Magnetic Resonance Tomography

M.F. Reiser · W. Semmler · H. Hricak (Eds.) Magnetic Resonance Tomography

M.F. Reiser · W. Semmler · H. Hricak (Eds.)

Magnetic Resonance Tomography With 1260 Figures and 175 Tables

123

Maximilian F. Reiser, Univ.-Prof. Dr. med. Dr. h.c. Institute for Clinical Radiology University Hospitals Grosshadern Ludwig-Maximilian University of Munich Marchioninistr. 15 81377 Munich Germany Wolfhard Semmler, Univ.-Prof. Dr. rer. nat. Dr. med. Division of Medical Physics in Radiology German Cancer Research Center Im Neuenheimer Feld 280 69120 Heidelberg Germany Hedvig Hricak, MD, PhD, Dr. h. c. Professor of Diagnostic Radiology and Chairman Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10065 USA

Parts of this book have been translated from the German original: M. Reiser, W. Semmler (eds) Magnetresonanztomographie 3rd ed. Springer 2002 ISBN 978-3-540-29354-5    e-ISBN 978-3-540- 29355-2 DOI 10.1007/b135693 Library of Congress Control Number: 2007933311 © 2008 Springer-Verlag Berlin Heidelberg This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broad-casting, reproduction on microfilm or any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registed names, trademarks etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product liability: the publishers cannot guarantee the accuracy of any information about dosage and application contained in this book. In every individual case the user must check such information by consulting the relevant literature. Cover design: Frido Steinen-Broo, eStudio Calamar, Spain Printed on acid-free paper 987654321 springer.com

Preface

This textbook—which describes the entirety of MRI in a single volume—is now a more than fifteen-year-old tradition. First published in German in 1992, it was updated every five years to keep up with the rapid advancement of the technology and clinical applications of MR tomography. Because it covered its subject in great breadth and detail, it became one of the most popular textbooks on MR tomography in German-speaking parts of the world. Each subsequent edition not only summarized well-established facts about MR tomography for practical application, but also discussed new procedures and insights acquired during the years since the previous edition. The present, 4th edition maintains this tradition—only it does so in the English language. Today, experts in science and medicine are distributed throughout the world, and English is gaining acceptance as the “lingua franca” of these fields. The idea of publishing the book in English was discussed multiple times over the years and was actively supported by Springer as represented by Dr. Ute Heilmann. Our goal was not simply to produce an English translation of a German book, but to produce a volume geared to the interests of an international community. We could not have reached this goal without the collaboration of Dr. Hedvig Hricak, who



agreed to come on board as an editor from an Englishspeaking country. Dr. Hricak introduced new ideas and topics, recruited additional authors who are experts in their fields of study, and with her enthusiasm and persistence substantially enriched and advanced this project. We are now extremely pleased to be able to present this English-language volume covering all aspects of MR imaging. We hope that this book, like the German editions preceding it, will become a daily companion and adviser to medical students, practicing radiologists and other physicians, and that it will give them an even stronger sense of the vast potential of MR imaging as it is being developed around the world. We want to take this opportunity to thank the authors, who generously contributed their knowledge and insights to this book. Special thanks go to Ms. Ada Muellner for her language editing. Finally, we are grateful to Springer Publishing—and particularly Dr. Ute Heilmann and Ms. Wilma McHugh—for supporting this project. Maximilian F. Reiser Wolfhard Semmler Hedvig Hricak

Contents

1

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1

2

Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3

2.1 

Overview  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5

2.2 

Physical Basics  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Brix Nuclear Spin and Magnetic Moment  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nucleus in a Magnetic Field  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macroscopic Magnetization  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic of Magnetization I: Resonance Excitation  . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic of Magnetization II: Relaxation  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The MR Experiment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Standard Pulse Sequences  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of the Electron Shell on the Local Magnetic Field  .. . . . . . . . . . . . . . . . . References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.2.1  2.2.2  2.2.3  2.2.4  2.2.5  2.2.6  2.2.7  2.2.8 

2.3  2.3.1  2.3.2  2.3.3  2.3.4   2.3.5  2.4  2.4.1  2.4.2  2.4.3  2.4.4  2.4.5  2.4.6 

  8   8   9   11   12   13   18   19   22   25   25

Image Reconstruction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Brix Magnetic Gradient Fields  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slice-Selective Excitation  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principle of Spatial Encoding within a Partial Volume: Projections  . . . . . . . . . . . Methods of Image Reconstruction in MRI  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiple-Slice Technique  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

  26

Image Contrasts and Imaging Sequences  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Brix, H. Kolem, and W.R. Nitz Image Contrasts  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Classical Imaging Sequences  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gradient-Echo Techniques  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modification of k-Space Sampling  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preparation Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sequence Families  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

  36

           

               

26 27 28 30 34 35

36 37 44 53 56 57 74 75

Technical Components  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   76 M. Bock 2.5.1  Magnet  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   76 2.5.2  Gradients  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   83 2.5 

VIII

Contents

2.5.3  2.5.4  2.5.5  2.5.6  2.5.7 

Shim   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radiofrequency System  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Computer System  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patient Monitoring  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.6 

Contrast Agents  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   92 A. Huppertz and C.J. Zech Physicochemical Properties of MR Contrast Agents  .. . . . . . . . . . . . . . . . . . . . . . .   92 Dependency of Contrast Agents from the Magnetic Field Strength  . . . . . . . . . .   94 Safety of MR Contrast Agents  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   95 Value of Contrast Agents in Clinical Practice  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   97 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   108

2.6.1  2.6.2  2.6.3  2.6.4  2.7  2.7.1  2.7.2  2.7.3  2.7.4  2.7.5  2.7.6  2.7.7  2.7.8  2.7.9  2.8  2.8.1  2.8.2  2.8.3  2.8.4  2.8.5  2.9  2.9.1  2.9.2  2.9.3  2.9.4  2.9.5 

3

Flow Phenomena and MR Angiographic Techniques  .. . . . . . . . . . . . . . . . . . . . .   M. Bock Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MR Properties of Blood  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Time-of-Flight MRA  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Arterial Spin Labeling  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Native-Blood Contrast  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Black-Blood MRA  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Velocity-Dependent Phase  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Contrast-Enhanced MRA  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Summary  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Diffusion-Weighted Imaging and Diffusion Tensor Imaging  .. . . . . . . . . . . . . .   O. Dietrich Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Physics of Diffusion  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MR Measurement of Diffusion-Weighted Images  . . . . . . . . . . . . . . . . . . . . . . . . . .   MR Measurement of Diffusion Tensor Data  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Visualization of Diffusion Tensor Data  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Risks and Safety Issues Related to MR Examinations  .. . . . . . . . . . . . . . . . . . . . .   G. Brix Safety Regulations and Operating Modes   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Static Magnetic Fields  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Time-Varying Magnetic Gradient Fields  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Radiofrequency Electromagnetic Fields  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Special Safety Issues, Contraindications  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

84 85 87 88 90 91

114 114 114 114 117 118 118 120 121 127 128 130 130 130 136 141 145 149 153 153 153 156 161 164 165

Brain, Head, and Neck  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   169

Brain: Modern Techniques and Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   M. Wintermark, M.D. Wirt, P. Mukherjee, G. Zaharchuk, E. Barbier, and W.P. Dillon 3.1.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.1.2  Diffusion-Weighted Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.1.3  Diffusion Tensor Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.1 

           

172 172 173 175

Contents

3.1.4  Dynamic Susceptibility Contrast Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.1.5  Arterial Spin Labeling  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.1.6  Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.2  3.2.1  3.2.2  3.2.3  3.2.4  3.2.5  3.2.6  3.2.7  3.3  3.3.1  3.3.2  3.3.3  3.3.4  3.3.5  3.3.6  3.3.7  3.3.8  3.4  3.4.1  3.4.2  3.4.3  3.4.4  3.4.5  3.5  3.5.1  3.5.2  3.5.3  3.5.4  3.6  3.6.1  3.6.2  3.6.3  3.6.4  3.6.5 

Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics  .. . . . . . .   B.B. Ertl-Wagner and C. Rummeny Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Examination Technique  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Normal Development of the Brain  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Congenital Disorders of the Brain  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Phakomatoses  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Hypoxic–Ischemic Injuries to the Pediatric Brain  .. . . . . . . . . . . . . . . . . . . . . . . . .   Metabolic Diseases of the Pediatric Brain  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

178 179 183 189 193 193 193 194 197 216 226 231 241

Intracranial Tumors   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   M. Essig Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   The WHO Classification of Brain Tumors  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Practical Aspects of MR Imaging in Brain Tumors  .. . . . . . . . . . . . . . . . . . . . . . . .   Blood–Brain Barrier and Tumor Enhancement: Mechanisms and Applications  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intra-Axial Cerebral Tumors   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Extra-Axial Cerebral Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Non-Tumorous Changes  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Functional Imaging in Intracranial Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

243

Cerebrovascular Disease   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   D.C. Bergen, J.M. Fagnou, and R.J. Sevick Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MR Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Acute Ischemic Stroke  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intracerebral Hemorrhage  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intracerebral Hemorrhage Etiology  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

310

Intracranial Infections  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   E. Turgut Tali and Serap Gültekin Meningitis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Empyema  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Cerebritis and Abscess  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Encephalitis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

348

Neurodegenerative Disorders  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   S. Karimi and A.I. Holodny Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Dementia  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Disorders with Prominent Motor Disability  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Hydrocephalus  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Mesial Temporal Sclerosis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

381

243 243 244 245 248 273 288 290 302

310 310 311 327 334 344

348 355 357 373 378

381 381 387 389 395 396

IX



Contents

3.7  3.7.1  3.7.2  3.7.3  3.7.4  3.8  3.8.1  3.8.2  3.8.3  3.8.4  3.8.5  3.8.6  3.9  3.9.1  3.9.2  3.9.3  3.9.4 

Pituitary Gland and Parasellar Region  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   M. Kanagaki, N. Sato, and Y. Miki Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Examination Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Pathological Conditions  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   The Orbits  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   N. Hosten, C. Zwicker, and M. Langer Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Examination Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Pathological Lesions  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Differential Diagnosis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Diagnostic Procedures  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Magnetic Resonance of the Skull Base and Petrous Bone  .. . . . . . . . . . . . . . . . .   R. Maroldi, D. Farina, A. Borghesi, E. Botturi, and C. Ambrosi Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Lesions of the Skull Base  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Lesions of the Temporal Bone   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

3.10  Head and Neck  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   H.E. Stambuk and N.J. Fischbein 3.10.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.10.2  Mucosal Diseases of the Head and Neck  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3.10.3  Non-Mucosal Diseases of the Head and Neck   .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

399 399 399 400 404 429 433 433 433 434 434 442 442 444 445 445 450 454 474 481 483 483 484 509 533

4

Spine and Spinal Canal  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   535

4.1 

Extradural Diseases of the Spine  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   C.S. Poon, J. Doumanian, G. Sze, M. Johnson, and C.E. Johnson Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Degenerative Spine Disease  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Extradural Spine Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Vertebral Column Trauma  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

