):179–186

299

22. Outwater E, Schiebler ML, Owen RS, Schnall MD. Characterization of hemorrhagic adnexal lesions with MR imaging: blinded reader study. Radiology 1993;186(2):489–494 23. Togashi K, Nishimura K, Kimura I, et al. Endometrial cysts: diagnosis with MR imaging. Radiology 1991;180(1):73–78 24. Kido A, Togashi K, Konishi I, et al. Dermoid cysts of the ovary with malignant transformation: MR appearance. AJR Am J Roentgenol 1999;172(2):445–449 25. Stevens SK, Hricak H, Campos Z. Teratomas versus cystic hemorrhagic adnexal lesions: differentiation with proton-selective fat-saturation MR imaging. Radiology 1993;186(2):481–488 26. Togashi K, Nishimura K, Itoh K, et al. Ovarian cystic teratomas: MR imaging. Radiology 1987;162(3):669–673 27. Yamashita Y, Torashima M, Hatanaka Y, et al. Adnexal masses: accuracy of characterization with transvaginal US and precontrast and postcontrast MR imaging. Radiology 1995;194(2): 557–565 28. Troiano RN, Lazzarini KM, Scoutt LM, Lange RC, Flynn SD, McCarthy S. Fibroma and fibrothecoma of the ovary: MR imaging findings. Radiology 1997;204(3):795–798 29. Kim SH, Kim SH. Granulosa cell tumor of the ovary: common findings and unusual appearances on CT and MR. J Comput Assist Tomogr 2002;26(5):756–761 30. Morikawa K, Hatabu H, Togashi K, Kataoka ML, Mori T, Konishi J. Granulosa cell tumor of the ovary: MR findings. J Comput Assist Tomogr 1997;21(6):1001–1004 31. Tumorregister Munich http://www.med.uni-muenchen.de/trm 32. Czernobilsky B. Common epithelial tumors of the ovary. In: Kurman RJ, ed. Blaustein's Pathology of the Female Genital Tract. New York–Berlin: Springer; 1987, pp. 639–646 33. Granberg S. Relationship of macroscopic appearance to the histologic diagnosis of ovarian tumors. Clin Obstet Gynecol 1993;36(2):363–374 34. Scoutt LM, McCarthy SM, Lange R, Bourque A, Schwartz PE. MR evaluation of clinically suspected adnexal masses. J Comput Assist Tomogr 1994;18(4):609–618 35. Stevens SK, Hricak H, Stern JL. Ovarian lesions: detection and characterization with gadolinium-enhanced MR imaging at 1.5 T. Radiology 1991;181(2):481–488 36. Wagner BJ, Buck JL, Seidman JD, McCabe KM. From the archives of the AFIP. Ovarian epithelial neoplasms: radiologic-pathologic correlation. Radiographics 1994;14(6):1351–1374, quiz 1375–1376 37. Kurtz AB, Tsimikas JV, Tempany CMC, et al. Diagnosis and staging of ovarian cancer: comparative values of Doppler and conventional US, CT, and MR imaging correlated with surgery and histopathologic analysis—report of the Radiology Diagnostic Oncology Group. Radiology 1999;212(1):19–27 38. Tempany CMC, Zou KH, Silverman SG, Brown DL, Kurtz AB, McNeil BJ. Staging of advanced ovarian cancer: comparison of imaging modalities—report from the Radiological Diagnostic Oncology Group. Radiology 2000;215(3):761–767 39. Pfisterer J, du Bois A. Das Ovarialkarzinom: Therapeutische Standards–klinische Empfehlungen. Stuttgart: Thieme; 2002 40. Brown DL, Zou KH, Tempany CMC, et al. Primary versus secondary ovarian malignancy: imaging findings of adnexal masses in the Radiology Diagnostic Oncology Group Study. Radiology 2001;219(1):213–218 41. Mitsumori A, Joja I, Hiraki Y. MR appearance of non-Hodgkin’s lymphoma of the ovary. AJR Am J Roentgenol 1999;173(1):245 42. Organkommission Ovar der Arbeitsgemeinschaft für Gynäkologische Onkologie (AGO). Diagnostische und therapeutische Standards beim Ovarialkarzinom (Langversion), December 2000

301

14

Magnetic Resonance Pelvimetry M.R. Muehler

Introduction The aim of pelvimetry is to objectively measure pelvic dimensions and assess the configuration of the maternal bony pelvis to help obstetricians recognize fetopelvic disproportion and identify women who should have a primary cesarean section rather than attempt vaginal delivery. Ultrasound is the standard imaging modality for determining the weight and size of an unborn child. Conventional pelvimetry was a routine examination until the end of the 1950s. The number of pelvimetries rapidly declined with an increasing awareness of the risks of intrauterine radiation exposure and reports on the cancer risk associated with the examination.1,2 New techniques such as digital image enhancement radiography and CT have considerably reduced radiation exposure and led to a revival of maternal pelvic measurement. The use of MRI for pelvimetry was first described in 1985, and it has since evolved into the standard modality.2,3 MR pelvimetry is a highly accurate method, having a measurement error of only 1 %.3 Both fetal ultrasound and MR pelvimetry have very low interobserver and intraobserver variability.2,4 Different obstetric indices are determined to identify fetopelvic disproportion, but their use has increasingly come into question as they are limited in predicting the course of labor.1,2,5,6 As a consequence, the use of pelvimetry has shifted from the prenatal to the postnatal period to rule out a small pelvis in women with a history of protracted labor.1 It was shown that pelvimetry significantly reduces the rate of secondary cesarean sections in breech presentation (its most common indication).5 Performing an MRI examination during the first trimester (organogenesis) is generally discouraged.7–9 A deleterious effect on the fetus is not known; however, potential effects of the static magnetic field, the gradient system, and the high-frequency field used in MRI have not been investigated systematically.8 For this reason, the specific absorption rate (SAR) should be as low as possible. Although both GRE and SE sequences are considered harmless, the SAR of an SE or TSE sequence is several times higher than that of a GRE sequence, while both yield the same diagnostic image quality.6

MR pelvimetry not only obviates the need for exposure to ionizing radiation but also has several additional advantages: · Obese patients can be examined with high image quality. · There is no need for repositioning when a second imaging plane is needed. · MRI directly measures pelvimetric values, while conventional pelvimetry uses a ruler. · The standard MRI protocol simultaneously enables measurement of the bony pelvis and evaluation of the pelvic soft tissues. Despite its limited predictive value, there is a role for MR pelvimetry in the evaluation of the maternal pelvis. The radiologist’s task is to carry out the examination and measurement, while it is left to the experienced obstetrician to judge the clinical relevance of the findings.

Indications There is no generally accepted set of indications for MR pelvimetry. The importance attached to pelvimetry is at the discretion of the individual obstetrician and reflects the fact that the dimensions of the bony pelvis are only one of many factors that can lead to protracted labor. MR pelvimetry may be justified in the following situations: · breech presentation · history of protracted labor (dystocia) · suspected pelvic deformity (after trauma, poliomyelitis, or rickets) · fetal–maternal pelvis size mismatch suspected on clinical examination, e. g., narrow outer pelvic diameter, hypermobility of the fetal head in a term primigravida, positive Zangemeister maneuver, or large fetal head. Several risk factors for protracted labor (dystocia) have been reported in the literature and may be regarded as relative indications for pelvimetry: mother’s height ≤ 164 cm10 and macrosomia with a fetal weight ≥ 90th percentile as of the 32nd week of gestation.11

302

14 Magnetic Resonance Pelvimetry

Definition of Pelvic Measures Obstetricians divide the inner pelvis into three planes: the pelvic inlet, the midpelvis, and the pelvic outlet, for each of which a transverse (coronal) and an anteroposterior (sagittal) diameter are measured. · Pelvic inlet: – transverse diameter (Fig. 14.1a), measured between the arcuate lines of the iliac bones

·

– sagittal diameter or obstetric conjugate (Fig. 14.1b), measured as the shortest distance between the symphysis pubis and the sacral promontory. Midpelvis: – transverse diameter (interspinous distance) (Fig. 14.1c), which is the narrowest portion between the ischial spines; in the axial plane, the ischial spines are at the same or a lower level than the easily identifiable femoral fovae – sagittal diameter (Fig. 14.1 d), measured as the shortest distance between the bottom of the symphysis pubis and the tip of the sacrum (S5).

a

b

c

d

e

Fig. 14.1a–e Postpartum MR pelvimetry. a Transverse diameter of the pelvic inlet, measured on oblique axial GRE image in the plane of the sagittal pelvic inlet diameter. b Sagittal diameter of the pelvic inlet (obstetric conjugate), measured on sagittal GRE image. c Transverse diameter of the midpelvis (interspinous distance), measured on axial GRE image. d Sagittal diameter of the midpelvis, measured on sagittal GRE image. e Transverse diameter of the pelvic outlet, measured on axial GRE image.

13.1–13.6 cm 11.7–12.1 cm

Midpelvis Transverse Sagittal

11.9 cm 11.6 cm

11.7–12.2 cm 11.3–11.8 cm

Pelvic outlet Transverse

12.3 cm

12.0–12.6 cm

2 2 2 8 8 8 2 2 2 5 5 10 280–340 280–340 280–340 256 × 192 256 × 192 256 × 192 No No No – – – 60 60 60 1.4–4 1.4–4 1.4–4

No. of slices FOV (mm) Matrix FS ETL Flip (°) TE (ms) TR (ms) Sequence type

Sequence No. 1: midsagittal scan to measure the sagittal diameter of the pelvic inlet and outlet. Sequence No. 2: axial scan angled parallel to the sagittal pelvic inlet diameter (obstetric conjugate) to measure the transverse pelvic inlet diameter. Sequence No. 3: axial scan (downward) starting ca. 5 cm above the upper border of the symphysis pubis to measure the transverse midpelvis and pelvic outlet diameters.

13.3 cm 11.9 cm

T1 T1 T1

Pelvic inlet Transverse Sagittal

No. 1 No. 2 No. 3

95 % confidence interval

Weight- Plane ing

Mean

Sequence

Table 14.1 Reference values for pelvimetry11

Table 14.2 Recommended pulse sequences and imaging parameters for MR pelvimetry

MR pelvimetry is performed with the patient in the supine position. Before the examination, it is important to inform the patient about vena cava compression syndrome and its early symptoms of nausea and sweating. To minimize the risk of this serious adverse effect of the examination, it should be established how long the patient can lie still on her back, and she should be encouraged to use the emergency button as soon as she notices any early symptoms. Careful education of a pregnant woman before an MRI examination is important not only in order to meet the medicolegal requirements but particularly to reassure the patient in this stressful situation. Fast T1w GRE sequences enable adequate identification of the landmarks used in MR pelvimetry and comparison of the results with those obtained using standard pelvic SE sequences. Clinical studies comparing MR pelvimetry with conventional or digital image enhancement radiography have shown MRI to yield the same results for most pelvic measures,6,13,14 except for the interspinous distance, for which a deviation of up to 2.3 cm was found between the different modalities. This deviation may be of clinical relevance, since the interspinous distance is the most prognostically relevant pelvic measure besides the obstetric conjugate. The error is systematic and has been confirmed in all comparative studies, suggesting that it is due to underestimation of the interspinous distance by conventional X-ray pelvimetry.7 This assumption has been corroborated by the results of a phantom study of X-ray and MR pelvimetry.13 In fact, the rate of random errors should be lower for MR pelvimetry, which uses electronic measurement, than for conventional pelvimetry, which uses a ruler and conversion factors (Table 14.2)

No. of acquisitions

Imaging Technique and Findings

120–150 120–150 120–150

Slice thickness (mm)

Pelvimetric reference values for a Central European population are given in Table 14.1.12

GRE GRE GRE

Scantime (min)

– Transverse diameter (intertuberous distance) (Fig. 14.1e), measured between the ischial tuberosities.

Sagittal Oblique axial Axial

Breathhold

· Pelvic outlet:

No No No

Imaging Technique and Findings

303

304

14 Magnetic Resonance Pelvimetry

The interspinous distance and obstetric conjugate are the two most prognostically relevant pelvic measures. Nevertheless, the predictive value of pelvimetry is limited and replacing the measurement of individual diameters by an overall evaluation of maternal pelvic proportions has been proposed. Validated and reliable methods for assessing the relationship of the pelvic inlet to the pelvic outlet based on pelvimetric data are not currently available.2

6.

7.

8. 9.

References 1. 2. 3. 4.

5.

Dixon AK. Pelvimetry revisited. Eur Radiol 2002;12(12): 2833–2834 Spörri S, Thoeny HC, Raio L, Lachat R, Vock P, Schneider H. MR imaging pelvimetry: a useful adjunct in the treatment of women at risk for dystocia? AJR Am J Roentgenol 2002;179(1):137–144 Stark DD, McCarthy SM, Filly RA, Parer JT, Hricak H, Callen PW. Pelvimetry by magnetic resonance imaging. AJR Am J Roentgenol 1985;144(5):947–950 Keller TM, Rake A, Michel SC, et al. Obstetric MR pelvimetry: reference values and evaluation of inter- and intraobserver error and intraindividual variability. Radiology 2003;227(1): 37–43 van Loon AJ, Mantingh A, Serlier EK, Kroon G, Mooyaart EL, Huisjes HJ. Randomised controlled trial of magnetic-resonance

10. 11.

12. 13.

14.

pelvimetry in breech presentation at term. Lancet 1997; 350(9094):1799–1804 Wentz KU, Lehmann KJ, Wischnik A, et al. [Pelvimetry using various magnetic resonance tomography techniques vs. digital image enhancement radiography: accuracy, time requirement and energy exposure]. Geburtshilfe Frauenheilkd 1994;54(4): 204–212 Ertl-Wagner B, Lienemann A, Strauss A, Reiser MF. Fetal magnetic resonance imaging: indications, technique, anatomical considerations and a review of fetal abnormalities. Eur Radiol 2002;12(8):1931–1940 Levine D, Barnes PD, Edelman RR. Obstetric MR imaging. Radiology 1999;211(3):609–617 Levine D, Zuo C, Faro CB, Chen Q. Potential heating effect in the gravid uterus during MR HASTE imaging. J Magn Reson Imaging 2001;13(6):856–861 Read AW, Prendiville WJ, Dawes VP, Stanley FJ. Cesarean section and operative vaginal delivery in low-risk primiparous women, Western Australia. Am J Public Health 1994;84(1):37–42 Brost BC, Goldenberg RL, Mercer BM, et al. The Preterm Prediction Study: association of cesarean delivery with increases in maternal weight and body mass index. Am J Obstet Gynecol 1997;177(2):333–337, discussion 337–341 Pfammatter T, Marincek B, von Schulthess GK, Dudenhausen JW. [MR pelvimetric reference values]. Rofo 1990;153(6):706–710 Bauer M, Schulz-Wendtland R, De Gregorio G, Sigmund G. [Obstetric pelvimetry using nuclear magnetic resonance tomography (MRI): clinical experiences with 150 patients]. Geburtshilfe Frauenheilkd 1992;52(6):322–326 Wright AR, English PT, Cameron HM, Wilsdon JB. MR pelvimetry—a practical alternative. Acta Radiol 1992;33(6):582–587

305

15

Magnetic Resonance Angiography of the Abdomen R. Vosshenrich and P. Reimer

Introduction MRI has been used for evaluation of blood vessels since the 1980s. At that time, vascular MRI was performed without contrast media using techniques such as black blood angiography, time-of-flight (TOF) MRI, or phase-contrast imaging. Acquisition with any of these techniques took several minutes, and the best results were achieved when vessels with directed flow were imaged against a stationary background. A full scan of the abdominal vessels during breath-holding was unthinkable. Even longer scan times resulted when electrocardiographic triggering was used to synchronize image acquisition with the patient’s cardiac cycle. Therefore, imaging of the abdominal vessels using conventional MR techniques could only be done by way of triggered acquisition of single slices with fast pulse sequences and was hampered by variable signal intensities and loss of signal due to dephasing. The long examination times resulting from the small field of view and poor spatial resolution limited the clinical utility of these techniques. Technical advances and the introduction of intravenous paramagnetic contrast media have since expanded the clinical utility of magnetic resonance angiography (MRA) dramatically. Prince et al. reported the first contrastenhanced MRA of the abdominal vessels in 1993.1 Today, contrast-enhanced MRA with acquisition of 3D datasets during a single breath-hold is indispensable in the preoperative evaluation of different vascular regions and follow-up after surgical or interventional treatment.

Indications With the development of contrast-enhanced techniques, MRA has evolved into a clinically relevant imaging modality for the abdominal vessels. Contrast-enhanced 3D MRA has become the primary imaging modality in patients with different conditions involving the major abdominal arteries and veins and is also increasingly being used to evaluate smaller vessels. Patients with acute dissecting or perforating aortic aneurysm are initially examined by multislice spiral computed tomography (MSCT) because it is more widely

available and faster and also allows better monitoring of vital functions; in these patients, MRA is useful for treatment planning and follow-up. It enables exact evaluation of the main branches of the celiac trunk and superior mesenteric artery, whereas optimal visualization of the inferior mesenteric artery is not achieved in all patients. The combination of contrast-enhanced 3D MRA with MR urography and nephrography has the potential to replace several conventional diagnostic tests in kidney disease and may turn MRI into a “one-stop shop” modality for these patients. Contrast-enhanced 3D MRA is well established for imaging of the portal venous system before liver transplant and in portal hypertension or as part of the preoperative diagnostic work-up of abdominal tumors. In the follow-up of patients with portal hypertension, contrast-enhanced MRA should be combined with a nonenhanced pulse sequence for quantifying blood flow. The abdominal veins can be reliably evaluated with both unenhanced and contrast-enhanced MRA techniques.

Imaging Technique The creation of an MR image relies on several properties that are unique to individual tissues, in particular T1 and T2 relaxation times and proton density. Another important phenomenon is proton motion. The susceptibility of MRI to motion causes flow artifacts and may degrade image quality. In MRA, flow-related effects are deliberately exploited to image moving spins. The two basic phenomena that are relevant here are flow-related signal enhancement (inflow effects) and flow-related signal loss (phase shifts). The MR techniques that make use of such flow effects are collectively known as unenhanced MRA and do so by specifically enhancing one effect while maximally suppressing the others.

Time-of-Flight MRA Time-of-flight (TOF) or inflow techniques are based on the motion of spins with a longitudinal magnetization component. Characteristically, the magnetization of flowing blood is manipulated at one site and sampled at another. In contrast, stationary spins within the imaging slice are

306

15 Magnetic Resonance Angiography of the Abdomen

subjected to many high-frequency pulses and thus become highly saturated when using a pulse sequence with a short repetition time (TR) compared with the T1 relaxation time of blood. Saturated spins are relatively insensitive to the excitation pulse and therefore yield little signal. In a vessel with blood flow perpendicular to the excited slice, the excited spins are constantly being replaced by fully relaxed inflowing spins, which give a high signal (bright blood imaging). As a result, contrast between stationary tissue and flowing blood is high.2 To make use of the inflow technique in a clinical setting, phase shifts, which occur simultaneously in the body and contribute to signal loss, have to be minimized. This is accomplished by using flow compensation (gradient moment rephasing, GMR) or a short echo time.3 Shorter TEs reduce the time in which spins can build up phase shifts. A TOF angiogram can be generated from 2D single slices, or from multiple small 3D slabs or one large one.

Phase-Contrast MRA The second non-contrast-enhanced MRA technique is based on the fact that excited spins, i. e., spins that have undergone transverse magnetization, experience a phase shift when moving along a gradient (phase contrast, PC). To illustrate the phase shift, let us assume that we apply a bipolar gradient pulse. Every time a gradient pulse is applied, this imparts a shift in the Larmor frequency at which spins precess. Following application of the pulse, the magnetization of the spins points in a different direction; in other words, the spins have undergone a change in phase. The amount of the phase shift depends on the gradient strength and the spin position. If after some time, we apply a second pulse of opposite polarity, the initial phase shift is reversed as long as the spins have not changed their position. Under ideal conditions, stationary spins will therefore have exactly the same magnetization as before application of the bipolar gradient. In contrast, the phase shift imparted to moving spins is not compensated for, precisely because the spins are no longer in the same position when the second pulse is applied. Thus, application of a bipolar gradient results in a net phase accumulation for moving spins. As the phase shift is greater for faster-flowing protons, PC MRA is an inherently quantitative method and can be used to calculate blood flow velocity.4

Contrast-Enhanced MRA Two major technical innovations have made possible the development of fast contrast-enhanced 3D MRA: highperformance gradient systems and dedicated phasedarray surface coils. The resulting high-gradient amplitudes with short rise times enable acquisition of complex 3D datasets with ultrashort TRs and TEs during a single breath-hold. In conjunction with large flip angles, very

short TRs also effectively suppress background signal. Use of paramagnetic contrast media, which shorten T1 relaxation time, ensures excellent delineation of vascular structures from surrounding tissue. Finally, ultrashort TEs minimize signal dephasing, thereby enabling more accurate evaluation of stenosis. Phased-array surface coils improve MRA by combining greater anatomic coverage with high signal-to-noise ratio (SNR).

Contrast Media Contrast-enhanced 3D MRA of the abdominal and pelvic vessels is performed using paramagnetic Gd chelates. Paramagnetic contrast media shorten the T1 relaxation times of the tissues into which they are taken up. Following intravenous bolus administration, a paramagnetic substance reduces T1 to < 50 ms during arterial passage (first pass). MR contrast media are characterized by a low rate of adverse effects,5 which is 2–3 times less than that of conventional X-ray contrast agents. Most MR contrast media on the market are approved only for specific applications or organ systems. The most comprehensive approval exists for the low-molecular-weight Gd chelate Magnevist (Bayer Schering Pharma) and also encompasses its use for MRA. Preparations specifically approved for MRA are Gd-DOTA (Dotarem, Guerbet), gadodiamide (Omniscan, General Electric), Gd-BOPTA (MultiHance, Bracco), and the 1 mol/L formulation of gadobutrol (Gadovist, Bayer Schering Pharma). Administration of other contrast media for MRA is considered an off-label use, placing the ultimate responsibility on the radiologist.6 Extracellular MR contrast media are chelates of Gd. The most important member of this class is Gd-DTPA, a metal–chelate complex of the rare earth element Gd and diethylenetriamine penta-acetic acid as chelating agent.7 The commercially available preparation of Gd-DTPA is a meglumine salt in aqueous solution at a concentration of 500 mmol/L. Following intravenous administration, the substance distributes in the intravascular space and then rapidly extravasates into the extracellular space. After an early peak, the plasma concentration drops by 70 % during the first 5 min of injection. Only in the brain do Gd-based contrast media have prolonged selective intravascular relaxivity because the blood–brain barrier prevents extravasation. The substance is excreted renally via glomerular filtration.7 Gd3 + ions have high magnetic moment and interact with protons, thereby considerably altering the local magnetic field. While relaxation rates, 1/T1 and 1/T2, increase in proportion to the Gd-DTPA concentration, signal intensity does not. T1-shortening predominates at lower concentrations, and signal intensity increases. In this way, the T1 relaxation time of blood is temporarily shortened from ca. 1000 ms to 30–100 ms. At high concentrations, T2shortening predominates, and signal intensity decreases.8 Another group of contrast media is low-molecularweight Gd chelates with weak protein binding. Although

Imaging Technique

weak, protein binding results in a prolonged and pronounced effect on intravascular signal intensity. These substances are eliminated in urine and bile.9 This group comprises two contrast agents approved for liver MRI: GdBOPTA (Multihance, Bracco-Byk Gulden) and Gd-EOBDTPA (Eovist, Bayer Schering Pharma). Of note are higher-concentration formulations of Gd chelates. Such formulations can be given as a more compact contrast bolus because only half the volume is needed and they offer advantages for perfusion imaging. The first Gd-based contrast medium marketed in a 1.0 mol/L concentration was gadobutrol (Gadovist, Bayer Schering Pharma).10 Another important class of contrast media is blood pool agents, currently in the clinical trial phase. These have longer intravascular residence times with little or no extravasation compared with the nonspecific Gd-based agents. This group comprises very small superparamagnetic iron oxide particles with increased T1 relaxivity as well as macromolecular and low-molecular-weight Gdbased agents, which are retained in the vasculature after intravenous injection because of strong plasma protein binding. The first preparation specifically approved for MRA was gadofosveset (Vasovist, Bayer Schering Pharma). Experimental and clinical data show that the excellent quality of contrast-enhanced 3D MRI during first pass of nonspecific Gd-based contrast media is very hard to surpass using blood pool contrast agents. Although the ultimate role of blood pool contrast media in MRA is still open, it has been suggested that the prolonged and strong enhancement of blood may cause problems because venous enhancement may interfere with evaluation of arteries.11 This problem may be overcome by the use of high-spatialresolution steady-state MRA.12

k-Space and Timing of Contrast-Enhanced MRA In an MRI experiment, the echo measured represents signal from the entire volume excited. To create an MR image, it is necessary to analyze the frequency information in a space-resolved manner. Spatial encoding is accomplished by means of a gradient field, which is weaker than the main magnetic field and is applied during readout to impart a linear increase in field strength along the gradient.13 As a result, the Larmor frequency at which protons precess increases along the axis of the gradient, e. g., the z-axis, in a space-dependent manner. In this way, only protons at a specific location can be excited and contribute to the final echo, which is why this gradient is also known as the slice-select gradient. To localize a point within the selected slice, a second gradient is applied along the y-axis. This gradient remains constant for each echo. In this way, a higher frequency is imparted to all nuclei along a line of the y-axis, and these points all have the same phase after the high-frequency pulse is switched off (phase-encoding gradient). Subsequently, a third gradient field is applied along the x-axis at the time

307

the echo is collected (readout gradient). Thus, all nuclei in the x-axis rotate faster. As a consequence, any one point within a given xy-plane has a unique phase and a frequency that depends on its location. In this way, each point in an excited volume is defined and can be localized in space. The plane defined by the x- and y-axis is known as the Fourier space or k-space, which is a raw data matrix that is converted into the image matrix using a Fourier transform. In a pulse sequence with linear phase encoding, the high and intermediate frequencies are sampled first, starting at the anterior edge of k-space (0 %). Toward the center of k-space, the low frequencies are sampled; after 50 % has been sampled, frequencies increase again toward the posterior edge of k-space (100 %). The higher frequencies determine detail resolution, while the lower frequencies define coarse structures and thus image contrast. Proper timing of image acquisition is critical for contrast-enhanced 3D MRA. To achieve optimal contrast, the central lines of k-space, which contribute most to image contrast, must be sampled at the time of peak enhancement of the target vessel. If the phase of optimal enhancement is missed, only the higher frequencies (peripheral k-space lines) contribute to image generation. Such an image has high spatial resolution but very poor contrast. In addition, premature acquisition may result in ringing artifacts. If acquisition starts too late after contrast injection, arterial evaluation may be impaired by undesired enhancement of tissues and/or veins. The time between contrast injection and initiation of data acquisition is the scan delay and varies with the patient’s circulation time. Individual determination of the scan delay is crucial to achieving optimal results.14 Individual circulation time is usually determined by intravenous injection of a test bolus of 1 mL of Gd-based contrast medium. Alternatively, automatic triggering techniques can be used.15,16 The use of MR-compatible automatic injectors ensures injection of defined amounts of contrast medium at defined rates for optimal homogeneous enhancement of the target vessel.17 The recommended injection rate is 1–5 mL/s.18 The usual clinical dose of Gd-DTPA is 0.1 mmol/kg body weight, and the total dose should not exceed 0.3 mmol/kg.19–21

Pulse Sequences MRA is performed using fast 3D gradient echo (GRE) sequences with short TR (< 5 ms) and TE (< 2 ms) together with relatively large flip angles of 30–60°. The field of view (FOV) is 360–450 mm with a matrix of 512 in the frequency-encoding direction and 126–256 in the phaseencoding direction. The volume slab thickness ranges from 60 to 120 mm, depending on the target anatomy. Images should be acquired with an effective slice thickness of less than 1.5 mm. A short scan time of < 20 s is desirable to enable image data acquisition during a single breathhold (Table 15.1). Selective fat saturation pulses are help-

308

15 Magnetic Resonance Angiography of the Abdomen

Table 15.1 Parameters for acquisition of a contrast-enhanced 3D MRA dataset Repetition time Echo time Flip angle FOV in readout direction FOV in phase-encoding direction Slice thickness Volume slab thickness Volume slab orientation Matrix resolution (frequency) Matrix resolution (phase) Fat suppression Scan time Other

< 5 ms < 2 ms 30–60° 320–450 65–75 % < 1.5 mm 60–120 mm Coronal/oblique coronal 512 40–75 % Yes < 20 s Breath-hold imaging

ful to eliminate interfering background signal,22,23 but acquisition will take 3–5 s longer. In general, one dataset is acquired after contrast injection. A biphasic protocol is recommended for MRA of the hepatic arteries and portal vein.14,24 Image subtraction can be used before other postprocessing techniques are applied to the MRA data. Parallel imaging techniques (marketed as ASSET, General Electric; SENSE, Philips; iPAT, Siemens; Speeder, Toshiba) have several advantages over conventional MRI pulse sequences. They can be used to improve spatial resolution in the same scan time or to reduce scan time while retaining resolution. In combination with phased-array multicoils and special reconstruction algorithms, parallel imaging techniques improve image quality by reducing distortion and blurring.

Postprocessing Techniques Various reconstruction techniques are available to analyze and display MRA data, the most common ones being multiplanar reconstruction (MPR), maximum intensity

a

b

projection (MIP), and different surface rendering techniques (e. g., shaded surface display, SSD). MPR is routinely used for interpreting MRA data because it is fast and straightforward and allows rapid assessment in any plane, which need not be aligned with the three axes of the volume slab. MPR images can be created using the software available on most MR scanners, without the need for a dedicated console. However, the reconstructed images may occasionally mimic pathology, which is why the radiologist should always re-examine the source images for confirmation of a suspected pathology.22,25 MIP is the most widely used postprocessing algorithm for MRA datasets. A MIP image represents the highest signal intensity voxel within each projection ray through a volume of data. These voxels are identified by means of a user-defined signal intensity threshold, below which all voxels are discarded. The retained high-intensity voxels are projected into a 2D plane. MIP algorithms have intrinsic shortcomings. Low-signal-intensity voxels (e. g., vessel edges or small-caliber vessels with slow blood flow) may be eliminated if an overlying structure has higher signal intensity, resulting in underestimation of vascular diameter and overestimation of stenosis. Vessel diameter and the degree of stenosis should therefore always be estimated from the source images.26 Surface rendering is another technique for displaying vascular structures. For this, the outer or inner surface of the vessel needs to be labeled in the dataset in a binary fashion to be recognizable by the visualization system. Threshold-based techniques are simple algorithms for segmenting vessels from background. A 3D impression is created by illuminating the object displayed with a virtual light source. Again, depending on the threshold selected, the resulting images may mimic or obscure pathology.27 Surface renderings are useful for displaying complex vascular anatomy and separating structures overlying each other,28 but they have little diagnostic value compared with other postprocessing techniques (Fig. 15.1).

c

Fig. 15.1a–c Contrast-enhanced MRA of a spontaneous splenorenal shunt. Illustration of different postprocessing techniques: MPR (a), MIP (b), surface rendering (c).

Imaging Technique

Coil Concept Selection of a suitable coil concept is critical for optimal contrast-enhanced 3D abdominal MRA covering the entire vascular tree from the proximal abdominal aorta to the pelvic arteries (ca. 60 cm in the z-direction) while at the same time ensuring adequate evaluation of the small hepatic and renal arteries. Conventional coils with a large FOV (e. g., whole-body coil) have a poor SNR, whereas a small coil such as a head coil has limited anatomic coverage. Phased-array surface coil systems combine increased anatomic coverage (up to 48 cm along the z-axis) with high SNR.29 A phased array system consists of multiple independent surface coils (receive coils), each with its own amplifier and channel. The individual coil elements are simultaneously switched for echo sampling with each element generating an image with a small FOV. The resulting images are then combined to form a composite image (array image). The good SNR of phased-array coils compared with the whole-body coil or conventional surface coils is attributable to the fact that image noise is limited to the small FOV of each individual coil element.

309

Multielement arrays of four or more coils are now available for MRA of abdominal vessels. Phased-array multicoils should provide maximum flexibility by allowing the radiologist to tailor the number of coil elements used to the individual situation. Moreover, it should be possible to use the abdominal array in combination with other phased-array systems to image the entire vascular tree comprising the abdominal, pelvic, and leg arteries if necessary.30 It is also desirable to combine individual coils of different array systems with each other. Powerful state-of-the-art MR imagers enable the combination of several surface phased-array systems with 32 or more coil elements.

Patient Examination Contrast-enhanced MRA can be performed in a short time using state-of-the-art imaging technology and a standardized protocol to optimize workflow. An example of an abdominal MRA protocol for clinical use is presented in Table 15.2.

Table 15.2 Protocol for contrast-enhanced 3D MRA 1. 2. 3.

4.

5.

6.