4.1.1  4.1.2  4.1.4  4.1.5  4.2  4.2.1  4.2.2  4.2.3  4.2.4  4.2.5 

Intradural Extramedullary Spine   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   D. Lin Introduction   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Examination Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Pathological Findings  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Indications and Value of MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Acknowledgement  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

536 536 536 562 574 587 590 590 590 591 592 613 614 614

Contents

4.3  4.3.1  4.3.2  4.3.3  4.3.4  4.3.5  4.3.6  4.3.7 

5

Intramedullary Diseases of the Spinal Cord  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   P. Pawha, C. Shen, J. Doumanian, F. Lin, M. Johnson, R. Ashton, and G. Sze MRI Techniques for Spinal Cord Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intramedullary Neoplasms  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Vascular Diseases of the Spinal Cord  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Demyelinating Disease  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Radiation Myelopathy  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intramedullary Infectious and Inflammatory Diseases  .. . . . . . . . . . . . . . . . . . . . .   Intramedullary Traumatic Injury  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

617 617 618 625 640 647 649 653 659

Thorax and Vasculature  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   663

Lungs, Pleura, and Mediastinum  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   G. Layer and H.U. Kauczor 5.1.1  General Requirements for Imaging of Thoracic Organs  .. . . . . . . . . . . . . . . . . . . .   5.1.2  Basic MR Sequences for Imaging of the Chest  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.3  MRI of Ventilation  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.4  Use of Contrast Agents  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.5  Signal Intensities and Contrast Behavior  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.6  Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.7  Lung Diseases  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.8  Diseases of the Pleura  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.9  Mediastinal Disease   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.10  Differential Diagnosis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1.11  Value of MRI with Regard to Other Imaging Modalities  .. . . . . . . . . . . . . . . . . . .   5.1.12  Diagnostic Procedure  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.1 

666 666 666 670 671 672 672 674 688 690 693 694 696 696

High-Risk Screening Breast MRI  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   E.A. Morris 5.2.1  Importance of Early Detection  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.2  Pathology of Breast Cancer: What Are We Looking for?  .. . . . . . . . . . . . . . . . . . .   5.2.3  Why Consider MRI?  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.4  Defining the High-Risk Population  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.5  Overview of High-Risk MRI Screening Studies  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.6  Description of High-Risk Screening MRI Studies  . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.7  National Guidelines  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.8  Current Issues with Using MRI for Screening  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.9  Increased Call-Backs and Biopsies  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.10  Inconsistency of DCIS Detection  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.11  MRI Interpretation   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.12  MRI Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.13  Research Needed  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.2.14  Summary  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

700

5.3  Heart  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.1  Acquisition Techniques and Protocols  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   B.J. Wintersperger References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.2  Congenital Heart Disease: Cardiac Anomalies and Malformations  . . . . . . . . . .   T.R.C. Johnson References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

711 711

5.2 

700 701 701 701 704 704 706 706 707 707 707 708 708 709 709

717 718 732

XI

XII

Contents

5.3.3  Primary Cardiomyopathies  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   K. Nikolaou References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.4  Secondary Cardiomyopathies and Specific Heart Muscle Diseases  .. . . . . . . . . .   A. Huber References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.5  Valvular Heart Diseases  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   A. Huber References   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.6  Pericardial Diseases  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   K. Bauner References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.7  Ischemic Heart Disease  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   K. Nikolaou References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.3.8  Cardiac Tumors  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   B.J. Wintersperger References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.4  MR Angiography  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5.4.1  MRA Techniques and Acquisition Techniques  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   H.J. Michaely and S.O. Schönberg 5.4.2  Pulmonary MRA  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   C. Fink 5.4.3  MRA of the Supra-Aortic and Intracranial Vasculature  .. . . . . . . . . . . . . . . . . . . .   C. Fink, U. Attenberger, H.J. Michaely, and S.O. Schönberg 5.4.4  MRA of the Renal Arteries  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   S.O. Schönberg and H.J. Michaely 5.4.5  Diseases of the Aorta  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   D. Theisen 5.4.6  Peripheral MRA  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   H.J. Michaely and H. Kramer 5.4.7  Whole-Body MRA  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   H. Kramer and H. Schlemmer References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

734 742 744 750 751 760 761 766 766 775 778 786 788 788 804 809 817 829 837 846 851

6

Abdomen and Retroperitoneum  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   863

6.1 

Abdominal MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   N.C. Balci, E. Altun, K. Hermann, and R.C. Semelka Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Imaging Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Liver  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Pancreas  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Gallbladder and Bile Ducts  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Bile Ducts  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Spleen  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Peritoneum  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Gastrointestinal Tract  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

6.1.1  6.1.2  6.1.3  6.1.4  6.1.5  6.1.6  6.1.7  6.1.8  6.1.9 

864 864 864 864 886 893 895 897 903 906 909

Kidneys, Adrenals, and Retroperitoneum  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   912 H.J. Michaely, M. Laniado, and S.O. Schönberg 6.2.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   912 6.2.2  General Examination Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   912 6.2 

Contents

6.2.3  6.2.4  6.2.5  6.2.6 

Kidney  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Adrenal Gland  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Lymph Nodes and Retroperitoneal Tumors  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Psoas Muscle  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

7

Pelvis  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   963

7.1 

Female Pelvis  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Sala and H. Hricak Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Resonance Imaging Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Normal Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uterus and Vagina  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adnexa  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miscellaneous Conditions  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7.1.1  7.1.2  7.1.3  7.1.4  7.1.5  7.1.6 

913 943 953 958 959

  964              

964 964 966 967 986 996 997

7.2  Male Pelvis  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   999 7.2.1  Urinary Bladder  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   999 U.G. Müller-Lisse and U.L. Müller-Lisse References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1016 7.2.2  Male Pelvis: Prostate  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1018 U.G. Müller-Lisse, M.K. Scherr, and U.L. Müller-Lisse References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1037 7.2.3  Male Pelvis: Scrotum  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1039 U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1054 7.2.4  Male Pelvis: Penis  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1055 U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1067 7.3  7.3.1  7.3.2  7.3.3  7.3.4  7.3.5 

Pelvic Floor Assessment by Magnetic Resonance Imaging  .. . . . . . . . . . . . . . .   A. Maubon, C. Servin-Zardini, M. Pouquet, Y. Aubard, and J.P. Rouanet Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MRI Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Functional Anatomy of the Pelvic Floor  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MRI Techniques for Pelvic Floor Dysfunction  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   Conclusion   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1069 1069 1069 1069 1070 1075 1076

8

Musculoskeletal System   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1079

8.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1081 C. Glaser, S. Weckbach, and M. Reiser

8.2  8.2.1  8.2.2  8.2.3  8.2.4 

Examination Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Patient Preparation and Positioning   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Coil Selection   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Imaging Planes   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   MRI Sequences   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

8.3 

Relaxation Times, Signal Intensities, and Contrast Behavior  . . . . . . . . . . . . .   1086 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1086

1081 1081 1082 1082 1082

XIII

XIV

Contents

Diseases of the Bone Marrow and Hematopoietic System  . . . . . . . . . . . . . . . .   Leukemia   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Multiple Myeloma   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Metastases and Malignant Lymphomas   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Storage Diseases   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Osteomyelofibrosis   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Aplastic Anemia and Sequelae of Chemotherapy and Radiotherapy   .. . . . . . .   Hemosiderosis, Hemochromatosis, and Sickle Cell Anemia   . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1086 1087 1088 1090 1091 1093 1093 1093 1094

8.5  Inflammatory Diseases of Bone and Joints  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.5.1  Bacterial/Viral Infections  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1094 1094 1106 1106

8.6  Avascular Necrosis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.1  Avascular Necrosis of the Hip and Transient Bone Marrow Edema Syndrome  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.2  Perthes Disease and Coxitis Fugax   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.3  Bone Infarction   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.4  Kienboeck’s Disease   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.5  Necrosis of the Scaphoid Bone   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.6.6  Osteochondritis Dissecans and Spontaneous Osteonecrosis of the Knee  . . . .   8.6.7  Osteonecrosis and Osteochondritis in Other Locations  .. . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1106

8.7  Imaging of Internal Joint Derangement  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.1  Imaging of Normal Joint Structures   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.2  Knee   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.3  Shoulder   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.4  Wrist   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.5  Temporomandibular Joint   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.6  Ankle and Foot   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.7.7  Elbow   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1117 1117 1117 1129 1130 1140 1141 1145 1146 1147 1151 1151 1155

Bone and Soft Tissue Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Intraosseous Tumor Extension  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Compact Bone  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Extraosseous Tumor Extension  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Soft Tissue Tumors  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Evaluation of Tumor Nature  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Tumor Classification  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Therapy Control  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1155 1155 1158 1158 1161 1162 1164 1167 1168

8.4  8.4.1  8.4.2  8.4.3  8.4.4  8.4.5  8.4.6  8.4.7 

8.8  8.8.1  8.8.2  8.8.3  8.8.4  8.8.5  8.8.6  8.8.7 

1107 1110 1113 1113 1114 1114 1116 1116

8.9  Posttraumatic Alterations  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1169 8.9.1  Bone Injuries  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1169 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1172 8.10  Differential Diagnosis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1173

Contents

8.11  Diagnostic Value of MRI and Comparison with Other Imaging Modalities  . . . . . . . . . . . . . . . . . . . . . . . .   8.11.1  Bone Marrow   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.2  Bone Tumors and Bone Metastases   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.3  Infections   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.4  Aseptic Osteonecroses   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.5  Joints   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.6  Bone and Soft Tissue Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   8.11.7  Traumatology   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1173 1174 1174 1174 1174 1174 1174 1175

8.12  Diagnostic Process  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1175 9

Skeletal Muscle  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1177

9.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1177 C. Born

9.2  9.2.1  9.2.2  9.2.3  9.2.4 

Examination Techniques  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Patient Preparation   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Patient Positioning and Selection of Coils  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Examination Sequences and Imaging Planes  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Whole-Body MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

9.3 

Value of MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1181

9.4 

Normal Anatomy  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1181

9.5  9.5.1  9.5.2  9.5.3  9.5.4  9.5.5  9.5.6 

MR Imaging Findings  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Muscle Edema  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Fatty Infiltration  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Mass Lesions and Calcification  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Contrast Enhancement  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Lesion Patterns  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Distribution  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

9.6 

Muscle Biopsy  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1188

9.7  9.7.1  9.7.2  9.7.3  9.7.4 

Skeletal Muscle Diseases  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Muscle  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Neuromuscular Junction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Nerves  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Central Nervous System: Motor Neuron Diseases  .. . . . . . . . . . . . . . . . . . . . . . . .  