Patient is informed; all ferromagnetic objects and digital devices are removed from the scanner room. Cannulation of a peripheral vein (as central as possible, e. g., a cubital vein; a cannula in the back of the hand is not very secure). Alternatively, a central indwelling catheter or port-catheter system can be used after disinfection and manual rinsing with saline. Patient is positioned supine and head first using a torso or body phased-array coil; if possible, in combination with a spinal phased-array coil. Make sure the patient is comfortable. Hands close to the body without contact to exposed skin. Tattoos in the target area are covered with damp cloths to prevent burns. The cannula is connected to an MR-compatible injector via an extension tube. Localizer scans (e. g., 3 coronal, 3 sagittal, 15 axial) using, e. g., FLASH, TrueFISP, or balanced FFE sequences (short scan times, good vessel conspicuity), acquired with or without breath-hold; breath-hold acquisition preferred in patients with respiratory insufficiency (age, cardiac or pulmonary disease, ascites, etc.) to adjust acquisition of 3D data set to patient’s respiratory cycle. Selection of the slice and level for determining circulation time (e. g., TurboFLASH sequence with acquisition of one image/s; duration, 30–40 s) and test bolus administration (1 mL of paramagnetic contrast medium, 20–30 mL saline). Bolus arrival can be followed in any plane (coronal, sagittal, axial). Circulation time is determined at the level of the target vasculature: · Visceral and renal arteries—at the origins of these arteries from the abdominal aorta · Abdominal aorta—at the level of the common iliac artery · Abdominal aorta and pelvic arteries—at the level of the common femoral artery (Fig. 15.2) · In patients with aortic aneurysm – distal to the aneurysm (to ensure homogeneous enhancement). Timing run to determine scan delay after contrast injection by means of time/signal intensity curve if test bolus method is used (Fig. 15.3): · Scan delay (time between contrast injection and start of acquisition) = circulation time minus center of k-space · Circulation time: peak arterial enhancement or plateau of enhancement (1 s after first enhancement peak) · Center of k-space: depends on sequence used (ask manufacturer or look up in menu).

7. Acquisition of unenhanced 3D dataset (plane, coronal; TR, < 5 ms; TE, < 2 ms; flip angle, 40°; volume slab, 60–120 mm; matrix, 192–256 × 512; slice thickness, < 1.5 mm) with fat saturation pulse (note: will increase scan time by 3–5 s) during breath-hold (scan time, < 20 s). Prescription of scan volume according to target region: Abdominal aorta, superior mesenteric artery · craniocaudal: celiac trunk to iliac bifurcation · sagittal: parallel to abdominal aorta Renal artery · craniocaudal: celiac trunk to iliac bifurcation · sagittal: parallel to abdominal aorta · anteroposterior: including both kidneys (Fig. 15.4) Celiac trunk and portal vein · craniocaudal: celiac trunk to iliac bifurcation · paracoronal: perpendicular to portal vein · anteroposterior: including portal vein, superior mesenteric vein, and splenic vein

310

15 Magnetic Resonance Angiography of the Abdomen

Table 15.2 (continued) 8. Acquisition of contrast-enhanced datasets (same parameters as for unenhanced sequences); contrast medium dose: 0.1 mmol/kg body weight (arterial) or 0.2 mmol/kg (arterial and portal venous); saline chaser: 20–30 mL; same flow rate as for test bolus. First dataset = arterial phase · 15–20 s breathing normally Second dataset = portal venous phase · 15–20 s breathing normally Third dataset = venous phase For MR urography: additional sequences every 3–5 min until contrast medium appears in the collecting system. In patients with aneurysm: fat-saturated axial T1w images after contrast administration (to identify perfused and thrombosed lumen). 9. Move patient out of bore; ask about unpleasant sensations and intolerance reactions. 10. Automatic or manual image subtraction of datasets: Arterial phase first contrast-enhanced dataset minus nonenhanced dataset Portal venous phase second contrast-enhanced dataset minus first contrast-enhanced dataset Venous phase third contrast-enhanced dataset minus first contrast-enhanced dataset Urographic phase depends on enhancement of collecting system 11. Postprocessing of datasets by generating MIPs (overviews and targeted volumes). Vessel origins and branchings in at least two planes (coronal, axial, sagittal = celiac trunk, superior mesenteric artery) or three projections (anteroposterior, oblique, sagittal). MPRs for determining vessel diameters (aneurysm, dissection, stenosis). Surface renderings if necessary to document aberrant vascular anatomy (variants of celiac trunk and portal venous system, renal arterial supply).

a

b Fig. 15.2a, b Contrast-enhanced MRA of an infrarenal aneurysm of the abdominal aorta. a Inhomogeneous enhancement of the abdominal aorta due to inadequate bolus timing (at the level of the

renal arteries). b Homogeneous enhancement using correct bolus timing technique (at the level of the iliac arteries).

MRI Appearance of Normal Anatomy

311

a

Fig. 15.3 Signal/time diagram of a test bolus of contrast medium. Acquisition of one image/s over 40 s using a TurboFLASH sequence (image range, 25–64). Image No. 44 shows peak arterial enhancement, from which a circulation time of 19 s is calculated (image No. 44 minus image No. 25).

Fig. 15.4a, b Positioning of the volume slab for contrast-enhanced e MRA of the renal arteries. a Sagittal image showing the slab prescribed parallel to the abdominal aorta and extending from the celiac trunk to the iliac bifurcation. b Axial view showing the slab dimensions for coverage of the kidneys.

MRI Appearance of Normal Anatomy

b

· Lateral branches:

The vessels examined in the abdomen and pelvis are the abdominal aorta and its branches, the pelvic arteries and their branches, the portal venous system, and the inferior vena cava and its tributaries.

Abdominal and Pelvic Arteries The abdominal aorta extends from the aortic hiatus (at about the T12/L1 vertebral level) to the aortic bifurcation (at about the L4/L5 level), where it divides into the right and left common iliac arteries (Fig. 15.5). The aorta courses anterior to the vertebral spine and slightly to the left of the midline. The following branches arise from the abdominal aorta:

·

·

– inferior phrenic arteries (paired arteries arising from the anterior surface of the abdominal aorta at the level of the celiac trunk) – middle suprarenal arteries (paired arteries arising from the lateral wall of the abdominal aorta at the level of the superior mesenteric artery) – renal arteries (largest lateral branches arising from the lateral wall of the abdominal aorta below the origin of the superior mesenteric artery) Posterior branches: – lumbar arteries (four pairs of lumbar arteries, of which the most inferior one arises from the iliolumbar arteries, rarely from the middle sacral artery) – middle sacral artery (from the posterior surface of the abdominal aorta above the aortic bifurcation) Anterior branches: – celiac trunk (from the anterior surface of the abdominal aorta, just below the diaphragm; three

312

15 Magnetic Resonance Angiography of the Abdomen

main branches: splenic, common hepatic, left gastric) – superior mesenteric artery (from the anterior surface of the abdominal aorta, ca. 1 cm below the origin of the celiac trunk; branches: inferior pancreaticoduodenal, 10–14 jejunal and ileal branches, middle colic, right colic, and ileocolic) – inferior mesenteric artery (from the left anterolateral surface of the abdominal aorta, ca. 3 cm above the aortic bifurcation; branches: left colic, sigmoid branches, superior rectal).

Fig. 15.5 Normal appearance of the abdominal aorta and its anterior, posterior, and lateral branches on contrast-enhanced MRA.

The pelvic arteries begin with the common iliac arteries at the L4/L5 vertebral level, which divide into the external and internal iliac arteries at the sacroiliac level. The external iliac artery courses along the medial edge of the psoas muscle and gives off the deep circumflex iliac and inferior epigastric arteries above the inguinal ligament. Below the ligament, the external iliac continues as the common femoral artery, which is only a few centimeters long and divides into the superior and deep femoral arteries. The internal iliac artery gives variable rise to anterior and posterior branches: superior and inferior vesicle branches, middle rectal artery, obturator artery, internal pudendal artery, inferior gluteal artery anteriorly; superior gluteal and iliolumbar arteries, and 2–4 sacral arteries posteriorly.

Portal Venous System The portal vein arises posterior to the pancreatic head as the confluence of the splenic vein with the superior mesenteric vein and the inferior mesenteric vein (Fig. 15.6). The splenic vein courses along the posterior aspect of the pancreas and receives the short gastric, pancreatic, duodenal, and left gastroepiploic veins. The superior mesenteric vein receives blood from the ileocolic vein, the iliac and jejunal veins, the pancreatic vein, the pancreaticoduodenal veins, and the right gastroepiploic vein. The inferior mesenteric vein receives the left colic vein, sigmoid veins, and the superior rectal vein.

Abdominal and Pelvic Veins

Fig. 15.6 Normal appearance of the splenic vein, superior mesenteric vein, and portal vein on contrast-enhanced MRA of the portal venous system.

The common femoral vein becomes the external iliac vein above the inguinal ligament. The latter unites with the internal iliac vein to form the common iliac vein at the level of the sacroiliac joint. The common iliac veins join to form the inferior vena cava to the right of the abdominal aorta at the L4/L5 vertebral level. The inferior vena cava receives the paired lumbar veins, the renal veins, the right suprarenal vein, the right gonadal vein, the phrenic veins, and the liver veins. The left gonadal and suprarenal veins usually drain into the left renal vein.

MRI Appearance of Pathologic Entities

313

MRI Appearance of Pathologic Entities Abdominal and Pelvic Arteries The most common indications for MRA of the abdominal aorta and pelvic arteries are aneurysm, dissection, and stenosis or occlusion.

Aortic Aneurysm Aneurysms of the abdominal aorta are common, with a reported incidence of 1.8–6.6 % in autopsy series. Over 90 % of abdominal aneurysms arise secondary to atherosclerosis. They affect the subrenal aorta in > 90 % of patients, and 30 % have concomitant involvement of the iliac arteries. Less common causes are inflammation, trauma, syphilis, and cystic medial necrosis. A true aneurysm is defined as dilatation of the aorta to > 3 cm and involves all three wall layers (intima, media, and adventitia), distinguishing if from pseudoaneurysm or false aneurysm. Fusiform aneurysm typically occurs on the basis of atherosclerosis, while saccular aneurysm may also be of mycotic origin (less common). Most aneurysms of the abdominal aorta are asymptomatic. Aneurysms 3–6 cm in diameter increase in size by 4 mm on average every year. The risk of rupture increases in proportion to size. Autopsy series suggest that the rupture risk is ca. 50 % for aneurysms > 7 cm. Smaller aneurysms are considered to be at risk of rupture if serial studies show a significant increase in size (> 5 mm/year). Emergency interventions have a high mortality of 50 % compared with 2–3 % for elective interventions.31 Conventional catheter angiography was the gold standard for preoperative evaluation of abdominal aortic aneurysms before the advent of ultrasound and CT and accurately defines the anatomic relationship of an aneurysm to visceral, renal, and pelvic arteries. Cross-sectional imaging modalities have the added benefit of enabling simultaneous evaluation of the perfused lumen and the aortic wall including wall abnormalities (thrombus, calcification). Because of continuous coverage, spiral and multislice CT (SCT/MSCT) allow more accurate localization of an aneurysm in relation to the renal and visceral arteries compared with conventional CT and ultrasound. SCT/MSCT and contrast-enhanced 3D MRA (to be supplemented by 2D slices if necessary) are currently the preferred modalities for preoperative imaging in aortic aneurysm (Fig. 15.7). Contrast-enhanced 3D MRA even detects rare complications of aortic aneurysm such as aortocaval fistula (ca. 1 %). As a rule, the source images

Fig. 15.7a–c Infrarenal aneurysm of the abdominal aorta not extending into the iliac arteries. a Contrast-enhanced MRA. b, c Unenhanced image (b) and contrast-enhanced fat-saturated (c) T1w image show partial thrombosis.

a

b

c

314

15 Magnetic Resonance Angiography of the Abdomen

Contrast-enhanced 3D MRA is also a reliable and suitable modality for follow-up of vascular prostheses. Whether artifacts will be present and obscure the stent lumen depends on the stent material. Stents made of steel or stainless steel (Palmaz or Wall stents) produce severe artifacts on MRI, which may cause complete signal void and preclude adequate assessment of the vessel lumen (Fig. 15.8). Nitinol stents (e. g., Sinus Flex, Smart, Luminex) or endoprostheses (e. g., Talent, Hemobahn, Excluder) are associated with minimal signal voids on pulse sequences with short TEs and can be elegantly imaged using contrast-enhanced 3D MRA.36–38

Aortic Intramural Hematoma

Fig. 15.8 Abdominal and pelvic arteries after stenting. Contrastenhanced MRA does not allow evaluation of stent patency because the stents cause variable signal loss.

of the 3D MRA dataset should be supplemented by conventional multiplanar sequences (e. g., axial and coronal fat-saturated T1w GRE sequences acquired during breathhold) for accurately estimating aneurysm size because, as with conventional angiography, contrast-enhanced 3D MRA will only show the perfused lumen and may underestimate the true extent of the aneurysm if partial thrombosis is present.32,33 Contrast-enhanced 3D MRA is superior to CT because it does not involve radiation exposure and intravenous Gdbased MR contrast media are less nephrotoxic. A disadvantage is the longer examination time, which is why ultrasound and/or SCT/MSCT tend to be preferred in acute situations and suspected aortic perforation. There is no conservative treatment for aneurysm, except for control of blood pressure. Surgical repair consists in placement of a tube or bifurcation prosthesis with the aneurysm left in place and sutured over to prevent intestinal erosion. A new approach is the placement of endoluminal prostheses. The choice of the most suitable treatment strategy is based on the morphologic type of the aneurysm (according to the Heidelberg classification) and accurate information on its extent and size. Studies have shown that this information can be reliably obtained using contrast-enhanced 3D MRA.34,35

Intramural hematoma (IMH) of the aorta was first described as a special form of aortic dissection by Krukenberg in 1920.39 The exact incidence of noncommunicating IMH of the aorta is not known. Clinical and autopsy studies identified IMH without a demonstrable intimal flap in 10–14 % of patients with aortic dissection.40 IMH may be due to trauma, typically involving the aortic arch at the level of the aortic isthmus, or occur spontaneously. One third of patients have traumatic hematoma with rupture of the vasa vasorum into the outer medial coat.41 Most patients have spontaneous IMH due to various underlying mechanisms such as rupture of an atherosclerotic plaque in the intima with subsequent dissection of blood into the media or a penetrating atherosclerotic ulcer (PAU) causing spontaneous localized hemorrhage into the wall of the thoracic aorta.42,43 The spontaneous form typically occurs in older patients with longstanding arterial hypertension and severe generalized atherosclerosis. Other predisposing factors are connective tissue disorders (Marfan syndrome, Ehlers–Danlos syndrome) with medial damage.44 Little is known about how common these mechanisms are. About 50 % of patients with IMH have severe atherosclerosis. The diagnosis of IMH is based on the clinical presentation and typical imaging findings. The signs and symptoms of IMH are similar to those of classic aortic dissection. These include acute or subacute chest, abdominal, or back pain in association with hypotension, tachycardia, syncope, and a tendency to collapse.45 Although intramural hemorrhage with secondary dissection or rupture of the aorta is a life-threatening event, this condition went unnoticed in the era of conventional angiography. Today, various semi-invasive imaging modalities are available for primary evaluation and characterization of aortic diseases: transesophageal echocardiography (TEE), CT, and MRI. In TEE, echoes reflected by sequestered blood in the aortic wall point to the diagnosis.46 Advantages of TEE are its wide availability and portability. It enables evaluation of the aortic wall and lumen with high spatial resolution. Although many investigators consider TEE the method of choice for detecting IMH, it has some serious limitations:

MRI Appearance of Pathologic Entities

315

a

b Fig. 15.9a, b Subacute intramural hematoma of the aorta. T1w (a) and T2w (b) SE images show areas of high SI.

it does not enable characterization of aortic wall thickening, and intramural edema or calcifications can lead to false positive results. Moreover, TEE does not allow complete evaluation of all segments of the ascending aorta or identification of hematoma extension to or concomitant IMH of the abdominal aorta. CT depicts IMH as concentric or eccentric smooth, crescentic thickening of a longer segment of the aorta.47 Unenhanced CT scans will characteristically show a ring of increased attenuation with attenuation values of ca. 80 Hounsfield units, indicating recent hemorrhage. Other CT findings that may be seen in IMH are intense contrast enhancement and thickening of the aortic wall around the hematoma, which is likely due to adventitial inflammation. Comparing the diagnostic accuracy of different imaging modalities, von Kodolitsch and Nienaber clearly demonstrated CT to be superior to TEE in diagnosing IMH.48 At MRI, the signal intensity of extracerebral hematoma is highly variable and is mainly determined by the age of the hemorrhage and the pulse sequence used. Although the different stages of MR signal intensities are basically the same for extracerebral and intracranial hemorrhage, there is more variation, probably due to chronic recurrent bleeding. Acute IMH is hyperintense on T2w images and isointense to muscle on T1w SE images. Subacute intramural hemorrhage is of high signal intensity on both T1w and T2w images (Fig. 15.9), which is attributable to extracellular methomoglobin formation.49 The different signal intensities usually allow the radiologist to give a rough estimate of when the IMH has occurred.50 Use of electrocardiography-triggered sequences will suppress pulsation artifacts and thus enable accurate assessment of the aortic wall. Acquisition of single slices during breath-holding may be useful in some patients. Fat suppression is recom-

mended to improve demarcation from mediastinal fat, especially for evaluation of the ascending aorta. Intravenous administration of paramagnetic contrast medium is helpful to differentiate hemorrhage from inflammatory aortic disease and can be used to acquire a 3D MRA sequence, which will provide a good overview of the thoracic and abdominal aorta and their main branches. MRI has several advantages—excellent soft-tissue contrast for evaluation of the aortic wall, use of special techniques to exclude slow blood flow, and multiplanar evaluation of the aortic arch and the origins of the supra-aortic arteries arising from it. On the other hand, MRI has some important limitations compared with CT, including failure to demonstrate intimal calcifications, long examination time, and limited monitoring of vital functions in patients with an acute and potentially life-threatening condition. IMH must be differentiated from other causes of aortic wall thickening. These include thickening due to atherosclerosis, long intraluminal thrombus, chronic aortic dissection, and inflammatory disease of the aorta. Atherosclerotic Wall Thickening Atherosclerotic wall thickening differs from IMH in that the lesions typically protrude into the lumen and tend to be ill-defined. They are usually only a few millimeters in size and may occur in locations other than the predilection sites of aortic dissection. TEE continues to be the method of first choice to screen patients for the presence of atherosclerosis of the thoracic aorta in the nonacute setting.51 Long Intraluminal Thrombus Long intraluminal thrombosis of a true aneurysm can mimic thickening of the aortic wall. The localization of intimal calcifications is crucial for the differential diagno-

316

15 Magnetic Resonance Angiography of the Abdomen

sis. If there is thickening of the wall of the ascending aorta, intraluminal thrombi are extremely unlikely due to the fast blood flow in this artery.52 Chronic Aortic Dissection An occasional IMH can be difficult to differentiate from chronic aortic dissection with thrombosis of the false lumen. The decisive criterion in such cases is that in classic dissection there may be “ballooning” of the perfused false lumen with partial compression of the true lumen. The resulting decrease in the diameter of the true lumen will also persist in the presence of thrombosis of the false lumen.52 However, complete thrombosis of a false lumen over the entire length is rare. In case of IMH, the size and shape of the perfused lumen usually remain unchanged. Acute IMH can be differentiated from a completely thrombosed false lumen in chronic dissection using CT and/or MRI,53 while it is not always possible to differentiate old IMH from chronic type B dissection with thrombosis of the false lumen. The differentiation has no therapeutic or prognostic implications.

Aortitis Inflammatory conditions of the aortic wall include syphilitic and rheumatoid aortitis, giant cell aortitis, and Takayasu aortitis (Fig. 15.10), which must be differentiated from inflammatory aneurysm. Syphilitic Aortitis Syphilitic aortitis is a manifestation of tertiary syphilis and is an incidental radiologic finding occurring 15–30 years after the initial infection. Etiologically, the condition is an obliterating endarteritis of the vasa vasorum, mainly of the adventitia but also of the media. As the condition progresses, destruction of collagenous and elastic tissue leads to aortic dilatation, scar formation, and calcification. These changes give rise to typical aneurysms with calcifications of the aortic wall, mainly involving the ascending aorta and aortic arch.

Rheumatoid Aortitis Rheumatoid aortitis has been associated with different forms of arthritis, recurrent polychondritis, and inflammatory bowel disease. It predominantly affects the wall of the ascending aorta and may involve the aortic sinus and mitral valves. Giant Cell Aortitis Giant cell aortitis primarily involves not only the great arteries but also medium-sized vessels such as the supraaortic vessels of the neck. Pathologically, the condition is characterized by focal granulomatous damage of all wall layers. Complications include segmental stenosis, inflammatory wall lesions, and aortic root dilatation. Morphologically, giant cell aortitis is indistinct from Takayasu aortitis.54 Takayasu Aortitis Takayasu aortitis predominantly occurs in women in the second and third decades of life and is relatively uncommon in Western countries. It was first described by the ophthalmologist M. Takayasu in 1908, who reported a case of unusual changes of central vessels of the ocular fundus.55 The etiology is still unknown, the clinical symptoms are nonspecific, and diagnostic criteria continue to be a matter of debate.56 Our experience in a small patient population suggests that one should consider aortitis in the differential diagnosis in patients with persistent fever of unclear origin and elevated inflammatory markers. Takayasu aortitis is characterized by multisegmental involvement of the great vessels with wall thickening and development of obliterating and stenotic lesions or aneurysms. Preferred sites are the aortic arch and origins of the supra-aortic vessels.57,58 During active inflammation, which may be suggested by contrast enhancement of the aortic wall, vascular interventions have an increased complication rate and only emergency interventions should be performed.

a, b

c Fig. 15.10a–c Inflammation of the aortic wall. a High-SI areas on T2w SE image. b Low-SI areas on T1w SE image. c Postcontrast image shows intense enhancement of the thickened aortic wall.

MRI Appearance of Pathologic Entities

317

a

b Fig. 15.11a, b Inflammatory aortic aneurysm. a Thickening of the aortic wall and a perivascular cuff of inflammatory tissue. b Intense enhancement on postcontrast image.

Inflammatory Aneurysm Another condition associated with inflammatory aortic wall lesions is inflammatory aneurysm. It is characterized by wall thickening, typically in conjunction with cufflike perivascular inflammatory infiltration with intense enhancement after contrast administration (Fig. 15.11). An autoimmune reaction to extravasated fluid has been proposed as a possible underlying mechanism. Inflammatory aneurysm may affect the abdominal or thoracic aorta and accounts for ca. 4.5–12 % of all aneurysms. The differential diagnosis contains retroperitoneal fibrosis in patients presenting with nonspecific abdominal symptoms and hydronephrosis with compression and medial deviation of the ureters.59

Aortic Dissection Aortic dissection originates in the thoracic aorta in virtually all patients. The Stanford classification distinguishes two types by site of origin: type A dissection begins in the ascending aorta, type B distal to the origin of the left subclavian artery. Dissection of the abdominal aorta usually represents distal extension of dissection of the thoracic aorta. Isolated dissection of the abdominal aorta is rare (0.02–4 %) and is usually due to trauma. Acute (< 14 days) and chronic (> 14 days) dissection are distinguished. The initiating event is an intimal tear resulting in blood entering the media and creating a false lumen. The false lumen is usually distinct from the true lumen, i. e., the normal conduit of blood in the aorta, if the latter can be followed throughout the length of the dissection. In the aorta, the true lumen is smaller than the false lumen in chronic dissection and in most cases of acute dissection. The true and false lumina may be difficult to distinguish if there are several intimal tears (Fig. 15.12).

SCT/MSCT angiography and contrast-enhanced 3D MRA are the modalities of choice for the imaging of patients with aortic dissection. Both are noninvasive and usually enable accurate evaluation of the intimal flap and localization of arterial origins to the true or false lumen on multiplanar reconstructions. However, use of contrast-enhanced MRA is logistically problematic in acute emergencies. MRA is superior to CTA in determining flow velocity in the true and false lumina (cine GRE sequences) and in diagnosing secondary aortic insufficiency.

Aortic Stenosis and Occlusion Aortic stenosis develops in patients with advanced atherosclerosis. Atherosclerotic occlusions of the aorta account for 8–28 % of all atherosclerotic occlusions and predominantly occur in the infrarenal aorta and iliac arteries. Manifestations of arterial occlusive disease range from flat plaques causing only mild wall irregularities to complete occlusion. Three types of occlusion are distinguished according to Allenberg60: · type 1 (segmental type): occlusion of a short segment of the infrarenal abdominal aorta and pelvic arteries (37 %) · type 2 (so-called bifurcation type, Leriche syndrome): occlusion of the aortic bifurcation (55 %) · type 3 (high aortic thrombosis): occlusion extends to the renal artery origins (8 %). Contrast-enhanced 3D MRA will show the extent of occlusion (Fig. 15.13) as well as possible involvement of visceral and renal arteries. With automatic table movement, the peripheral arteries can be imaged to evaluate their suitability for anastomosis in patients considered for iliac or iliacofemoral bypass surgery.61,62 The results of

318

15 Magnetic Resonance Angiography of the Abdomen Fig. 15.12a–c Type B aortic dissection. Contrast-enhanced MRA of the thoracic and abdominal aorta. a MIP image. b, c The true and false lumina are difficult to distinguish on axial MPR image (b) and precontrast image (c).

b

a

c

a

b

c

Fig. 15.13a–c Contrast-enhanced MRA of the abdominal and pelvic arteries. a Segmental pelvic artery occlusion on the right side. b Occlusion of the aortic bifurcation. c High aortic thrombosis without involvement of the renal arteries.

MRA are comparable to those of diagnostic catheter angiography, which has therefore lost its significance in this setting.

Visceral Arteries Contrast-enhanced 3D MRA has replaced diagnostic catheter angiography of the proximal visceral arteries for identifying aberrant vessels (Fig. 15.14) or for mapping vascular anatomy before surgery (Fig. 15.15). The results are comparable to those of SCT/MSCT angiography. How-

MRI Appearance of Pathologic Entities

319

ever, selective catheter angiography is still superior for evaluating the peripheral vessels supplying the abdominal organs, especially those arising from the hepatic and superior mesenteric arteries. Visceral Aneurysm Most aneurysms of the visceral arteries are detected incidentally. Visceral aneurysms most commonly occur in the splenic and hepatic arteries and in the superior mesenteric artery, less commonly in the gastroduodenal and pancreaticoduodenal arteries. While aneurysms of the hepatic artery are typically due to trauma, aneurysms of the splenic artery occur secondary to acute pancreatitis or atherosclerosis. Superior mesenteric artery aneurysm is most commonly mycotic in origin or is due to acute pancreatitis. Visceral aneurysms can be diagnosed with contrast-enhanced 3D MRA or SCT/MSCT angiography. Mesenteric Ischemia Many conditions can cause mesenteric ischemia, including aneurysms, inflammatory disease, bleeding, and arteriovenous malformations. The presentation is acute or chronic. The acute form is generally due to cardiac embolism. Duplex or Doppler ultrasound is the first-line noninvasive imaging technique to diagnose mesenteric occlusion but is often limited by overlying bowel gas. Although CTA and contrast-enhanced 3D MRA provide superior evaluation of mesenteric vessels, patients with clinically suspected acute occlusive or nonocclusive mesenteric ischemia should undergo prompt selective catheter angiography and intervention if necessary. Chronic occlusion of visceral vessels develops secondary to atherosclerotic lesions in 90 % of cases. In young patients, it may also be due to inflammatory disease (Takayasu arteritis) or fibromuscular dysplasia. In chronic intestinal ischemia, contrast-enhanced 3D MRA provides information on vascular morphology.63 Initial promising results were published by Meany et al., who reported 100 % sensitivity and 95 % specificity for evaluation of the proximal visceral arteries in patients with chronic mesenteric ischemia (Figs. 15.16 and 15.17), but resolution is still too poor for evaluation of peripheral vessels.64 Although functional information can be obtained by acquiring a phase-contrast MRA sequence for flow quantification, catheter angiography continues to be the method of choice for the peripheral visceral arteries and in patients with suspected nonocclusive mesenteric ischemia.65

Fig. 15.16 Contrast-enhanced MRA of the abdominal aorta. Lateral MIP view demonstrates a stenosis of the celiac trunk (arrow).

Fig. 15.14 Aberrant right hepatic artery arising from the superior mesenteric artery (arrow).

Fig. 15.15 Contrast-enhanced MRA showing normal enhancement of the main trunk of the superior mesenteric artery and inadequate visualization of side branches.

320

15 Magnetic Resonance Angiography of the Abdomen

have atherosclerotic lesions of renal arterioles and small arteries. In detail, the epidemiologic data are as follows: · Atherosclerosis is the underlying cause of RAS in 75 % of cases. · Fibromuscular dysplasia is the cause in 15–25 %. · RAS is more common in men (60 %). · Fibromuscular dysplasia is far more common in women (80 %). · RAS is bilateral in 25 % of patients.66

Fig. 15.17 Atherosclerotic stenosis at the origin of the superior mesenteric artery (arrow).

Fig. 15.18 Bilateral double renal arteries on contrast-enhanced MRA. Image shows relevant stenosis of the upper renal artery on the right.

Renal Arteries The main indications for MRA of the renal arteries are exclusion of renal artery stenosis (RAS), evaluation of living kidney donors, and follow-up after interventions and kidney transplant. Renal Artery Stenosis Stenosis or occlusion of a renal artery can lead to renovascular hypertension and progressive renal failure. Hypertension is caused by unilateral or bilateral RAS in 2–5 % of cases. At autopsy, more than 70 % of those over 60, especially individuals with diabetes or hypertension,

Radiologic tests, in particular noninvasive ones, have a central role in diagnosing renovascular hypertension. Two established screening tests, though not imaging modalities in the true sense, are captopril renal function scintigraphy and color duplex ultrasound. The latter has been gaining in importance in recent years. The low prevalence of hemodynamically relevant RAS does not justify the use of invasive catheter angiography for screening. Apart from color duplex ultrasound and spiral/multislice CTA, contrast-enhanced 3D MRA is the most important semi-invasive modality for diagnosing renovascular hypertension and has been found to have a sensitivity of > 90 % in identifying hemodynamically relevant RAS at or near the renal artery origin.67,68 It also reveals accessory renal arteries and fibromuscular dysplasia.69 However, its spatial resolution limits its accuracy in assessing intrarenal arteries, which is why selective catheter angiography should be preferred for these indications. In general, MRA tends to overestimate the degree of RAS since vessel lumina are small, especially when high-grade or very-high-grade stenosis is present (Figs. 15.18 and 15.19). Published data on the accuracy of RAS grading appear doubtful in light of routine clinical experience. The radiologist must always look at the source images, and a meticulous technique to optimize spatial resolution and contrast enhancement is even more important than in other vascular territories. Excellent quantification of RAS is possible using phase-contrast imaging, though the technique is a little more demanding and not widely used.70 Living Kidney Donors Imaging of candidates for living kidney donation is another recent addition to the indications for contrastenhanced 3D MRA (Figs. 15.20, 15.21, 15.22). The combination of MRA, urography, and nephrography provides all the information that is needed in a living kidney donor and is also cost-effective because it can serve as a “onestop shop” and replace several conventional radiologic modalities (see Chapter 7). Initial results are promising.71,72 Renal Artery PTA and Stenting Contrast-enhanced 3D MRA is theoretically well suited for postoperative or postinterventional follow-up. However, the stents typically used in the renal arteries (Palmaz, Corinthian, Genesis, Herku-Link, Devon, AVE, etc.) cause

MRI Appearance of Pathologic Entities

321

a

b Fig. 15.19a, b Renal artery stenosis near the origin of the left renal artery. Contrast-enhanced MRA overestimates the degree of stenosis (a) compared with arterial DSA (b).

signal voids due to susceptibility effects, precluding evaluation of the stent lumen (see Aortic Aneurysm, p. 313 f). Renal Transplants Another application of contrast-enhanced 3D MRA is in evaluating the blood supply (arteries and veins including anastomoses) and perfusion of renal transplants (Figs. 15.23, 15.24, 15.25, 15.26). Scrutiny of the source images of the 3D datasets is necessary to detect localized perfusion defects in a transplant kidney and parenchymal or vascular damage secondary to trauma.73

Portal Venous System Detailed and accurate information on the portal venous system is crucial for comprehensively evaluating the vascular status and planning therapy in patients with portal hypertension. This comprises vascular anatomy, detection of abnormalities such as stenosis and thrombosis, and identification of relevant portosystemic collateral pathways. Postinterventional or postoperative follow-up MRA serves to obtain information on vascular morphology, anastomoses, and shunt patency.

Fig. 15.20 Early branching of the right renal artery close to its origin from the abdominal aorta. Also seen is a lower polar artery arising from the left common iliac artery.

Portal Veins and Portosystemic Collaterals MRI of the portal vein was performed as early as 1985 in patients with portal hypertension using SE sequences.74 In this study, the investigators rated little or no signal in the portal veins as normal and marked intraportal signal intensity as thrombosis. In a later study, abnormal signal intensity was defined as being higher than that of surrounding liver tissue on T2w images and isointense to liver on T1w images.75 Zirinsky et al. reported MRI to have higher sensitivity and specificity than CT or sonography for confirming or excluding portal vein thrombosis.76 Conventional SE sequences were also found to enable demonstration of cavernous transformation.77

Fig. 15.21 Contrast-enhanced MRA revealing three right renal arteries and two left renal arteries.

322

15 Magnetic Resonance Angiography of the Abdomen

b

a

c Fig. 15.22a–c Contrast-enhanced MRA of a left renal artery aneurysm. Illustration of different postprocessing techniques: MIP (a), axial MPR (b), and surface-rendered display (c).

Fig. 15.24 Clamp stenosis of the left external iliac artery proximal to the arterial anastomosis of a transplanted kidney (arrows).

Fig. 15.23 Normal appearance of a transplant kidney on contrastenhanced MRA. Anastomosis at the level of the left iliac bifurcation.

MRI Appearance of Pathologic Entities

323

Fig. 15.26 Hemodynamically relevant arteriovenous fistula (arrow) after biopsy. The fistula leads to premature enhancement of the pelvic veins on the left side.

v Fig. 15.25 High-grade transplant renal artery stenosis. Contrastenhanced MRA image fails to depict the residual lumen (arrow).

Early optimism about portal venous MRI was dampened when it became apparent that intraluminal dephasing phenomena in tortuous vessels can cause signal inhomogeneity and lead to misinterpretation. A reduction of the flow-related artifacts degrading SE images was achieved by using improved GRE sequences.78 On GRE images, flowing blood is hyperintense (bright), while thrombus is hypointense. Supplementing GRE sequences with bolus tracking techniques or PC MRA, one can obtain semiquantitative data on portal venous blood flow.79,80 These MRA techniques depict the portal veins directly and hence provide more information on the vascular system compared with invasive, indirect splenoportography.81,82 Contrast-enhanced 3D MRA has evolved into the method of choice for imaging the portal veins. It is superior to nonenhanced MRA because it affords a higher contrast-to-noise ratio (CNR), better spatial resolution, and a larger FOV. More recent studies suggest that contrast-enhanced 3D MRA regularly enables good evaluation of the portal venous system with high spatial resolution, reliably demonstrating portal vein thrombosis, its extent, and collateral vessels.83,84 Moreover, a biphasic imaging protocol after single contrast injection allows evaluation of the arterial and portal venous blood supply to the liver. The use of new time-resolved or MP-RAGE (magnetization-prepared rapid acquisition gradient echo) techniques has been reported to further improve vascular conspicuity by increasing intravascular signal compared with routinely used 3D FLASH. This advantage would be especially beneficial for evaluation of intrahepatic portal branches and hepatic veins.85 The left gastric vein is the most commonly encountered portosystemic collateral.86 Dilatation of this vein to > 5–6 mm is considered an indirect sign of portal hyper-

Fig. 15.27 Contrast-enhanced MRA in a patient with liver cirrhosis. Image shows dilatation of the left gastric vein (arrows) as an indirect sign of portal hypertension.

tension (Fig. 15.27). The same holds true for the short gastric veins, which drain the gastric fundus.87 Esophageal and paraesophageal varices receive blood from the left gastric vein and drain into the azygos and hemiazygos veins. The splenoportal venous axis and left renal vein are connected via the left and short gastric veins or other

324

15 Magnetic Resonance Angiography of the Abdomen

veins may not be depicted because they are frequently anterior to the imaging volume. Also common in patients with portal hypertension are retroperitoneal collateral pathways. Relevant retroperitoneal collaterals form between duodenal veins and the renal vein and/or the inferior vena cava.90 These collaterals are conspicuous because of considerable shunt volumes. Dilated retroperitoneal veins are consistently seen with contrast-enhanced 3D MRA (Fig. 15.28).

Hepatic Veins and Collaterals

Fig. 15.28 Contrast-enhanced MRA in a patient with portal hypertension. Image reveals a spontaneous splenorenal shunt with bizarre pattern of retroperitoneal collaterals.

The hepatic veins and inferior vena cava have been imaged using unenhanced TOF MRA since the early 1990s.91,92 Intrahepatic venous collaterals can form in patients with hepatic venous outflow obstruction. The major hepatic veins, which typically course obliquely upward through the liver, become dilated and have a more bizarre configuration when they are affected by occlusive disease, as in Budd–Chiari syndrome. The venovenous shunts forming between the hepatic veins have a characteristic hockey-stick configuration.93 Further venous drainage is via subcapsular veins into the azygos and hemiazygos veins. Both the large intrahepatic venovenous collateral vessels and extrahepatic shunts are readily depicted by MRA.94,95

Liver Transplant

Fig. 15.29 Contrast-enhanced MRA before liver transplant. Image shows invasion of the portal vein by a pancreatic head tumor (T).

veins draining into the splenic vein. In a study of 460 patients with portal hypertension, Kimura et al. identified gastrorenal collaterals in 18–23 % of cases, while they saw splenorenal shunts in only 7 %.88 The paraumbilical veins vary in number and course. They anastomose with the superior or inferior vena cava via the upper or lower epigastric veins. These portosystemic collaterals are detectable with unenhanced MRA techniques.89 However, the large FOV afforded by contrast-enhanced 3D MRA along the patient’s long axis improves overall evaluation compared with TOF or PC MRA, the only possible drawback being that the paraumbilical

Evaluation of arterial and portal venous anatomy is important before liver transplant because the surgical approach has to be adjusted if thrombosis and collaterals are present.96,97 A change in treatment also becomes necessary in those 5 % of transplant candidates in whom preoperative work-up reveals an occult tumor (Fig. 15.29). The advantages of contrast-enhanced 3D MRA before liver transplantation are well established.98 Moreover, MRA can be supplemented by magnetic resonance cholangiopancreatography (MRCP) for evaluation of intra- and extrahepatic bile ducts (see Chapter 2).99 Vascular complications occurring after liver transplant threaten patient and graft survival.100 Hepatic artery thrombosis is a much-feared complication (3–14 % in adults, up to 26 % in children) with ca. 75 % mortality if left untreated.101,102 Prompt and reliable diagnosis is important because the clinical symptoms are multifarious.103,104 Stenosis of the hepatic artery or portal vein is less common and less critical.105,106 Ultrasonography plays an important role in post-transplant follow-up because it is a bedside technique and generally available. It is limited because it is examiner-dependent and documentation of findings is often inadequate. A study by StaffordJohnson confirms the potential of contrast-enhanced 3D MRA in the follow-up of liver transplant recipients.107 This imaging modality also enables evaluation of the liver parenchyma and detection of extrahepatic fluid collections,

MRI Appearance of Pathologic Entities

325

a, b

c Fig. 15.30a–c Contrast-enhanced MRA of living donor liver transplant. Images show normal arterial (a), portal venous (b), and venous anastomoses (c). Focal signal void at the level of the termi-

nation of the liver veins into the inferior vena cava is an artifact caused by a metal clip (arrow).

and the findings can guide interventional procedures (Fig. 15.30).

Currently available MRA techniques do not allow direct evaluation of stent lumina because the metal alloys cause artifacts, limiting morphologic evaluation and causing errors in flow measurement.84

Transjugular Intrahepatic Portosystemic Shunt Evaluation of the portal and hepatic venous system is helpful in planning a transjugular intrahepatic portosystemic shunt (TIPS) procedure.108 Several studies have shown that prior MRA can reduce TIPS-related complications and significantly shorten the intervention.109–111 Supplementary phase-contrast sequences have been shown to enable accurate determination of blood flow direction and velocity in the portal venous system112; however, these sequences must be acquired during breath-holding. This method can also be used to assess outcome after the procedure.113 In a study of 20 patients with portal hypertension investigated before and after TIPS, Debatin et al. found a 96 % increase in portal venous blood flow after TIPS.114 The method also allows accurate flow measurement in the azygos vein. Flow is measured by acquisition of a cine phase-contrast MRA sequence in the axial plane at the level of the 6th–7th thoracic vertebral body. Electrocardiography-triggered acquisition is necessary because blood flow in the azygos vein reflects the varying pressure in the right atrium. Flow quantification in the azygos vein is clinically relevant because it serves as an indirect measure of collateral blood flow in the gastroesophageal veins, which drain into it.115 Mean azygous blood flow is ca. 90 mL/min in healthy subjects and, unsurprisingly, is significantly increased to a mean of 424 mL/min in portal hypertension. In their study, Debatin et al. showed that azygos blood flow is reduced by ca. 45 % in patients with patent TIPS and increases again when the shunt becomes occluded.114

Surgical Portosystemic Shunts MRA is also suitable to follow up patients after surgical creation of a portosystemic shunt. The potential of MRA for this indication was confirmed by several studies performed as early as the 1990s.4,116–118 Unenhanced MRA does not always enable direct, complete evaluation of anastomotic sites, which is why secondary signs have to be taken into account to determine whether or not a portosystemic shunt is patent. These include an abrupt change in diameter of the vena cava at the site of anastomosis and an atypical course of other veins such as the splenic vein. Thrombosis or occlusion of the shunt can be demonstrated by contrast-enhanced 3D MRA.84

Abdominal and Pelvic Veins The most common indications for MRA of the inferior vena cava and renal veins are evaluation for congenital anomaly, compression, and thrombus. Most imaging examinations of the pelvic veins are performed to exclude or confirm thrombosis and its sequelae.

Inferior Vena Cava and Renal Veins The incidence of inferior vena cava anomalies at autopsy is ca. 1.9–2.4 %. Retroaortic left renal veins are about as frequent.119 These anomalies generally do not cause any symptoms but should be known to the surgeon in patients

326

15 Magnetic Resonance Angiography of the Abdomen Fig. 15.31a, b Patient with known tumor of the left kidney. a Coronal TrueFISP image reveals a low-SI tumor thrombus filling the inferior vena cava and extending into the right atrium. b Cavogram confirms the MRI findings.

a

b

scheduled for retroperitoneal surgery. MRA will depict complex vascular structures and communications with other vessels, which may be obscured by overlying structures on conventional angiograms. This vascular territory can be imaged using 2D TOF or contrast-enhanced MRA. With a fast GRE sequence (TrueFISP: fast imaging with steady-state precession; balanced FFE: fast field echo), the vena cava and proximal pelvic veins can be imaged in one breath-hold. Fat saturation pulses will improve vessel conspicuity. Tumor-related obstruction of the renal veins and inferior vena cava is a frequent complication of renal cell carcinoma (RCC). Invasion of the renal veins occurs in 18–35 % of cases, invasion of the inferior vena cava in 3–16 %.120 Complete surgical resection is the only curative treatment for RCC and crucially relies on accurate information on the tumor extent, which determines the surgical approach (Fig. 15.31). As early as 1992, Arlart et al. demonstrated that the entire inferior vena cava including its termination in the right atrium can be assessed on a non-contrast-enhanced coronal 2D FLASH sequence. The authors found a sensitivity of > 90 % and a specificity of > 75 % for determining intravascular tumor extent.121 The exact extent of tumor in the renal veins is difficult to define because intravascular signal voids due to overlap from the splenic, superior mesenteric, or portal vein can give rise to misinterpretation. Interfering overlap can be avoided by acquiring unenhanced TrueFISP or balanced FFE sequences. However, these sequences cannot differentiate between intravascular tumor thrombus and appositional thrombus due to inherently limited contrast of stationary tissue. Contrast-enhanced 3D MRA usually allows identification of malignant tumor thrombi because they show the same enhancement as the tumor, while

benign thrombi do not enhance. An organized thrombus with ingrowth of vessels must be considered in the differential diagnosis. Vascularization of intravascular thrombi begins after 10–14 days (Fig. 15.32). Compression of the inferior vena cava or renal veins is usually due to retroperitoneal lymphadenopathy in patients with a metastatic, granulomatous, or lymphatic underlying disease. Other possible causes include hepatomegaly; hepatic, pancreatic, or adrenal tumors; retroperitoneal hemorrhage, inflammation, or fibrosis; and aortic aneurysm. In these cases, T1w and T2w pulse sequences acquired before and after contrast-enhanced 3D MRA will help narrow the differential diagnosis.

Pelvic Veins Pelvic vein thrombosis is a serious complication of trauma or surgery as well as during and after pregnancy. Other factors that can compromise venous drainage are congenital or acquired anomalies and compression by a pelvic mass. Historically, patients with suspected pelvic vein thrombosis were examined with invasive contrastenhanced techniques. Noninvasive tests such as Doppler and duplex ultrasound are now successfully used for assessing peripheral vessels but may occasionally have limited accuracy for pelvic vessels due to overlying air or obesity. In most cases, MR venography using 2D TOF MRA (Table 15.3) enables evaluation of the external, internal, and common iliac veins. Among the tributaries of the internal iliac veins, the superior and inferior gluteal veins are visualized in virtually all cases, the internal pudendal veins in over half of cases. Visualization of the ovarian and

MRI Appearance of Pathologic Entities

327

Fig. 15.32a, b Tumor of the right kidney with invasion of the inferior vena cava. The intravascular tumor component is of low SI on unenhanced TrueFISP image (a) and shows the same enhancement as the intrarenal tumor on contrast-enhanced MRA image (b).

a

b

testicular veins is often limited to the proximal thirds. During pregnancy and postpartum, the diameter of the ovarian veins is considerably increased and they are virtually always seen.122 Evaluation of vascular anatomy is usually possible regardless of patient position. Imaging in the supine position may occasionally impair assessment because of compression of the iliac veins. Pregnant women should be imaged in the oblique prone position so as to avoid compression of the veins of interest by the enlarged uterus, which might otherwise be misinterpreted as vascular occlusion. A signal loss in the proximal left common iliac vein may be due to compression by the right common iliac artery and should not be mistaken for a venous spur. The following are considered criteria for venous thrombosis at 2D TOF MRA123: · absence of a vein (e. g., left pelvic veins) · irregular vessel discontinuity · intraluminal signal loss (axial slices) · identification of collaterals (Fig. 15.33).

Table 15.3 Parameters for 2D TOF MRA of pelvic and ovarian veins Repetition time Echo time Flip angle FOV in readout direction FOV in phase-encoding direction Slice thickness Slice overlap No. of slices Imaging plane Matrix resolution (frequency) Matrix resolution (phase) Presaturation pulse Scan time Other

25 ms 6.72 ms 40°–60° 350 87.5 % 3–5 mm 33–50 % 100 Axial 256 75 % Saturation band superior to the imaging slice < 6 min Flow compensation

b Fig. 15.33a, b Left pelvic vein thrombosis. a 2D TOF MRA image fails to demonstrate left-sided pelvic veins. b Adjunct T1w image shows dilatation of the left external iliac vein without intraluminal signal.

a

328

15 Magnetic Resonance Angiography of the Abdomen

MR venography is also performed to assess the pelvic veins in women with suspected septic puerperal ovarian vein thrombosis (SPOVT), which is a possible cause of puerperal fever and, if left untreated, is associated with a high mortality. Ovarian vein thrombosis presents with fever and right or left lower abdominal pain, nausea, and tenderness. The differential diagnosis includes appendicitis, adnexal torsion, pyelonephritis, hematoma in the broad ligament of the uterus, and abscess formation, which is why imaging has an important role. Contrastenhanced CT was the method of choice in the 1980s. Color-coded duplex ultrasound is widely available and can be performed at the bedside, but has limited accuracy in women with postpartum obesity.124

References 1. Prince MR, Yucel EK, Kaufman JA, Harrison DC, Geller SC. Dynamic gadolinium-enhanced three-dimensional abdominal MR arteriography. J Magn Reson Imaging 1993;3(6):877–881 2. Moran PR, Moran RA, Karstaedt N. Verification and evaluation of internal flow and motion. True magnetic resonance imaging by the phase gradient modulation method. Radiology 1985;154(2): 433–441 3. Laub GA, Kaiser WA. MR angiography with gradient motion refocusing. J Comput Assist Tomogr 1988;12(3):377–382 4. Dumoulin CL, Hart HRJr. Magnetic resonance angiography. Radiology 1986;161(3):717–720 5. Runge VM. Safety of approved MR contrast media for intravenous injection. J Magn Reson Imaging 2000;12(2):205–213 6. Knopp MV, Lodemann M, Kage U, Runge VM. [Administration of MR contrast agents outside of approved indications (off-label use]. Radiologe 2001;41(3):296–302 7. Weinmann HJ, Brasch RC, Press WR, Wesbey GE. Characteristics of gadolinium-DTPA complex: a potential NMR contrast agent. AJR Am J Roentgenol 1984;142(3):619–624 8. Stehling MK, Niedermeyer M, Laub G. [Contrast medium enhanced magnetic resonance angiography. Theory, technique and practical implementation]. Radiologe 1997;37(7):501–507 9. Kirchin MA, Pirovano GP, Spinazzi A. Gadobenate dimeglumine (Gd-BOPTA). An overview. Invest Radiol 1998;33(11):798–809 10. Tombach B, Heindel W. Value of 1.0- M gadolinium chelates: review of preclinical and clinical data on gadobutrol. Eur Radiol 2002;12(6):1550–1556 11. Port M, Corot C, Rousseaux O, et al. P792: a rapid clearance blood pool agent for magnetic resonance imaging: preliminary results. MAGMA 2001;12(2-3):121–127 12. Hadizadeh DR, Gieseke J, Lohmaier SH, et al. Peripheral MR angiography with blood pool contrast agent: prospective intraindividual comparative study of high-spatial-resolution steadystate MR angiography versus standard-resolution first-pass MR angiography and DSA. Radiology 2008;249(2):701–711 13. Lauterbur PC. Magnetic resonance zeugmatography. Pure Appl Chem 1974;40(1–2):149–157 14. Kopka L, Rodenwaldt J, Vosshenrich R, et al. Hepatic blood supply: comparison of optimized dual phase contrast-enhanced three-dimensional MR angiography and digital subtraction angiography. Radiology 1999;211(1):51–58 15. Earls JP, Rofsky NM, DeCorato DR, Krinsky GA, Weinreb JC. Breath-hold single-dose gadolinium-enhanced three-dimensional MR aortography: usefulness of a timing examination and MR power injector. Radiology 1996;201(3):705–710 16. Riederer SJ, Bernstein MA, Breen JF, et al. Three-dimensional contrast-enhanced MR angiography with real-time fluoroscopic triggering: design specifications and technical reliability in 330 patient studies. Radiology 2000;215(2):584–593 17. Kopka L, Vosshenrich R, Rodenwaldt J, Grabbe E. Differences in injection rates on contrast-enhanced breath-hold three-dimensional MR angiography. AJR Am J Roentgenol 1998;170(2): 345–348

18. Frayne R, Grist TM, Swan JS, Peters DC, Korosec FR, Mistretta CA. 3D MR DSA: effects of injection protocol and image masking. J Magn Reson Imaging 2000;12(3):476–487 19. Czum JM, Ho VB, Hood MN, Foo TK, Choyke PL. Bolus-chase peripheral 3D MRA using a dual-rate contrast media injection. J Magn Reson Imaging 2000;12(5):769–775 20. Hany TF, McKinnon GC, Leung DA, Pfammatter T, Debatin JF. Optimization of contrast timing for breath-hold three-dimensional MR angiography. J Magn Reson Imaging 1997;7(3): 551–556 21. Lee VS, Rofsky NM, Krinsky GA, Stemerman DH, Weinreb JC. Single-dose breath-hold gadolinium-enhanced three-dimensional MR angiography of the renal arteries. Radiology 1999;211(1):69–78 22. Ho KY, de Haan MW, Kessels AG, Kitslaar PJ, van Engelshoven JM. Peripheral vascular tree stenoses: detection with subtracted and nonsubtracted MR angiography. Radiology 1998;206(3): 673–681 23. Siegelman ES, Charafeddine R, Stolpen AH, Axel L. Suppression of intravascular signal on fat-saturated contrast-enhanced thoracic MR arteriograms. Radiology 2000;217(1):115–118 24. Gaa J, Wendl K, Tesdal IK, et al. [Combined use of MRI and MR cholangiopancreatography and contrast enhanced dual phase 3D MR angiography in diagnosis of pancreatic tumors: initial clinical results]. Rofo 1999;170(6):528–533 25. Schuhmann D, Seemann M, Schoepf UJ, et al. [Computer-assisted diagnosis based on computer-based image interpretation and 3D-visualization]. Radiologe 1998;38(10):799–809 26. Stringer WA. MRA image production and display. Clin Neurosci 1997;4(3):110–116 27. Dillon EH, van Leeuwen MS, Fernandez MA, Mali WP. Spiral CT angiography. AJR Am J Roentgenol 1993;160(6):1273–1278 28. Hany TF, Schmidt M, Davis CP, Göhde SC, Debatin JF. Diagnostic impact of four postprocessing techniques in evaluating contrast-enhanced three-dimensional MR angiography. AJR Am J Roentgenol 1998;170(4):907–912 29. Roemer PB, Edelstein WA, Hayes CE, Souza SP, Mueller OM. The NMR phased array. Magn Reson Med 1990;16(2):192–225 30. Janka R, Fellner FA, Fellner C, et al. [A hybrid technique for the automatic floating table MRA of peripheral arteries using a dedicated phased-array coil combination]. Rofo 2000;172(5): 477–481 31. Loscalzo J, Craeger M, Dzau V. Vascular Medicine: A Textbook of Vascular Biology and Diseases. Boston: Little Brown; 1996 32. Gilfeather M, Holland GA, Siegelman ES, et al. Gadolinium-enhanced ultrafast three-dimensional spoiled gradient-echo MR imaging of the abdominal aorta and visceral and iliac vessels. Radiographics 1997;17(2):423–432 33. Yucel EK. MR angiography for evaluation of abdominal aortic aneurysm: has the time come? Radiology 1994;192(2):321–323 34. Biederer J, Link J, Steffens JC, Fronius M, Heller M. [Contrast media-enhanced 3D MR angiography before endovascular treatment of aneurysm in the abdominal aorta, iliac artery and peripheral vessels]. Rofo 2000;172(12):985–991 35. Thurnher SA, Dorffner R, Thurnher MM, et al. Evaluation of abdominal aortic aneurysm for stent-graft placement: comparison of gadolinium-enhanced MR angiography versus helical CT angiography and digital subtraction angiography. Radiology 1997;205(2):341–352 36. Baum F, Vosshenrich R, Fischer U, Castillo E, Grabbe E. [Stent artifacts in 3D MR angiography: experimental studies]. Rofo 2000;172(3):278–281 37. Hilfiker PR, Quick HH, Debatin JF. Plain and covered stent-grafts: in vitro evaluation of characteristics at three-dimensional MR angiography. Radiology 1999;211(3):693–697 38. Lenhart M, Völk M, Manke C, et al. Stent appearance at contrastenhanced MR angiography: in vitro examination with 14 stents. Radiology 2000;217(1):173–178 39. Krukenberg E. Beiträge zur Frage des Aneurysma dissecans. Beitr Pathol Anat Allg Pathol 1920;67:329–351 40. Wilson SK, Hutchins GM. Aortic dissecting aneurysms: causative factors in 204 subjects. Arch Pathol Lab Med 1982;106(4): 175–180 41. Kodolitsch Y, Spielmann RP, Petersen B, et al. [Intramural hemorrhage of the thoracic aorta as a precursor of dissection]. Z Kardiol 1995;84(11):939–946

MRI Appearance of Pathologic Entities 42. Banning AP. Aortic intramural hematoma. N Engl J Med 1997;337(20):1476–1477 43. Hayashi H, Matsuoka Y, Sakamoto I, et al. Penetrating atherosclerotic ulcer of the aorta: imaging features and disease concept. Radiographics 2000;20(4):995–1005 44. Sommer T, Abu-Ramadan D, Busch M, et al. [Intramural hematoma of the thoracic aorta: diagnostic imaging and differential diagnosis]. Rofo 1996;165(3):249–256 45. Aicher D, Graeter T, Langer F, Schäfers HJ. [Progression of an intramural hematoma to dissection]. Z Kardiol 2000;89(10): 965–968 46. Willens HJ, Kessler KM. Transesophageal echocardiography in the diagnosis of diseases of the thoracic aorta: part 1. Aortic dissection, aortic intramural hematoma, and penetrating atherosclerotic ulcer of the aorta. Chest 1999;116(6):1772–1779 47. Dieckmann C, von Kodolitsch Y, Habermann CR, et al. [ Intramural hemorrhage of the thoracic aorta]. Rofo 1998;169(4): 370–377 48. von Kodolitsch Y, Nienaber CA. [Intramural hemorrhage of the thoracic aorta: diagnosis, therapy and prognosis of 209 in vivo diagnosed cases]. Z Kardiol 1998;87(10):797–807 49. Oliver TB, Murchison JT, Reid JH. Serial MRI in the management of intramural haemorrhage of the thoracic aorta. Br J Radiol 1997;70(840):1288–1290 50. Robbins RC, McManus RP, Mitchell RP. Management of patients with intramural hematoma of the thoracic aorta. Circulation 1993;88(part 2):1–10 51. Tunick PA, Krinsky GA, Lee VS, Kronzon I. Diagnostic imaging of thoracic aortic atherosclerosis. AJR Am J Roentgenol 2000; 174(4):1119–1125 52. Wolff KA, Herold CJ, Tempany CM, Parravano JG, Zerhouni EA. Aortic dissection: atypical patterns seen at MR imaging. Radiology 1991;181(2):489–495 53. Fattori R, Nienaber CA. MRI of acute and chronic aortic pathology: pre-operative and postoperative evaluation. J Magn Reson Imaging 1999;10(5):741–750 54. Choe YH, Kim DK, Koh EM, Do YS, Lee WR. Takayasu arteritis: diagnosis with MR imaging and MR angiography in acute and chronic active stages. J Magn Reson Imaging 1999;10(5): 751–757 55. Takayasu M. A case with peculiar changes of the retinal central vessels. Acta Soc Ophthal Jpn 1908;12:554–555 56. Sharma BK, Jain S, Suri S, et al. Diagnostic criteria for Takayasu arteritis. Int J Cardiol 1996;54(suppl):141–147 57. Meller J, Grabbe E, Becker W, Vosshenrich R. Value of F-18 FDG hybrid camera PET and MRI in early takayasu aortitis. Eur Radiol 2003;13(2):400–405 58. Meller J, Strutz F, Siefker U, et al. Early diagnosis and follow-up of aortitis with [(18)F]FDG PET and MRI. Eur J Nucl Med Mol Imaging 2003;30(5):730–736 59. Vosshenrich R, Zöller G. [Radiologic diagnosis and treatment of Ormond’s disease]. Rontgenblatter 1990;43(12):489–494 60. Vollmar J, Laubach K, Campana JM. [Surgical management of chronic arterial occlusion in the aorto-iliac vascular segment]. Thoraxchir Vask Chir 1965;13(6):453–478 61. Hany TF, Debatin JF, Leung DA, Pfammatter T. Evaluation of the aortoiliac and renal arteries: comparison of breath-hold, contrast-enhanced, three-dimensional MR angiography with conventional catheter angiography. Radiology 1997;204(2): 357–362 62. Snidow JJ, Johnson MS, Harris VJ, et al. Three-dimensional gadolinium-enhanced MR angiography for aortoiliac inflow assessment plus renal artery screening in a single breath hold. Radiology 1996;198(3):725–732 63. Meaney JF. Non-invasive evaluation of the visceral arteries with magnetic resonance angiography. Eur Radiol 1999;9(7): 1267–1276 64. Meaney JF, Prince MR, Nostrant TT, Stanley JC. Gadoliniumenhanced MR angiography of visceral arteries in patients with suspected chronic mesenteric ischemia. J Magn Reson Imaging 1997;7(1):171–176 65. Li KC, Whitney WS, McDonnell CH, et al. Chronic mesenteric ischemia: evaluation with phase-contrast cine MR imaging. Radiology 1994;190(1):175–179 66. Diehm C, Allenberg JR, Nimura-Eckert K. Farbatlas der Gefäßkrankheiten. Berlin: Springer; 1999, pp.115–116

329

67. De Cobelli F, Venturini M, Vanzulli A, et al. Renal arterial stenosis: prospective comparison of color Doppler US and breathhold, three-dimensional, dynamic, gadolinium-enhanced MR angiography. Radiology 2000;214(2):373–380 68. Shetty AN, Bis KG, Kirsch M, Weintraub J, Laub G. Contrastenhanced breath-hold three-dimensional magnetic resonance angiography in the evaluation of renal arteries: optimization of technique and pitfalls. J Magn Reson Imaging 2000;12(6): 912–923 69. Bakker J, Beek FJ, Beutler JJ, et al. Renal artery stenosis and accessory renal arteries: accuracy of detection and visualization with gadolinium-enhanced breath-hold MR angiography. Radiology 1998;207(2):497–504 70. Debatin JF, Ting RH, Wegmüller H, et al. Renal artery blood flow: quantitation with phase-contrast MR imaging with and without breath holding. Radiology 1994;190(2):371–378 71. Low RN, Martinez AG, Steinberg SM, et al. Potential renal transplant donors: evaluation with gadolinium-enhanced MR angiography and MR urography. Radiology 1998;207(1):165–172 72. Winterer JT, Strey C, Wolffram C, et al. [Preoperative examination of potential kidney transplantation donors: value of gadolinium-enhanced 3D MR angiography in comparison with DSA and urography]. Rofo 2000;172(5):449–457 73. Gibson M, Cook G, Gedroyc WM. Case report: renal transplant artery stenosis—three cases where magnetic resonance angiography was superior to conventional arteriography. Br J Radiol 1995;68(805):89–92 74. Williams DM, Cho KJ, Aisen AM, Eckhauser FE. Portal hypertension evaluated by MR imaging. Radiology 1985;157(3): 703–706 75. Levy HM, Newhouse JH. MR imaging of portal vein thrombosis. AJR Am J Roentgenol 1988;151(2):283–286 76. Zirinsky K, Markisz JA, Rubenstein WA, et al. MR imaging of portal venous thrombosis: correlation with CT and sonography. AJR Am J Roentgenol 1988;150(2):283–288 77. Torres WE, Gaylord GM, Whitmire L, Chuang VP, Bernardino ME. The correlation between MR and angiography in portal hypertension. AJR Am J Roentgenol 1987;148(6):1109–1113 78. Silverman PM, Patt RH, Garra BS, et al. MR imaging of the portal venous system: value of gradient-echo imaging as an adjunct to spin-echo imaging. AJR Am J Roentgenol 1991;157(2):297–302 79. Burkart DJ, Johnson CD, Morton MJ, Ehman RL. Phase-contrast cine MR angiography in chronic liver disease. Radiology 1993; 187(2):407–412 80. Tamada T, Moriyasu F, Ono S, et al. Portal blood flow: measurement with MR imaging. Radiology 1989;173(3):639–644 81. Finn JP, Kane RA, Edelman RR, et al. Imaging of the portal venous system in patients with cirrhosis: MR angiography vs duplex Doppler sonography. AJR Am J Roentgenol 1993;161(5): 989–994 82. Hughes LA, Hartnell GG, Finn JP, et al. Time-of-flight MR angiography of the portal venous system: value compared with other imaging procedures. AJR Am J Roentgenol 1996;166(2):375–378 83. Kreft B, Strunk H, Flacke S, et al. Detection of thrombosis in the portal venous system: comparison of contrast-enhanced MR angiography with intraarterial digital subtraction angiography. Radiology 2000;216(1):86–92 84. Leyendecker JR, Rivera EJr, Washburn WK, Johnson SP, Diffin DC, Eason JD. MR angiography of the portal venous system: techniques, interpretation, and clinical applications. Radiographics 1997;17(6):1425–1443 85. Van Hoe L, De Jaegere T, Bosmans H, et al. Breath-hold contrastenhanced three-dimensional MR angiography of the abdomen: time-resolved imaging versus single-phase imaging. Radiology 2000;214(1):149–156 86. Doehner GA, Ruzicka FFJr, Rousselot LM, Hoffman G. The portal venous system: on its pathological roentgen anatomy. Radiology 1956;66(2):206–217 87. Cho KC, Patel YD, Wachsberg RH, Seeff J. Varices in portal hypertension: evaluation with CT. Radiographics 1995;15(3):609–622 88. Kimura K, Ohto M, Matsutani S, Furuse J, Hoshino K, Okuda K. Relative frequencies of portosystemic pathways and renal shunt formation through the “posterior” gastric vein: portographic study in 460 patients. Hepatology 1990;12(4 Pt 1):725–728 89. Johnson CD, Ehman RL, Rakela J, Ilstrup DM. MR angiography in portal hypertension: detection of varices and imaging techniques. J Comput Assist Tomogr 1991;15(4):578–584

330

15 Magnetic Resonance Angiography of the Abdomen

90. Ibukuro K, Tsukiyama T, Mori K, Inoue Y. Veins of Retzius at CT during arterial portography: anatomy and clinical importance. Radiology 1998;209(3):793–800 91. Arlart IP, Guhl L, Fauser L, Edelman RR, Kim D, Laub G. [MR angiography of the abdominal veins]. Radiologe 1991;31(4): 192–201 92. Finn JP, Edelman RR, Jenkins RL, et al. Liver transplantation: MR angiography with surgical validation. Radiology 1991;179(1): 265–269 93. Kane R, Eustace S. Diagnosis of Budd-Chiari syndrome: comparison between sonography and MR angiography. Radiology 1995;195(1):117–121 94. Miller WJ, Federle MP, Straub WH, Davis PL. Budd-Chiari syndrome: imaging with pathologic correlation. Abdom Imaging 1993;18(4):329–335 95. Soyer P, Rabenandrasana A, Barge J, et al. MRI of Budd-Chiari syndrome. Abdom Imaging 1994;19(4):325–329 96. Ferris JV, Marsh JW, Little AF. Presurgical evaluation of the liver transplant candidate. Radiol Clin North Am 1995;33(3):497–520 97. Pieters PC, Miller WJ, DeMeo JH. Evaluation of the portal venous system: complementary roles of invasive and noninvasive imaging strategies. Radiographics 1997;17(4):879–895 98. Glockner JF. Three-dimensional gadolinium-enhanced MR angiography: applications for abdominal imaging. Radiographics 2001;21(2):357–370 99. Reinhold C, Bret PM. Current status of MR cholangiopancreatography. AJR Am J Roentgenol 1996;166(6):1285–1295 100.Tzakis AG. The dearterialized liver graft. Semin Liver Dis 1985; 5(4):375–376 101. Holbert BL, Campbell WL, Skolnick ML. Evaluation of the transplanted liver and postoperative complications. Radiol Clin North Am 1995;33(3):521–540 102. Yokoyama I, Tzakis AG, Imventarza O, et al. Pediatric liver transplantation from neonatal donors. Transpl Int 1992;5(4): 205–208 103. Dalen K, Day DL, Ascher NL, et al. Imaging of vascular complications after hepatic transplantation. AJR Am J Roentgenol 1988;150(6):1285–1290 104. Zajko AB, Chablani V, Bron KM, Jungreis C. Hemobilia complicating transhepatic catheter drainage in liver transplant recipients: management with selective embolization. Cardiovasc Intervent Radiol 1990;13(5):285–288 105. Bechstein WO, Blumhardt G, Ringe B, et al. Surgical complications in 200 consecutive liver transplants. Transplant Proc 1987;19(5):3830–3831 106. Wozney P, Zajko AB, Bron KM, Point S, Starzl TE. Vascular complications after liver transplantation: a 5-year experience. AJR Am J Roentgenol 1986;147(4):657–663 107. Stafford-Johnson DB, Hamilton BH, Dong Q, et al. Vascular complications of liver transplantation: evaluation with gadoliniumenhanced MR angiography. Radiology 1998;207(1):153–160 108. Puttemans T, Agneessens E, Mathieu J. T.I.P.S.: follow-up imaging and revision procedure. Acta Gastroenterol Belg 2000;63(2): 174–178