9.8 

Follow-Up  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1209

9.9  Future Developments  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   9.9.1  Diffusion-Weighted Imaging and Blood-Oxygen Level-Dependent Imaging    9.9.2  Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10

1178 1178 1178 1178 1179

1184 1184 1184 1184 1184 1185 1186

1188 1188 1204 1206 1208

1209 1209 1210 1210

MRI of the Fetal Body  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1213

10.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1213 A.D. McKenna and F.V. Coakley

XV

XVI

Contents

10.2  Historical Development of Fetal MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1213 10.3  Safety of MRI in Pregnancy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1213 10.4  Fetal MRI Technique  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1214 10.5  Normal MRI Findings of the Fetal Body  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1215 10.6  MRI Findings in Pathologic Conditions of the Fetal Body  .. . . . . . . . . . . . . . .   10.6.1  Airway Compromise  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.2  Pulmonary Sequestration  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.3  Congenital Cystic Adenomatoid Malformation  .. . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.4  MRI for Fetal Lung Maturity  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.5  Congenital Diaphragmatic Hernia  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.6  Abdominal Masses  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   10.6.7  Gastrointestinal Tract Anomalies  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1218 1218 1219 1221 1222 1222 1226 1227

10.7  Summary: Indications for MRI for the Fetal Body  . . . . . . . . . . . . . . . . . . . . . . .   1228 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1229 11

Whole-Body MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1231

11.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1231 H.-P. Schlemmer 11.2  Examination Techniques  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1232 11.2.1  Hardware  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1232 11.2.2  Sequence Protocols  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1232 11.3  Postprocessing/Reading  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1235 11.4  Pathologic Findings  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   11.4.1  Cardiovascular Diseases  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   11.4.2  Oncology  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   11.4.3  Musculoskeletal   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   11.4.4  Adipose-Tissue Quantification  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12

1235 1235 1237 1243 1246 1254

Interventional MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1257

12.1  Technical Implementation  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   M. Bock and F. Wacker 12.1.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.2  Interventional MR Systems   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.3  Instrument Tracking   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.4  Instrument Navigation   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.5  Real-Time Imaging  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.6  Functional Imaging  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.7  Safety Aspects  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.1.8  Summary  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1258 1258 1258 1262 1265 1267 1269 1272 1274 1274

12.2  Clinical Applications of Interventional and Intraoperative MRI  . . . . . . . . . .   1277 F.A. Jolesz and E. Samset 12.2.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1277

Contents

12.2.2  Interventional MRI and Computer-Assisted Surgery  .. . . . . . . . . . . . . . . . . . . . .   12.2.3  Interventional MRI and MR Imaging Techniques  .. . . . . . . . . . . . . . . . . . . . . . . .   12.2.4  MRI-Guided Thermal Ablations   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   12.2.5  Current Clinical Applications of Interventional MRI  .. . . . . . . . . . . . . . . . . . . . .   12.2.6  Conclusion  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13

1278 1280 1280 1284 1285 1286

Functional MRI  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1291

13.1  Basics  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   L.R. Schad 13.1.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.2  Physiology  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.3  Contrast  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.4  Measuring Techniques  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.5  Sequence Optimization  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.6  Data Analysis  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.1.7  Comparison of Methods  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1292

13.2  Clinical Applications  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   K.K. Peck and A.I. Holodny 13.2.1  Introduction   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.2.2  Normal Functional Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.2.3  Functional MRI in Disease   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   13.2.4  Specific Pathology  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1308

14 

1308 1308 1309 1314 1318

Computer Aided 3D Radiation Planning Using MRI  . . . . . . . . . . . . . . . . . . . .   1323

14.1  Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain  . . . . . . . . . . . . . . . . . . . . . . . . .   L.R. Schad 14.1.1  2D-Phantom Measurement  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   14.1.2  3D-Phantom Measurement  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15

1292 1292 1292 1294 1295 1298 1303 1306

1323 1323 1325 1329

Clinical Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1331

15.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1331 W. Semmler; H.-P. Schlemmer 15.2  Basics  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.2.1  Nuclei for MRS in vivo  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.2.2  Technical Requirements  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.2.3  Spectra Analysis  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.2.4  Localization Techniques  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.2.5  Spin–Spin Coupling and Double-Resonance Techniques  . . . . . . . . . . . . . . . . . .   15.2.6  Limitations of MRS in vivo  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1332 1332 1332 1333 1334 1336 1339

15.3  Clinical MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.3.1  1H MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.3.2  31P MRS  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.3.3  13C MRS  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1342 1342 1345 1348

XVII

XVIII Contents

15.3.4  19F MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1348 15.4  Clinical Applications in Selected Organs and Tumors  .. . . . . . . . . . . . . . . . . . .   15.4.1  Brain  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.4.2  Liver  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.4.3  Heart  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.4.4  Skeletal Muscles  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.4.5  Urogenital Tract  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   15.4.6  Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   Suggested Reading  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16

1350 1350 1354 1356 1358 1360 1360 1369 1379

Molecular Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1381

16.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1381 W. Semmler 16.2  Molecular Imaging Modalities  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.2.1  Optical Imaging  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.2.2  Computed Tomography  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.2.3  Positron Emission Tomography and Single-Photon Emission Tomography    16.2.4  Magnetic Resonance Imaging and Spectroscopy  .. . . . . . . . . . . . . . . . . . . . . . . . .   16.2.5  Ultrasound  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.2.6  Multimodal Imaging Approaches: PET–CT and PET–MRI  . . . . . . . . . . . . . . . .  

1384 1384 1386 1387 1389 1389 1390

16.3  Molecular Imaging Probes  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1390 16.3.1  Signal Generators  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1392 16.3.2  Molecular Targeting for Imaging  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1397 16.4  Applications  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.1  Antibodies  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.2  Protein Amplification  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.3  Reporters  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.4  Smart Probes  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.5  Cell Tracking  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   16.4.6  Imaging Using Hyperpolarization  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   17

1399 1399 1399 1400 1403 1403 1407 1407

Systems Biology and Nanotechnology  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1411

17.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1411 M. S. Bradbury, H. Hricak, and J. R. Heath 17.2  Linking Cancer Biology, P4 Medicine, and Molecular Imaging  . . . . . . . . . . .   1413 17.3  Application of a Systems Biology Approach to Cancer  . . . . . . . . . . . . . . . . . . .   1414 17.4  Molecular Diagnostics and Personalized Therapies for Cancer  . . . . . . . . . . .   1418 17.4.1  Cancer Biomarkers for Clinical Use   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1419 17.5  Role of Cancer Nanobiotechnology in Personalized Medicine  . . . . . . . . . . . .   17.5.1  Nanodevices as Biosensors  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   17.5.2  Nanoparticles for Cancer Diagnostics and Therapeutics  . . . . . . . . . . . . . . . . . . .   17.5.3  Nanobiotechnology for Drug Discovery  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1420 1420 1421 1424

Contents

17.6  Role of Radiology in Personalized Medicine  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1425 17.7  Clinical and Near-Clinical Molecular Imaging Applications Using Targeted Probes  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   17.7.1  Oncologic Molecular Imaging for Detection of Nodal Metastases and Cancer Staging  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   17.7.2  Molecular Imaging of Dendritic Cell Migration to Regional Nodes for Tumor Immunotherapy   .. . . . . . . . . . . . . . . . . . . . . . . . . .   17.7.3  Molecular Imaging of Thrombosis in Atherosclerotic Disease  .. . . . . . . . . . . . .   References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  

1426 1426 1428 1428 1430

Glossary  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1435 List of Abbreviations  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1465 Subject Index  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1479

XIX

Contributors

Altun, Ersan, MD Department of Radiology University of North Carolina 101 Manning Dr, Old Clinic Bldg Chapel Hill, NC 27599-9677 USA

Barbier, Emmanuel, PhD INSERM, Université Joseph Fourier NeuroImagerie Fonctionnelle et Métabolique CHU Michallon-Pav. B Grenoble France

Ambrosi, Claudia, MD Department of Radiology Neuroradiology Section University of Brescia Piazzale Spedali Civili I 25123 Brescia, BS Italy

Bauner, Kerstin, Dr. med. Institute of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Ashton, Robert, MD Yale Diagnostic Radiology Yale New Haven Hospital 20 York St. New Haven, CT 06510 USA

Bergen, Daryl C., MD Department of Radiology Diagnostic Imaging Foothills Medical Centre 1403-29th St. N.W. Calgary, AB, T2N 2T9 Canada

Attenberger, Ulrike, Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Bock, Michael, Dr. rer. nat., Dipl.-Phys. Division of Medical Physics in Radiology German Cancer Research Center Im Neuenheimer Feld 280 69120 Heidelberg Germany

Aubard, Yves, MD Servide de Gynécologie-Obstétrique CHU Dupuytren 2 Ave Martin Luther King 87042 Limoges France

Borghesi, Andrea, MD Department of Radiology University of Brescia Piazzale Spedali Civili I 25123 Brescia, BS Italy

Balci, Numan Cem, MD Department of Radiology University of North Carolina 101 Manning Dr., Old Clinic Bldg. Chapel Hill, NC 27599-9677 USA

Born, Christine, Dr. med. Department of Clinical Radiology University Hospitals Innenstadt Ludwig-Maximilian University Ziemssenstraße 1 80336 Munich Germany

XXII

Contributors

Botturi, Elisa, MD Department of Radiology University of Brescia Piazzale Spedali Civili I 25123 Brescia, BS Italy

Doumanian, John, MD Department of Radiology Yale University School of Medicine P.O. Box 208042 (CB-30) New Haven, CT 06520-8042 USA

Bradbury, Michelle S., MD Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10065 USA

Ertl-Wagner, Birgit B., PD Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Brix, Gunnar, Prof. Dr. rer. nat., Dipl.-Phys. Division of Medical Radiation Hygiene Federal Office for Radiation Protection Ingolstädter Landstr. 1 85764 Oberschleißheim Germany Coakley, Fergus, MD, Professor Department of Radiology University of California San Francisco 505 Parnassus Ave., M372 San Francisco, CA 94143-0712 USA Degenhart, Christoph, Dr. med. Department of Clinical Radiology University Hospitals Innenstadt Ludwig-Maximilian University Ziemssenstraße 1 80336 Munich Germany Dietrich, Olaf, Dipl. Phys. Institute of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Dillon, William P, MD, Professor Department of Radiology Neuroradiology Section University of California, San Francisco Box 0628, L-358 505 Parnassus Avenue San Francisco, CA 94143-0628 USA

Essig, Marco, Prof. Dr. med. Department of Radiology German Cancer Research Center Im Neuenheimer Feld 280 69120 Heidelberg Germany Fagnou, John M., MD Department of Radiology Diagnostic Imaging Foothills Medical Centre 1403-29th St. N.W. Calgary, AB T2N 2T9 Canada Farina, Davide, MD Department of Radiology University of Brescia Piazzale Spedali Civili I 25123 Brescia BS, Italy Fink, Christian, PD Dr. Section Chief Thoracic Imaging Associate Chair of Clinical Operations University Hospital Mannheim Medical Faculty Mannheim–University of Heidelberg Theodor-Kutzer-Ufer 1-3 68167 Mannheim Germany Fischbein, Nancy, MD, PhD, Associate Professor Department of Radiology Stanford University 300 Pasteur Drive Stanford, CA 94305 USA

Contributors XXIII

Glaser, Christian, Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Huber, Armin, Dr. med. Institute for Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Gültekin, Serap, MD Department of Radiology Division of Neuroradiology Gazi University School of Medicine Besevler, Ankara 06510 Turkey

Huppertz, Alexander, Dr. med. Imaging Science Institute Charité Berlin-Siemens Robert-Koch-Platz 7 10115 Berlin Germany

Heath, James R., PhD Professor of Chemistry Eliszabeth W. Gilloo, Professor Division of Chemistry and Chemical Engineering California Insitute of Technology Pasadena, CA 91125 USA

Johnson, Carl E., MD Weill Cornell Department of Radiology New York Presbyterian Hospital Cornell Medical Center 525 York Ave. New York, NY 10021 USA

Herrmann, Karin, PD Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Johnson, Michele H., MD Department of Radiology Yale University School of Medicine P.O. Box 208042 (CB-30) New Haven, CT 06520-8042 USA

Holodny, Andrei I., MD, Associate Professor, Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10065 USA

Johnson, Thorsten R.C., Dr. med. Institute of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany

Hosten, Norbert, Univ.-Prof. Dr. med. Institute of Diagnostic Radiology and Neuroradiology University Clinic Greifswald Friedrich-Löffler-Str. 23 17487 Greifswald Germany

Jolesz, Ferenc, MD, Professor Department of Radiology Brigham & Women’s Hospital 75 Francis St. Boston, MA 02115 USA

Hricak, Hedvig, MD, PhD, Dr. h. c. Professor of Diagnostic Radiology and Chairperson Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10065 USA

Kanagaki, Mitsunori, MD, PhD Department of Diagnostic Imaging &Nuclear Medicine Kyoto University Hospital 54 Shogoin Kawahara-cho Sakyo-ku, Kyoto 606-8507 Japan

XXIV Contributors

Karimi, Sasan, MD, Assistant Professor Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10065 USA Kauczor, Hans-Ulrich, Prof. Dr. med. Department of Radiology German Cancer Research Center Im Neuenheimer Feld 280 69120 Heidelberg Germany Kolem, Heinrich, Dr. rer. nat., Dipl.-Phys. President and Chief Executive Officer Siemens Medical Solutions USA, Inc. 51 Valley Stream Parkway Malvern, PA 19355 USA Kramer, Harald, Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Langer, Mathias, Univ.-Prof. Dr. med. Department of Radiological Diagnosis Radiological Clinic University Hospitals Albrecht Ludwig University Hugstetter Str. 55 79106 Freiburg Germany Laniado, Michael, Univ.-Prof. Dr. med. Institute and Policlinic for Radiological Diagnostic University Hospitals Carl Gustav Carus at the Technical University Dresden Fetcherstr. 74 01307 Dresden Germany Layer, Günter, Prof. Dr. med. Institute for Diagnostic and Interventional Radiology Clinic of Ludwigshafen Bremserstr. 79 67063 Ludwigshafen Germany

Lin, Doris, MD, PhD Radiology Department Radiology and Radiologic Science Institute Johns Hopkins University School of Medicine 600 N. Wolfe St. Baltimore, MD 21287 USA Lin, Felix, MD Department of Radiology Yale University School of Medicine P.O. Box 208042 (CB-30) New Haven, CT 06520-8042 USA Maroldi, Roberto, MD Professor Department of Radiology University of Brescia Piazzale Spedali Civili I 25123 Brescia, BS Italy Maubon, Antoine, MD Service de Radiologie et Imagerie Médicale CHU Dupuytren 2 Ave Martin Luther King Limoges 87042 France McKenna, Arthur D., MD, MB BCh Department of Radiology University of California San Francisco 505 Parnassus Ave., M372 San Francisco, CA 94143-0712 USA Michaely, Henrik J., Dr. med. Department of Clinical Radiology University Hospital Mannheim Medical Faculty Mannheim–University of Heidelberg Theodor-Kutzer-Ufer 1-3 68167 Mannheim Germany Miki, Yukio, MD, Assistant Professor Department of Diagnostic Imaging & Nuclear Medicine Kyoto University Hospital 54 Shogoin Kawahara-cho Sakyo-ku, Kyoto 606-8507 Japan Morris, Elizabeth A., MD, Assistant Professor Department of Radiology Memorial Sloan-Kettering Cancer Center 1275 York Ave. New York, NY 10021 USA

Contributors

Mukherjee, Pratik, MD, PhD Department of Radiology Neuroradiology Section University of California San Francisco Box 0628, L-358 505 Parnassus Avenue San Francisco, CA 94143-0628 USA Müller-Lisse, Gerd Ullrich, PD Dr. med. Department of Clinical Radiology University Hospitals Innenstadt Ludwig-Maximilian University Ziemssenstraße 1 80336 Munich Germany Müller-Lisse, Ulrike-L., Dr. med. Department of Urology University Hospitals Innenstadt Ludwig-Maximilian University Nussbaumstraße 20 80336 Munich Germany Nikolaou, Konstantin, Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Nitz, Wolfgang R., Dr. rer. nat., Dipl.-Ing. MREA Medical Solutions Siemens AG Henkelstr. 127 91052 Erlangen Germany Pawha, Puneet S., MD, Assistant Professor Yale Diagnostic Radiology Yale New Haven Hospital 20 York St. New Haven, CT 06510 USA Peck, Kyung K., MD Functional MRI Lab Radiology & Medical Physics Memorial Sloan-Kettering Cancer Center New York, 10021 USA

Poon, Colin S., MD, PhD, Assistant Professor Department of Radiology State University of New York Upstate Medical University 750 East Adams St. Syracuse, NY 13210 USA Pouquet, M., MD Service de Radiologie et Imagerie Médicale CHU Dupuytren 2 Ave Martin Luther King Limoges 87042 France Reiser, Maximilian, Univ.-Prof. Dr. med. Dr. h.c. Institute for Clinical Radiology University Hospitals Grosshadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Rouanet, Jean-Pierre, MD, Professor Service de Radiologie CMC Beau Soleil 119, Ave. de Lodève 34000 Montpellier France Rummeny, Claudia, Dr. med. Department of Clinical Radiology University Hospitals Innenstadt Ludwig-Maximilian University Ziemssenstraße 1 80336 Munich Germany Sala, Evis, MD, PhD, University Lecturer Department of Radiology University of Cambridge Addenbrooke’s Hospital Box 219, Hills Road Cambridge CB22QQ UK Samset, Eigil, PhD Department of Radiology Brigham & Women’s Hospital 75 Francis St. Boston, MA 02115 USA

XXV

XXVI Contributors

Sato, Noriko, MD, PhD Department of Radiology National Center Hospital for Mental Nervous and Muscular Disorders National Center of Neurology and Psychiatry 4-1-1 Ogawahigashi-chyo Kodaira, Tokyo 187-8551 Japan Schad, Lothar, Prof. Dr. rer. nat., Dipl.-Phys. Department of Computer Assisted Clinical Medicine University Clinic Mannheim Medical Faculty Mannheim – University of Heidelberg Theodor-Kutzer-Ufer 1–3 68167 Mannheim Germany Scherr, Michael K., Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilians-University Marchioninistraße 15 81377 Munich Germany Schlemmer, Heinz-Peter, Prof. Dr. med. Dipl.-Phys. Department of Radiology Eberhard Karl University Hoppe-Seyler-Str. 72076 Tübingen Germany Schönberg, Stefan O., Univ.-Prof. Dr. Chairman, Department of Clinical Radiology University Hospital Mannheim Medical Faculty Mannheim – University of Heidelberg Theodor-Kutzer-Ufer 1–3 68167 Mannheim Germany Semelka, Richard, MD, Professor Department of Radiology University of North Carolina 101 Manning Dr, Old Clinic Bldg. Chapel Hill, NC 27599-9677 USA Semmler, Wolfhard, Univ.-Prof. Dr. rer. nat. Dr. med., Dipl.-Phys. Division of Medical Physics in Radiology German Cancer Research Center Im Neuenheimer Feld 280 69120 Heidelberg Germany

Servin-Zardini, C., MD Service de Radiologie et Imagerie Médicale CHU Dupuytren 2 Ave. Martin Luther King Limoges 87042 France Sevick, Robert, MD, PhD Department of Radiology Foothills Hospital 1403 29 St. N.W. Calgary, AB T2N 2T9 Canada Shen, Charles, MD Department of Radiology Yale University School of Medicine P.O. Box 208042 (CB-30) New Haven, CT 06520-8042 USA Stambuk, Hilda, MD, Assistant Professor Memorial Sloan-Kettering Cancer Center Department of Radiology 1175 York Ave. New York, NY 10021 USA Sze, Gordon, MD, PhD, Professor Department of Radiology Yale University School of Medicine P.O. Box 208042 (CB-30) New Haven, CT 06520-8042 USA Tali, E. Turgut, MD, Professor Division of Neuroradiology Gazi University School of Medicine Besevler, Ankara 06510 Turkey Theisen, Daniel, Dr. med. Department of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Wacker, Frank K., Prof. Dr. med. Klinik und Poliklinik f. Radiologie u. Nuklearmedizin Campus Benjamin Franklin Charité-Universitätsmedizin Berlin Hindenburgdamm 30 12203 Berlin Germany

Contributors XXVII

Weckbach, Sabine, Dr. med. Institute of Clinical Radiology University Hospital Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Wintermark, Max, MD Department of Radiology Neuroradiology Section University of California San Francisco, CA 94143-0628 USA Wintersperger, Bernd J., Dr. med. Institute of Clinical Radiology University Hospitals Großhadern Ludwig-Maximilian University Marchioninistraße 15 81377 Munich Germany Wirt, Michael D., MD Department of Radiology Neuroradiology Section University of California San Francisco, CA 94143-0628 USA

Zaharchuk, Greg, MD, PhD Neuroradiology Section Department of Radiology Stanford University Stanford, CA USA Zech, Christoph J., Dr. med. Department of Clinical Radiology University Hospitals Innenstadt Ludwig-Maximilian University Ziemssenstraße 1 80336 Munich Germany Zwicker, B. PD, Dr. med. Department of Radiology Krankenhaus Singen Virchowstr. 10 78224 Singen Germany

Chapter 1

Introduction W. Semmler, M. Reiser, and H. Hricak

Although the principle of nuclear magnetic resonance was discovered in 1946 and then used intensively in physics and chemistry, magnetic resonance imaging (MRI) did not emerge until the 1970s. The advent of MRI was as important as the discovery of X-ray beams or the development of X-ray computed tomography. From the very beginning, the technical development of MRI progressed quickly, and it seems to be continuing at an ever-increasing pace. Interest in MRI has been especially strong because of its ability to show not only anatomy but also metabolism and function. As a result, over the last 25 years there has been an explosion in the number of clinical applications of MRI. Initially, MRI was mostly applied to solve special clinical problems related to the central nervous system. Today, however, MRI is the method of choice with numerous diseases. The uses of MRI range from early diagnosis to treatment planning and follow-up. MRI systems are increasingly used in early stages of disease, and in many cases, MRI has become the only diagnostic test applied. The chapters in this book were written by a multitude of international experts who are experienced in their scientific fields and whose outstanding work has contributed to progress in the development and clinical application of MRI. The book provides a comprehensive overview of both the physics and the clinical applications of MRI, including practical guidelines for imaging. The authors try to define the importance of MRI in the diagnosis of several disease groups in comparison or in

1

combination with other methods. Chapters dealing with basic principles of MRI, MR spectroscopy (MRS), interventional MRI, and functional MRI (fMRI) as well as the application of MRI in radiotherapy treatment planning (RTP) illustrate the broad range of applications for MR systems. Both standard and cutting-edge applications of MR systems are included. Furthermore, chapters on molecular imaging and nanotechnology give glimpses into the future of the field and outline the potential role of MRI in molecular medicine. This book is intended for a wide audience, including radiologists in general practice, those who specialize in MRI, physicists, and clinicians in other fields. Its clear explanations of the basic science of MR should prove helpful to students preparing for examinations. Moreover, as an in-depth and wide-ranging reference source, it should be of service to residents, young physicians just starting out, and even experienced physicians. To help the reader use the book effectively, the authors were asked to apply the same structure to each chapter wherever possible. As a result, the reader can always look in the same place for statements concerning, for example, examination technique, normal and pathological findings, and the effectiveness of the method. We would like to thank the authors for their contributions to this book and for making their knowledge available to all those who are interested in the optimized use of MRI.