109. Eustace S, Buff B, Kruskal J, et al. Magnetic resonance angiography in transjugular intrahepatic portosystemic stenting: comparison with contrast hepatic and portal venography. Eur J Radiol 1994;19(1):43–49 110. Lee JP. Variation in portal and hepatic venous anatomy as shown by magnetic resonance imaging: implication for transjugular intrahepatic portosystemic shunt. Clin Radiol 1995;50(2): 108–110 111. Müller MF, Siewert B, Stokes KR, et al. MR angiographic guidance for transjugular intrahepatic portosystemic shunt procedures. J Magn Reson Imaging 1994;4(2):145–150 112. Burkart DJ, Johnson CD, Ehman RL, Weaver AL, Ilstrup DM. Evaluation of portal venous hypertension with cine phase-contrast MR flow measurements: high association of hyperdynamic portal flow with variceal hemorrhage. Radiology 1993;188(3): 643–648 113. Thomsen C, Stahlberg F, Henriksen O. Quantification of portal venous blood flow during fasting and after a standardized meal: a MRI phase-mapping study. Eur Radiol 1993;3:242–247 114. Debatin JF, Zahner B, Meyenberger C, et al. Azygos blood flow: phase contrast quantitation in volunteers and patients with portal hypertension pre- and postintrahepatic shunt placement. Hepatology 1996;24: 1109–1115 115. Mahl TC, Groszmann RJ. Pathophysiology of portal hypertension and variceal bleeding. Surg Clin North Am 1990;70(2):251–266 116. Lewis WD, Finn JP, Jenkins RL, Carretta M, Longmaid HE, Edelman RR. Use of magnetic resonance angiography in the pretransplant evaluation of portal vein pathology. Transplantation 1993;56(1):64–68 117. Nghiem HV, Winter TCIII, Mountford MC, et al. Evaluation of the portal venous system before liver transplantation: value of phase-contrast MR angiography. AJR Am J Roentgenol 1995; 164(4):871–878 118. Wegmüller H, Vock P. [MR angiography of the abdomen]. Radiologe 1993;33(2):81–86 119. Royal SA, Callen PW. CT evaluation of anomalies of the inferior vena cava and left renal vein. AJR Am J Roentgenol 1979;132(5): 759–763 120. Roubidoux MA, Dunnick NR, Sostman HD, Leder RA. Renal carcinoma: detection of venous extension with gradient-echo MR imaging. Radiology 1992;182(1):269–272 121. Arlart IP, Guhl L, Edelman RR. [Magnetic resonance angiography of the renal veins and the vena cava inferior for the staging of renal cell carcinoma]. Rofo 1992;157(6):584–590 122. Richter CS, Duewell S, Krestin GP, et al. [Three-dimensional imaging of the pelvic veins using magnetic resonance angiography]. Rofo 1993;159(2):161–166 123. Arlart IP, Bongartz GM, Marchal G. Magnetic Resonace Angiography. Berlin: Springer; 1996, 321–336 124. Savader SJ, Otero RR, Savader BL. Puerperal ovarian vein thrombosis: evaluation with CT, US, and MR imaging. Radiology 1988;167(3):637–639

331

16

Intra-abdominal Lymph Nodes M. Taupitz and D. Beyersdorff

Introduction

Indications

The prognosis and treatment of a patient with malignant disease depend on the tumor stage, which is determined by the size of the primary tumor, the presence of metastases, and possible metastatic lymph node involvement. Tumor staging is especially relevant for predicting whether curative treatment, which usually means radical surgical resection, will be possible or whether palliative measures should be initiated. Despite its exquisite softtissue contrast, MRI unfortunately cannot differentiate between benign and malignant lymph node enlargement because normal and malignant lymph nodes have similar T1 and T2 relaxation times and proton densities.1 Nodal size therefore remains the sole MRI criterion for identifying metastatic involvement, and it is generally accepted that lymph nodes with a short-axis diameter ≥ 10 mm are likely to be malignant. Metastases in normal-sized lymph nodes escape detection with current MRI technology. Reported MRI sensitivities for detecting lymph node metastasis with currently available techniques and without specific contrast agents depend on the patient population investigated and the size threshold used and range from 0–89 % with specificities of 44–100 %.2–10 Despite these limitations, which also hamper the other crosssectional imaging modalities, MRI has some advantages over CT. These derive not from the higher soft-tissue contrast or multiplanar capability of MRI but from the fact that it is superior in defining local tumor stage, especially in patients with malignant neoplasms of the true pelvis. It has been shown that nodal metastatic spread is unlikely in tumors confined to the organ of origin but should be expected once a tumor has spread beyond that organ. A radiologist interpreting the MRI appearance of lymph nodes must therefore always take into account the local tumor stage (see Chapters 10–13). Ultra-small superparamagnetic iron oxide (USPIO) nanoparticles for intravenous injection hold promise for improving noninvasive lymph node evaluation in the future.11,12 These particles have completed clinical trials and are expected to be approved for clinical use.

Abdominal lymph node evaluation alone is not an absolute indication for MRI. Apart from lymph node size, which can also be estimated by CT, there are no proven MRI criteria for identifying metastatic abdominal lymph nodes. Suspiciously enlarged lymph nodes are accurately depicted with CT, which is the primary staging modality for most abdominal tumors, assessment of abdominal lymph nodes in patients with extra-abdominal primary tumors (e. g., malignant melanoma, breast cancer), and systemic disease (e. g., Hodgkin disease). The following presentation therefore refers to those lymph node stations that are depicted on MR images acquired for the local staging of primary tumors in the true pelvis or abdomen (Table 16.1). These are specifically tumors for which local staging by MRI is superior to CT, i. e., tumors of the female reproductive system and some urologic tumors (prostate, urinary bladder) (see Chapters 10–13). In testicular cancer, no local staging is done and the retroperitoneal lymph nodes, the first site of nodal metastasis, are assessed in the setting of an abdominal CT scan. In patients with upper abdominal malignancy, the retroperitoneal lymph nodes have to be evaluated for metastasis from tumors of the kidneys and upper urinary tract (renal cell carcinoma, transitional cell carcinoma of the renal pelvis or ureter), adrenal glands, pancreas, and stomach. Lymph nodes in the hepatoduodenal ligament and celiac region may be involved in patients with tumors of the liver and bile ducts, stomach, or pancreas. Ovarian tumors tend to metastasize to omental and mesenteric lymph nodes. All of these lymph node regions must be evaluated for lymphadenopathy in patients with malignant melanoma or lymphoma. A dedicated MRI examination for lymphadenopathy is not indicated in patients with Hodgkin or non-Hodgkin lymphoma. In these patients, the lymph nodes are only evaluated if MRI is performed to identify possible organ involvement. Imaging is indicated for follow-up and monitoring of treatment outcome after surgery and/or radiotherapy or chemotherapy. In this setting, MRI is the preferred imaging modality, especially for gynecologic tumors. The images are evaluated for possible tumor recurrence after

332

16 Intra-abdominal Lymph Nodes

Table 16.1 Indications and MRI techniques for intra-abdominal lymph nodes Indication

Sequence

Plane

Comment

Tumors of the true pelvis (uterus, in particular uterine cervix; bladder; prostate)

PD/T1w TSE

Axial Coronal (if indicated)

Evaluation of pelvic lymph nodes

T2w TSE

Axial

Evaluation for necrosis in metastatic lymph nodes

Tumors of the upper abdomen (liver, kidneys, pancreas, etc.)

T1w GRE

Axial Coronal/sagittal (if indicated)

Evaluation of retroperitoneal, mesenteric, omental lymph nodes

T2w TSE (single-shot)

Lymphoma

According to site of organ involvement

Axial Coronal/sagittal (if indicated) According to site of organ involvement

surgery or the response to radiotherapy or chemotherapy. Again, the radiologist should also note any enlarged lymph nodes. The spectrum of indications for nodal MRI is changing. State-of-the-art MRI scanners that support whole-body imaging (automatic table feed, multichannel phasedarray coil systems, parallel imaging techniques) have expanded the role of MRI in tumor evaluation and lymph node staging.13 With these new techniques, it is possible to image the area of the primary tumor, e. g., the pelvis in cervical cancer, as well as the entire abdomen to search for lymphadenopathy, especially in the retroperitoneum, with a minimum of extra time. Finally, the use of USPIO as a lymphotropic contrast agent for intravenous injection might improve the accuracy of nodal MRI in the future. One such substance has undergone extensive clinical trials and approval for clinical application is expected (see below).

Imaging Technique Within the abdomen, different imaging strategies are required for assessing lymph nodes in the upper and midabdomen as opposed to those in the pelvis. Administration of an antispasmodic agent (e. g., butylscopolamine or glucagon) is important to minimize peristaltic artifacts. A body or torso phased-array coil is recommended for assessment of both abdominal and pelvic lymph nodes.5

Imaging Planes Axial images allow good evaluation of abdominal and pelvic lymph nodes. These may be supplemented by coronal images for evaluation of retroperitoneal nodes or oblique coronal images for the nodes along the external iliac vessels; the second plane serves to define the relationship to the major vessels and to estimate the ratio of short-to-long axis diameter. The coronal plane also ena-

Lymph node evaluation only if dedicated MRI examination is performed to evaluate for organ involvement

bles very good visualization of mesenteric and omental lymph nodes. The peripancreatic or perigastric lymph nodes can be evaluated on sagittal images obtained in patients undergoing dedicated MRI of these organs. Alternatively, secondary reconstructions in any plane can be generated from a 3D dataset. Very rarely, a double oblique imaging plane may be needed to improve delineation of a suspicious lymph node (Table 16.1).

Pulse Sequences Table 16.2 lists the pulse sequences that enable good evaluation of lymph nodes in different regions. Since it is usually not necessary to obtain both T1w and T2w images for detecting enlarged lymph nodes, the weighting that affords the best anatomic detail resolution should be chosen. For a high SNR, breath-hold imaging should be performed without fat suppression; however, free-breathing T2w sequences with multiple signal averages should be acquired with fat suppression (spectral saturation or inversion prepulse) to reveal lymph nodes as very-high-signal-intensity structures. Upper abdomen: Axial T1w or T2w sequences of the kidneys, adrenal glands, pancreas, liver, and stomach usually allow good identification of enlarged lymph nodes, especially when performed during breath-holding (T1w GRE, T2w single-shot TSE). Alternatively, the upper and mid-abdomen can be depicted with excellent quality using respiratory-triggered T2w sequences (e. g., respiratory bellows, respiratory cushion, or navigator echo technique) after administration of a spasmolytic agent. Additional coronal or sagittal images should also be acquired with a breath-hold sequence to shorten scan time and eliminate motion artifacts, which may particularly degrade coronal images (e. g., T1w GRE, T2w single-shot TSE). Pelvis: Pelvic MRI should include a high-resolution sequence with T1 to proton density (PD) weighting acquired after administration of a spasmolytic agent and covering the region from the aortic bifurcation to the pelvic floor. This sequence offers excellent depiction of pelvic lymph

T2 T2

T1

* The echo time (TE) of the T1w GRE sequence can be adjusted to obtain in-phase (IP) and opposed-phase (OP) images (see Chapter 1, Table 1.3). On OP images, lymph nodes are delineated from surrounding fat by a black line. Always use slice distance of 20 % of slice thickness (distance factor, 0.2). Use of a body or torso phased-array coil is recommended for acquiring high-resolution TSE/FSE sequences and breath-hold sequences. Note: The suggested parameters are only examples and have to be adjusted for use on different brands of scanners. Parallel imaging techniques can be used to shorten scan time (for sequences with one signal average) but may come with a penalty in SNR.

No Yes 3–5 min ca. 20 s 8 8 TSE 80–120 – ~5000 Single-shot TSE (e. g., HASTE), parameters fixed

7–15

Yes/no 128 × 256 128 × 256

320 (75 %) 320 (75 %)

23 21

3 1

Yes ca. 20 s 8 4–5* GRE

165

90

–

Yes/no 128 × 256

320 (75 %)

19–23

1

No No ca. 5 min ca. 8 min 8 8 3 4 23 19 320 (75 %) 320 (75 %) 228 × 512 192 × 256 No No 3 – – – 10–15 15 TSE SE

Axial Axial (alternative) Axial (sagittal/ coronal) Axial Axial (sagittal/ coronal) PD/T1 T1

~1500 500

Flip (°) TE (ms) TR (ms) Sequence type Weighting Plane

Table 16.2 Recommended pulse sequences for MRI of intra-abdominal lymph nodes

ETL

FS

Matrix

FOV (mm)

No. of slices

No. of acquisitions

Slice thick- Scan time ness (mm)

Breathhold

Imaging Technique

333

nodes. If available, the sequence should be performed as a fast or turbo spin-echo sequence (FSE or TSE) with a short effective TE (Table 16.2).

Contrast Media Both oral and intravenous contrast media can improve delineation of abdominal and pelvic lymph nodes. One lymph-node-specific agent has completed the clinical trial phase and holds promise for improving lymph node MRI in the future (see p. 336 ff).

Oral Contrast Media The use of oral contrast medium depends on the recommendations for MRI of the primary tumor under investigation or on the region imaged and the pulse sequences used if oral contrast is given specifically to improve lymph node conspicuity. For details on timing of oral contrast administration, see Chapter 5. No oral contrast medium is needed for evaluating abdominal lymph nodes if the system hardware (1.0–1.5 T) supports fast breath-hold imaging (e. g., T1w GRE, T2w single-shot TSE), which minimizes respiratory and peristaltic artifacts. In the true pelvis, respiratory artifacts are less of a problem and the FSE or TSE sequences typically used to image this region provide excellent image quality when an antispasmodic agent has been given to reduce peristalsis. Although scanning takes a few minutes, good delineation of pelvic lymph nodes is usually achieved, except in very slender or cachectic patients, in whom the lack of intra-abdominal fat may preclude differentiation of lymph nodes from bowel loops. In these cases oral contrast administration is indicated for abdominal and pelvic scans, even with state-of-the-art MRI techniques. Image quality may be poorer on less sophisticated MR scanners with longer scan times for a complete abdominal examination because multiple signal averaging sequences are necessary. Thus, oral contrast administration is indicated for imaging of both the abdomen and the pelvis.

Nonspecific Intravenous Contrast Media There is no indication for intravenous administration of nonspecific, extracellular Gd-based MR contrast medium (e. g., Magnevist, Dotarem, Gadovist) for MR imaging of the lymph nodes. Nonspecific agents can improve differentiation of lymph nodes from vessels on T1w sequences, but this is also possible with T2w sequences. If intravenous contrast administration is indicated for evaluation of the primary tumor, the images can improve identification of necrosis in metastatic lymph nodes.

334

16 Intra-abdominal Lymph Nodes

Tissue-Specific Intravenous Contrast Media USPIO particles that accumulate in normal lymph node tissue have been extensively investigated in clinical trials but have not yet been approved for clinical use (USA: Combidex, Advanced Magnetics, Cambridge; Europe: Sinerem, Guerbet, Paris). What follows is therefore only a brief outline of the prospects these agents may offer in the future. USPIO particles have a diameter of ca. 20 nm. After injection into a peripheral vein, the particles extravasate from capillaries throughout the body into the interstitium and are taken up into the lymph fluid and transported to the lymph nodes.14–16 The particles are taken up by macrophages in normal lymph nodes, thereby causing a loss of signal by markedly shortening T2 relaxation time, while metastatic lymph nodes retain their signal intensity because they do not accumulate the particles. The optimal imaging time point is 24–36 h after contrast medium infusion. The selective signal decrease of normal lymph nodes is best appreciated on high-resolution T2*w GRE images (512 matrix). The mechanism of action is comparable to that of similar preparations used for detecting focal liver lesions (e. g., Endorem, Resovist; see Chapter 1). There will be an indication for MRI of lymph nodes as soon as a specific USPIO preparation for abdominal and pelvic lymph node imaging has been approved.

MRI Appearance of Normal Lymph Nodes The smallest lymph nodes depicted with conventional MR pulse sequences are 1.0–1.5 cm in size.8 Lymph nodes as small as ca. 3–5 mm can be identified with optimized techniques (body or torso phased-array coil, 512 matrix, 3D acquisition).5,6 Lumbar lymph nodes are usually discernible along the straight major vessels.

Since lymph nodes that are neither activated nor enlarged by metastasis have a mean diameter of only a few millimeters (3–5 mm for abdominal nodes, ca. 3 mm for pelvic nodes, measured by CT17,18), depiction by MRI is next to impossible. If they are depicted, normal lymph nodes are markedly hypointense to surrounding fat on T1w images, moderately hypointense on PD images, and isointense or moderately hyperintense on T2w images. At times, the fatty hilum is distinct from the surrounding stroma in a normal lymph node (Fig. 16.1). Retroperitoneal lymph nodes usually have an ovoid shape, while lymph nodes near the pelvic sidewalls (internal iliac nodes, obturator nodes) may appear as elongated, cordlike structures several centimeters in length (Fig. 16.2). A diagram of the abdominal and pelvic lymph node stations is presented in Fig. 16.3. Pelvic lymph nodes may be difficult to distinguish from elongated iliac vessels, especially on T1w images, because both have high signal intensity. Differentiation is improved after intravenous administration of nonspecific, Gd-based contrast medium. The problem may also arise if conventional T2w SE sequences with flow rephasing are used, which also depict both lymph nodes and vessels with high signal intensity. This problem does not occur on TSE sequences, which are increasingly being used and depict blood with low signal intensity. Differentiation is optimal on PD images with intermediate signal intensity of lymph nodes and no signal from vessels. In patients without a known primary tumor, lymph nodes with a short-axis diameter ranging between 5 and 10 mm should only be reported as such in the MRI report because this is above the average size of “normal” lymph nodes.17–19 As a general rule, a short-axis diameter of ≥ 10 mm is considered suspicious for malignancy (Fig. 16.4).

a

b Fig. 16.1a, b Normal bilateral pelvic lymph nodes at the level of the iliac bifurcation. Axial T1w (a) and T2w (b) TSE images. Straight arrows indicate two lymph nodes with a round cross-section on the right side. The nodes are hypointense to surrounding fat on T1w

image (a) and nearly isointense on T2w image (b). Curved arrow indicates a single lymph node on the left side with fatty hilum, which is clearly seen on T1w image (a). The lymph node is poorly demarcated on T2w image (b) due to chemical shift artifact.

MRI Appearance of Normal Lymph Nodes

335

a

b Fig. 16.2a, b Normal pelvic lymph nodes along the pelvic sidewall on both sides (obturator chain). Axial T1w (a) and T2w (b) TSE images (1.5 T). Elongated lymph nodes (arrows) with low SI relative to surrounding fat on T1w image (a) and near isointensity on T2w image (b).

a

!$ 4#'& 5 %4#' & 17 % 4#' &

% 4#'& Fig. 16.3 Diagram of retroperitoneal lymph nodes.

b Fig. 16.4a, b Enlarged and metastatic para-aortic lymph node below the renal pedicle in a patient with renal cell carcinoma. Axial T1w GRE (a) and T2w single-shot TSE (b) images (1.5 T). In this case, the short-axis diameter of the lymph node of ca. 1.5 cm (arrow) suggests metastatic involvement, while its SI is similar to that of the normal lymph nodes in Figs. 16.1 and 16.2.

336

16 Intra-abdominal Lymph Nodes

MRI Appearance of Abnormal Lymph Nodes For lymph node staging in general, the reader is referred to a standard reference work.20 In general, imaging findings of lymph nodes are of little relevance in therapeutic decision making. However, if imaging demonstrates enlarged lymph nodes outside the primary sites of nodal metastasis of a given primary tumor, this is an important clue and such lymph nodes should be additionally sampled for histologic work-up when lymphadenectomy is performed. Most abdominal tumors initially metastasize along the draining lymphatics, involving first the regional lymph nodes and later the distal lymph node stations. In rare cases of occlusion of the draining lymphatics by metastases, there will be atypical metastatic spread with primary involvement of distal or contralateral nodes via lymphatic collaterals. Cancer of the true pelvis (prostate, urinary bladder, uterus, cervix, upper third of the vagina) typically spreads first to the lymph node groups at the pelvic wall (obturator group, internal and external iliac nodes). In patients with a primary tumor in one of these locations, marginally enlarged lymph nodes (5–10 mm) are usually not classified as malignant unless the primary tumor extends beyond the organ of origin. If images in two planes are available (axial plus coronal or sagittal), the radiologist can assess the configuration of enlarged lymph nodes. Malignant lymph nodes tend to be more rounded, a phenomenon known as spherical transformation in radiographic lymphography. Conversely, reactively enlarged lymph nodes tend to be elongated and oval in shape. Images obtained with a high-resolution 3D technique instead of a conventional 2D sequence improve conspicuity of individual lymph nodes and facilitate determination of the short-to-long axis ratio (S/L ratio). Using a 3D MP-RAGE sequence, Jager and coworkers6 were able to identify pelvic lymph nodes as small as 3 mm in patients with cancer of the prostate or urinary bladder. The malignancy criteria used in this study were a short-axis diameter of over 8 mm and an S/L ratio of over 0.8 (rounded lymph node). However, even the optimized morphologic detail resolution afforded by the sequence used in this study resulted in sensitivities for malignant lymph nodes of only 60 % (prostate cancer) and 83 % (urinary bladder cancer), while specificity was high for both tumors (98 %). In another study investigating pelvic lymph nodes in women with cervical cancer, 75 % sensitivity and 88 % specificity were achieved using a threshold of 1.5 cm. Another criterion, asymmetry in lymph node size relative to the opposite side, also failed to identify metastatic involvement of lymph nodes < 10 mm.8 The poor sensitivity is attributable to the fact that normal-sized lymph nodes frequently contain small metastases and reactively enlarged nodes are nonmetastatic. A histopathologic study of 310 pelvic lymphadenectomy specimens from prostate cancer patients identified lymph node metasta-

ses in 40 cases (12.9 %).21 In 6 of these patients the lymph nodes were grossly involved and in 34 there were only microscopic metastases. Renal cell carcinoma tends to be associated with reactive nodal enlargement and lymph node sizes of up to ca. 2 cm, which is due to the fact that necrotic degeneration of the primary tumor is common. Studer et al.22 detected metastatic involvement in only 42 % of regional lymph nodes with diameters between 1.0 and 2.2 cm in patients with renal cell carcinoma. Large reactive lymph nodes may also be detected on dedicated MR images obtained in patients with gastric cancer.23 As just outlined, the size criterion is rather unreliable for differentiating between benign and malignant lymph nodes, and other MR criteria such as T1 or T2 signal intensities are lacking.1 Both reactively enlarged and metastatic lymph nodes have low T1 signal intensity and high T2 signal intensity relative to surrounding fat. More reliable MR criteria are only available for identifying advanced nodal metastasis, e. g., central necrosis, which is suggested by very high signal intensity on T2w images. Other findings highly suggestive of lymph node metastasis are conglomerate lymph nodes and multiple enlarged lymph nodes, especially in patients with a known primary tumor (Figs. 16.5 and 16.6). Melanotic metastases from malignant melanoma are a notable exception in terms of identification by MRI. Because of the relaxivity-shortening effects of melanin, melanotic metastases have high T1 signal intensity and are moderately hyperintense to hypointense on T2w images, depending on the amount of melanin present (Fig. 16.7)24,25. Amelanotic melanoma metastases behave in the same way as metastases from other primary tumors. Malignant systemic diseases (Hodgkin disease, nonHodgkin lymphoma) tend to involve organs and lymph nodes. If present in these patients, lymphadenopathy can be clearly identified by MRI (Fig. 16.8). MR images of the upper abdomen also depict the retrocrural area, and the radiologist should therefore also scrutinize this region for enlarged lymph nodes, which— strictly speaking—are intrathoracic lymph nodes (Fig. 16.9).

Use of Contrast Media Nonspecific contrast media do not improve the differentiation between metastatic and reactive lymph nodes because both show enhancement on postcontrast images.5 Central necrosis in a metastatic node is more conspicuous on contrast-enhanced T1w images.26 USPIO holds great promise for improving the detection of malignant lymph nodes (Figs. 16.10 and 16.11). One USPIO preparation (Sinerem, Guerbet, Paris) has undergone clinical trials for lymph node imaging in different body regions (pelvis, abdomen, mediastinum, head–neck

Use of Contrast Media

337

a

b Fig. 16.5a–c Conglomerate metastatic lymph nodes along the right external iliac artery in a patient with malignant melanoma (amelanotic) of the right leg. a Coronal T1w SE image. b Coronal T2w TSE image. c Coronal T1w SE image after IV injection of Gdbased contrast medium. Images acquired at 1.5 T. The conglomerate lymph nodes (black arrow) are of homogeneous hypointensity on T1w image (a) and of moderate hyperintensity on T2w image (b) with inhomogeneous enhancement on postcontrast image (c). Additional small common iliac node (white arrow).

c

a

b Fig. 16.6a, b Multiple enlarged retroperitoneal lymph nodes in a woman with breast cancer. a Axial T1w GRE image (IP). b Axial T2w TSE image obtained with fat suppression and navigator echo technique. Multiple enlarged para-aortic, preaortic, retrocaval, and interaortocaval lymph nodes measuring up to 2 cm in diameter. On

T1w image (a), hypointense lymph nodes are difficult to differentiate from each other and from low-SI vessel cross-sections. On T2w image (b), the lymph nodes have high SI and are clearly delineated from the vessel cross-sections. Straight arrow = aorta; curved arrow = inferior vena cava.

338

16 Intra-abdominal Lymph Nodes

a

b Fig. 16.7a–c Melanotic lymph node metastases from malignant melanoma. a Axial T1w GRE image. b Axial T2w single-shot TSE image. c Coronal T1w GRE image. Due to the presence of melanin, the para-aortic lymph node conglomerate (straight arrows) has high SI on T1w images (a, c) and low SI on T2w image (b). Also present is a low-SI retrocaval lymph node (curved arrow). High SI of the inferior vena cava (arrowhead) on T1w image (a) due to inflow effects.

c

a

b Fig. 16.8a, b Intra-abdominal lymphadenopathy in a patient with Hodgkin disease. a Axial T1w GRE image. b Axial fat-suppressed T2w TSE image. Large lymph nodes in the portocaval space with low T1 SI and high T2 SI (arrows).

region, axilla). In a study of 58 patients with urinary bladder carcinoma, USPIO was shown to increase sensitivity for metastatic lymph node detection from 76 % to 96 % compared with the size criterion, while specificity decreased slightly from 99 % to 95 %.11 It is noteworthy that metastases were identified in 10 of 12 lymph nodes

that were not pathologically enlarged (< 10 mm). Harisinghani and coworkers reported an increase in sensitivity from 35 % to 90 % when comparing USPIO-enhanced MRI with lymph node size on a node-by-node basis in patients with prostate cancer.12

Use of Contrast Media

339

a

b Fig. 16.9a, b Retrocrural lymph nodes in metastatic renal cell carcinoma. Axial T1w GRE (a) and T2w single-shot TSE (b) images. Patient had a left tumor nephrectomy with subsequent displacement of the pancreatic tail into the left renal compartment. Arrow

a

indicates a retrocrural lymph node ca. 1 cm in size on the right. Increase in size compared with prior examination (not shown) suggests a metastatic lymph node.

b Fig. 16.10a–c Normal lymph node on MR lymphography with IV infusion of ultra-small superparamagnetic iron oxide particles (USPIO). a Unenhanced axial T1w TSE image. b, c Axial images obtained with T2*w GRE sequence before (b) and 24 h after (c) IV USPIO infusion. Small, normal lymph node at the right pelvic wall (arrow) isointense to muscle on T1w image (a) and hyperintense on precontrast T2*w GRE image (b). Due to the T2- and T2*-shortening effect of the USPIO particles, there is loss of signal of the lymph node on the postcontrast image (c).

c

340

16 Intra-abdominal Lymph Nodes

c

a, b Fig. 16.11a–c Metastatic lymph node in a patient with urinary bladder carcinoma on MR lymphography performed after IV infusion of ultrasmall superparamagnetic iron oxide particles (USPIO). a Axial T1w SE image. b, c Axial T2*w GRE images obtained before (b) and 24 h after (c) IV USPIO infusion. Marginally enlarged lymph

node at the left pelvic wall (arrows) isointense to muscle on T1w image (a) and moderately hyperintense on precontrast T2*w GRE image (b). Unchanged SI after USPIO administration (c) is consistent with metastatic involvement of the lymph node; metastatic nodal tissue does not take up the contrast medium.

References

13. Schmidt GP, Schmid R, Hahn K, Reiser MF. [Whole-body MRI and PET/CT in tumor diagnosis.]. Radiologe 2004;44(11):1079–1087 14. Taupitz M, Wagner S, Hamm B. [Contrast media for magnetic resonance tomographic lymph node diagnosis (MR lymphography)]. Radiologe 1996;36(2):134–140 15. Wagner S, Pfefferer D, Ebert W, et al. Intravenous MR lymphography with superparamagnetic iron oxide particles: Experimental studies in rats and rabbits. Eur Radiol 1995;5:640–646 16. Weissleder R, Elizondo G, Wittenberg J, Lee AS, Josephson L, Brady TJ. Ultrasmall superparamagnetic iron oxide: an intravenous contrast agent for assessing lymph nodes with MR imaging. Radiology 1990;175(2):494–498 17. Dorfman RE, Alpern MB, Gross BH, Sandler MA. Upper abdominal lymph nodes: criteria for normal size determined with CT. Radiology 1991;180(2):319–322 18. Vinnicombe SJ, Norman AR, Nicolson V, Husband JE. Normal pelvic lymph nodes: evaluation with CT after bipedal lymphangiography. Radiology 1995;194(2):349–355 19. Delorme S, van Kaick G. Imaging of abdominal nodal spread in malignant disease. Eur Radiol 1996;6(3):262–274 20. Wittekind Ch, Sobin LH. TNM classification of malignant tumours. New York: Wiley-Liss; 2002 21. Epstein JI, Oesterling JE, Eggleston JC, Walsh PC. Frozen section detection of lymph node metastases in prostatic carcinoma: accuracy in grossly uninvolved pelvic lymphadenectomy specimens. J Urol 1986;136(6):1234–1237 22. Studer UE, Scherz S, Scheidegger J, et al. Enlargement of regional lymph nodes in renal cell carcinoma is often not due to metastases. J Urol 1990;144(2 Pt 1):243–245 23. Fukuya T, Honda H, Hayashi T, et al. Lymph-node metastases: efficacy for detection with helical CT in patients with gastric cancer. Radiology 1995;197(3):705–711 24. Atlas SW, Braffman BH, LoBrutto R, Elder DE, Herlyn D. Human malignant melanomas with varying degrees of melanin content in nude mice: MR imaging, histopathology, and electron paramagnetic resonance. J Comput Assist Tomogr 1990;14(4): 547–554 25. Premkumar A, Sanders L, Marincola F, Feuerstein I, Concepcion R, Schwartzentruber D. Visceral metastases from melanoma: findings on MR imaging. AJR Am J Roentgenol 1992;158(2): 293–298 26. Steinkamp HJ, Heim T, Schubeus P, Schörner W, Felix R. [The magnetic resonance tomographic differential diagnosis between reactively enlarged lymph nodes and cervical lymph node metastases]. Rofo 1992;157(4):406–413

1. Dooms GC, Hricak H, Moseley ME, Bottles K, Fisher M, Higgins CB. Characterization of lymphadenopathy by magnetic resonance relaxation times: preliminary results. Radiology 1985; 155(3):691–697 2. Beer M, Schmidt H, Riedl R. [The clinical value of preoperative staging of bladder and prostatic cancers with nuclear magnetic resonance and computerized tomography]. Urologe A 1989; 28(2):65–69 3. Hammerer P, Huland H. [Diagnosis of localized prostate cancer: screening and preoperative staging]. Urologe A 1991;30(6): 378–386 4. Hawnaur JM, Johnson RJ, Buckley CH, Tindall V, Isherwood I. Staging, volume estimation and assessment of nodal status in carcinoma of the cervix: comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994;49(7):443–452 5. Heuck A, Scheidler J, Kimmig R, et al. [Lymph node staging in cervix carcinomas: the results of high-resolution magnetic resonance tomography (MRT) with a phased-array body coil]. Rofo 1997;166(3):210–214 6. Jager GJ, Barentsz JO, Oosterhof GO, Witjes JA, Ruijs SJ. Pelvic adenopathy in prostatic and urinary bladder carcinoma: MR imaging with a three-dimensional TI-weighted magnetizationprepared-rapid gradient-echo sequence. AJR Am J Roentgenol 1996;167(6):1503–1507 7. Perrotti M, Kaufman RPJr, Jennings TA, et al. Endo-rectal coil magnetic resonance imaging in clinically localized prostate cancer: is it accurate? J Urol 1996;156(1):106–109 8. Roy C, Le Bras Y, Mangold L, et al. Small pelvic lymph node metastases: evaluation with MR imaging. Clin Radiol 1997; 52(6):437–440 9. Scheidler J, Hricak H, Yu KK, Subak L, Segal MR. Radiological evaluation of lymph node metastases in patients with cervical cancer. A meta-analysis. JAMA 1997;278(13):1096–1101 10. Wolf JS Jr, Cher M, Dall’era M, Presti JC Jr, Hricak H, Carroll PR. The use and accuracy of cross-sectional imaging and fine needle aspiration cytology for detection of pelvic lymph node metastases before radical prostatectomy. J Urol 1995;153(3 Pt 2):993–999 11. Deserno WM, Harisinghani MG, Taupitz M, et al. Urinary bladder cancer: preoperative nodal staging with ferumoxtran-10enhanced MR imaging. Radiology 2004;233(2):449–456 12. Harisinghani MG, Barentsz J, Hahn PF, et al. Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 2003;348(25):2491–2499

341

17

Abdominal MRI in Children B. Stoever

Introduction Minimizing radiation exposure is a primary concern when imaging children. Ultrasonography (US) is the first-line imaging modality for examining the abdomen in pediatric patients. In all cases requiring further diagnostic work-up, MRI should be preferred whenever it is expected to provide the same level of diagnostic information as CT.

Indications The indications for pediatric MRI are discussed in relation to the specific disease entities presented in the remainder of this chapter.

Imaging Technique Children can be examined at 0.2–0.5 T, but, as in adults, image quality is better at higher field strength (1.0–1.5 T).

Coils and Pulse Sequences A considerable size discrepancy can exist between the coil and the imaging volume in newborns and infants. Therefore, the smallest coil available to image the target anatomy should be employed to optimize the signal-to-noise ratio (SNR). To compensate for the size discrepancy and to increase image quality, the filling factor of the coil can be improved by placing saline infusion bags inside the coil together with the child, even when imaging neonates with a head or knee coil. Flexible wraparound coils are most suitable to ensure optimal visualization of very small volumes in neonates and infants. Body or torso phasedarray coil systems should only be used in older children. Conventional spin echo (SE) sequences or their rapid counterparts—fast or turbo SE sequences (FSE or TSE)—acquired with T1, T2, or proton density (PD) weighting are the mainstay of abdominal imaging in pediatric patients. The SE sequences can be supplemented by gradient echo (GRE) sequences (e. g., FLASH, FFE). Markedly shorter scan

times are achieved by using an inversion-prepared T1w GRE sequence (TurboFLASH) or a very heavily T2-weighted sequence such as single-shot TurboSE or RARE. Heavily T2weighted sequences accentuate fluid and provide useful information for tissue characterization and narrowing the differential diagnosis within a few seconds. In older children, additional information for tissue characterization can be obtained by acquisition of in-phase (IP) and opposed-phase (OP) images during breath-holds. Fat suppression techniques are especially beneficial in abdominal MRI because they reduce artifacts and improve contrast. Magnetic resonance angiography (MRA) sequences are useful for detecting abdominal vascular malformations and angiomatous tumors, and MRA is increasingly being used to assess the vascular supply of a tumor before surgery or to define vascular anatomy and identify vascular anomalies prior to transplant. More specific information on the most suitable pulse sequences will be provided with the disease entities discussed in this chapter.

Positioning Careful positioning is especially important in children to compensate for the aforementioned size discrepancy between coils and scan volumes. For most examinations, some fixation inside and outside the coil will be necessary to help prevent spontaneous movement.

Sedation Anesthesia is rarely needed for an abdominal MRI scan in children. Newborns and infants in the first 3 months of life rarely require sedation. They will be fed immediately before imaging and wrapped up warmly. In contrast, most children aged between 4 months and ca. 4–5 years need to be sedated for the examination. Various sedation regimens, including oral or rectal chloral hydrate and oral diazepam, have shown to be effective. If intravenous sedation is used, reliable venous access must be ensured throughout the examination. To minimize the period of sedation and the risk of adverse effects, including erethic states, intravenous sedation

342

17 Abdominal MRI in Children

should not be started until after the child has been positioned on the scanner table. Pediatric patients under sedation should be monitored by electrocardiography and pulse oximetry. Depending on the situation, a pediatrician or anesthetist should be present, or at least a nurse. Children will tolerate an MRI examination best if a parent or some other familiar adult stays with them in the scanner room.

Contrast Media Although no serious adverse reactions have been observed in children after intravenous administration of contrast medium, strict criteria apply and must be followed in every case. Some of the nonspecific Gd-based MR contrast media (e. g., Dotarem) are approved for all ages. The dose is 0.1 mmol Gd per kg body weight, corresponding to a dose of 0.2 mL/kg for contrast preparations formulated at the usual concentration of 0.5 mmol Gd/mL. Currently, a higher dose of 0.2 mmol Gd/kg is used only for MRA. Manual injection is recommended for babies and young children. Oral contrast media are also beneficial in pediatric patients undergoing abdominal MRI. Gd-based signal-enhancing (positive) agents are no longer on the market. The only commercially available negative oral contrast medium (eliminating signal) is a ferrite preparation (Lumirem) that is not approved for children. Chloral hydrate, which is used for sedation, often also produces good contrast of bowel segments. Whole milk can serve as a positive contrast agent and ensures good delineation of the bowel, especially on T2w images. Perfluorooctyl bromide also acts as a negative oral contrast agent and has been investigated in studies but is not approved as a contrast medium either. Since no serious adverse events are known for oral contrast media in adults, their use in children is justified in individual cases, especially in cancer patients.

Contraindications In children as in adults, a pacemaker is the most important contraindication to MRI. Other implants that may be made of ferromagnetic material include spinal stabilizers or older shunt systems for cerebrospinal fluid drainage.1 Most implants used in interventional procedures are made of nonferromagnetic materials. However, so-called umbrellas used for closure of congenital cardiac defects may be partially ferromagnetic and preclude MRI. In all other cases children with implants can be examined by MRI if the implants are made of materials approved for MRI. Metal splinters may preclude an MRI examination in children who have had an accident. Careful screening for shrapnel is warranted in children from war zones.

MRI Appearance of Pathologic Entities Gastrointestinal Tract US is the primary imaging modality for the pathomorphologic evaluation of the gastrointestinal (GI) tract and will diagnose the vast majority of congenital anomalies, infectious diseases, and tumors. MRI can supplement US if it is expected to provide additional information (Fig. 17.1). MRI is performed using SE sequences or GRE sequences with fat suppression. In older children, breath-hold T1w IP and OP images will yield excellent results. Presaturation bands are usually needed to eliminate motion artifacts. For some examinations, an antispasmodic agent may be required to suppress peristalsis. During the first year of life, MRI of the GI tract is limited by the high physiologic respiratory rate and the presence of large amounts of air throughout the bowel.

Congenital Anomalies MRI is not the first imaging test in children with a suspected congenital anomaly of the upper GI tract, which is typically diagnosed by prenatal US and confirmed after birth by US and radiography of the chest and abdomen. Intestinal Duplication MRI is rarely used to evaluate intestinal duplication, which results from incomplete separation of the intestine from the primitive neurenteric canal and is typically an incidental finding on prenatal US or presents with a mass effect after birth. MRI is occasionally used to demonstrate the spatial relationship between the mass and the bowel, thereby confirming the sonographically suspected diagnosis. In older children, bowel obstruction can also result from encasement of a bowel loop by a benign mass such as an inflammatory pseudotumor. The tumor extent and its obstructive effects can be evaluated by MRI (Fig. 17.2). Anorectal Anomalies Imperforate anus, or anal atresia, is the most common anorectal malformation, occurring in ca. 1 of 5000 births. The level of the obstruction relative to the levator ani muscle—high, intermediate, or low—can be diagnosed by US. MRI provides information on the developmental status of the sphincter muscles and will also identify fistulas, which may develop between the rectum and the urogenital tract.2,3 The anatomic changes can be diagnosed by T1w SE sequences acquired in all three orthogonal planes with the thinnest slices possible. Fluid-filled fistulous tracts are better revealed on T2w images. Axial slices for demonstration of an imperforate anus are acquired at two levels: one slice through the pubococcygeal muscle where the rectum lies behind the pros-

MRI Appearance of Pathologic Entities

343

Fig. 17.1a–c Abdominal lymphangioma in a 14-year-old girl with basal cell nevus syndrome (Gorlin syndrome) (1.5 T). Axial (a), coronal (b), and sagittal (c) single-shot T2w TSE images. Large, well-defined cystic tumor of predominantly high SI. Some clusters of small cysts are seen in the tumor periphery. The large mass occupies nearly the entire right abdomen. The organ of origin cannot be determined.

a

b

c

tate or cervix and is surrounded by the puborectalis sling of the levator ani muscle and the second slice through the ramus of the ischium and the ischial tuberosity to encompass the external anal sphincter (Fig. 17.3). MRI thus provides preoperative information on the level of the defect and its relationship to the levator ani sling as well as on the developmental status of the external anal sphincter muscle. In addition, MRI will reveal suspected fistulas and enables evaluation for concomitant anomalies, e. g., of the genital tract and spinal canal, in the same session. MRI is also indicated in children with fecal incontinence following surgical correction of imperforate anus. In this setting, unless already revealed by preoperative MRI, the examination will provide information on the developmental state of the anal sphincter and other muscles. MRI will also demonstrate anterior displacement of the rectum, which can be corrected in most cases. Displacement of the rectum can cause incontinence if the rectum lies outside the levator sling and external anal sphincter muscle or if the rectum is displaced anteriorly but still within the external sphincter.

Inflammatory Bowel Disease Crohn disease is the most common inflammatory bowel disease in childhood, with a prevalence of 1 in 5000. Children are on average 11 years old when the condition is first diagnosed. Ulcerative colitis is less common, affecting 4 in 100 000 children. The mean age at diagnosis is 10.4 years. In children, changes of the bowel wall and mucosa can be diagnosed using US or small-bowel follow-through. However, technical refinements of MRI increasingly enable localization of affected bowel segments and assessment of inflammatory activity. If an MRI examination is performed, unenhanced T1w and T2w imaging is supplemented by a fat-suppressed, contrast-enhanced T1w study using a GRE or SE/TSE sequence. A study of MRI with oral administration of polyethylene glycol in children with Crohn disease has shown that MRI criteria of disease activity—wall thickness of involved bowel, mural signal enhancement after intravenous contrast medium, and length of involved bowel segment— correlate well with the clinical activity index.4 The MRI

344

17 Abdominal MRI in Children

a

b

c Fig. 17.2a–c Inflammatory pseudotumor in a 10-year-old girl (1.5 T). a Axial T2w image. b Axial T1w image after IV administration of nonspecific contrast medium. c Coronal T2w image. All images acquired with single-shot sequence. The tumor is hypointense to

liver, has ill-defined margins, and displaces surrounding structures. There is encasement of a bowel loop by the tumor with marked wall thickening and compression of the lumen.

a Fig. 17.3a, b Anterior ectopic anus in an 8-year-old boy (1.5 T). Axial T1w SE image (a) and sagittal T2w single-shot TSE image (b). Anteriorly displaced anus and normal appearance of muscles.

b

MRI Appearance of Pathologic Entities

345

a

b Fig. 17.4a, b Pelvic abscess in a 6-year-old boy (1.5 T). Axial T1w SE images before (a) and after (b) IV contrast administration. Rounded abscess formation abutting the right ala. There is marked, irregular rim enhancement of the mass, in keeping with a more recent abscess.

differentiation between Crohn disease and ulcerative colitis is primarily based on which bowel segments are involved. An advantage of MRI is that it enables transmural assessment.5,6 MRI has a special role in evaluating complications of inflammatory bowel disease such as fistulas and abscesses. Even an abscess arising from an abdominal organ other than the bowel may occasionally be demonstrated. MRI should therefore always be preferred to CT in children. On T2w images, an abscess has a bright center surrounded by a wall of variable signal intensity. Inflammatory stranding and edema indicate an acute abscess. Delineation from surrounding tissues can be improved by administration of a contrast agent. The enhancement pattern provides clues to the age of an abscess: rim enhancement suggests an acute abscess and uniform enhancement an older abscess (Fig. 17.4). Most fistulas can be adequately assessed without contrast administration on images acquired with a fat suppression technique. A fistula is best revealed if it contains fluid, which has high signal intensity on T2w images. If no fluid is present, only a hypointense or isointense contour is seen. Supralevator fistulas are more clearly depicted than infralevator fistulas. A fistula extending to the anus is detected by an increase in anal signal intensity. A healing fistula may escape detection by MRI. Non-Hodgkin lymphoma (NHL), the most common differential diagnosis of inflammatory bowel disease, is usually suggested by an US examination and confirmed by CT. MRI is less commonly used since oral contrast media are currently not generally administered when examining children.

Hepatobiliary System As in adults, the liver parenchyma has short T1 and T2 relaxation times. Consequently, the child’s liver has higher T1 and lower T2 signal intensity than the spleen. Fatsuppressed sequences may provide additional information for characterizing focal liver lesions.

Benign Tumors Focal liver lesions are not uncommon in children.7 Hemangioma or hemangiomatosis is one of the most important benign conditions of the liver in childhood. The diagnosis can be made by US or MRI. A definitive diagnosis is possible if multiple hemangiomas can be demonstrated throughout the liver.8 Hemangiomas have a long T2 relaxation time, resulting in low signal intensity on T1w images and very high signal intensity on T2w images (Fig. 17.5). A single hemangioma is sometimes difficult to distinguish from a liver cyst on T2w images. In such cases, a RARE or TSE sequence will establish the diagnosis by depicting a cyst as a homogeneous lesion. If multiple hyperintense lesions are demonstrated throughout the liver in young infants, hemangioendotheliomatosis must be considered in the differential diagnosis. A histologic examination is required for a definitive diagnosis. If the sonographic appearance is indeterminate regarding the diagnosis of a liver abscess, MRI is the next imaging modality employed in children. The same holds true for granuloma. On postcontrast MR images, an abscess shows rim enhancement or, if the lesion is older, complete uniform enhancement. Benign liver tumors such as hamartoma are delineated from the surrounding liver as circumscribed lesions, but their MR signal characteristics do not allow specific characterization, which is only possible histologically.

346

17 Abdominal MRI in Children

a

b Fig. 17.5a, b Hemangiomatosis of the liver in a 2-year-old girl (1.5 T). Axial (a) and coronal (b) T2w single-shot TSE images. Hemangiomas predominantly located in the right hepatic lobe. All hemangio-

mas are depicted as focal lesions of high SI and are clearly demarcated from surrounding liver parenchyma.

a

b Fig. 17.6a, b Echinococcosis of the liver in a 14-year-old boy (0.23 T). Axial T1w (a) and T2w (b) images. Large lesions of low T1 SI and high T2 SI in both hepatic lobes, consistent with cysts.

Another rare benign lesion, adenoma, is hypointense on T2w images and slightly hyperintense on T1w images. MRI is rarely used to diagnose a simple liver cyst. Echinococcosis can also be diagnosed by US, but MRI may provide important information on the extent of abdominal echinococcosis and involvement of other organs (Fig. 17.6).

Malignant Tumors Two thirds of all childhood liver tumors are malignant with hepatoblastoma predominating before age 3 and hepatocellular carcinoma (HCC) becoming more common after age 5.

Hepatoblastoma The tumor is typically revealed as a well-circumscribed, isolated liver lesion with a tendency to invade the portal vein and liver veins. In conjunction with elevated serum α-fetoprotein, this MRI appearance makes the diagnosis of hepatoblastoma very likely. Hepatocellular Carcinoma HCC is more common in children with bile duct atresia or a metabolic disorder involving the liver such as cystinosis or Wilson disease. Children in whom HCC is suspected on the basis of US findings are examined by MRI. Most HCCs have irregular borders, and a tumor capsule can be demonstrated in 40 % of cases. The presence of a central scar at MRI differen-

MRI Appearance of Pathologic Entities

347

tiates HCC from hepatoblastoma. The MR signal intensity is usually nonspecific and inhomogeneous. A characteristic finding is displacement and compression of intrahepatic vessels with secondary invasion. After contrast administration, HCC enhances rapidly, but the degree of enhancement is highly variable (Fig. 17.7). Mesenchymal Sarcoma Hepatic undifferentiated mesenchymal sarcoma is a rare pediatric malignancy diagnosed between 5 and 15 years. The neoplasm is characterized by rapid growth, cystic degeneration, hemorrhage, and necrosis. Degenerative changes within the tumor have low T1 signal intensity. On T2w images, these areas appear hyperintense and typically contain low-signal-intensity septa. A fibrous pseudocapsule also has low signal intensity, while myxoid tumor components have very high signal intensity. The differential diagnosis includes metastases, lymphoma, granuloma, and hamartoma.9

a

Liver Metastases US continues to be the first-line imaging modality for demonstrating liver metastases in pediatric patients. The MR appearance of liver metastases is highly variable in children. Although they may be indistinct from normal parenchyma on T1w images, most hepatic metastases are of high signal intensity on T2w images. These include metastases from neuroblastoma, a tumor with a tendency to diffusely invade the liver in young children. Hepatic metastases must be differentiated from hamartoma, hemangioma, and abscess.

b

Angiomyolipoma A rare liver tumor in children, angiomyolipoma is depicted by MRI as a well-circumscribed lesion with high signal intensity on non-fat-suppressed T1w and T2w images. Biopsy of Liver Lesions All liver lesions in children require confirmation by biopsy even if the MR appearance and signal characteristics suggest a specific diagnosis. The only exceptions are hemangiomas and cysts.

Wilson Disease Wilson disease is a disorder of copper metabolism with deposition of ceruloplasmin–copper complexes in hepatocytes, causing inflammation, fatty transformation, fibrosis, and cirrhosis. Severe cirrhosis is associated with formation of regenerative nodules. Such nodules can accumulate excess iron (siderotic nodules) and are then seen as hypointense foci within the liver on T2w images. Fatty degeneration increases MR signal intensity on nonenhanced images. After contrast administration, inflammatory tissue enhances more intensely than normal liver, and regenerative nodules are revealed as low-signal-

c Fig. 17.7a–c Hepatocellular carcinoma in a 15-year-old boy (1.5 T). Axial fat-suppressed T2w image (a) and T1w images acquired after IV contrast administration (b and c). Ill-defined, lobulated tumor of inhomogeneous SI located medially. Tumor displaces intrahepatic vessels and enhances rapidly. Contrast enhancement is heterogeneous and more intense in the tumor periphery. Multiple metastases of variable size in the right hepatic lobe. (Images courtesy of Dr. F. Heinisch, Berlin.)

intensity lesions. It is still unclear how hepatic signal intensity is affected by copper deposits.10–12 MRI of the head is indicated to identify degenerative changes of the brain, particularly the basal ganglia, in children with Wilson disease who present with neurologic symptoms.

348

17 Abdominal MRI in Children

Hepatic Iron Overload

Trauma

Primary hemochromatosis is a rare disease in childhood and is characterized by abnormal gastrointestinal iron absorption and ferritin accumulation in the liver parenchyma, causing cell death and cirrhosis. Secondary hemochromatosis is due to increased iron intake and accumulation of hemosiderin in the cells of the mononuclear phagocyte system (MPS) in the liver, spleen, and bone marrow. The secondary form is observed in children who require repeated blood transfusions because they suffer from thalassemia or other bone marrow diseases such as myelodysplastic syndrome, dyserythropoietic anemia, or aplastic anemia. The paramagnetic effect of hemosiderin shortens T2 relaxation time, resulting in marked hypointensity of the liver, spleen, and bone marrow on T1w sequences (Fig. 17.8). T2 signal intensity of the liver and spleen is also reduced. Measurement of T1 and T2 relaxation times shows that a correlation exists between the amount of excess iron in the liver and the signal intensity difference between liver and muscle. Using this method, iron concentrations of < 100 µg/mg of liver can be differentiated from concentrations > 100 µg/mg. The method is inaccurate at higher iron concentrations.13 A noninvasive modality for monitoring hepatic iron levels is desirable in children undergoing chelation therapy with desferrioxamine. It has been shown that a good intraindividual correlation exists between hepatic iron content and ferritin or the amount of blood received. MRI appears to be more suitable for monitoring chelation therapy compared with measurement of serum ferritin. However, there exist considerable interindividual discrepancies.13,14 Results from larger studies are not available for children. Also, studies investigating whether hepatocellular iron storage can be differentiated from deposition in the MPS on the basis of effects on MR signal intensities have only been performed in adults. The same holds true for the use of MR spectroscopy in iron overload.15

CT is much faster than MRI and is therefore preferred as the second-line imaging modality in children with acute liver trauma. MRI including MRA can be used after the acute event. MRA may obviate the need for invasive angiography.

Fig. 17.8 Iron overload of the liver after multiple transfusions in a 13-year-old boy with thalassemia. Axial T1w image shows marked hypointensity of the entire liver.

Bile Ducts Biliary anomalies involving the large bile ducts or gallbladder can be diagnosed using magnetic resonance cholangiopancreatography (MRCP). With MRCP, it is straightforward to visualize choledochal cysts, which are circumscribed ductal dilatations (Fig. 17.9). The most common type is tubular dilatation of the hepatocholedochal duct, which may affect the entire duct including the bifurcation into the right and left hepatic ducts. Caroli syndrome is characterized by cystic dilatation of the intrahepatic bile ducts. These anomalies can occur in association with the common channel syndrome, which is defined as the union of the biliary and pancreatic ducts outside the duodenal wall. This union can also be demonstrated by MRCP. Gallbladder disorders are rare in children. They are easily diagnosed with US, and MRI is performed only if complications are present.

Pancreas Most children with pancreatic disease can be examined by US alone. If MRI is indicated, axial images provide the most diagnostic information. As in adults, the signal intensity of the pancreas is intermediate between that of the liver and the spleen. A normal, nondilated pancreatic duct is not visualized.

Inflammation Pancreatitis in children usually develops secondary to a congenital anomaly. Therefore, a diagnostic technique should demonstrate both the underlying anomaly and the inflammatory changes. Edema is the most important feature of early pancreatitis and is identified on MRI by reduced T1 signal intensity and increased signal on T2w images. An inflamed pancreas has blurred contours. A non-fat-suppressed SE sequence will demonstrate peripancreatic changes. With progressive inflammation, the pancreas becomes more heterogeneous in signal intensity, but this is a nonspecific finding. Complications of pancreatitis such as pseudocysts can be diagnosed sonographically. An additional MRI examination is only necessary to determine organ involvement or evaluate for suspected abscess formation or hemorrhage in children with large multilocular pseudocysts.16

MRI Appearance of Pathologic Entities

a

b

349

c

Fig. 17.9a–c Choledochal cyst in a 2-year-old girl (1.5 T). a Coronal T2w image. b, c Coronal (b) and axial (c) MRCP images. Dilatations of the choledochal and common hepatic ducts with small cystic outpouchings. (Images courtesy of Dr. M. Sinzig and Dr. H. Umschad, Klagenfurt.)

a

b Fig. 17.10a–c Recurrent pancreatitis, stenosis, and stones in the pancreatic duct in a 13-year-old girl (1.5 T). a Axial T1w image. b MRCP image. c MRCP image obtained 4 months later. Stenosis of the pancreatic duct at the junction of the head and tail with prestenotic dilatation. Additional stenoses in the tail portion of the duct. Three stones are depicted. Follow-up MRCP (c) shows deterioration with dilatation of first-order branches of the pancreatic duct in the body and tail.

c

MRCP is also increasingly used in children (Fig. 17.10). As a result of technical advances, the method now allows noninvasive detection of underlying anomalies in pancreatitis.17 It is therefore good practice to perform MRCP to rule out or confirm an underlying anomaly in all children with pancreatitis of unknown etiology.

Tumors Pancreatic tumors are rare in children. Insulinoma is a hormone-secreting islet cell tumor that may be present immediately after birth and should be suspected in newborns with severe, difficult to control hypoglycemia. In this setting, MRI may have an important role as it will

350

17 Abdominal MRI in Children

demonstrate insulinomas as small as 0.8–1 cm. These tumors are identified as very hyperintense lesions within the pancreas on T2w images, while their low signal intensity on T1w images makes them difficult to differentiate from surrounding pancreas. Surgery is facilitated if the exact tumor size can be determined by preoperative MRI. Data on the accuracy of MRI in localizing insulinoma vary. The fact that the reported accuracy of MRI is still below that of the other cross-sectional imaging modalities may also be attributable to the lack of larger studies.18 In most cases, MRI is now replaced by PET CT. Abdominal lymphoma involving the pancreas is often difficult to demonstrate by MRI because it is similar in signal intensity to the pancreas. Pancreatic involvement is suggested only indirectly by enlargement of the organ. Pancreatic carcinoma and cystadenoma are extremely rare in childhood. The MRI appearance of these malignancies is the same as in adults.

Spleen Anomalies and Rupture MRI is not needed to demonstrate splenic anomalies such as asplenism or polysplenism. Either of these anomalies is suggested by US and absence of the spleen is confirmed by scintigraphy. An ectopic spleen can be found in different locations and must be considered in the differential diagnosis whenever MRI demonstrates a smoothly marginated abdominal mass that is similar to spleen in signal intensity and the spleen is not detected in its expected location. Splenic rupture is the most common organ injury occurring in blunt abdominal trauma. This is not an indication for MRI in the acute setting, where either US alone or US in conjunction with CT is the primary imaging modality.

Splenic Tumors Most splenic cysts (Fig. 17.11) are congenital epidermoid cysts. Splenic cysts developing secondary to trauma are less common. If a purely cystic lesion is demonstrated, an echinococcal cyst must be considered in the differential diagnosis. Splenic abscess may pose a diagnostic challenge. Hematogenous spread is the cause in 85 % of cases and trauma in 15 %. MRI provides useful information if US is indeterminate and does not enable reliable differentiation between abscess and splenic infarction. However, in most cases, these two entities can be differentiated on clinical grounds. Splenic infarction can occur as a complication in children with sickle cell anemia. In the acute phase, the MRI appearance is very similar to that of the normal splenic parenchyma. In children with low T1 signal intensity of the spleen due to iron overload secondary to repeated blood transfusion, T2w images are necessary, which will reveal infarcted areas as circumscribed lesions of high signal intensity. Primary tumors such as hemangiolymphangioma and teratoma of the spleen are uncommon. Also reported have been hamartomas.19 Secondary involvement of the spleen in systemic malignancy such as Hodgkin disease, leukemia, and Langerhans cell histiocytosis is more common and can be demonstrated by MRI.

Abdominal Vascular Malformations Vascular MRI in children is performed using contrastenhanced MRA with a fast 3D GRE sequence and a contrast medium dose of 0.2 mmol/kg Gd (see Chapter 15). Vascular malposition associated with situs inversus or complex cardiac defects can be confirmed by MRA (Fig. 17.12). MRI is superior to US in evaluating generalized malformations such as lymph-hemangiomatosis (Fig. 17.13) or Takayasu arteritis.

Portal Vein Thrombosis

Fig. 17.11 Splenic cyst in an 18-year-old girl with MEN syndrome. Axial T1w image (0.23 T) shows a cyst in the center of the spleen. The cyst is slightly less hypointense than would be expected for a cyst containing watery fluid; SI is consistent with proteinaceous cyst fluid.

MRI is generally not needed in assessing portal vein thrombosis, the most common cause of a prehaptic block. However, it may be indicated if portal vein thrombosis occurs secondary to an intra-abdominal inflammatory process such as intraperitoneal abscess. The MR signal intensity of a thrombus in the portal vein depends on thrombus age. A thrombus can be demonstrated in all three orthogonal planes. Another sign is stagnant blood in the portal vein. So-called cavernous transformation is present if partial recanalization has occurred following portal vein thrombosis and multiple paraportal collaterals are present, which can be demonstrated by duplex US or MRI. Convoluted vessels of low signal intensity are seen on SE images.

MRI Appearance of Pathologic Entities

351

a

b

Fig. 17.12 Pulmonary sequester in a newborn girl (1.5 T). MRA image shows the pulmonary sequester to be supplied by an artery arising from the abdominal aorta.

Fig. 17.13a–c Lymph-hemangiomatosis in a 14-year-old boy. w a, b Axial T1w andT2w images. c Coronal T2w image. Extensive lymph-hemangiomatosis involving the gluteus muscle and the flank with posterior extension on the left side. The process continues intra-abdominally, involving the left psoas muscle and extending from the splenic hilum into the true pelvis.

The improved MRA techniques available today also enable demonstration of the sequelae of an intrahepatic block, which is most commonly due to idiopathic thrombosis or occurs secondary to cirrhosis, bile duct atresia, or a metabolic disorder such as Wilson disease. In these cases, MRI will demonstrate cavernous transformation and may occasionally also provide information on extrahepatic, perisplenic, or paraesophageal collaterals. MRA is also increasingly being used for postoperative follow-up, e. g., to assess shunt volume and flow after creation of a splenorenal shunt.20

c

Congestive hepatopathy secondary to obstruction of hepatic venous outflow (posthepatic block) is known as Budd–Chiari syndrome. A posthepatic block may be congenital or occur secondary to compression by a tumor, myelopoietic diseases, or clotting disorders. A less common cause is veno-occlusive disease after bone marrow transplantation. The latter is more difficult to demonstrate by MRI because the liver veins are patent and only small hepatic vein radicles are obliterated.

352

17 Abdominal MRI in Children

a

b Fig. 17.14a, b Right duplex kidney with reflux in a 2-year-old girl (1.5 T). a Coronal T2w single-shot TSE image. b MR urogram. There is dilatation of the ureter draining the hydronephrotic upper moiety with low ectopic insertion into the bladder.

Kidneys

Fig. 17.15 Caliceal cysts in an 11-year-old girl with nephrolithiasis (1.5 T). Coronal T2w image (MR urogram) shows a large caliceal cyst in the right kidney with marked reduction of renal parenchyma and smaller caliceal cysts in the left kidney.

Retroperitoneum Although only a little fat is present in the retroperitoneal space in children, MRI enables good visualization of retroperitoneal structures in all planes. Artifacts due to peristalsis or respiratory motion are minimal. Image quality can be improved by applying presaturation bands.

T1w images enable somewhat better corticomedullary differentiation than T2w images, but the latter are superior in delineating the kidneys from surrounding fatty tissue. The contours of the renal pelvis are also more conspicuous on T2w images, especially as the amount of fatty tissue increases with age (Fig. 17.14). Contrast medium administration is necessary to characterize intrarenal pathology, ideally by means of a dynamic contrast-enhanced study. Renal anomalies such as agenesis, hypoplasia, dysplasia, ectopia, malrotation, and abnormalities of form are diagnosed by US. MRI may be used if the US findings are inconclusive (Fig. 17.15). The same holds true for evaluating renal parenchymal disease. MRI can be used to evaluate children with multicystic renal dysplasia. A complete multicystic dysplastic kidney is associated with additional anomalies of the contralateral side in 20–30 % of cases (Fig. 17.16). Multicystic renal dysplasia is characterized by massive renal enlargement due to the presence of multiple cysts of variable size and can be demonstrated by US or MRI. MRI is only needed for identification of complications such as hemorrhage or tumor or for vascular evaluation. Multicystic renal dysplasia can be diagnosed and followed up by US. MRI does not provide additional information for differentiating the forms of polycystic kidney disease.

MRI Appearance of Pathologic Entities

353

a

b Fig. 17.16a, b Left multicystic dysplastic kidney in a newborn girl (1.5 T). Coronal single-shot T1w (a) and T2w (b) images. The left kidney is massively enlarged by the presence of multiple cysts

completely replacing normal renal structures. There is no evidence of tumor. The right kidney appears normal.

Renal Tumors

Table 17.1 SIOP staging system for Wilms tumors

Wilms tumor, or nephroblastoma, accounts for 10–12 % of all malignancies in childhood. The tumor is bilateral in ca. 15 % of cases and becomes clinically apparent at 2–3 years. Wilms tumor is rare in neonates and older children. It is associated with other congenital anomalies such as hemihypertrophy, aniridia, Beckwith–Wiedemann syndrome, and neurofibromatosis. Extrarenal Wilms tumors in perirenal location or in the pelvis are extremely rare. Nephroblastoma is a triphasic tumor with blastemal, epithelial, and mesenchymal components. Imaging has a crucial role in Wilms tumor because the decision for preoperative chemotherapy is based solely on clinical parameters and imaging findings. At the end of the first cycle, the tumor is surgically resected and the diagnosis confirmed by histologic examination. The Société Internationale d’Oncologie Pédiatrique (SIOP) and national societies (e. g., GPOH in Germany) have defined subtypes of Wilms tumor based on malignancy and response to therapy. Staging is done according to the SIOP classification system (Table 17.1). US alone may result in understaging because it may fail to accurately depict the extrarenal tumor extent, lymph node involvement, or bilateral tumors.21 MRI provides useful additional information. Large tumors appear inhomogeneous due to the presence of older hemorrhage and areas of necrosis. Necrotic tissue has low T1 signal intensity and high T2 signal intensity (Fig. 17.17). Demonstration of a pseudocapsule consisting of compressed renal parenchyma is diagnostic of nephro-

Stage 1 Stage 2 Stage 3

Stage 4 Stage 5

Tumor is confined to the kidney and is completely removed by surgery Tumor extends locally beyond the kidney but is completely removed by surgery Residual tumor is present after surgery but is confined to the abdomen; no distant metastases, no intraoperative rupture Distant metastases Both kidneys are involved

blastoma. A pseudocapsule is of low signal intensity on nonenhanced images and enhances intensely after intravenous contrast administration (Fig. 17.18). Intravenous administration of contrast medium increases the sensitivity of MRI for tumor characterization from 43 % to 58 % because the heterogeneous character of nephroblastoma is highlighted, and viable tumor can be differentiated from degenerative tissue. However, even MRI fails to detect small tumors (< 4 mm) in the contralateral kidney.22,23 MRI is more accurate than US in determining the tumor volume of large Wilms tumors. For this reason MRI is used to monitor the response to chemotherapy in children with large tumors. Intratumoral areas of necrosis induced by treatment are identified by MRI as increases in signal intensity on T2w images (Fig. 17.18). Displacement of the vena cava by nephroblastoma is identified on coronal and sagittal images throughout the course of the vein. MRI also demonstrates invasion of the

354

17 Abdominal MRI in Children

b Fig. 17.17a, b Wilms tumor in an 8-year-old girl (1.5 T). a Coronal fat-saturated T2w image. Tumor consisting of multiple nodular components in predominantly anterior location and displacing the large vessels. b Axial fat-saturated T2w image shows a tumor surrounded by a hypointense capsule and containing some necrotic foci.

a

a

c

b

Fig. 17.18a–d Wilms tumor in a 4-year-old girl during therapy (1.5 T). a, b Axial T1w images obtained before (a) and after (b) contrast administration. c, d Coronal (c) and sagittal (d) T2w singleshot TSE images. Large tumor occupying most of the left abdomen on axial images. The tumor appears heterogeneous due to numerous necrotic areas. Fig. 17.18d e

MRI Appearance of Pathologic Entities

355

renal vein or vena cava and adjacent organs. Moreover, MRI can suggest metastatic lymph node involvement but may fail to identify tumor infiltration of the renal capsule. The following entities must be considered or ruled out in the differential diagnosis: · cystic nephroblastoma · neuroblastoma · renal cell carcinoma · renal lymphoma · cystic adenoma · teratoma and hamartoma. Nephroblastomatosis Nephroblastomatosis is the persistence of multiple rests of embryonic tissue within the kidney. Small foci of metanephric tissue may persist through infancy or childhood and may be diffusely dispersed throughout the renal cortex (Fig. 17.19). Metanephric rests can give rise to Wilms tumor but usually regress spontaneously.

a

Congenital Mesoblastic Nephroma Congenital mesoblastic nephroma is a primarily benign renal tumor that tends to grow invasively and is already present at birth. The involved kidney is not functional.

b

c

d Fig. 17.18d

Fig. 17.19a–c Nephroblastomatosis in a 3-year-old girl (1.5 T). a, b Coronal T1w images obtained before (a) and after (b) contrast administration. Multiple low-SI lesions of different size on both sides without enhancement on postcontrast image (b). c Follow-up axial fat-suppressed T1w image showing regression of the lesions.