Chapter 2

Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2

2.1 

Overview  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   5

2.3.3 

2.2 

Physical Basics  . . . . . . . . . . . . . . . . . . . . . . . .   8 G. Brix

2.3.3.1  Frequency Encoding  . . . . . . . . . . . . . . . . . . .   28

2.2.1 

Nuclear Spin and Magnetic Moment  .. . . .   8

2.3.3.2  Phase Encoding  .. . . . . . . . . . . . . . . . . . . . . . .   30

2.2.2 

Nucleus in a Magnetic Field  .. . . . . . . . . . . .   9

2.3.4  

2.2.2.1  Quantum Mechanical Description  .. . . . . .   9 2.2.2.2  Semiclassical Description  .. . . . . . . . . . . . . .   10 2.2.3 

Macroscopic Magnetization  .. . . . . . . . . . . .   11

2.2.4 

Dynamic of Magnetization I: Resonance Excitation  .. . . . . . . . . . . . . . . . . .   12

2.2.5 

Dynamic of Magnetization II: Relaxation    13

Principle of Spatial Encoding within a Partial Volume: Projections  .. . . .   28

Methods of Image Reconstruction   in MRI  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   30

2.3.4.1  Projection Reconstruction Method  . . . . . .   31 2.3.4.2  2D Fourier Method  . . . . . . . . . . . . . . . . . . . .   33 2.3.4.3  3D Fourier Method   .. . . . . . . . . . . . . . . . . . .   33 2.3.4.4  Alternative k-Space Sampling  . . . . . . . . . . .   34 2.3.5 

2.2.5.1  Physical Model of Relaxation Processes  . .   14

Multiple-Slice Technique  . . . . . . . . . . . . . . .   34 Suggested Reading  . . . . . . . . . . . . . . . . . . . . .   35

2.2.5.2  Phenomenological Description of Relaxation Processes  .. . . . . . . . . . . . . . . .   15

2.4 

Image Contrasts and Imaging Sequences    36 G. Brix, H. Kolem, and W.R. Nitz

2.2.5.3  Proton Relaxation Times of Biological Tissues  .. . . . . . . . . . . . . . . . . . .   17

2.4.1 

Image Contrasts  . . . . . . . . . . . . . . . . . . . . . . .   36

2.2.6 

The MR Experiment  .. . . . . . . . . . . . . . . . . . .   18

2.4.1.1  Contrast Determinants   and Optimization in MRI  .. . . . . . . . . . . . . .   36

2.2.7 

Standard Pulse Sequences  .. . . . . . . . . . . . . .   19

2.4.1.2  Definition of Image Contrast  .. . . . . . . . . . .   36

2.2.7.1  Saturation Recovery Sequence  .. . . . . . . . . .   19

2.4.2 

2.2.7.2  Inversion Recovery Sequence  . . . . . . . . . . .   20

2.4.2.1  Spin-Echo Sequence  .. . . . . . . . . . . . . . . . . . .   37

2.2.7.3  Spin-Echo Sequence  .. . . . . . . . . . . . . . . . . . .   21

2.4.2.2  Inversion Recovery Sequence  . . . . . . . . . . .   39

Influence of the Electron Shell   on the Local Magnetic Field  .. . . . . . . . . . . .   22

2.4.2.3  Limiting Factor: Acquisition Time  .. . . . . .   44

2.2.8 

2.2.8.1  Macroscopic Effect: Diamagnetism  . . . . . .   22

2.4.3 

Classical Imaging Sequences  . . . . . . . . . . . .   37

Gradient-Echo Techniques  .. . . . . . . . . . . . .   44

2.2.8.2  Microscopic Effect: Chemical Shift  . . . . . .   24

2.4.3.1  Low–Flip Angle Excitation and Gradient Echoes  . . . . . . . . . . . . . . . . . . .   44

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   25

2.4.3.2  Example FLASH  . . . . . . . . . . . . . . . . . . . . . . .   48

Suggested Reading  . . . . . . . . . . . . . . . . . . . . .   25

2.4.3.3  Example trueFISP  .. . . . . . . . . . . . . . . . . . . . .   48

2.3 

Image Reconstruction  . . . . . . . . . . . . . . . . .   26 G. Brix

2.4.3.4  Influence of Magnetic Field Inhomogeneities  . . . . . . . . . . . . . . . . . . . . . . .   51

2.3.1 

Magnetic Gradient Fields  . . . . . . . . . . . . . . .   26

2.4.3.5  Influence of the Chemical Shift   . . . . . . . . .   52

2.3.2 

Slice-Selective Excitation  . . . . . . . . . . . . . . .   27

2.4.4 

Modification of k-Space Sampling  . . . . . . .   53



2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 2.4.4.1  Half-Fourier Technique  .. . . . . . . . . . . . . . . .   53

2.5.6.4  Electroencephalogram  .. . . . . . . . . . . . . . . .   90

2.4.4.2  Fourier Interpolation  .. . . . . . . . . . . . . . . . . .   54

2.5.7 

2.4.4.3  Parallel Imaging  . . . . . . . . . . . . . . . . . . . . . . .   55 2.4.4.4  Segmented k-Space Sampling  . . . . . . . . . . .   55

Summary  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   90 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   91

2.6 

Contrast Agents  . . . . . . . . . . . . . . . . . . . . . .   92 A. Huppertz and C.J. Zech

2.6.1 

Physicochemical Properties of MR Contrast Agents  . . . . . . . . . . . . . . . .   92

2.6.2 

Dependency of Contrast Agents from the Magnetic Field Strength  .. . . . . .   94

2.4.6.1  Sequence Classification   . . . . . . . . . . . . . . . .   59

2.6.3 

Safety of MR Contrast Agents  . . . . . . . . . .   95

2.4.6.2  Fast Spin-Echo Techniques  .. . . . . . . . . . . . .   61

2.6.4 

Value of Contrast Agents in Clinical Practice  .. . . . . . . . . . . . . . . . . . .   97

2.4.5 

Preparation Techniques  .. . . . . . . . . . . . . . . .   56

2.4.5.1  Fat Saturation  .. . . . . . . . . . . . . . . . . . . . . . . . .   56 2.4.5.2  Magnetization Transfer  .. . . . . . . . . . . . . . . .   56 2.4.6 

Sequence Families  .. . . . . . . . . . . . . . . . . . . . .   57

2.4.6.3  Gradient-Echo Techniques  .. . . . . . . . . . . . .   65 2.4.6.4  Single-Shot Gradient-echo Imaging  .. . . . .   72 2.4.6.5  Single-Shot Gradient-echo Imaging with Preparation of Magnetization (Diffusion-Weighted Imaging)  . . . . . . . . . .   73 2.4.6.6  Hybrid Techniques  .. . . . . . . . . . . . . . . . . . . .   74 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   74 Suggested Reading  . . . . . . . . . . . . . . . . . . . . .   75 2.5 

Technical Components  .. . . . . . . . . . . . . . . .   76 M. Bock

2.5.1 

Magnet  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   76

2.6.4.1  Contrast Agents in Neuroimaging   .. . . . .   97 2.6.4.2  Contrast Agents in MR Angiography  . . .   98 2.6.4.3  Contrast Agents for Soft-Tissue Lesions     99 2.6.4.4  Hepatobiliary Imaging  .. . . . . . . . . . . . . . . .   101 2.6.4.5  Lymph Node Imaging  . . . . . . . . . . . . . . . . .   104 2.6.4.6  Gastrointestinal Imaging  .. . . . . . . . . . . . . .   105 2.6.4.7  Cardiac Imaging  . . . . . . . . . . . . . . . . . . . . . .   106 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   108 2.7 

Flow Phenomena and MR Angiographic Techniques  .. . . .   114 M. Bock

2.5.1.2  Resistive Magnets  . . . . . . . . . . . . . . . . . . . . . .   78

2.7.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   114

2.5.1.3  Superconducting Magnets  . . . . . . . . . . . . . .   79

2.7.2 

MR Properties of Blood  .. . . . . . . . . . . . . . .   114

2.5.2 

Gradients  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   83

2.7.3 

Time-of-Flight MRA  . . . . . . . . . . . . . . . . . .   114

2.5.3 

Shim   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   84

2.7.4 

Arterial Spin Labeling  . . . . . . . . . . . . . . . . .   117

2.5.4 

Radiofrequency System  .. . . . . . . . . . . . . . . .   85

2.7.5 

Native-Blood Contrast  .. . . . . . . . . . . . . . . .   118

2.5.4.1  RF Cabin  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   85

2.7.6 

Black-Blood MRA  . . . . . . . . . . . . . . . . . . . .   118

2.5.4.2  Transmitter  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   85

2.7.7 

Velocity-Dependent Phase  .. . . . . . . . . . . .   120

2.5.4.3  RF Coils  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   85

2.7.7.1  Flow Measurements  .. . . . . . . . . . . . . . . . . .   120

2.5.4.4  Receiver  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   87

2.7.7.2  Phase-Contrast MRA  .. . . . . . . . . . . . . . . . .   121

2.5.1.1  Permanent Magnets  .. . . . . . . . . . . . . . . . . . .   78

2.5.5 

Computer System  .. . . . . . . . . . . . . . . . . . . . .   87

2.7.8 

Contrast-Enhanced MRA  .. . . . . . . . . . . . .   121

2.5.5.1  Host Computer  . . . . . . . . . . . . . . . . . . . . . . . .   87

2.7.8.1  First-Pass Studies  . . . . . . . . . . . . . . . . . . . . .   123

2.5.5.2  Hardware-Control Computer  . . . . . . . . . . .   88

2.7.8.2  Intravascular Contrast Agents  .. . . . . . . . .   127

2.5.5.3  Image-Reconstruction Computer  .. . . . . . .   88

2.7.9 

2.5.6 

Patient Monitoring  .. . . . . . . . . . . . . . . . . . . .   88

2.5.6.1  Electrocardiogram  . . . . . . . . . . . . . . . . . . . . .   89

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   128 2.8 

Diffusion-Weighted Imaging and Diffusion Tensor Imaging  .. . . . . . . .   130 O. Dietrich

2.8.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   130

2.5.6.2  Pulse Oximetry  . . . . . . . . . . . . . . . . . . . . . . . .   89 2.5.6.3  Breathing Synchronization  .. . . . . . . . . . . . .   90

Summary  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   127

2.1  Overview 2.8.2 

Physics of Diffusion  . . . . . . . . . . . . . . . . . . .   130

2.9 

Risks and Safety Issues Related to MR Examinations  .. . . . . . . . . . . . . . . . .   153 G. Brix

2.9.1 

Safety Regulations and Operating Modes   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   153

2.9.2 

Static Magnetic Fields  . . . . . . . . . . . . . . . . .   153

2.8.2.1  Brownian Molecular Motion  .. . . . . . . . . .   130 2.8.2.2  Diffusion Tensor  .. . . . . . . . . . . . . . . . . . . . .   132 2.8.2.3  Diffusion Anisotropy  .. . . . . . . . . . . . . . . . .   134 2.8.3 

MR Measurement of Diffusion-Weighted Images  .. . . . . . . . .   136

2.8.3.1  Diffusion Gradients and Diffusion Contrast  . . . . . . . . . . . . . . . .   136 2.8.3.2  Pulse Sequences for Diffusion MRI  . . . . .   138 2.8.3.3  Artifacts in Diffusion MRI: Motion and Eddy Currents  .. . . . . . . . . . . .   140

2.9.2.1  Magnetic Properties of Matter  .. . . . . . . . .   153 2.9.2.2  Biophysical Interaction Mechanisms  .. . .   154 2.9.2.3  Biological Effects  .. . . . . . . . . . . . . . . . . . . . .   155 2.9.2.4  Exposure Limits  . . . . . . . . . . . . . . . . . . . . . .   156 2.9.3 

Time-Varying Magnetic Gradient Fields    156

MR Measurement   of Diffusion Tensor Data  .. . . . . . . . . . . . . .   141

2.9.3.1  Electric Fields and Currents Induced by Time-Varying Magnetic Fields   .. . . . .   156

2.8.4.1  Diffusion Trace Imaging  .. . . . . . . . . . . . . .   141

2.9.3.2  Biophysical Interaction Mechanisms  .. . .   158

2.8.4.2  Basic Diffusion Tensor Imaging  . . . . . . . .   141

2.9.3.3  Biological Effects  .. . . . . . . . . . . . . . . . . . . . .   158

2.8.4.3  Optimizing Diffusion Tensor Imaging  . .   143

2.9.3.4  Exposure Limits  . . . . . . . . . . . . . . . . . . . . . .   160

2.8.4.4  Beyond Diffusion Tensor Imaging  .. . . . .   144

2.9.4 

2.8.4 

Radiofrequency Electromagnetic Fields    161

Visualization of Diffusion Tensor Data     145

2.9.4.1  Biophysical Interaction Mechanisms  .. . .   161

2.8.5.1  Scalar Diffusion Quantities  . . . . . . . . . . . .   145

2.9.4.2  Biological Effects  .. . . . . . . . . . . . . . . . . . . . .   161

2.8.5.2  Vector Diffusion Quantities  .. . . . . . . . . . .   145

2.9.4.3  Exposure Limits  . . . . . . . . . . . . . . . . . . . . . .   163

2.8.5.3  Full Tensor Visualization  . . . . . . . . . . . . . .   146

2.9.5 

2.8.5 

2.8.5.4  Fiber Tracking  . . . . . . . . . . . . . . . . . . . . . . . .   147

Special Safety Issues, Contraindications    164 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   165

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   149

2.1  Overview In this chapter, the basic principles of magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) (Sects. 2.2, 2.3, and 2.4), the technical components of the MRI scanner (Sect. 2.5), and the basics of contrast agents and the application thereof (Sect. 2.6) are described. Furthermore, flow phenomena and MR angiography (Sect. 2.7) as well as diffusion and tensor imaging (Sect. 2.7) are elucidated. The basic physical principles of the nuclear magnetic resonance (NMR in medical literature: magnetic resonance [MR]) can be understood in depth and in detail based on quantum mechanics. In Sect. 2.2, however, another description is attempted that is almost physically exact and uses only a few simple arguments of quantum mechanics. In turn, the presentation will be more complex, but still can be understood with only basic knowledge in physics. For this reason, this synopsis should precede the detailed description in the following sections to guide the reader.

MR examinations are possible if atomic nuclei of tissue of interest possess a nuclear magnetic moment µ. Atomic nuclei with odd numbers of nucleons (here: protons, neutrons) do possess such magnetic moments. The nucleus of the hydrogen atom consisting of only one proton is the simplest atomic nucleus with an odd number of nucleons and thus has the biggest magnetic moment of all nuclei. Its natural abundance of almost 100% and its ubiquitous occurrence and the high mobility of water protons in living matter are further prepositions for using low-sensitivity NMR method for imaging in human subjects. This low sensitivity compared with other imaging methods—e.g., positron emission tomography—cannot be emphasized enough. The sensitivity difference of this both methods is several orders of magnitude (~105–106). This fact has to be taken into account when magnetic resonance imaging is envisioned for specific probe imaging, nowadays known as molecular imaging. In spite of the abovementioned low sensitivity of MR, proton imaging is possible in humans because of the high magnetic moment, ~ 100% abundance, high concentra-





2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

tion, and high mobility of protons in tissue. The following consideration will be restricted to the hydrogen nuclei only. The basis of the magnetic resonance imaging is a simple resonance phenomenon. In a magnetic field, free environmental magnetic moments of a specimen are not oriented at all; however, in an external magnetic field the magnetic moments are no longer randomly oriented. The application of an external magnetic field B0 forces the magnetic moments µ to align along the magnetic field. Due to basic physics principles, the orientation has two quantum states with respect to the external magnetic field: first the parallel, and second the antiparallel state, both of which have different magnetic energies Em, and its energy difference being ∆Em = γ · ħ · B0, and γ, ħ being the gyromagnetic ratio and Planck’s constant, respectively. In thermal equilibrium, both states possess different occupation numbers, with the low-energy parallel state having higher probability of occupation than does the low-energy antiparallel state, resulting in a macroscopic and therefore measurable net magnetization parallel to the orientation of the external magnetic field. This thermal equilibrium state can be distorted by irradiation with alternating electromagnetic field having a radiation energy ERF identically to the energy-splitting ∆Em caused by the magnetic field, and the radiation energy being ERF = ħ · ω0, and ω0 being the resonance frequency of the spin system—the so-called Larmor frequency. Due to the resonant irradiation, the spin system takes up additional energy that can be dissipated only if the system is coupled to its microenvironment. This coupling strength is described by the so-called T1 relaxation time (also known as longitudinal or spin-lattice relaxation time). An equivalent for the coupling of the spins to each other is the T2 relaxation time (also known as transversal or spin-spin relaxation time). For tissues, typical T1 relaxation times for tissues are between 100 and 2,000 ms and T2 relaxation times between 10 and 1,000 ms. MR imaging utilizing pulsed NMR—this means the alternating electromagnetic field, the so-called radiofrequency (RF) field—is applied only for a short period of time (in general, pulses are some milliseconds). The short RF pulse excites the spin system via a transmitter coil. After irradiation of the nuclear spin system, a receiver coil can detect a damped time-dependent signal with a frequency of ω0. This signal is called the free induction decay (FID). The damping of the signal is ruled by the T2 relaxation times, and the period by the strength of the external magnetic field (constant magnetic moment assumed). In practical terms, not only does the T2 relaxation time influence the damping of the signal, but also the technically related inhomogeneity of the external magnetic field. The signal damping caused by the inhomogeneity is called T2* relaxation time, and is in general much stronger than that caused by T2 relaxation times. Only special pulse sequences (e.g., spin-echo sequences)

can eliminate the influence of the inhomogeneity of the external magnetic field and thus allow the measurement of the T2 relaxation times specific to the substance/tissue. The influence of T2 relaxation times is mainly limited to the amplitude of the signal. Preposition for the image reconstruction (Sect. 2.3) is the exact information about the MR signal’s origin. This spatial information can be generated by space-dependent magnetic fields additionally applied along the three space coordinates. These space-dependent magnetic fields— called magnetic field gradients—are small as compared with the main external field and are generated by special coils mounted in the bore of the magnet. Due to these additional magnetic field gradients, the total magnetic field is slightly different in each volume element (voxel) and in turn, so is the resonance frequency of the spin system in each voxel. As a result, irradiation with a RF pulse of defined frequency ω′ excites only those nuclei in such voxels where the Larmor frequency ω0 given by the field strength matches the resonance condition. Suitable changes of the field gradients allow moving a volume element in space, fulfilling this condition. Keeping in mind that the signal intensity of a volume element is given by the number of the spins in the volume element, the relaxation times of the tissue and the specific measurement parameters (e.g., pulse repetition time, echo time etc.), this signal intensity is assigned to the corresponding picture element (pixel). In this manner, the region of interest can be sampled by moving the volume element through space, and successively, an image with respect to pixels can be constructed. This method requires a long time to acquire images, assuming every experiment needs about 1 s to measure a voxel and a pixel, respectively. Thus, the measurement of an image 128 × 128 pixels will require more than 16,000 s to complete. Nowadays, 2D-, 3D-, and/or phase encoding methods as well as half-Fourier methods are applied, allowing data acquisition times of minutes or even less. Special fast imaging techniques (e.g., FLASH, RARE, EPI sequences) allow further reduction of the acquisition time (cf. Sect. 2.4). In contrast to X-ray computed tomography, where the attenuation is governed purely by the electron density, as mentioned above, in MRI the signal intensity is a complex function of the proton density and the T1, T2, and T2* relaxation times. Additionally, the signal intensity—and hence the image contrast—can be influenced by the measurement parameters (e.g., echo time, repetition time) set at the scanner. The knowledge of these interrelations of the different parameters influencing the signal intensity and hence the image contrast is mandatory in interpreting MR images correctly. The MR scanner is a complex system (Sect. 2.5). Its main components are the magnet, the RF system, and the gradient coils. The entire system is controlled and supervised by a computer. The development of MR imaging was only possible after the development of Fourier trans-