356

17 Abdominal MRI in Children

Renal Columnar Hypertropyhy The diagnosis of renal columnar hypertrophy is suggested by US. Prominent renal columns can mimic a tumor, which is why MRI is performed if follow-up US demonstrates any changes in appearance. MRI can exclude a Wilms tumor by demonstrating normal renal tissue, especially on contrast-enhanced images. Angioma and Angiomyolipoma Over 50 % of children and adolescents with tuberous cerebral sclerosis have renal angiomas or angiomyolipomas. Both tumors have a characteristic MR appearance based on their tissue composition such as fatty components in the case of angiomyolipoma. If extensive hemorrhage is present, MRI is the imaging modality of first choice before interventional angiography is performed. Lymphoproliferative Disease Renal involvement in lymphoproliferative disease is diffuse or focal. MRI is rarely indicated for a dedicated examination of the kidneys in this setting but provides more accurate information on focal renal involvement when performed for evaluation of the entire abdomen.

Diseases of the Renal Vessels Fig. 17.20 Intermittent hydronephrosis of vascular origin in a 6-year-old girl. Contrast-enhanced MRA (1.5 T). The left kidney is supplied by a large renal artery and a second artery entering the lower pole. The second artery caused intermittent postural obstruction of the ureteropelvic junction.

Like other cross-sectional imaging modalities, MRI will only reveal a solid mass in a nonfunctioning kidney, which does not enhance. Multilocular Cystic Nephroma This renal tumor consists of primitive mesenchymal tissue and contains glomeruli and tubules within fibrous cyst walls. The cysts measure a few millimeters to several centimeters. They are focal, unilateral, and do not communicate with each other. The appearance resulting from the mixed composition does not allow confident differentiation from Wilms tumor with any imaging modality. Clear Cell Tumor Clear cell tumors account for 4–6 % of all renal tumors in children. This malignant tumor is usually diagnosed later than Wilms tumor, between 3 and 5 years of age. All clear cell tumors are unilateral and cannot be differentiated from Wilms tumors on the basis of their MRI appearance. Histologic demonstration of fibrovascular septa, hyalinization, and osteosarcomatous components is necessary to establish the diagnosis. Unlike other renal tumors of childhood, a clear cell tumor can give rise to both osteolytic and osteoblastic skeletal metastases, which is why bone scintigraphy is mandatory.

In the pediatric population, renal vein thrombosis and renal infarction are most common in newborns and are typically due to hypovolemia. Renal artery occlusion is less common in this age group. MRI is rarely used in newborns for these indications because renal vein thrombosis is easily demonstrated by duplex US (Fig. 17.20). In older children, MRA is gaining in importance for demonstrating renal artery stenosis. If RAS is suggested by duplex US, MRA should be used to confirm the diagnosis. Invasive angiography is indicated if an interventional procedure is considered. MRI can also be used if US yields indeterminate findings after renal transplant (Fig. 17.21).

Adrenal Glands The adrenal glands are readily imaged with US and MRI at all ages.24 The adrenal glands are very large during embryonic life and have the adult size immediately after birth (ca. 12 mm in diameter). Up to 67 % of the volume is lost due to rapid involution in the first weeks of life.25 Adrenal hemorrhage is the most common cause of an adrenal mass in newborns (Fig. 17.22). Hemorrhage affects the right adrenal gland in 70 % of cases and both adrenals in 5–10 %. An acute hematoma is revealed by US as an echogenic mass at the upper pole of the kidney. The mass is avascular and clearly distinct from the renal vessels. After 2 weeks, the hematoma is hypoechoic and develops a capsule in the further course of organization. If serial US does not show

MRI Appearance of Pathologic Entities

357

a

b Fig. 17.21a–c Urinoma in a 14-year-old girl after kidney transplant (1.5 T). a, b Axial T1w images obtained before (a) and after (b) contrast administration. c Coronal T2w image. Leakage of contrast medium into the perirenal fluid (arrow in b) is indicative of a urinoma and differentiates urinoma from a lymphocele. The kidney graft in the left abdomen is surrounded by fluid (c).

c

a

b Fig. 17.22a, b Adrenal hemorrhage in a newborn boy (1.5 T). Coronal T1w (a) and fat-suppressed T2w (b) images. Heterogeneous formation with mass effect extending along the left kidney. Older

adrenal hemorrhage with large hyperintense areas; peripheral stripes of low SI on T1w and T2w images correspond to hemosiderin.

358

17 Abdominal MRI in Children

Table 17.2 Staging of neuroblastoma Stage 1 Stage 2 Stage 3 Stage 4 Stage 4S

Tumor is localized to the adrenal gland or sympathetic ganglia Tumor extends locally but not across the midline Tumor crosses the midline Distant metastases to bone, lymph nodes, or other tissues Primaries with metastases to liver, skin, and/or bone marrow in children < 1 year old

a

b Fig. 17.23a, b Neuroblastoma in a 5-year-old boy (1.5 T). Axial T1w image (a) and contrast-enhanced coronal T1w image (b). Images show a large neuroblastoma of relatively homogeneous SI. The tumor extends far anteriorly and displaces and encases vessels.

regression of a suspected renal hemorrhage within 4 weeks and echogenicity does not increase again, a neuroblastoma should be considered as a differential diagnosis. In such cases, MRI will be helpful. The acute stage is characterized by high T1 and T2 signal intensity, which decreases over the following 3 weeks. A larger and older

hematoma will show the characteristic hemosiderin rim on both T1w and T2w images. Thrombosis of the renal vein or vena cava may rarely occur in the presence of extensive hemorrhage and can be demonstrated by duplex US or MRI.26

Neuroblastoma Neuroblastomas constitute 10 % of all malignant tumors in children. They are the most common adrenal tumors of childhood and the most common tumors of the first years of life in general. About 5–8 % of neuroblastomas are already present at birth.27–29 Two thirds of all neuroblastomas arise from the adrenal glands and one third from the sympathetic chain. Neuroblastomas of the adrenal medulla have a better prognosis. Extra-abdominal neuroblastomas may occur in the chest, pelvis, and neck region. A tumor of uniform echogenicity demonstrated by US at the upper pole of the kidney, which displaces the kidney inferiorly and laterally but does not infiltrate the organ, is diagnostic of neuroblastoma in the newborn period and early infancy. There are four stages of neuroblastoma (Table 17.2). Stage 4S has a favorable prognosis. Seventy-six percent of neuroblastomas secrete catecholamines. Most neuroblastomas can therefore be diagnosed by laboratory tests including elevated LDH (lactate dehydrogenase) levels and sonographic demonstration of a mass in the area of the adrenal glands. MRI is the preferred second-line imaging modality unless CT is needed to demonstrate calcifications, which are characteristic of neuroblastoma and differentiate this entity from Wilms tumor (Fig. 17.23). MRI is preferable to CT because its multiplanar capability allows reliable evaluation of spinal canal invasion (Fig. 17.24). On MR images, neuroblastoma is clearly delineated from surrounding structures. Most neuroblastomas have the same signal intensity as the kidneys, and they are isointense to muscle on T1w images and hyperintense on T2w images. Their appearance is homogeneous unless hemorrhage or necrosis is present. Many neuroblastomas are rather large at the time of diagnosis, crossing the midline and invading adjacent structures. Axial and above all coronal images are suitable for defining the organ of origin, tumor extent, and relationship to surrounding structures. Intravenous contrast administration is helpful to differentiate neuroblastoma from normal renal tissue. Invasion of surrounding lymph nodes is typically identified as conglomeration. MRI provides good visualization of the great vessels without contrast administration and is superior to CT in identifying vessel displacement and invasion. MRI enables straightforward demonstration of liver metastases and the extent of metastatic spread throughout the abdomen in one session.

MRI Appearance of Pathologic Entities

359

Stage 4S disease is characterized by metastases throughout the liver (formerly known as Pepper syndrome), which are depicted as focal lesions of lower signal intensity than the liver on T1w images and higher signal on T2w images. Spinal metastases can be detected or ruled out using a RARE or TurboFLASH sequence. If there is extensive continuous extension of a neuroblastoma into the spinal canal, it is important to acquire fat-suppressed sagittal or axial SE images. Intraspinal growth is variable: it may be flat, and intraspinal lesions may skip the spinal segments directly adjacent to the main tumor. MRI is the method of choice in this setting and has replaced CT myelography. Bone marrow metastases are best appreciated on T1w SE images. If brain metastases are present, imaging must demonstrate or rule out dural involvement. In all neonates and infants presenting with opsoclonus, neuroblastoma in the typical location as well as ectopic neuroblastoma (Kinsbourne syndrome) must be ruled out or confirmed. Neuroblastomas occur in 1–3 % of children with infantile myoclonic encephalopathy. The most accurate diagnostic modality is 99Tc-MIBG scintigraphy, especially for identifying small tumors.30 The response to treatment can be monitored using MRI in conjunction with US. In most cases, MRI will demonstrate a decrease in tumor size. Alternatively, tumor regression may be suggested by a change in signal intensity, which is best appreciated on T1w images but is highly variable. A cerebral MRI examination may become necessary if complications such as cerebral infarction or subdural hematoma occur during treatment.

Ganglioneuroma Ganglioneuroma is the benign form of neuroblastoma and arises from the sympathetic ganglia or through maturation of neuroblastoma (Fig. 17.25) in the abdomen, chest, or true pelvis. It is usually diagnosed in older children because it is a very slow-growing tumor and may not be noticed until it has become very large and its paravertebral extension causes scoliosis or other vertebral changes. Ganglioneuromas typically enlarge unilaterally, displacing rather than infiltrating surrounding structures. However, there may be encasement of the major abdominal vessels. The MR signal intensity and growth pattern do not allow differentiation between ganglioneuroma and neuroblastoma. The benign form is suggested if MRI shows displacement rather than invasion of vessels and there is no early enhancement of the tumor in a dynamic contrastenhanced study.31

Ganglioneuroblastoma Ganglioneuroblastomas, like neuroblastomas, are of ganglion cell origin and contain both malignant neuroblast cells and mature ganglion cells.32 They are also predom-

a

b Fig. 17.24a, b Neuroblastoma in a 9-month-old infant (1.5 T). Coronal (a) and axial (b) T1w images obtained after IV contrast administration. Left paraspinal tumor with extensive invasion of the spinal canal and mass effect.

inantly found in the abdomen but may also arise in the mediastinum, the neck region, and the extremities. Ganglioneuroblastomas are typically diagnosed in children aged between 2 and 4 years and rarely after age 10. The prognosis is more favorable than that of neuroblastoma. The tumors have a variable MR appearance due to their mixed composition. Sudden growth of an assumed ganglioneuroma, which typically grows slowly, should raise suspicion of a ganglioneuroblastoma.

Pheochromocytoma and Other Adrenal Tumors Pheochromocytoma is rare in childhood; it is typically diagnosed after age 10 and is more common in boys. Most pheochromocytomas occur in the adrenals (70 %), but ectopic forms can be found in the pelvis, chest, and neck. In 30 % of cases, they are multiple and bilateral. Less than 10 % of pheochromocytomas are malignant. Close screening is required in hypertensive children with a family history of pheochromocytoma and multiple

360

17 Abdominal MRI in Children

endocrine neoplasia (MEN) or von Hippel–Lindau disease. Both the thorax and abdomen must be screened for the timely detection of a pheochromocytoma. A pheochromocytoma is a well-defined mass of very high and homogeneous signal intensity due to prolonged T2 relaxation. Size may be variable but is typically ca. 2 cm. The high signal intensity distinguishes pheochromocytoma from other adrenal masses and allows excellent delineation from surrounding structures. Larger pheochromocytomas may have a heterogeneous appearance and also show irregular enhancement if they take up contrast medium. MRI provides the same diagnostic information as CT and is the preferred imaging modality in children with clinically and sonographically suspected pheochromocytoma since it does not involve ionizing radiation. The entire abdomen is imaged to rule out additional tumor manifestations. Other adrenal tumors such as adenoma, carcinoma, and metastases are extremely uncommon in childhood. MRI is preferred to CT for diagnosing adenoma in children. An adenoma is typically seen as a rather small, rounded lesion with well-defined margins but variable enhancement on dynamic contrast-enhanced MRI. In children with portal hypertension, tortuous intra-adrenal vessels may mimic a tumor on CT scans, while MRI enables reliable differentiation between an adrenal mass and vessels.

a

b

Pelvis Genital Organs

c Fig. 17.25a–c Ganglioneuroma in a 15-year-old boy (1.5 T). Coronal T2w image (a), CT scan (b), and US view (c). Large, smoothly marginated mass of homogeneous high SI directly adjacent to the vertebral column; intraothoracic and intra-abdominal tumor extension; no spinal canal invasion.

Anomalies of the Male and Female Genitalia MRI plays an increasing role in the diagnostic evaluation of neonates in whom a disorder of sexual differentiation is suspected. Rudimentary female internal genitalia are present in hermaphroditism but are absent in testicular feminization. MRI can correctly localize the gonads and characterize the type of defect including gonadal dysgenesis in a large percentage of cases. MRI enables accurate diagnosis of müllerian duct anomalies such as dilatation of the prostatic utricle and hypospadias in boys or bicornuate uterus and vaginal duplication or atresia in girls. These anomalies can be easily and noninvasively diagnosed with MRI, especially in older children.33 A uterus didelphys (double uterus) with unilateral vaginal atresia and hematometra may be difficult to demonstrate using US. In these cases, MRI is useful and will additionally identify the associated ipsilateral renal anomaly (agenesis). MRI is also used if US provides no definitive diagnosis in Mayer–Rokitansky–Küster–Hauser syndrome, a müllerian duct agenesis characterized by congenital absence of the uterus and vagina in the presence of normal ovaries.

MRI Appearance of Pathologic Entities

361

Girls with hydrometrocolpos or hematosalpinx are examined by US, reserving MRI for cases in which additional anomalies are expected (Fig. 17.26).34 Male Genitalia Undescended Testis. About 1 % of boys have an undescended testis (cryptorchidism) at 1 year. In 80 % of these cases, the undescended testis is located in the inguinal canal or high in the scrotum and can be palpated. In the remaining 20 % of boys, the undescended testis is not palpable and only these boys require imaging. Crosssectional imaging also has limitations, and there is considerable variation in the sensitivities and specificities reported for US, CT, and MRI in identifying undescended testicles. A testis high in the inguinal canal is hyperintense on T2w images and is clearly identified by MRI, while atrophied testes in ectopic intra-abdominal location have intermediate T1 and T2 signal intensity and will also be missed by MRI because they cannot be differentiated from lymph nodes. Testicular Tumors. Testicular tumors comprise only 1 % of all tumors in children. Most testicular tumors are malignant, the most common type being germ cell tumors. These include embryonal carcinoma and endodermal sinus tumors (yolk sac tumors). Twenty-five percent of solid testicular tumors in children are teratomas. MRI is sensitive but not specific in detecting testicular tumors and is not able to distinguish between benign and malignant tumors or predict the histologic type. Testicular tumors are isointense or hypointense to normal testicular tissue on T1w images and hypointense on T2w images. Secondary testicular tumors are far more common. They occur in the setting of leukemia, lymphoma, and neuroblastoma, less commonly in Ewing sarcoma. Secondary tumors involve one or both testes and their diagnosis is straightforward in most cases because the underlying disease is known. MRI is usually not needed in these cases and is also inferior to duplex US in boys presenting with acute scrotum. Female Genitalia Inflammatory ovarian disease is a clinical diagnosis. Ovarian torsion and ovarian cysts are diagnosed by US, and additional imaging is only needed if a teratoma has to be ruled out (Figs. 17.27 and 17.28).35 Ovarian tumors are rare in girls, and < 30 % are malignant. The vast majority of ovarian tumors are germ cell tumors, followed by mesenchymal and epithelial tumors. Ovarian germ cell tumors include dysgerminoma, teratoma, endodermal sinus tumor (yolk sac tumor), embryonal carcinoma, and choriocarcinoma.36 Dysgerminomas tend to occur bilaterally and have a rather favorable prognosis. Endodermal sinus tumors have a much poorer prognosis. Only 10 % of ovarian teratomas are malignant (Fig. 17.27). Mesenchymal tumors, in particular granulosa cell tumors, produce insulin and become clinically apparent through precocious puberty.

Fig. 17.26 Hematosalpinx in a 15-year-old girl with periodic abdominal symptoms after repair of bladder exstrophy (1.5 T). Sagittal T2w single-shot image. Hematosalpinx is seen as a hyperintense structure above the bladder.

Only teratomas have characteristic MR features. They usually have a large fatty component, which appears bright on T1w and T2w images, while it is dark on a fatsuppressed T1w sequence. MRI will also demonstrate cystic components and intracystic hemorrhage, but CT is superior in demonstrating intralesional calcifications. All other ovarian masses are mixed solid and cystic lesions, and their MRI appearance does not allow a specific diagnosis to be made. Cystadenocarcinoma is rare in childhood. The tumor contains cystic areas with thickened walls and septa, which enhance after contrast medium administration.

Urinary Bladder As in adults, the thickness of the bladder wall varies considerable with bladder filling. The bladder wall is best evaluated on PD or T1w sequences, on which it is intermediate in signal intensity between urine and perivesical fat.

362

17 Abdominal MRI in Children

a

b Fig. 17.27a, b Ovarian cyst in a 16-year-old girl (1.5 T). a Coronal T1w image. b Axial T2w image. Large, smooth-walled cyst of the right ovary; polycystic ovary syndrome.

b Fig. 17.28a, b Ovarian teratoma in a 15-year-old girl (1.5 T). Coronal (a) and axial (b) T2w images. A smoothly marginated mass of inhomogeneous SI is demonstrated to the right of the uterus. a

Rhabdomyosarcoma is the most frequent childhood tumor arising from the urinary bladder, urethra, or prostate. It is more common in boys and has two age peaks, one during the first 4 years of life and one in adolescence between 15 and 20 years. Bladder rhabdomyosarcoma is usually located in the neck or trigone and invades the wall. Prostate rhabdomyosarcoma invades the bladder neck, the posterior urethra, and the perirectal space. MRI with image acquisition in all three orthogonal planes is mandatory in children with suspected rhabdomyosarcoma because it reliably demonstrates both thickening of the bladder wall and the extent of bladder wall infiltration by the tumor. Rhabdomyosarcoma is depicted as a solid lesion of intermediate signal intensity on T1w images and high signal intensity on T2w images. Intense enhancement on postcontrast images allows excellent differentiation from surrounding tissue. MRI is indispen-

sable prior to surgery of urogenital rhabdomyosarcoma (Fig. 17.29). MRI is also used to follow up the response to chemotherapy for precisely defining regression of the tumor and lymph node involvement.

Presacral Tumors MRI is optimally suited for imaging presacral masses. An anterior meningomyelocele is a cystlike protrusion of the spinal cord containing cerebrospinal fluid (CSF). It is associated with a low-lying conus and spinal anomalies (Fig. 17.30). A presacral lipoma is reliably diagnosed based on its MR signal intensity, as is presacral chordoma, which is rare in childhood. Common presacral tumors are extragonadal germ cells tumors such as endodermal sinus tumors and teratomas. The MR signal intensity of endoder-

MRI Appearance of Pathologic Entities

a

363

b Fig. 17.29a, b Vaginal rhabdomyosarcoma in a 2-year-old girl (1.5 T). Coronal (a) and sagittal (b) T2w single-shot TSE images. Tumor of predominantly high signal intensity (T) occupying the entire true pelvis with upward displacement of the urinary bladder (B) and uterus (U).

mal sinus tumors is nonspecific, while that of sacrococcygeal teratomas is defined by their mixed solid and cystic tissue composition with characteristic fatty components. Presacral teratomas can become very large and their extent can be accurately determined using MRI (Fig. 17.31).

Systemic Neoplastic Disease Systemic neoplastic diseases include leukemias and all malignant proliferative diseases of the lymphatic tissue and histiomonocytes. Infiltration of a parenchymal organ by leukemia or lymphoma results in enlargement of the affected organ. There is no primary indication for MRI, but organ enlargement or bone marrow involvement may be incidentally revealed on MRI performed for other reasons. A specific diagnosis of systemic neoplastic disease on the basis of MRI signal intensities is rarely possible. Lymphoma is hypointense to fat and slightly hyperintense to muscle on T1w images and hyperintense to muscle and isointense to fat on T2w images.37 Benign atypical lymphoproliferative disorders such as Castleman disease or Rosai–Dorfman syndrome can be detected by MRI, but the specific diagnosis needs to be confirmed by biopsy. Fig. 17.30 Currarino triad in a 3-month-old girl (1.5 T). Sagittal T2w image. Sacral anomaly, anterior meningocele of predominantly high SI, and anal atresia.

364

17 Abdominal MRI in Children

References

a

b Fig. 17.31a, b Teratoma in the presacral area and coccygeus muscle in a 6-month-old infant (1.5 T). a Sagittal T2w image. b Paracoronal T1w image after contrast administration. Predominantly cystic mass with solid components inferiorly and additional cystic components anteroinferiorly. The tumor displaces the rectum, uterus, and urinary bladder anteriorly and superiorly. There is enhancement of the entire tumor and invasion of pelvic muscles on the postcontrast image (b).

1. Shellock FG. Pocket Guide to MR Procedures and Metallic Objects: Update 1994. New York: Raven Press; 1994 2. Grebe P, Kreitner KF, Kersjes W, Würfel A, Schild H. [MRT for surgical planning in anal atresia]. Rofo 1993;159(6):528–532 3. Tissot O, Bodnar D, Henry L, Dubreuil A, Valette PJ. [Ano-perineal fistula in MRI. Contribution of T2 weighted sequences]. J Radiol 1996;77(4):253–260 4. Laghi A, Borrelli O, Paolantonio P, et al. Contrast enhanced magnetic resonance imaging of the terminal ileum in children with Crohn’s disease. Gut 2003;52(3):393–397 5. Kettritz U, Isaacs K, Warshauer DM, Semelka RC. Crohn’s disease. Pilot study comparing MRI of the abdomen with clinical evaluation. J Clin Gastroenterol 1995;21(3):249–253 6. Shoenut JP, Semelka RC, Magro CM, Silverman R, Yaffe CS, Micflikier AB. Comparison of magnetic resonance imaging and endoscopy in distinguishing the type and severity of inflammatory bowel disease. J Clin Gastroenterol 1994;19(1):31–35 7. Schweinitz V, Glüer P, Mildenberger H. Liver tumors in neonates and very young infants: diagnostic pitfalls and therapeutic problems. Eur J Pediatr Surg 1995;5:72–76 8. Stöver B, Laubenberger J, Niemeyer C, Stahl F, Brandis M, Langer M. Haemangiomatosis in children: value of MRI during therapy. Pediatr Radiol 1995;25(2):123–126 9. Martí-Bonmatí L. MR imaging characteristics of hepatic tumors. Eur Radiol 1997;7(2):249–258 10. King AD, Walshe JM, Kendall BE, et al. Cranial MR imaging in Wilson’s disease. AJR Am J Roentgenol 1996;167(6):1579–1584 11. Mortele KJ, Ros PR. MR imaging in chronic hepatitis and cirrhosis. Semin Ultrasound CT MR 2002;23(1):79–100 12. Vogl TJ, Steiner S, Hammerstingl R, et al. [MRT of the liver in Wilson’s disease]. Rofo 1994;160(1):40–45 13. Jensen PD, Jensen FT, Christensen T, Ellegaard J. Evaluation of transfusional iron overload before and during iron chelation by magnetic resonance imaging of the liver and determination of serum ferritin in adult non-thalassaemic patients. Br J Haematol 1995;89(4):880–889 14. Siegelman ES, Mitchell DG, Rubin R, et al. Parenchymal versus reticuloendothelial iron overload in the liver: distinction with MR imaging. Radiology 1991;179(2):361–366 15. Wang ZJ, Haselgrove JC, Martin MB, et al. Evaluation of iron overload by single voxel MRS measurement of liver T2. J Magn Reson Imaging 2002;15(4):395–400 16. Helmberger T, Holzknecht N, Gauger J, Beuers U, Reiser M. [MRI of the pancreas: radiologic-pathologic correlation]. Radiologe 1996;36(5):419–426 17. Hirohashi S, Hirohashi R, Uchida H, et al. Pancreatitis: evaluation with MR cholangiopancreatography in children. Radiology 1997;203(2):411–415 18. Kisker O, Bastian D, Frank M, Rothmund M. [Diagnostic localization of insulinoma. Experiences with 25 patients with solitary tumors]. Med Klin (Munich) 1996;91(6):349–354 19. Ramani M, Reinhold C, Semelka RC, et al. Splenic hemangiomas and hamartomas: MR imaging characteristics of 28 lesions. Radiology 1997;202(1):166–172 20. Renard TH, Andrews WS, Rollins N, et al. Use of distal splenorenal shunt in children referred for liver transplant evaluation. J Pediatr Surg 1994;29(3):403–406 21. Stöver B. [Abdominal magnetic resonance tomography in the child]. Radiologe 2001;41(5):418–426 22. Gylys-Morin V, Hoffer FA, Kozakewich H, Shamberger RC. Wilms tumor and nephroblastomatosis: imaging characteristics at gadolinium-enhanced MR imaging. Radiology 1993;188(2): 517–521 23. Zoeller G, Pekrun A, Lakomek M, Ringert RH. Wilms tumor: the problem of diagnostic accuracy in children undergoing preoperative chemotherapy without histological tumor verification. J Urol 1994;151(1):169–171 24. Dunnick NR. Hanson lecture. Adrenal imaging: current status. AJR Am J Roentgenol 1990;154(5):927–936 25. Hoeffel C, Legmann P, Luton JP, Chapuis Y, Fayet-Bonnin P. Spontaneous unilateral adrenal hemorrhage: computerized tomography and magnetic resonance imaging findings in 8 cases. J Urol 1995;154(5):1647–1651

MRI Appearance of Pathologic Entities 26. Brill PW, Jagannath A, Winchester P, Markisz JA, Zirinsky K. Adrenal hemorrhage and renal vein thrombosis in the newborn: MR imaging. Radiology 1989;170(1 Pt 1):95–98 27. Berdon WE, Ruzal-Shapiro C, Abramson SJ, Garvin J. The diagnosis of abdominal neuroblastoma: relative roles of ultrasonography, CT, and MRI. Urol Radiol 1992;14(4):252–262 28. Forman HP, Leonidas JC, Berdon WE, Slovis TL, Wood BP, Samudrala R. Congenital neuroblastoma: evaluation with multimodality imaging. Radiology 1990;175(2):365–368 29. Tschäppeler H. [The abdominal neuroblastoma]. Radiologe 1993;33(12):675–678 30. Parisi MT, Hattner RS, Matthay KK, Berg BO, Sandler ED. Optimized diagnostic strategy for neuroblastoma in opsoclonusmyoclonus. J Nucl Med 1993;34(11):1922–1926 31. Ichikawa T, Koyama A, Fujimoto H, et al. Retroperitoneal ganglioneuroma extending across the midline: MR features. Clin Imaging 1993;17(1):19–21

365

32. Rha SE, Byun JY, Jung SE, Chun HJ, Lee HG, Lee JM. Neurogenic tumors in the abdomen: tumor types and imaging characteristics. Radiographics 2003;23(1):29–43 33. Hochreiter W, Stenzl A, Altermatt HJ, Kraft R, Spiegel T. [Urogenital malformation complex including the Mullerian system]. Urologe A 1994;33(2):154–158 34. Reinhold C, Hricak H, Forstner R, et al. Primary amenorrhea: evaluation with MR imaging. Radiology 1997;203(2):383–390 35. Lauszus F, Balslev T, Johansen A. The role of ultrasound scan in childhood ovarian tumors. Acta Obstet Gynecol Scand 1994; 73(1):67–69 36. Davidoff AM, Hebra A, Bunin N, Shochat SJ, Schnaufer L. Endodermal sinus tumor in children. J Pediatr Surg 1996;31(8): 1075–1078, discussion 1078–1079 37. Negendank WG, al-Katib AM, Karanes C, Smith MR. Lymphomas: MR imaging contrast characteristics with clinical-pathologic correlations. Radiology 1990;177(1):209–216

366

Index

Notes: Page numbers in italics refer to figures; page numbers in bold refer to tables

A abdominal aorta aneurysm see aortic aneurysm (abdominal aorta) anterior branches 311–312, 312 atherosclerotic occlusions 317–318, 318 dissection see aortic dissection intramural hematoma see aortic intramural hematoma (IMH) lateral branches 311, 312 long intraluminal thrombosis 315–316 normal MR angiography 311–312, 312 posterior branches 311, 312 thrombosis 313, 317, 318 wall inflammation (aortitis) 316, 316–317 wall thickening 315, 315–316 abdominal arteries normal MR angiography 311–312, 312 pathologic conditions, MRA 313–321 abdominal belt 219 abdominal lymph nodes see intraabdominal lymph nodes abdominal trauma kidneys 151, 152 liver, children 348 spleen 102, 350 abdominal vascular malformations 350–351 see also under MR angiography (MRA) abdominal veins normal MR angiography 312 pathologic entities, MRA 325–328, 350–351 abscess see individual types of abscess ACTH, ectopic secretion 185 Addison disease 184 adenoma adrenal see adrenal adenoma colon 119, 131–132 hepatocellular 11, 14, 18, 346 renal 155, 161 adenomyoma 261 adenomyosis, uterus 258, 260, 260–261 adnexa 285–299 normal anatomy 287 see also ovaries; uterine tubes adrenal adenoma 181, 185, 186 aldosterone-producing 185, 186

atypical, inhomogeneous images 186, 188 children 360 functioning 185, 186 lipid-containing 182, 186, 187, 193 malignant tumors vs 186 nonfunctioning 184, 186, 187, 188 SIrel cut-off values 183 adrenal glands 181–194 abscess 185 aim of MRI 193 angiomyolipoma/myelolipoma 188, 189 benign conditions 184–189 benign vs malignant lesions 183, 184, 186, 193 central liquefaction 185 cortex, destruction 184 cysts 186, 186 ganglioneuroma in children 359 hemangioma 188, 190 hemorrhage 356, 357 hyperplasia 185, 185 indications for MRI 181 inflammatory conditions 184–185, 193 lipid-containing tumors see adrenal adenoma malignant tumor 184, 189–193 carcinoma see adrenocortical carcinoma, primary children 358–360, 359 lymphoma 190 metastases 181, 190, 191 neuroblastoma see neuroblastoma pheochromocytoma 190, 192, 192, 193, 359–360 masses 181, 184 characterization, flowchart 193–194, 194 MRI imaging technique 181–184 contrast media 182, 184 image analysis 182–184 index of signal intensities (SI) 183 in-phase/opposed-phase imaging 181, 181, 182, 183, 193 pulse sequences 181–182, 182 normal MRI appearance 183, 184 pathologic entities, MRI appearance 184–193 children/infants 356, 358–360 sizes in children/adults 356 tuberculosis 184–185 adrenal incidentaloma 181, 184 adrenocortical adenoma see adrenal adenoma

adrenocortical carcinoma, primary 189–190, 191 children 360 adrenocortical hyperplasia/hypersecretion 185, 185 adrenocortical insufficiency 184 ampullary cancer 77, 77 anal canal 127–141, 128, 129 indications for MRI 127 MRI imaging technique 127–128 normal MRI appearance 128–129, 129, 130, 131 pathologic conditions, MRI 136 see also rectum anal cancer 120, 127, 136, 137 anal cushions (corpora cavernosa recti) 129 anal sphincters 128–129, 130, 131 androgen, excess production 185 aneurysm abdominal aorta see aortic aneurysm (abdominal aorta) inflammatory (aortic) 317, 317 renal arteries 322 visceral arteries 319 angiography, MR see MR angiography (MRA) angioma, renal 356 angiomatosis, systemic cystic 96 angiomyolipoma adrenal 189 hepatic, children 347 renal 160, 160, 161 children 356 angiosarcoma renal 171 spleen 100 anocutaneous line 128 anorectal abnormalities, children/infants 342–343 anorectal abscess, MRI pulse sequences/parameters 129 anorectal angle 140 anorectal fistula children/infants 342, 343, 345 MRI pulse sequences/parameters 129 anorectal flexure 128, 129 anorectal line 128 antispasmodic agents gastrointestinal MRI 110 intra-abdominal lymph node MRI 332, 333 liver MRI 8 pancreas MRI 69 pelvic MRI 241, 285 rectal/anal canal MRI 127