2.1  Overview

form NMR as well as fast computers calculating fast Fourier transformations within minutes. The development of large-bore superconducting magnets of ≥ 0.3–1.5 T in the 1980s accelerated the development and the application of MRI in clinical practice. Nowadays, 3-T scanners are in routine clinical use. Scanners with ≥ 7 T are installed and will further accelerate the development of MRI and MRS. Most of the magnets are made of solenoid coils. Other magnet types, like scanners with Helmholtz coils configuration, give better access to the patients; however, are installed mostly for special purposes, e.g., in an operation suite. MR scanners with conventional resistivity magnets and fields smaller than 0.5 T are rarely used, except in countries with short supplies of helium or other restrictions that may not allow installation of a superconducting system. The risk of side effects is assumed low if the magnetic fields are ≤ 1.5 T, except for the danger caused by ferromagnetic subjects accelerated into the magnet. Nevertheless, at fields of 1.5 T and even ≥ 3 T, the knowledge about side effects is rare, especially the long-term exposure due to high static magnetic fields, gradient fields, and RF fields to organisms. The problems concerning safety are extensively discussed in Sect. 2.9. In the early days of MRI, the simplicity and wide range with which to manipulate contrasts in MRI by changing the imaging parameters led to the conclusion that development of MR contrast agents is dispensable. However, experience taught that contrast media significantly improve MR diagnostics, not only in the central nervous system, but also in other diagnostic procedures. In contrary to X-ray contrast agents, where absorption is the dominating physical effect producing the contrast, MR contrast media are based on other principles. The paramagnetic and/or super-paramagnetic properties of the contrast media influence the relaxation times of tissue, or change contrast by obliterating the signal of protons and thus increase contrast. Whereas in X-ray the contrast is proportional to the concentration of the contrast medium, in MR the dependency on the concentration is in general much stronger than linear —most often exponential. MR-contrast media are described in Sect. 2.6. The intrinsic sensitivity of NMR to motion was already observed early in the 1950s. In MR imaging, motion, in particular flow, is often recognized as artifacts. However, these phenomena can be used to measure flow and/or

represent the vascular system. Two effects are used for these kinds of measurements, the time-of-flight phenomenon (or the wash-in/wash-out effect) or the spin-phase phenomenon. In time-of-flight measurements, moving spins are excited at one location (in the vessel), and detection of the spins is performed downstream at another known location (slice). The delay time between excitation and detection can be used to calculate the flow velocity. Several modifications of the method exist (e.g., presaturation, bolus tracking), and are used depending on the setup of the measurement and sequences used. The spin-phase phenomenon can be used for angiographic imaging as well. The phase of the transverse magnetization of moving spins along a field gradient changes according to the Larmor equation. These phase-shift effects are observed for flow in all directions. The phase changes are prone to different flow parameters (e.g., velocity, turbulences, acceleration, etc.) and on the pulse sequences used. The signal variations produced by the two effects can be used to produce images of the vascular structures. Using phase-sensitive effects, magnitude subtraction is a common procedure: dephased and rephrased image are acquired sequentially and are subtracted. Using time-offlight effects, mostly maximum-intensity projection is used to construct images of the vasculature. The angiographic techniques are described in detail Sect. 2.7. Diffusion-weighted and -tensor imaging is a method applied first for clinical problems in brain, e.g. stroke, characterization of brain tumors, multiple sclerosis, etc. Molecules in gases and fluids undergo microscopic random motions due to the thermal energy proportional to the temperature of the gas or fluid. If the molecules—in this context only water molecules are considered—are imbedded in a structure, for instance in tissues, the random walk motion may be restricted by the cellular tissue structure and hence reduce diffusion constants. If the structure of tissue has a preferred direction, diffusion will no longer isotropic; the diffusion will have higher components in the preferred direction of tissue. This kind of diffusion is called anisotropic diffusion. In mathematical terms, the anisotropic diffusion can be represented by a tensor. The so-called apparent diffusion coefficient can be measured, and the anisotropy of the diffusion can be determined and contains information about the structure of tissue. The basics of diffusion imaging are elucidated in Sect. 2.8.





2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.2  Physical Basics G. Brix 2.2.1  Nuclear Spin and Magnetic Moment All nuclei with an odd number of protons and/or neutrons possess in their ground state a non-zero angular momentum or nuclear spin I, which results from the intrinsic angular momentums and the orbital angular momentums of the constituent protons and neutrons. As with any other angular momentum at the atomic and nuclear level, the angular momentum vector I is quantized. This quantization is described by the following fundamental postulates of quantum physics: • Quantization of the magnitude: The magnitude (length) |I | of the angular momentum vector can only take the discrete values |I| = ħ I(I + 1), with ħ being the Planck’s constant (ħ = 1.05 × 10–34 Js) and I the spin quantum number, which is either integer or half-integer. • Quantization of the direction: The component Iz of the angular momentum vector I along the direction of an external magnetic field is quantized. For a given value of I, only the discrete values of Iz = mħ are admitted, where m is the magnetic quantum number which is limited to the values –I, –I + 1, . . . , I – 1, I. In total, there are thus only 2I + 1 orientations of the angular momentum vector I allowed. Example: Figure 2.2.1 illustrates spin quantization in form of a vector diagram for a nucleus with the spin quantum number I = 3/2. In this case, there are 2I + 1 = 2 · 3/2 + 1 = 4 orientations of the spin vector I with the magnitude (length) |I| ћ I(I+1) ћ 3/2 · (3/2+1) ћ 15/4 allowed.

μ = γ I.

(2.2.1)

The proportionality constant γ is denoted as gyromagnetic ratio and is a characteristic property of a nuclide. Whereas all nuclei with I ≠ 0 can be used in principle for spectroscopic MR examinations, the nucleus of the hy-

Fig 2.2.1  Quantization of the nuclear spin. Vector diagram for a nucleus with the quan­tum numbers I = 3/2 and m = 1/2. The three other possible orientations of the spin vector I are drawn thinly

Remark: The spin quantum number I is frequently referred to as “nuclear spin,” which means that the maximum (minimum) component of the vector I along the chosen axes is ħI (– ħI).

The angular momentum I of an atomic nucleus is always related with a magnetic moment μ. This nuclear magnetism forms the basis of magnetic resonance. Remark: An atomic nucleus can be imagined as a rotating, positively charged sphere (Fig. 2.2.2). The rotation of the charge results in a circular electric current, inducing a magnetic dipolar field. Both the direction and magnitude of the magnetic field are characterized by the magnetic moment μ. In the simple model considered, the vector μ is collinear with the mechanical angular momentum of the sphere. Surprisingly, in quantum physics this simple relationship is even valid when the angular momentum is an inherent property of a particle (e.g., an electron or a nucleus) which is not associated with a mechanic rotation.

As shown by a large number of experiments, there is a linear relationship between the nuclear magnetic moment and the nuclear spin

Fig 2.2.2  Magnetic moment of a charged sphere. In the classical model, the rotation of a charged particle, described by its angular momentum I, results in an electric current, which induces a magnetic dipolar field. Direction and magnitude of this field are described by the magnetic moment μ. The vector μ is directed collinear to the angular momentum I of the sphere (magnetomechanic parallel­ism)

2.2  Physical Basics Table 2.2.1  MR-relevant properties of nuclei which are important for biological MR examinations (Harris 1986 [3]) Isotope

Spin quantum number l

Gyromagnetic 7 ratio γ/10 –1

–1

(rad T s ) 1

H

1/2

H

1

C

0

C

1/2

N

1

N

1/2

0

0

0

5/2

0

0

F

1/2

Na

3/2

P

1/2

2

12

13 14

15

16 17

18

19 23

31

26.752 4.1066

Resonance frequency υ  = ω/2π at B0  = 1T (MHz)

Natural abundance (%)

42.577

99.985

6.536

0.015

–

Relative MR sensitivity compared 1 to H (%) 100.0 0.96

–

98.89

–

6.7283

10.708

1.11

1.59

1.9338

3.078

99.63

0.10

– 2.7120

4.316

0.37

0.10

–

–

0.04

2.91

0.20

–

– – 3.6279

99.76 5.774

–

–

25.181

40.077

100.0

83.34

11.268

100.0

9.25

17.254

100.0

6.63

7.0801 10.841

drogen atom, which has a spin quantum number of I =1/2, is almost exclusively used in MRI due to two reasons: • It is the most abundant nucleus in biological systems. • It has the largest gyromagnetic ratio of all stable nuclei.

m = –I, –I + 1, … , I – 1, I). Consequently, there are 2I + 1 equidistant energy levels, which are denoted as nuclear Zeeman levels (Fig. 2.2.3)

Table 2.2.1 summarizes MR-relevant properties of the most important nuclei in biological tissue.

Remark: Numerous books use the magnetic field strength H instead of the magnetic flux density B. Within matter, however,

Em = –γħmB0.

(2.2.3)

2.2.2  Nucleus in a Magnetic Field 2.2.2.1  Quantum Mechanical Description In the absence of a magnetic field, all allowed orientations of the magnetic moment μ = γ I are energetically equal. This corresponds to the well-known fact that a bar magnet can be positioned arbitrarily within the field-free space; its potential energy is independent of its orientation. However, if the nucleus is located in a homogenous static magnetic field with the magnetic flux density B 0 (magnitude, B 0 = |B 0|) directed along the z-axis of a coordinate system, the nucleus has the additional potential energy E = –μz B0 ,

(2.2.2)

where μz is the z-component of the magnetic moment, which can only take the discrete values μz = γħm (with

Fig. 2.2.3  Nuclear Zeeman levels. Splitting of the energy levels of a nucleus with the spin quantum number I = 3/2 in an external magnetic field with the flux density B 0 . The energy difference between the four equidistant nuclear Zeeman levels is ΔE = ħω0 = γħB 0



10

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy the B field represents the “real” magnetic field that interacts with the magnetic moments of the nuclei. The relation between the two magnetic field quantities is explained in Sect. 2.2.8.1.

When considering an isolated magnetic moment within a static magnetic field, one will find that transitions between the different energy levels are prohibited due to the law of energy conservation. Transitions can exclusively be induced by an additional time-dependent electromagnetic RF field that interacts with the magnetic moment, the effect is known as magnetic resonance (MR). In MR, transitions are induced by a magnetic RF field B1(t) with the angular frequency ωRF, which is irradiated perpendicular to the direction of the static magnetic field B 0 . Such a time-dependent magnetic field, however, can only induce transitions fulfilling the selection rule ∆m = ±1, i.e., transitions between neighboring energy levels. As a consequence, the energy ERF = ħωRF of a photon of the RF field must be identical with the energy difference ΔE = ħω0 = γħB 0 between two neighbored energy levels, which yields the resonance condition

not only be described by quantum physics, but also by a classical approach, which is mediated by the intuitive semi-classical model described in the next section. 2.2.2.2  Semiclassical Description

(2.2.4)

In an external magnetic field, a cylindrical permanent magnet—characterized by a magnetic moment μ—experiences a mechanical torque that tends to align the permanent magnet parallel to the external magnetic field and thus minimize the poten­tial energy of the system. However, in the case that the permanent magnet rotates around its longitudinal axis and thus possesses an angular momentum (“magnetic gyroscope”), it cannot align parallel to the external field due to the conservation of the angular momentum. In this situation, it experiences a torque perpendicular to both the direction of the magnetic field and the angular momentum, which results in a rotation (precession) of the magnet on a cone about the direction of the external B 0 field (see Fig. 2.2.4b). The frequency of this precession, the Larmor frequency, corresponds to the resonance frequency ω0 given by Eq. 2.2.4.