Index

anus atresia 342 ectopic 343, 344 imperforate 342 aortic aneurysm (abdominal aorta) 313, 313–314 causes 313 contrast-enhanced 3D MRA 313–314 fusiform 313 inflammatory 317, 317 infrarenal 310, 313, 313 multislice spiral CT 305, 313–314 saccular 313 treatment follow-up by 3D MRA 314, 314 true aneurysm 313 aortic dissection 317, 318 chronic 316 contrast-enhanced 3D MRA 317 false and true lumen 316, 317 aortic intramural hematoma (IMH) 314–316, 315 acute/subacute 315, 315 aortic wall thickening vs 315–316 MR angiography 315, 315 aortic occlusion 317–318, 318 aortic stenosis 317–318, 318 aortic thrombosis 313, 317, 318 aortitis 316, 316–317 aortocaval fistula 314 appendicitis 123, 125, 125 MRI indication 109, 109 appendicoliths 123 arteriovenous fistula, MR angiography 323 artifacts on MRI imaging arms causing wraparound 143 motion see breathing motion artifacts; motion artifact(s) ascites 72 cirrhosis with 39, 41 endometrial carcinoma 265 ovarian cancer recurrence 298 rectouterine pouch 248 atherosclerosis aortic occlusions 317–318, 318 aortic wall thickening 315 penetrating ulcer 314 superior mesenteric artery 320 autoimmune hemolysis 36, 105 autosomal dominant polycystic kidney disease (ADPKD) 155, 154–156 autosplenectomy 103 azygos vein, blood flow 325

B Bartholin cysts 274 basal cell nevus syndrome 343 benign prostatic hyperplasia (BPH) 225, 225 MR spectroscopy 223 prostate cancer with 225, 226 bile ducts 47–64 anatomic variants 56 inflammatory conditions 55 MRCP see MRCP normal anatomy, MRI 51, 52, 52–53, 55 pathologic entities, MRI 53–63 children 348

small, normal 55 strictures/stenoses 53, 55, 55 biliary–enteric anastomosis 48 bladder (urinary) 209–217 benign tumors 211 carcinoma 211–214, 230 lymph node metastases 336, 340 papillary 213 recurrence 212, 214, 215 seminal vesicle involvement 212, 214, 215 T1 to T4 stages 212, 213, 214 TNM staging 212, 212, 212, 214 cervical cancer invasion 268, 269, 272, 273 recurrence after surgery 278, 281 congenital anomalies 210 diverticula 210, 211 hemangioma 211 hyperintense, motion artifacts 263 indications for MRI 209 inflammation 211 inhomogeneous signal intensity (layering) 210, 213 leiomyoma 211 lymphoma 215 malignant tumors 210, 211–215 carcinoma see bladder, carcinoma rare 214–215 recurrent 210, 212, 214, 215 rhabdomyosarcoma (children) 362, 363 MRI imaging technique 209–210 contrast media 210, 212 imaging planes 209 pulse sequences and protocol 209, 210 MRI vs CT 212, 214 normal MRI appearance 129, 210 optimal filling, before MRI 209, 241, 285 pathologic entities, MRI appearance 210–216 children 361–362 pheochromocytoma 211 prostate cancer invasion 231 sarcoma 215 wall, cancer invasion 212, 213, 231 blood, T1 relaxation time, MRA 306 blood flow, velocity, calculation 306 blood pool agents 307 blood transfusions, repeated, iron overload 348, 348 blueberry juice, as contrast medium 49, 69 bone marrow, neuroblastoma metastases 359 bowel cleansing 109, 110, 140 bowel obstruction 342 breast cancer, metastases 20, 20, 336, 337 breath-holding techniques colorectal cancer MRI 121 intra-abdominal lymph node MRI 332, 333 kidney and upper urinary tract MRI 143 liver MRI 1, 2, 3, 7, 31 MRCP 48, 49 pancreas MRI 65, 66 retroperitoneum MRI 196, 196 splenic MRI 87

367

see also VIBE sequence (volume-interpolated breath-hold examination) breathing motion artifacts gastrointestinal tract MRI 111 intra-abdominal lymph node MRI 333 liver MRI 3, 5–6 minimization 1, 3, 7, 31, 49, 219 pelvic MRI 241, 243, 285 see also breath-holding techniques broad ligament, normal anatomy 287 bronze diabetes 72 Budd–Chiari syndrome (BCS) 42, 42, 324, 351 Buscopan 110 butylscopolamine 110, 111, 112, 121, 127

C CA 19–9 80 calcification, tumors 19, 257, 358 Candida abscess, liver 25 cannula, indwelling 1 carcinoid tumors 113 cardinal ligaments (Mackenrodt ligaments) 246, 246 Caroli syndrome 53, 348 Castleman disease 363 catecholamines, secretion by neuroblastoma 358 cavernous transformation, portal vein 350 CD 117 (c-kit proto-oncogene) 113, 197 celiac trunk 311 stenosis 319 cerebral MRI 347, 359 ceruloplasmin, deficiency 36 cervical canal 245–246 in uterus didelphys 253, 254 cervical carcinoma 267–273 contrast-enhanced images 267, 268 extent, ligaments in evaluation 246, 246 incidence 267 metastatic lymph nodes 267, 273, 273 MRI appearance 267–273 parametrial invasion 271, 271, 272 pulse sequences/protocol 267, 267 radiotherapy, MRI after 277, 279 recurrence after surgery 278, 280, 281 residual tumor after radiotherapy 278, 279 stage IA and IB 268, 270, 271 stage IIA and IIB 268, 271, 271, 272 stage IIIA and IIIB 269, 272, 272 stage IV and IVA 269, 272, 273 staging 267–273, 270 MRI role, vs CT 273, 273 vaginal invasion 271, 272, 273, 275 viable vs necrotic areas 267–268, 269 cervical intraepithelial neoplasia 267 cervix, uterine carcinoma see cervical carcinoma cysts 262, 262 endometrial carcinoma invasion 266

368

Index

cervix, uterine fibrosis, postoperative 278, 280 mucus 249 normal anatomy 245, 245 normal MRI appearance 248, 248, 250, 251 stenosis 261, 277 T2 w images 248, 248, 250 zonal anatomy 249, 250 chemical shift imaging (CSI) 182, 210, 292 3D MRS see MR spectroscopy, 3D children, abdominal MRI 341–365 abdominal vascular malformations 350–351 adrenal gland disease 356, 358–360, 359 neuroblastoma 358, 358–359, 359, 359 contraindications for MRI 342 ganglioneuroblastoma 359 ganglioneuroma 359, 360 gastrointestinal tract 342–345 hepatobiliary system 345–348, 346, 347, 348 indications for MRI 341 kidneys 352–356, 357 MRI imaging technique 341–342 coils and pulse sequences 341 contrast media 342 positioning 341 sedation 341–342 pancreatic abnormalities 348–349 pathologic entities, MRI appearance 342–364 pelvic conditions/MRI 361–363, 363 retroperitoneum 352 splenic abnormalities 350 systemic neoplastic disease 363 chloral hydrate 341, 342 cholangiocarcinoma 11, 22, 25, 63 cholangiocellular carcinoma (Klatskin tumor) 63, 64 cholangitis acute 76 primary sclerosing (PSC) 61, 63 secondary sclerosing 61 choledochal cysts 348, 349 choledochal stent 51, 53 choledocholithiasis 58 choline, MR spectroscopy of prostate 222, 223, 223 prostate cancer 232, 233 chordoma, presacral 362 choriocarcinoma 261 chromaffinoma, of adrenal gland 190, 192, 192, 193 cirrhosis 38–42 advanced 40–42 ascites with 39, 41 confluence fibrosis, with 39, 40 contrast media use 32–33, 323 dysplastic nodules 40, 41 first MRI signs 38 indication for MRI 34 macronodular 39, 41 neoplastic lesions 39–40, 41 portal vein thrombosis 40 primary biliary 55 primary hemochromatosis and 36 regenerative lesions 40, 41, 42 siderotic nodules 39

splenic infarction with 104 splenomegaly with 104, 105 Wilson disease 347 citrate, MR spectroscopy of prostate 222, 222, 223, 233 prostate cancer 232, 233 c-kit proto-oncogene 113, 197 clear cell tumor, renal 356 coils body, pelvic MRI 243 conventional, in MR angiography 309 endorectal contraindications (acute prostatitis) 224 pelvic MRI (female) 242 prostate MRI 219, 219, 235 rectum MRI 128 flexible wraparound for neonates/ infants 341 phased-array bladder MRI 209 intra-abdominal lymph nodes 332 kidney and urinary tract 143–144 liver MRI 2 MR angiography 306, 309 MRCP 47 pelvic MRI 242, 243, 285 rectum and anal canal MRI 127–128 retroperitoneum MRI 195 spleen MRI 87 surface, bile/pancreatic duct MRI 47 colon 119–125 adenoma 119, 131–132 ascending, carcinoma 122 carcinoma see colorectal cancer Crohn disease 121, 123 diverticulitis 121, 123, 124, 124 inflammatory disease 121, 123–125 lymphatic drainage 120 MRI vs CT 124, 125 polyp 120 radiation-induced changes 121, 123 sigmoid 122, 123, 124 ulcerative colitis 121, 123, 123 see also gastrointestinal tract colonography, MR see MR colonography (MRC) colorectal cancer 119–121, 131–132 “apple core” appearance 119, 119 distant metastases 121 liver metastases 19, 19–20, 28–29, 121 local recurrence 121 MRI role, and vs PET/CT 121, 125 screening, MR colonography 109 spread 119–120 TNM staging 119, 119, 132 see also rectal cancer common bile duct diameter after cholecystectomy 55, 63 normal MRCP 51, 52, 54 stones 55, 58–59, 63 common channel syndrome 348 common hepatic duct, bifurcation, cholangiocarcinoma 63 computed tomography (CT) intramural hematoma of aorta 315 MRI comparison see individual conditions

spiral and multislice (SCT/MSCT) 305 aortic aneurysm 305, 313–314 aortic dissection 317 pancreas 65 renal trauma 151 congenital anomalies/diseases bladder (urinary) 210 gastrointestinal tract 342–343 kidney 150, 150, 352, 352 pancreas 71–72 prostate 223 uterus see under uterus vagina 274 congenital mesoblastic nephroma 355–356 contraindications for MRI 241, 342 contrast media administration methods 9, 10 low rate of adverse effects 306 nonspecific see gadolinium-based contrast media tissue-specific see iron oxide particles; superparamagnetic iron oxide particles (SPIO) see also specific anatomic regions (e. g. liver) copper intratumoral, hepatocellular carcinoma 21 toxicity, Wilson disease 42–43, 347 corpus luteum cysts 289 creatine, MR spectroscopy of prostate 222, 223, 223 prostate cancer 232, 233 Crohn disease 109 antrum 112 children 343, 345 colon 121, 123 complications 116, 124 enterocutaneous fistula 117 perianal fistulas 136, 137, 137, 138 rectum 136–138, 137, 138 small intestine 114–115, 116 ulcerative colitis vs 123 Crohn Disease Activity Index (CDAI) 114–115 cryptorchidism 361 CT during arterial portography (CTAP) 33 Currarino triad 363 Cushing syndrome 185 cyst(s) see individual cysts cystadenocarcinoma ovarian 295, 297, 361 pancreatic 79 cystadenofibroma, serous 294 cystadenoma mucinous 79, 81, 290, 291, 295 ovarian 290, 291, 295, 295 pancreatic 77, 79, 79, 81 serous 77, 79, 79, 290 cystic duct MRCP 55 stones 58 cystic fibrosis, pancreas 71 cystic pancreatic neoplasms 77, 79, 79 cystitis 211 cystoscopy, bladder cancer 214

Index

D defecation disorders 125, 140 Denonvilliers fascia 221 dermoid cysts 292–293 diazepam 341 diffusion-weighted imaging (DWI), rectal/anal canal imaging 128 dilatation and curettage, MRI after 276 diverticula duodenal 115 urinary bladder 210, 211 diverticulitis, colon 121, 123, 124, 124 diverticulosis 123 Dotarem (Gd-DOTA) 8–9, 34, 127, 306 DOTATOC 125 doughnut sign, liver metastases 19 duct-penetrating sign 55, 80, 84 duodenal C-loop 70 duodenal diverticula 115 duodenum, pancreatic head differentiation (MRI) 69, 70 dynamic contrast-enhanced MRI see contrast media dysgerminoma 297–298, 361 dystocia 301

E echinococcal infection 186, 346, 346 echo-planar imaging (EPI), liver 5 ectocervix 249, 250 ectopic pregnancy 288 edema intrahepatic 37, 38 pancreatitis in children 348 peritumoral (retroperitoneal tumors) 200 ejaculatory ducts 221, 221 embryonal carcinoma 361 endocervical glands 249 endocervix, endometrial carcinoma invasion 266 endodermal sinus tumors 361, 362 endometrial carcinoma 262–266 epidemiology and risk factors 262 MRI appearance 263–266 myometrial invasion 262, 263, 264, 266 pulse sequence/protocol 263, 263 recurrence after surgery 278, 280 residual tumor after radiotherapy 278, 280 stage I and IA 263, 263, 266 stage IB/IC 263, 264, 266 stage II, III and IV 265, 266 staging/FIGO classification 263–266, 266 errors and overstaging 263 MRI role 266–267 T2 w images (high intensity) 263 vaginal spread 262 endometrial cavity 253 blood clot in 247, 248 endometrioma (endometriotic cysts) 290, 292, 292 endometriosis 253, 254, 290 adenomyosis vs 258 endometriotic cyst 290, 292, 292 endometrium 247 abnormal thickening 264

menstrual cycle effects 247, 247, 248 polyps 261, 261 radiotherapy effect 276–277, 277 Endorem 8, 27 endoscopic retrograde cholangiopancreatography (ERCP) 47 enterocutaneous fistula 117 Eovist, 3D MR angiography 307 esophageal squamous cell carcinoma, metastases 112 esophageal varices 323

F fallopian tubes see uterine tubes fast spin echo sequence see FSE sequence fat saturation techniques 66 spectral/effective, liver imaging 7, 8, 35 fat suppression adrenal gland imaging 181, 183 for children/infants 341 liver imaging 2, 3, 4, 7–8, 9, 35 pancreas imaging 66 pelvic MRI 242–243 retroperitoneum imaging 196, 196 small intestinal MRI 115 splenic imaging 87 see also short tau inversion recovery (STIR) sequence fecal incontinence 343 female genitalia anomalies 360–361 disease/tumors in children 361, 362 see also specific anatomic structures fetopelvic disproportion 301 fibroids see uterus, leiomyoma (fibroids) fibromuscular dysplasia 319, 320 fibrosarcoma renal 171 retroperitoneal 203, 204 FIGO classification cervical carcinoma 268, 270 endometrial carcinoma 263, 266, 266 FLASH 144, 196 flow compensation (gradient moment nulling), liver imaging 7 flow-related signal enhancement, MR angiography 305–306 flow-related signal loss (phase shifts) 305, 306 focal nodular hyperplasia (FNH), liver see under liver follicular cysts 287–288, 289 Fourier space (k-space) 307 FSE sequence for children/infants 341 liver MRI 3, 5 pelvic MRI 242, 252 retroperitoneum MRI 196 fungal infections 25, 102, 185

G gadobutrol see Gadovist (gadobutrol) gadofosveset (Vasovist) 307 gadolinium-based contrast media 3D MR angiography 306–307

369

adrenal gland imaging 182, 184 bladder imaging 210 children/infants MRI 342 intra-abdominal lymph nodes 333, 334, 336 kidney and upper urinary tract 146, 148, 174 liver imaging 8, 9, 27, 29, 31–32, 34, 91 pancreas imaging 66, 68 prostate imaging 220, 238 rectal imaging 127 retroperitoneum imaging 196 splenic imaging 87, 91 uterine zonal anatomy 247, 248 gadolinium chelates 306, 307 gadolinium-DTPA 306, 307 gadolinium ions (Gd3 +) 306 Gadovist (gadobutrol) 8–9, 306, 307 gallbladder adenomyomatosis 55, 62 disorders in children 348 polyps 55, 61 stones 55, 58, 60 Gamna–Gandy bodies 39, 105, 106 ganglioneuroblastoma 359 ganglioneuroma 359, 360 Gartner duct, cysts 274, 275 gastric cancer see stomach, adenocarcinoma gastric leiomyoma 112 gastric leiomyosarcoma 112 gastric lymphoma 112 gastric vein, left, dilatation 323, 323 gastrinoma 22, 80, 82 gastrointestinal stromal tumor (GIST) 113–114, 115, 197 gastrointestinal tract 109–126 advantages of MRI, over CT 109, 121, 125 congenital anomalies 342–343 functional/dynamic MRI 125 inadequate distention and tumor mimic 111 indications for MRI 109, 109 motion artifacts and prevention 110 MRI imaging technique 109–112 abdominal imaging 110–111 contrast media 110, 111–112, 125 patient preparation 109–110 pelvic imaging 111 pulse sequences 110–111 pathologic conditions, MRI appearance 112–125 children/infants 342–345 PET/CT vs MRI 121, 125 see also colon; small intestine; stomach Gaucher disease, spleen 102 genitalia, anomalies 360–361 germ cell tumors, testicular 361 gestational trophoblastic disease (GTD) 261 ghosting/ghosts, liver imaging 6 giant cell aortitis 316 glucagon, rectal/anal canal imaging 127 glucagonoma 80 Gorlin syndrome 343 graafian follicles 287 gradient echo sequences see GRE sequences

370

Index

gradient moment nulling (flow compensation), liver 7 gradient moment rephrasing (GMR) 306 graft rejection, renal 175, 177 granuloma, hepatic 43, 345 granulosa cell tumors 293, 361 GRE sequences 3D adrenal gland 182, 182 gastrointestinal tract 110–111 intra-abdominal lymph nodes 336 kidney and upper urinary tract 144 liver lesions 3, 5 MR angiography using 307 pancreas 66 retroperitoneum 196 see also VIBE sequence (volumeinterpolated breath-hold examination) adnexa imaging 286 intra-abdominal lymph nodes 332, 332, 333 MR pelvimetry 301, 303, 303 pelvic MRI 243 portal venous MRI 323 in pregnancy 301 retroperitoneum imaging 196

H hamartoma liver, children 345 splenic 96 HASTE, imaging of adnexa 286 adrenal glands 181, 182, 183 bile/pancreatic duct 48, 49, 50 gastrointestinal tract 109, 110–111, 111, 115 colorectal cancer 120, 121, 122 inflammatory disease 124 kidney and upper urinary tract 144, 146 liver 3, 5, 5 rectal/anal canal 128 retroperitoneum 196, 196 splenic 89 hemangioendotheliomatosis, children 345 hemangioma adrenal 188, 190 bladder 211 capillary, splenic 93 cavernous, splenic 93 liver 12, 13, 13–14, 15 centripetal enhancement 30, 31 children 345, 346 cysts vs 12, 13, 13 dynamic contrast-enhanced MRI 13, 14, 15, 32 iron oxide particle injection 27, 29 metastases vs 13, 30, 31 retroperitoneal 200 spleen 93, 95, 95–96 hemangiomatosis liver, children 345, 346 splenic 93, 95, 96 hematocolpos 274, 274

hematoma acute adrenal 356, 358 hepatic 22 intramural, aortic see aortic intramural hematoma (IMH) perirenal 151, 152 renal see kidney, hematoma splenic 102 hematometra 278, 278 hematometrocolpos 253 hematosalpinx 278, 278, 290, 361, 361 hemochromatosis pancreatic MRI imaging 66 primary (hereditary) children 348 liver 36, 37, 106, 348 pancreas71, 71–72 spleen 105, 106 secondary 36, 37, 72, 348 hemoglobinuria, paroxysmal nocturnal 36, 105, 105 hemorrhage adrenal 356, 357 after biopsy for prostate cancer 227, 227, 228 intratumoral, liver metastases 19, 21, 22 uterine leiomyoma 258, 257 hemorrhagic lesions, liver 22 hemosiderin 292, 348 hemospermia 224–225 hepatic arteries MR angiography, protocol 308 right, aberrant 319 stenosis, and thrombosis 324 hepatic duct, right aberrant 53 normal MRCP 51 hepatic steatosis 26, 26, 27 hepatic veins, MR angiography 324 hepatic venous outflow obstruction 42, 42, 324 children 351 hepatitis, acute 37, 38 hepatobiliary contrast media 27, 30, 30 hepatobiliary system, children/infants 345–348 hepatoblastoma 346 hepatocellular adenoma 14, 18, 346 hepatocellular carcinoma (HCC) 11, 21–22, 23, 24 children/infants 346–347, 347 contrast media use 32–33 focal, in adenomatous hyperplasia 42 focal lesion in cirrhosis 39–40, 41 hepatocellular adenoma vs 14 metastases vs 27, 34 pseudocapsule 21, 22, 23, 24 hepatocellular nodules 39, 40–41 hepatocytes, iron uptake 36 hepatopathy, congestive 351 hermaphroditism 360 Hodgkin lymphoma lymph node metastases 331, 336, 338 spleen involvement/MRI 90, 99 see also lymphoma hormone therapy, prostate cancer 232–233

human chorionic gonadotropin (β-hCG) 261 hydatidiform mole 261 hydrometrocolpos 361 hydronephrosis, vascular origin 356, 356 hyperaldosteronism, primary 185, 186 hypercortisolism 185 hypospadias 360

I ileus 110 iliac arteries, left external, stenosis 322 iliac crest imaging 196 imatinab 113–114, 197 implants, cardiac 342 infants see children, abdominal MRI inferior mesenteric artery 305, 312 inferior mesenteric vein 312, 312 inferior phrenic arteries 311 inferior vena cava 312 anomalies, MRA 325–326 compression 326 obstruction, renal cell carcinoma 326, 326 occlusion 42, 42 renal tumor invasion 166, 326, 327 inflammatory bowel disease children 343, 345 MRI criteria for assessment 117, 343 see also colon; Crohn disease; rectum; ulcerative colitis inflammatory ovarian disease 361 inflammatory pseudotumor bowel obstruction, children 342, 344 splenic 96, 98, 99 in-phase (IP) imaging adrenal glands 181, 181, 182, 183, 193 liver 3, 5, 6, 6, 7, 12 insulinoma 80, 349–350 internal iliac artery, normal MRA 312 intersphincteric fistula 138 intestinal duplication 342 intra-abdominal lymph nodes 331–340 asymmetry, in metastases 336 benign vs malignant enlargement 331, 336 conglomerate 336, 337 contrast media 333–334, 336, 338 non-specific (gadolinium-based) 333, 334, 336 oral 333 tissue-specific 334 indications for MRI 331–332, 332 malignancy criteria 334, 336 metastases 331, 335, 336 melanotic 336, 337, 338 MRI advantages over CT 331 MRI appearance of abnormal nodes 336, 337, 338, 339 MRI imaging technique 332–334 imaging planes 332, 332 pulse sequences 332, 332–333, 333 MRI sensitivity 331, 334, 338, 339 multiple enlarged, in malignancy 336, 337

Index

normal MRI appearance 334, 334, 335, 339 pelvic see pelvic lymph nodes size 331, 334 malignancy diagnosis 334, 335, 336 spherical transformation 336 T1 w hyperintensity in melanoma 336, 337, 338 T1 w hypointensity and T2 w hyperintensity 334, 334, 336 upper abdomen, pulse sequences 332, 332, 333 intracervical mucus 249 intraductal papillary mucinous tumor (IPMT) 77, 79 intrahepatic block 351 intramural hematoma (IMH) of aorta see aortic intramural hematoma (IMH) intraperitoneal abscess, children 350 intrathoracic lymph nodes 336 intrauterine device (IUD) 241 iodine-based X-ray contrast media 8–9 iron chelation therapy 348 deposition/storage 36, 102, 105, 106 liver 35–36, 106 pancreas 71, 71–72 splenic 102, 103–104, 104, 105 iron gluconate 286 iron overload liver 35–36, 348, 348 pancreas 71, 71–72 iron oxide particles bile/pancreatic ducts contrast 48, 48–49, 51 pancreas imaging 69 see also superparamagnetic iron oxide particles (SPIO) iron storage disease 35–36, 71, 71–72 ischiorectal abscess 138, 139 islet cell tumors 80, 82, 83 liver metastases 28–29

K kidney 143–179 advantages of MRI 143 artifacts (wraparound) 143 atrophy 151, 154 benign conditions 150–161 congenital anomalies 150, 150, 352, 352 cysts see renal cysts duplex 352, 352 ectopic 150 functional imaging 174 hematoma 151, 152 in graft recipient 175, 176 horseshoe 150, 150 indications for MRI 143 inflammation/abscess 151, 152 living donor evaluation 174, 174 lymphoma 171, 172, 173 metastases in 171, 172, 173 MRI imaging technique 143–148 children/infants 352 coils 143–144 contrast-enhanced imaging 144, 147

contrast media 148, 174, 352, 353 precontrast imaging 144, 146 protocol for graft recipients 175 pulse sequences 144, 144, 145 for specific indications 144, 144 MRI vs CT 143 normal MRI appearance 146, 148 pathologic entities, MRI appearance 150–177 children 352–356 polycystic disease 154, 155–156, 352 stones 154, 154 trauma 151, 152 tumors see renal tumors vascular anatomy 174, 174 see also entries beginning renal Kinsbourne syndrome 359 Klatskin tumor 63, 64 Klippel–Trenaunay–Weber syndrome 93 Krukenberg tumor 112, 297, 297 k-space (Fourier space) 307 Kupffer cells 8, 42

L labor, protracted 301 lactate dehydrogenase (LDH), elevated 358 Larmor frequency 306, 307 left common iliac veins 327 leiomyoma bladder 211 gastric 112 small intestine 113 uterine see uterus, leiomyoma (fibroids) leiomyosarcoma epithelioid, myxoid/granular cell types 203 gastric 112 prostate 237, 238 renal 171 retroperitoneal 197, 202–203 leiosarcoma, small intestine 113 Leriche syndrome 317 leukemia, MRI of children 363 levator ani muscle 128–129, 130 lipoma presacral 362 retroperitoneal 202, 202 small intestine 119, 119 liposarcoma perirenal fatty tissue 160 renal 160, 171 retroperitoneal 200, 201, 202 littoral cell angioma 96, 97 liver 1–45 abscess 25, 26, 345 adenoma 11, 14, 18, 346 adenomatous hyperplasia, focal HCC in 42 advantages/benefits of MRI 1 atrophy 25, 38, 39 in cirrhosis 38, 39, 41 benign focal lesions 12–18 benign tumors 1, 12, 34 children/infants 345–346 see also hemangioma, liver biopsy, children 347 caudate lobe hypertrophy 43

371

cirrhosis see cirrhosis contrast-enhanced MRI 27–30 diffuse liver disease 34, 39 focal lesion characterization 34 focal lesion diagnosis 30–33, 31, 32, 33 hepatobiliary contrast media 8, 9, 27, 28–29, 30, 30 late postcontrast images 31–32 nonspecific contrast media 31–32 SPIO (iron oxide particles) 8–9, 27, 28, 29, 30 contrast inflow, temporal course 10, 10 cysts 12, 12, 13, 91 children 345, 346, 346 diffuse disease 34–43, 35 acute infections/inflammatory 37, 38 vascular 42, 42 edema (intrahepatic) 37, 38 eggshell calcification 12 fatty infiltration 3, 6 diffuse/generalized 35, 36 focal lesions 26, 26, 27 fibrosis 38–42 confluent, cirrhosis with 39, 40 see also cirrhosis focal lesions 1–34 benign 12–18 characterization 34 detection (contrast media use) 30–33, 31, 32, 33 fatty 26, 26, 27 hemorrhagic 22 malignant 19–22 role of MRI 30–34 focal nodular hyperplasia (FNH) 11, 14, 16, 27 dynamic MRI 30, 33 pedunculated 14, 17 granuloma 43, 345 hemangioma see hemangioma, liver hilum, enlarged lymph nodes 43 incidentally detected lesions, MRI for 1 indeterminate lesions, MRI techniques 2 indications for MRI 1, 2, 34 intraparenchymal hemorrhage 42 iron deposition 35–36, 106 malignant tumors 11, 12, 27 children 346–347 see also hepatocellular carcinoma (HCC) metastases 1, 2, 7, 11, 19–21 from breast cancer 20, 20, 101 calcification 19 children 347 from colorectal cancer 19, 19–20, 28–29, 121, 125 dynamic contrast-enhanced MRI 12, 19, 27, 31 fatty lesions resembling 26, 27 from gastrinoma 22 HCC vs 27, 34 hemangioma vs 13, 30, 31 hemorrhage in 19, 21, 22 hypervascular 12, 20, 21, 32 hypovascular 19, 19, 19–20, 30, 31 from islet cell carcinoma 28–29

372

Index

liver metastases melanotic (melanoma) 12, 20–21, 22, 101–102 neuroblastoma 358, 359 nonspecific contrast media 31–32, 91 from pancreatic carcinoma 77, 78 from rectal cancer 31, 132 from renal cell carcinoma 20, 21 tissue-specific contrast media 27, 28–29, 30, 31 MRI technique see liver, MRI technique nodule-in-nodule 42 nodules see hepatocellular nodules normal MRI appearance 11, 11 parasitic infections 37 pathology, MRI appearance diffuse liver disease 35–43 focal lesions 11–26 sarcoidosis 43 siderotic nodules 39, 40–41 transplantation 324, 324–325, 325 trauma, children 348 vascular disease 42, 42 Wilson disease 42–43 liver, MRI technique 1–11, 2, 5 coils 2 contrast media 1, 8–9 administration methods 9 nonspecific (gadolinium) 8, 9, 27, 29, 31–32, 34 tissue-specific see below fat suppression 2, 3, 4, 7, 35 imaging planes 1–2 motion artifacts and reduction of 5–8 see also motion artifact(s) patient positioning/preparation 1 protocols 9–11 dynamic 9–10, 10, 13, 14, 15, 34 multiphasic 10 static using tissue-specific contrast media 10–11 unenhanced 9 pulse sequences 2–8, 4, 5 3D GRE sequence 2–3, 5 in-phase and opposed-phase 3, 5, 6, 6, 7, 12 intermediate field strength T1 w 3 T1 w imaging 2–3, 5, 6, 7, 11, 12 T2 w imaging 3, 5, 5, 7, 12, 12 turbo/fast SE (TSE, FSE) 3, 5 tissue-specific contrast media 8–9, 10–11, 31 hepatobiliary (low-molecularweight) 8, 9, 27, 30, 30, 37 SPIO (iron oxide particles) 8–9, 27, 28, 29, 30, 34, 39 lumbar arteries 311 lumbar lymph nodes 334 Lumirem 342 lung metastases 121, 132 lymphadenopathy, retroperitoneal 326 lymphangioma abdominal, children 343 splenic 96 lymph-hemangiomatosis 350, 351 lymph nodes

hilum of liver 43 intra-abdominal see intra-abdominal lymph nodes pelvic see pelvic lymph nodes perirectal, metastases 133 sizes, normal vs metastases 334, 335, 336 lymphoma adrenal 190 bladder 215 children, MRI 350, 363 gastric 112 mesenteric 113 ovarian 298, 298 renal 171, 172, 173 small intestine 113, 119, 119 spleen involvement 90, 99, 99, 100 see also Hodgkin lymphoma; nonHodgkin lymphoma lymphoproliferative disease 356 benign atypical 363

M Mackenrodt ligaments (cardinal ligaments) 246, 246 magnetic resonance cholangiopancreatography (MRCP) see MRCP magnetic resonance elastography (MRE) 34, 38 magnetites see superparamagnetic iron oxide particles (SPIO) Magnevist (Gd-DTPA) 34, 66, 68, 127, 306 male genitalia, anomalies 361 malignant fibrous histiocytoma (MFH) 204, 205 maximum intensity projection (MIP), MRA 308, 322 Mayer–Rokitansky–Küster–Hauser syndrome 253, 360 Meigs syndrome 293 melanoma, malignant amelanotic, metastases 336 liver metastases 12, 20–21, 22 lymph node metastases 336, 337, 338 splenic metastases 100, 101 Ménétrier disease 112 meningomyelocele 362, 363 menstrual cycle endometrium thickness 247, 247, 248 ovarian changes 288 vaginal changes 249 mesenchymal tumors 197 granulosa cell 361 hemangioma of liver 13, 13–14 hepatic sarcoma 347 retroperitoneal see retroperitoneal soft-tissue tumors mesenteric ischemia 319, 319, 320 mesenteric lymphoma 113 mesorectal fascia 128, 131 mesorectum 130, 132 metastases see specific cancers and anatomic structures methemoglobin 292 MIBG scintigraphy, pheochromocytoma 192 microabscesses, splenic 102 middle sacral artery 311

middle suprarenal arteries 311 milk, whole, as contrast agent 342 mononuclear phagocytic system (MPS), magnetite uptake 8, 27 motion artifact(s) 3, 5–8 bowel movement 3, 5–6, 49, 110, 219, 241 causes 3, 5–6 ghosting/ghosts 6 minimization 6–8, 9 gastrointestinal MRI 110, 111 pancreatic MRI 66 prostate MRI 219 respiratory triggering 8, 9, 28–29 see also breath-holding techniques; fat suppression; presaturation band pelvic MRI 241 pulsation artifacts 3, 6, 7 respiratory see breathing motion artifacts urinary bladder, endometrial carcinoma MRI 263 MP-RAGE (magnetization-prepared rapid acquisition gradient echo) 323, 336 MR angiography (MRA) 305–330 contrast-enhanced 3D 305, 306–310 automatic injectors and rates 307 coils 309 contrast media 306–307 k-space and timing 307 parallel imaging techniques 308 patient examination 309, 309–310 postprocessing techniques 308, 308, 310, 311, 322 protocol 309–310 pulse sequences 307–308, 308 flow-related signal enhancement (inflow effects) 305–306 flow-related signal loss (phase shifts) 305, 306 imaging technique 305–310 indications 305 kidney and upper urinary tract 143, 144, 147, 147 liver transplantation evaluation 324, 324, 325 normal anatomy, appearance 311–312, 312 pancreas 66, 83 pathologic entities, MRI appearance 313–328 children/infants 341, 350, 351, 351 pelvic inflammatory disease 288 phase-contrast (PC) 306 renal transplantation evaluation 321, 322, 323 time-of-flight (TOF) 305–306, 327, 327 unenhanced 305–306 see also individual arteries/veins MR colonography (MRC) 109, 109 colorectal cancer 120, 121, 122 inflammatory disease of colon 123–125 water enema or air insufflation 110 MRCP 47–64, 65 advantages and cost-benefits 47 children 348, 349

Index

indications for 47 maximum intensity projection (MIP) algorithm 48, 49, 52 normal 51, 52, 55 oral contrast media 48–49 negative (iron oxide) 48, 48–49, 51 in pancreas MRI protocol 66 patient preparation 48–49 recommended protocol 49, 50, 51, 51, 52 technique 47–51 coils 47 pulse sequences 48, 49, 50 MR defecography 140, 140 MR enteroclysis 109, 109, 110 MR hysterosalpingography 288, 289 MR pelvimetry 301–304 advantages and indications 301 imaging technique and findings 303 pelvic measures 302, 302–303 postpartum 301, 302, 302–303 prenatal 301, 303 pulse sequences 303, 303 reference values 303 MR spectroscopy, 3D 222–223 free induction decay (FID) signal 223 frequency spectrum 222, 223 prostate see prostate; prostate cancer MR urography (MRU) 143, 144, 147, 148 3D MR angiography with 305 renal graft recipients 175 T1 w excretory, and T2 w static 148, 149 mucoviscidosis (cystic fibrosis) 71 müllerian duct, anomalies 252–255, 360 agenesis 360 cysts 223 incomplete/absent (unicornuate uterus) 252, 254, 254 partial nonfusion 252, 254, 254 multicystic renal dysplasia 352, 353 Multihance (Gd-BOPTA) 8, 9, 31, 306, 307 multilocular cystic nephroma 160 multiplanar reconstruction (MPR) 308, 322 multiple endocrine neoplasia type 1 (MEN 1) 80 multiple endocrine neoplasia type 2A/2B 192 myelolipoma, adrenal 188, 189 myometrium invasion by endometrial carcinoma 262, 263, 264, 266 invasion by endometrial tissue (adenomyosis) 258, 260, 260–261 normal MRI appearance 247, 247, 248

N Nabothian cysts 246, 249, 262, 262 navigator echo technique 8, 9, 48 neonates and infants, MRI imaging 92, 341 see also children, abdominal MRI

fibroma see ovarian fibroma follicular cysts 287–288, 289 functional cysts 289, 290, 361 functioning tumors 293, 294 hemorrhagic cysts 289, 290 indications for MRI 285 inflammatory disease in children 361 lymphatic drainage 287 lymphoma 298, 298 malignant tumors 294–296 children 361 metastases 295, 297 primary see ovarian cancer menstrual cycle effects 288 MRI imaging technique 285–286 contrast media 286 imaging planes 285 pulse sequences/protocol 285–286, 286 normal anatomy 287 normal MRI appearance 287, 287–288 pathologic entities, MRI appearance 288–298, 361 polycystic ovary syndrome 289–290, 361, 362 torsion 289, 292, 361 zonal anatomy 287, 288

nephroblastoma see Wilms tumor (nephroblastoma) nephroblastomatosis 355, 355 nephrogenic systemic fibrosis (NSF) 148 nephroma congenital mesoblastic 355–356 multilocular cystic 160, 356 neuroblastoma 192–193, 358, 358–359, 359 staging 359 neuroendocrine tumors 79, 80, 125 neurofibroma, and neurofibromatosis 204, 205 neurovascular bundle, prostate base see prostate Niemann–Pick disease 102 non-Hodgkin lymphoma adrenal 190 inflammatory bowel disease vs 345 lymph node metastases 331, 336 mesenteric 113 ovarian 298, 298 spleen involvement/MRI 99, 99, 100 see also lymphoma

O omentum, endometrial carcinoma invasion 265 Omniscan 8–9, 66, 68, 306 oncocytoma 155, 158, 159 opposed-phase (OP) imaging adrenal glands 181, 181, 182, 183, 193 liver 3, 5, 6, 7 opsoclonus 359 Optimark (gadoversetamide) 8–9 oral contraceptives 14, 248, 249 Ormond disease (retroperitoneal fibrosis) 204, 206 ovarian cancer 285, 294–295 borderline 295, 296, 298 criteria of malignancy 295, 296, 297 FIGO/UICC stages 297 histologic classes 294–295 lymph node metastases 331 recurrence 298, 298 serous papillary 296 staging laparotomy 295 ovarian fibroma 293, 294 malignant transformation 293, 294, 295 uterine leiomyoma vs 256 ovarian follicles 287–288, 288 ovarian fossae 287 ovarian ligament 245, 287 ovarian teratomas 361, 362 ovarian veins 287, 326–328 2D TOF MRA 327, 327 thrombosis, septic puerperal 328 ovaries 285–299 benign tumors 289–293 cystic 289–293, 361, 362 benign vs malignant masses 286, 295 blood supply and venous drainage 287 cortex, and medulla 287, 288 cystadenoma 290, 291, 295, 295 dermoid cysts 292–293, 293 endometriotic cysts 290, 292, 292

373

P pacemakers 342 PACE technique, bile/pancreatic duct imaging 48 pancreas 65–86 anatomy 65, 65, 70 annular 66, 71 atrophy 74 calcification 74, 76 chronic 75 congenital anomalies/diseases 71–72 see also pancreas divisum cystadenoma, children 350 cystic fibrosis 71 cysts 71 endocrine/exocrine secretions 49, 65 outflow obstruction 52 fibrosis 74 focal enlargement 80 functions 65 head 69, 70, 70 carcinoma 76, 76–77 enlargement 71, 72, 75 pseudocyst 73 indications for MRI 65, 66 inflammatory disease see pancreatitis lipomatosis 70, 70 lymphatic drainage 65, 70, 77 mass(es) 57, 61, 63 MRI imaging technique 66–69, 67 contrast media 66, 68–69 dynamic 66, 68, 69, 83 imaging planes 66 pulse sequences 66, 67, 68, 69 MRI vs CT 65, 72, 74, 77 neoplasms 76–80 ampullary cancer 77, 77

374

Index

pancreas neoplasms carcinoma see pancreatic carcinoma children 349–350 cystic 77, 79, 79 islet cell tumors 28–29, 80, 82, 83 pseudotumors vs 80, 83, 84 normal MRI 70 normal parenchyma 68, 69, 70 pathologic conditions, MRI appearance 71–83 children/infants 348–349 primary hemochromatosis 71, 71–72 pseudocysts see pancreatic pseudocysts pseudotumors, vs neoplasms 80, 83, 84 string of pearls appearance 73, 74 transplant 83 pancreas divisum 52–53, 54, 58, 71 MRI indication 66 pancreatic carcinoma 76–77 absence of duct-penetrating sign 83 benign lesions vs 65 children 350 classification 77 head of pancreas 76, 76–77 metastases and spread 77, 78 MRI indication 66 neuroendocrine 21, 79, 80 portal vein invasion 324 signal intensity 77, 78 tail of pancreas 76, 78 vascular encasement 77, 83 pancreatic duct 47–64 accessory 52, 71 anatomic variants 52–53, 55, 56, 71 carcinoma 76 dilatation 73, 75, 77 gradual narrowing vs abrupt obstruction 83 inflammatory conditions 55 MRCP protocol 49, 50, 51, 51, 65 see also MRCP normal anatomy, MRI 51, 52, 52–53, 55 normal MRCP 51, 52 pathologic entities, MRI appearance 53–63 small, normal 55 stones 55, 56, 58–59, 74 children 348, 349 strictures/stenoses 53, 55, 55 children 349 vs malignancy 55 pancreatic pseudocysts 55, 64, 72, 73, 74, 75–76 children 348 hemorrhagic 74 pancreatitis 66, 72–76 acute 54, 56, 72, 74 complications 72, 73, 74 edematous 72, 72 children 348, 349 chronic 52, 55, 65, 70, 73, 74–75 recurrent 48, 80, 348, 349 papilla of Vater 52, 55 papillary sclerosis 57 para-anal horseshoe abscess 139

paracolpium (paravaginal tissues) 248, 249, 251 paraesophageal varices 323 paraganglioma 192 parallel imaging 2, 128, 143 paramagnetic low-molecular-weight compounds 8, 9, 68 3D MR angiography 306, 307 parametrium 246, 270 cervical cancer involvement 271, 271, 272 normal MRI appearance 249, 251 parapelvic cysts 154, 156 parasitic cysts, adrenal 186 parasitic infections, liver 37 paraumbilical veins 324 paravaginal tissues 248, 249, 251 parovarian cyst 290, 290 paroxysmal nocturnal hemoglobinuria 36, 105, 105 pelvic abscess 345, 345 pelvic arteries normal MRA 311–312, 312 pathologic conditions, MRA 313–321, 318 pelvic floor, functional disorders 140 pelvic inflammatory disease (PID) 288, 289 pelvic inlet, measures 302, 302–303 pelvic lymph nodes contrast media for MRI 333, 334 metastatic, MRI appearance 336 MRI pulse sequences 128, 332, 332–333, 333 normal 334, 334, 335 size for malignancy diagnosis 336 pelvic MRI (female) 241–244, 285–286 artifacts 241 congenital uterine anomalies 252–253, 253 contraindications 241 contrast media 244 imaging planes 242, 285 imaging strategy/protocol 242, 244, 244 menstrual status 285 patient preparation/position 241, 285 pulse sequences 242, 285 adnexa imaging 285–286, 286 congenital anomalies 253, 253 uterus and vagina imaging 242–243, 244 pelvic outlet, measures 302, 303 pelvic veins 2D TOF MRA 327, 327 normal MRA 312 pathologic entities, MRA 326–328 thrombosis 326–328, 327 pelvimetry 301–304 MR see MR pelvimetry penetrating atherosclerotic ulcer (PAU) 314 Pepper syndrome 359 perfluorooctyl bromide, as contrast agent 342 perfusion-weighted imaging (PWI), rectal/anal canal imaging 128 perianal abscess 136, 137–138 perianal fistula 127, 128, 136, 137, 137

peripancreatic fluid 70, 72, 75 perirectal abscess 127 perirectal fat 131, 132, 133, 135 perirectal lymph nodes, metastases 132, 133 perirenal hematoma 151, 152 peristalsis, motion artifact 3, 5–6, 49, 110, 219, 241 peritoneal carcinosis, gelatinous 121 peritoneal cyst 298 peritoneum, covering uterus 246, 246 phase contrast (PC), MR angiography 306 phased-array coils see coils, phasedarray phase-encoding gradient 307 phase-ordering technique (ROPE) 8, 9 phase shifts, MR angiography 305, 306 pheochromocytoma 190, 192, 192, 193 bladder 211 children 359–360 polycystic kidney disease 154, 155–156, 352 polycystic ovary syndrome 289–290, 361, 362 polyethylene glycol electrolyte solution (PEG–ES) 109, 110, 343 polyps colon 120 endometrium 261, 261 gallbladder 55, 61 porphyria cutanea tarda 36 portal hypertension 34, 184, 323, 324 children 360 MR angiography 321, 323–324 retroperitoneal collaterals 324, 324 portal veins invasion by pancreatic tumor 324 normal MRA 312, 312 pathological MRA 308, 312, 312, 321, 323–324 children 350–351 stagnant blood in 350 stenosis 324 thrombosis 40, 91, 350–351 portal venous system contrast-enhanced 3D MRA 323, 323 normal MRA 312, 312 pathologic entities, MRA 321–325 intraluminal dephasing phenomena 323 portosystemic collaterals, MRA 321, 323–324 portosystemic shunts 325 positron emission tomography (PET), MRI vs, colorectal cancer 121 posthepatic block see hepatic venous outflow postmenopausal bleeding 262 postmenopausal women ovaries 288 uterus 248, 249, 277 vagina 249 pregnancy 261, 301 ectopic 288 MRI during first trimester 301 MR pelvimetry see MR pelvimetry ovarian veins 326 prehepatic block 350

Index

presacral teratoma 362–363, 364 presacral tumors, children 362–363, 363 presaturation band 241, 242, 243 liver imaging 6–7 pancreas imaging 66 presaturation pulses 242, 242, 243 PRESS (point-resolved spectroscopy) 222 primary biliary cirrhosis (PBC) 55 primary sclerosing cholangitis (PSC) 61, 63 Primovist (Gd-EOB-DTPA) 8, 9, 31, 31, 33 Prohance (gadoteridol) 8–9 prostate 219–240 abscess 224 anatomy 220–221, 221 benign hyperplasia see benign prostatic hyperplasia (BPH) cancer see prostate cancer (below) congenital anomalies 223 cysts 223, 224 indications for MRI 219 inflammatory conditions 224–225 malignant tumors 226–238, 237 adenocarcinoma see prostate cancer atypical carcinomas 236, 237 mucinous carcinoma 236–237 rhabdomyosarcoma 362 sarcoma 237, 237–238, 238 transitional cell carcinoma 237 MRI imaging technique 219–220 contrast media 220, 238 dynamic imaging 220 endorectal coil 219, 221, 235 imaging planes 219–220 pulse sequences and protocol 220, 220 MRI/MR spectroscopy combined 222, 232, 233 MR spectroscopy (3D) 222, 222–223, 223 citrate, choline and creatine 222–223, 232, 233 neurovascular bundle at base 221, 221, 222 cancer involvement 228, 229 nodules 225, 225 normal MRI appearance 129, 220–223 pathologic entities, MRI appearance 223–238 pseudocapsule 225, 225 scar formation 224 sensitivity of MRI 219 T1 w image homogeneity 221 T2 w images 221, 226, 235, 238 zonal subdivision 221, 221, 226 transitional zone 221, 221, 223, 225 prostate cancer 226–235, 235, 236 accuracy of MRI 230 benign prostatic hyperplasia with 225, 226 biopsy 226 hemorrhage after 227, 227, 228 MR-guided 234, 234 capsular penetration criteria 227–228, 228

dynamic contrast-enhanced 230, 232, 232 endorectal coil MR imaging 226–232 extent assessment 227–228, 232 extracapsular 227–228, 228, 229 lymph node involvement 235, 236, 336 MRI/MR spectroscopy combined 232, 233, 234 MRI vs CT or ultrasound 235 MR spectroscopy 232–234, 233 neovascularization 226 peripheral/transitional zone 226, 227 radiologist's experience importance 230, 232 radiotherapy, MR spectroscopy after 233, 235 recurrent 219, 235, 235, 236 routes of spread 228, 229, 230 seminal vesicle invasion 228, 229, 230, 230 small solitary tumors 228 stage T1 and T2 226 stage T2A 222 stage T3 and T3 b 226, 229, 230 stage T4 226, 228, 230, 231 staging/TNM classification 219, 226, 226 prostatectomy, prostate cancer recurrence after 235, 235, 236 prostate-specific antigen (PSA) 219, 230, 233, 235 prostatic capsule 222, 227 prostatic utricle 360 prostatitis 224 proton density (PD) sequence 90, 332, 341 proton motion 305 pseudocapsule hepatocellular carcinoma 21, 22, 23, 24 prostate 225, 225 soft-tissue sarcoma 200 uterine leiomyoma 256, 257, 260 Wilms tumor 353, 354–355 pseudocysts adrenal 186 pancreatic see pancreatic pseudocysts splenic 93, 102 pseudodiverticula, urinary bladder 210 pseudomyxoma peritonei 121 pubic symphysis, normal MRI appearance 129 pubococcygeal line 140 pulmonary sequester 351 pyosalpinx 288

R radiation cystitis 211 radiation enteritis 115, 121, 123 radiation-induced hepatitis 37 radiotherapy cervical carcinoma 277, 278, 279, 280 prostate cancer 233, 235 uterus MRI appearance after 276–277, 277, 280

375

vaginal MRI appearance after 277, 277–278 readout gradient 307 rectal ampulla 128 rectal cancer 131, 132 circumferential resection margin 132 distant metastasis 132 lymph node metastases 132, 133 MRI pulse sequences/parameters 129, 132 MRI role/appearance 132–133, 133, 134 recurrence 121, 135, 135–136 spread 120–121, 127 staging 132, 132 T1, T2 and T3 132, 133 T4 133, 134 total mesorectal excision 128, 132 see also colorectal cancer rectocele 140, 140 rectouterine pouch (pouch of Douglas) 248, 287, 288, 298 rectum 127–141 anatomy 128–129, 129, 130 anterior displacement, children 343 blood supply 129 cervical cancer invasion 272, 273, 278, 281 Crohn disease 136–138, 137, 138 functional disorders 140 indications for MRI 127 inflammatory bowel disease 127, 128, 136–138 lymphatic drainage 120, 132 MRI imaging technique 127–128 MRI vs CT 127 normal, MRI appearance 128–131, 129, 130 pathologic entities, MRI appearance 131–140 pelvic (rectal ampulla) 128 perineal see anal canal venous drainage 132 wall, layers 129, 130, 131 relaxation enhancement (RARE) technique, MRCP 48 renal adenoma 155, 161 renal angiosarcoma 171 renal arteries aneurysm 322 contrast-enhanced MRA 311, 320, 320–321 evaluation in living kidney donors 320, 321, 322 evaluation in transplant recipients 321, 322, 323 normal MR angiography 311 occlusion, children 356, 356 PTA and stenting 320–321 stenosis 320, 320, 321, 323 children 356, 357 variants 320, 321, 322 renal cell carcinoma (RCC) 161–163, 162 3D MRA 163, 168 cystic 168 extent determination by MRI 161, 163, 164–168 inferior vena cava invasion 166, 326, 326, 327

376

Index

renal cell carcinoma (RCC) kidney contour distortion 162, 165, 166–167 liver metastases 20, 21, 339 metastatic lymph nodes 163, 168, 335, 336 Robson/TNM classification 162 venous invasion 163, 166, 326 renal collecting system, dilatation 154 renal columnar hypertrophy 356 renal cysts 151, 152, 154, 155 complicated 154, 156, 157 hemorrhagic 154, 157 polycystic kidney disease 154, 155–156, 352 simple 154, 155 renal dysplasia, multicystic 352, 353 renal infarction 356 renal inflammation 151 renal lymphoma 171, 172, 173 renal oncocytoma 155, 158, 159 renal transplant 175 3D MRA evaluation after 321, 322, 323 complications 175, 176, 177 living donor evaluation 174, 174 3D MRA 320, 321, 322 MRI imaging protocol 175 normal MR angiography 322 rejection 175, 177 renal trauma 151, 152 renal tumors benign 155, 158, 158, 159 angiomyolipoma 160, 160, 161 multilocular cystic nephroma 160 oncocytoma 155, 158, 159 rare 158, 160–161 small solid tumors 161 children/infants 353–356 malignant 161–173, 162 lymphoma 171, 172, 173 metastases 171, 172, 173 primary 161–171 rare 162, 171 transitional cell carcinoma 163, 169, 170 see also renal cell carcinoma (RCC) renal veins compression 326 diseases in children 356, 356, 357 left 323–324 retroaortic left 325 thrombosis 326, 356, 358 tumor-related obstruction 326, 326 renovascular hypertension 320 repetition time (RT), MR angiography 306, 307 Resovist 9, 10, 27, 28–29 respiratory triggering, bile/pancreatic duct imaging 48 respiratory motion artifacts see breathing motion artifacts respiratory triggering 8, 9, 28–29 retrocrural lymph nodes 336, 339 retroperitoneal collaterals 324, 324 retroperitoneal fat 184 retroperitoneal fibrosis, idiopathic 204, 206 retroperitoneal fistula 195 retroperitoneal lymphadenopathy 326

retroperitoneal lymph nodes 335 metastases 331 normal 334 retroperitoneal soft-tissue tumors 197, 200 benign vs malignant 200, 201 characteristics (margination/ growth/necrosis) 200–202 classification 197, 198–199 enhancement patterns 200 epidemiology/incidence/staging 197, 199 fibrosarcoma 203, 204 initial diagnostic work-up 200–206 invasion/spread 200, 201, 202 leiomyosarcoma 197, 202–203 lipoma 202, 202 liposarcoma 200, 201, 202 malignant fibrous histiocytoma 204, 205 mixed benign 200, 201 MRI follow-up scheme 206 patterns of spread 195 rare tumors 204, 205 recurrence, MRI 206 sarcomas 197, 200–201 T2 w and T1 w sequence intensity 201 retroperitoneal xanthofibrogranulomatosis 204, 206 retroperitoneum 195–207 advantages of MRI 195, 197, 206 children/infants, MRI imaging 195, 352 diffuse (nontumorous) diseases 204, 206 indications for MRI 195, 196 metastatic disease 206 MRI imaging technique 195–197 coils 195 contrast media 196–197, 200 imaging planes 195–196 pulse sequences 196, 196 normal anatomy 197 tumorlike lesions 197 tumors see retroperitoneal soft-tissue tumors rhabdomyolysis 103–104 rhabdomyosarcoma bladder 215, 362 prostate 237, 237, 362 renal 171 vaginal 363 rheumatoid aortitis 316 rim enhancement 30 Rokitansky–Aschoff sinuses 62 Rokitansky protuberance 292 Rosai–Dorfman syndrome 363

S sacral flexure 128, 129 salpingitis 288 salpinx see uterine tubes santorinicele 52 Santorini duct 52, 71 sarcoidosis 43, 102, 103 sarcoma bladder 215 mesenchymal, hepatic 347 prostate 237, 237, 238 see also specific types of sarcoma

schistosoma mansoni 37 schwannoma 204, 205 secretin 49 sedation, children/infants 341–342 seminal colliculus 221, 221 seminal vesicle 219–240 agenesis 223 anatomy 221, 221 bladder cancer involvement 212, 214, 215 ectopic 223 inflammation 224–225 MRI imaging 222 prostate cancer invasion 228, 229, 230, 230 see also prostate sensitivity encoding technique (SENSE) 143 septic puerperal ovarian vein thrombosis (SPOVT) 328 Sertoli–Leydig cell tumors 293 shading sign 292 short tau inversion recovery (STIR) sequence colorectal cancer 121, 122 islet cell tumors 80 liver imaging 7–8 rectal/anal canal imaging 128, 137 sickle cell anemia 103, 350 siderotic nodules (Gamma–Gandy bodies) 39, 105, 106 sigmoidovesicular fistula 123, 124 signal-to-noise ratio (SNR), coils improving 2, 47, 127, 143, 219, 309, 341 Sinerem 8, 9, 334, 336, 339, 340 single-shot technique, liver imaging 3 situs inversus 350 slice-select gradient 307 small intestine 113–119 adenocarcinoma 113 Crohn disease 114–115, 116 diverticula 115 leiomyoma and leiosarcoma 113 lipoma 119, 119 lymphoma 113, 119, 119 MRI sequence 115, 117, 117–118 radiation enteritis 115 strictures 117 thickening of wall 114, 115, 116, 117 see also gastrointestinal tract soft-tissue tumors 197 classification 197, 198–199 epidemiology/incidence 197, 199 retroperitoneal see retroperitoneal soft-tissue tumors somatostatinoma 80, 83 spasmolysis see antispasmodic agents spatial encoding, MR angiography 307 specific absorption rate (SAR) 301 spectral presaturation with inversion recovery (SPIR) 8, 9 spherical transformation, lymph nodes 336 sphingomyelin, accumulation 102 spinal canal, neuroblastoma invasion 358, 359, 359 spindle cell tumors 203 spin echo (SE) sequences for children/infants 341 in pregnancy 301 retroperitoneum imaging 196

Index

spleen 87–108 abscesses 102, 350 angiosarcoma 100 atrophy 103 benign tumors 93–98, 350 cysts 93, 93, 94–95, 350 children 350 epidermoid 93 diffuse disease affecting 103, 105–106 ectopic 350 focal lesions, contrast media 89, 90 Gaucher disease 102 hamartoma 96 hemangioma 93, 95, 95–96 hemangiomatosis 95, 96 hematoma 102 hemorrhagic lesion 96 indications for MRI 87 infarction 103, 104, 350 infectious conditions 102 inflammatory pseudotumor 96, 98, 99 inhomogeneous appearance 95, 96, 100 iron deposition 102, 103–104, 104, 105 littoral cell angioma 96, 97 lymphangioma 96 lymphoma 90, 99, 99, 100 malignant tumors 88, 99–100 children 350 metastases 100, 101, 101–102, 350 MRI imaging technique 87–90, 89, 89 contrast media 87, 89–90, 91, 91, 95 imaging planes 87, 88, 89 pulse sequences 87, 89, 91 sequence protocol 90 SPIO, signal loss after 99 SPIO preparation use 90, 91–92, 92 neonatal, MRI appearance 92 noninfectious conditions 102 normal MRI appearance 88, 91, 91–92 pathologic conditions, MRI appearance 93–106 children/infants 350 pseudocysts 93, 102 rupture 350 sarcoidosis 102, 103 sphingomyelin accumulation 102 trauma 102, 350 splenic artery, pseudoaneurysm 74 splenic vein 312 thrombosis 66, 74 splenomegaly 103, 104, 105, 106 splenoportal venous axis 323–324 splenorenal shunt, spontaneous, contrast-enhanced MRA 308, 324 STEAM sequences (stimulated echo acquisition mode) 222 Stein–Leventhal syndrome 289–290 stenting abdominal/pelvic arteries, 3D MRA 314, 314 renal artery 320–321 stomach 112–113 adenocarcinoma 112, 112, 113, 114

metastases 112 staging 112 leiomyoma, leiomyosarcoma 112 lymphoma 112 MRI sequences/technique 112, 112–113, 113, 114 signet-ring carcinoma 112, 297, 297 see also gastrointestinal tract string of pearls appearance, pancreas 73, 74 subcutaneous fistula, Crohn disease 138, 138 superior mesenteric artery 305 atherosclerotic stenosis 320 normal MRA 312, 319 superior mesenteric vein 312, 312 superparamagnetic iron oxide particles (SPIO) 27, 28, 29, 30, 30, 31 liver imaging 8–9, 39 splenic imaging 90, 91–92, 92, 99 ultra-small see ultra-small superparamagnetic iron oxide particles (USPIO) supralevator abscess 138, 139, 140 surface rendering, MRA 308, 322 syphilitic aortitis 316

T Takayasu aortitis/arteritis 316, 319, 350 tampon use, vaginal MRI 241, 249 target sign 19, 19, 117 teratoma mature cystic 292–293, 293 ovarian 361, 362 presacral 362–363, 364 Teslascan (Mn-DPDP) 8, 9, 31, 34, 69 testicular feminization 360 testicular tumors 331, 361 testicular veins 326 testis, undescended 361 thecomas 293, 294 thorotrastosis 103 time of inversion (TI) 7 total mesorectal excision (TME) 128, 132 transesophageal echocardiography (TEE), aorta 314–315 transitional cell carcinoma (TCC) bladder 211–214 prostate 237 renal pelvis 163, 169, 170 transjugular intrahepatic portosystemic shunt (TIPS) 325 translevator fistula 140 transrectal prostate biopsy, MRguided 234, 234 transsphincteric fistula 138, 138 transvaginal ultrasound (TVUS) 285, 288 trimethylsilyl propionate (TSP) 222 TrueFISP, gastrointestinal tract 110–111 colorectal cancer 120, 121, 122 functional disorders of rectum 140 TSE sequence adnexa imaging 286 for children/infants 341 intra-abdominal lymph nodes 332, 332, 333, 334 liver MRI 3, 5

377

pelvic MRI 242, 252 rectum MRI 128 retroperitoneum MRI 196 tubal cancer/conditions see uterine tubes tuberculosis, adrenal 184–185 tubo-ovarian abscess 288 turbo inversion recovery sequence (TIR) 8 turbo spin echo sequence see TSE sequence tyrosine kinase inhibitor (imatinab) 113–114, 197

U ulcer, penetrating atherosclerotic 314 ulcerative colitis 121, 123, 123, 124 children 343, 345 Crohn disease vs 123 ultra-small superparamagnetic iron oxide particles (USPIO) 9, 307, 334 endometrial carcinoma staging 266 malignant lymph nodes 331, 332, 334, 336, 340 normal lymph nodes 339 optimal imaging time point 334 retroperitoneum imaging 196 ultrasound in children/infants 342 pelvic 241, 266–267 prostate cancer 235 transvaginal 285, 288 ureter 246 dilatation and obstruction 154 ectopic insertion 150, 151 stones 154, 154 see also kidney; urinary tract, upper urethra 221 carcinoma 212, 215 diverticula 210, 211 urinary bladder see bladder (urinary) urinary incontinence 150 urinary obstruction 154 urinary tract, upper 143–179 MRI imaging technique see under kidney obstruction 154 see also ureter urinoma 175, 175, 357 urothelial carcinoma see transitional cell carcinoma (TCC) uterine arteries 259, 287 uterine artery embolization (UAE) 255 uterine tubes blockage 288 cancer 297 dilated 290, 291 functioning tumors 293, 294 imaging technique 285–286 indications for MRI 285 MR hysterosalpingography 288, 289 MRI imaging technique 285–286 normal anatomy/MRI 287 uterosacral ligament 246, 246 uterovesical ligament 246, 246 uterus 241–283 acquired benign disorders 255–262 adenomyosis 258, 260, 260–261, 262

378

Index

uterus agenesis (complete/partial) 253, 253 arcuate 255 atrophy 277 benign vs malignant tumors 256 bicornuate 252, 254, 254–255, 360 congenital anomalies 252, 252–255 pulse sequences 253, 253 corpus 245, 245 dilatation and curettage, MRI after 276 double 252, 253, 254, 255, 360 dynamic contrast-enhanced 244, 257, 259 gestational trophoblastic disease 261 hormone treatment affecting 248, 249 hypoplasia 253 indications for MRI 241, 252 isthmus 245, 245 junctional zone 247 focal 260, 260–261 thickened, adenomyosis 260, 260 leiomyoma (fibroids) 255–257, 263, 279 adenomyosis with 258, 260 degenerative changes 258, 257 intramural 255, 256, 256 MRI appearance 256, 256–257, 257 ovarian fibroma vs 293 pedunculated 256, 257, 258, 293 pseudocapsule 256, 257, 260 pulse sequences 256, 256 submucosal 255, 256 subserosal 255, 256, 257, 258 ligaments supporting 246, 246 malignant tumor see endometrial carcinoma before menarche 247, 248 menstrual cycle effects 247, 247, 248 MRI imaging technique see pelvic MRI MRI vs CT/ultrasound 241, 266–267 neonatal 247 normal MRI appearance 245, 245–249, 246, 247 peritoneum covering 246, 246 postmenopausal women 248, 249

radiotherapy, MRI appearance after 276–277, 277 residual tumor after 278, 280 rudimentary horn 254, 254 septate 252, 255, 255 T1 w sequences 244, 246–247, 247, 253 T2 w sequences 244, 248, 249, 253 unicornuate 252, 254, 254 zonal anatomy 246–247, 247, 248 uterus didelphys 252, 253, 254, 360

V vagina 241–283 agenesis 254, 274 atresia 360 cervical cancer invasion 271, 272, 273, 275 congenital anomalies 274 cysts 274 double 254, 274, 274, 360 endometrial carcinoma invasion 262 indications for MRI 241 menstrual cycle effect 249 MRI imaging technique see pelvic MRI normal anatomy 245 normal MRI appearance 249, 251 pathologic entities, MRI 274–276 postmenopausal women 249 radiotherapy effect 277, 277–278 stenosis, radiation-induced 277, 277, 278, 278 T1 w images 249, 251 T2 w sequence (axial) 244, 249 tumors 274–276 pulse sequences/protocol 275–276, 276 rhabdomyosarcoma 363 staging 275, 275–276, 276 wall, carcinoma involvement 271, 276 vaginal bleeding 267 vaginal fornix, normal MRI 249, 250 vaginal septum 253–254 vaginal stump 278, 280 vaginal vault, ovarian cancer 298, 298 vascular malposition 350, 351 Vasovist (gadofosveset) 307

vena cava displacement 353, 355 thrombosis 358 see also inferior vena cava vena cava compression syndrome 303 VIBE sequence (volume-interpolated breath-hold examination) adrenal gland MRI 182, 182 gastrointestinal tract 109, 110, 111–112, 115 colorectal cancer 120, 121, 122 inflammatory disease of colon 124 kidney and upper urinary tract 144, 146 liver MRI 3 pancreas MRI 66 retroperitoneum MRI 196 see also GRE sequences, 3D VIPoma 80 visceral arteries aneurysm 319 mesenteric ischemia 319, 319, 320 MR angiography 318–319, 319

W Wilms tumor (nephroblastoma) 171, 171, 353–355, 354–355 benign variant 160 differential diagnosis 355, 358 extrarenal 353 monitoring, MRI 353, 354–355 subtypes and staging (SIOP) 353, 353 Wilson disease 42–43, 347

X xanthofibrogranulomatosis, retroperitoneal 204, 206

Y yolk sac tumors 361

Z Zollinger–Ellison syndrome 80