Remarkably, Planck’s constant ħ does not occur in this fundamental equation of magnetic resonance. This indicates that the basic principles of magnetic resonance can-

Remark: The precession of a magnetic gyroscope in an external magnetic field can be illustrated by a mechanic analog. When a child’s spinning top is deflected so that its axis is not parallel to

Fig 2.2.4  Analogy between an atomic nucleus and a top. a Precession of a rotating top in the gravitational field G of the earth. b Precession of a magnetic moment μ around the direction of a static magnetic field B 0. The fundamental difference between

the top and the nucleus is that the nucleus possesses an intrinsic angular momentum I, whereas the angular momentum L of the top has to be initiated mechanically

ωRF = ω0 = γB 0.

2.2  Physical Basics the direction of the gravitational field, it will continue rotating around its axis, but the axis itself will start rotating—the top precesses on a cone around the direction of the gravitational field (Fig. 2.2.4a). It should be mentioned, however, that the child’s top and the nucleus differ with regard to the fact that the child’s top has to be spun, whereas the nucleus possesses an intrinsic angular momentum.

The quantization of direction of the nuclear magnetic moment μ can be integrated into this classical description by limiting the angle between the field axis and the precession cone to the discrete values which relate to the 2I + 1 orientations of the angular momentum I permitted. For a spin-1/2 nucleus, this results in a double-precession cone as shown in Fig. 2.2.5. However, this semiclassical model is rendered questionable, because the classical concept of a continuous trajectory in space is hardly compatible with the quantization of physical quantities. For instance, what would the trajectory of the vector μ look like when transitions between the various precession cones, reflecting discrete energy levels, are induced by an RF field, such as for a spin-1/2 nucleus, the transition from the lower to the upper precession cone (cf. Fig. 2.2.5)? Is it possible to assign to the vector μ a well-defined direction in space at any point in time, and would this direction change over time? If so, then this negates the postulate of discrete energy and angular momentum levels. This aporime can only be solved by a rigorous quantum mechanical treatment of the system. However, when considering only the mean values of physical quantities averaged over a large ensemble of nuclei—which can only be measured in a real MR experiment—it becomes obvious that the models and laws of classical physics are valid.

However, as compared with the thermal energy, the difference between the two energy levels is extremely small, so that the difference in the occupation numbers of the two levels is very small. At body temperature of 37°C, the difference in the occupation numbers with respect to the total number of spins is as low as 0.000003!

Fig 2.2.5  Double-precession cone for a nucleus with the nuclear spin quantum number I = 1/2. The two permitted spin states (precession cones) are characterized by the magnetic quantum numbers m = ±1/2

2.2.3  Macroscopic Magnetization In field-free space, the magnetic moments of nuclei in a macroscopic sample are randomly oriented due to their thermal motion and thus mutually compensate each other. In a homogeneous static magnetic field B 0, however, only 2I + 1 discrete orientations of the magnetic moments with respect to the direction of the external field are permitted, the energy levels of which differ according Eq. 2.2.3. In thermal equilibrium, the population of the 2I + 1 levels (spin states) is described by the Boltzmann statistic: The lower the energy Em = –γħB0m of a state with the magnetic moment μz = γħm in the zdirection, the greater is the occupation number. Example: Let us consider an ensemble of hydrogen nuclei in a static magnetic field of the flux density B0 = 1 T. According to the Boltzmann statistic, more nuclei will occupy the state of the lower energy (m = +1/2, µz parallel to B 0) than the state of the higher energy (m = –1/2, µz antiparallel to B 0) (Fig. 2.2.6).

Fig 2.2.6  Origin of the nuclear magnetization. In thermal equilibrium, the distribution of an ensemble of spin-1/2 nuclei on the two allowed precession cones is described by the Boltzmann statistic. The occupation number of the state of the lower energy (m = +1/2, µz parallel to B 0) is somewhat higher than that of the state of the higher energy (m = –1/2, µz antiparallel to B 0) which leads to macroscopic (bulk) magnetization M0

11

12

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Although the difference in the occupation numbers is extremely small, it results in a measurable bulk magnetic moment along the direction of the B 0 field due to the large number of nuclei in a macroscopic sample (“nuclear paramagnetism”). The macroscopic magnetization in thermal equilibrium is described by the magnetization vector M0, which is defined as the vector sum of the nuclear magnetic moments per unit volume V. The magnitude of the equilibrium magnetization M0 is given by M 0 |M0|

1 V

N i=1

( µz )i

N I(I+1) y 2 ħ2 B0 V 3kT

(2.2.5)

where N is the total number of nuclei in the sample, T the absolute temperature of the sample, and k Boltzmann’s constant (k = 1.38 · 10–23 J/K). The ratio ρ = N/V is called spin density. As both the body temperature and the spin density cannot be altered in living beings, the equilibrium magnetization M0 can only be increased according to Eq. 2.2.5 by increasing the magnetic flux density B0. 2.2.4  Dynamic of Magnetization I: Resonance Excitation The equilibrium state of a spin system can be disturbed by a magnetic RF field B 1(t) with a frequency ωRF equal to the Larmor frequency ω0, which tilts the magnetization M. Whereas a nuclear magnetic moment μ can only take 2I + 1 discrete orientations relative to the static magnetic field B 0 (quantization of direction), the macroscopic magnetization M can take any direction in space and change it steadily. The action of a magnetic RF field B 1(t), which rotates with the Larmor frequency ω0 around the direction of the static B 0 field, can be analyzed most effectively in a rotating frame, i.e., a coordinate system that rotates with the Larmor frequency around the z-axis (Fig. 2.2.7). The change to a rotating frame with the axes (x′, y′, z) has two advantages: • As the x′–y′-plane of the rotating frame is synchronized with the RF field, the B 1 vector remains stationary in this frame. In the following analysis, we will assume that the static B 1 field points along the x′-axis (Fig. 2.2.7). • As shown in Sect. 2.2.2.2, a nuclear magnetic moment μ precesses with the Larmor frequency ω0 around the direction of the B 0 field (see Fig. 2.2.5). Of course, this holds equally for the sum of the nuclear magnetic moments, i.e., for the macroscopic magnetization M. Therefore, an observer observing the precession of the magnetization M from the rotating frame will come to conclude that the position of the magnetization does not change. From his point of view, the magnetization behaves as if the B 0 field is absent (Larmor’s theorem).

Fig 2.2.7  Radiofrequency field in a stationary and in a rotating frame of reference. In the stationary frame (x, y, z) the magnetic RF field B 1(t) rotates with the angular frequency ωRF in the x–yplane around the z-axis. If one observes this rotation from a rotating frame (x′, y′, z), which rotates with the angular frequency ωRF around the z-axis, the vector is stationary. Typically, the rotat­ing frame is chosen in such a way that the B 1 field points in the x′-direction

Summarizing both reflections, it can be concluded that the dynamics of the magnetization M in the rotating frame is determined only by the static B 1 field. If it points toward the x′-axis, then the magnetization M will precess around the x′-axis (Fig. 2.2.8a). Analogous to Eq. 2.2.4, the frequency ω1 of this precession is given by ω1 = γB1.

(2.2.6)

When looking at this simple rotation of the magnetization M in the y′–z-plane of the rotating frame from a laboratory frame of reference (x, y, z), the movement is superimposed by a markedly faster rotation (B0 > B1) around the z-axis. Thus, within the laboratory frame of reference, the tip of the vector M moves in a helical manner on the surface of a sphere around the B 0 field; the length of the vector M remains constant (Fig. 2.2.8b). If the magnetization M points toward the static field B 0 before the RF field B 1(t) is switched on, the magnetization M is rotated from the equilibrium position under the influence of the RF field during the duration tp by the flip angle: α = ω1tp = γB1tp.

(2.2.7)

If the duration tp of the RF field is chosen to rotate the magnetization in the rotating frame by 90°, then this

2.2  Physical Basics

pulse is denoted as 90° or π/2 pulse (Fig. 2.2.9a). Accordingly, the magnetization M is rotated by 180° when the duration of the RF pulse is doubled at the same flux density B1. This pulse, which inverts the magnetization from the positive to the negative z-direction, is called 180° or π pulse (Fig. 2.2.9b). Remark: Precisely speaking, a short RF pulse with the carrier frequency ωRF will excite not only the nuclei that exactly fulfill the resonance condition ωRF = ω0, but also nuclei whose resonance frequency slightly differs from ωRF. This is because the frequency spectrum of an RF pulse of finite duration consists of a continuous frequency band around the nominal frequency ωRF (Fig. 2.2.10). The width of the frequency distribution is inversely proportional to the duration tp of the pulse: the shorter the pulse, the broader the frequency spectrum is be distributed around ωRF. If the RF field is irradiated over a very long period (tp → ∞), the spectrum will be quasi-monochromatic.

To simplify the following analysis, the magnetization M is separated into two components: the longitudinal magnetization Mz, which is parallel to the direction of the static magnetic field B 0 , and the transverse magnetization Mxy, which is perpendicular to it (Fig. 2.2.11). In the laboratory frame the transverse magnetization Mxy precesses with the Larmor frequency ω0; in the rotating frame it remains stationary. It is instructive to describe the effect of a 90°/180° pulse on an ensemble of spin-1/2 nuclei within the semiclassi-

cal model described in Sect. 2.2.2.2. As can be shown, the magnetic RF field induces transitions between the two permitted spin states (precession cones) until the occupation numbers are either identical (90° pulse) or inverted (180° pulse). Furthermore, irradiation of a 90° pulse results in a phase synchronization of the nuclear magnetic moments of the sample, which yields a macroscopic transverse magnetization Mxy, the magnitude of which is equal to that of the equilibrium magnetization M0. Figuratively speaking, this means that the precession of the transverse magnetization Mxy can be described as a common (phase coherent) precession of a “spin package” (Fig. 2.2.12). 2.2.5  Dynamic of Magnetization II: Relaxation Up to this point, we have assumed that interactions of nuclear spins between one another and with their environment can be neglected. However, this assumption is not valid for real spin systems, as the magnetization returns to its equilibrium (Mxy = 0, Mz = M0) after RF excitation. This process is called relaxation. Two different relaxation processes have to be distinguished: • The relaxation of the longitudinal magnetization Mz characterized by the longitudinal or spin-lattice relaxation time T1 • The relaxation of the transverse magnetization Mxy characterized by the transverse or spin-spin relaxation time T2.

Fig 2.2.8  Resonance excitation. a In a rotating frame of reference, which rotates with the Larmor frequency ω0 around the direction of the B0 field, the magnetization M precesses with the frequency ω1 around the stationary B1 field. b In the stationary frame this simple rotation is superimposed by the markedly faster rotation around the z-axis. Therefore, the tip of the vector M moves in a helical manner on the surface of a sphere

Fig. 2.2.9  90° and 180° pulse. If one chooses the rotating frame so that the RF pulse is irra­diated along the x′-axis, the magnetization M will be rotated (a) by a 90° pulse along the y′-direction and (b) by a 180° pulse to the negative z-direction

13

14

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.2.12  Phase synchronization by a 90° pulse. The 90° pulse leads to a synchronization of the phases of the magnetic moments μ of the nuclei in the sample (spin packet), which results in a macroscopic transverse magnetization Mxy, the magnitude of which corresponds to that of the longitudinal magnetization before irradiation of the 90° pulse. In the figure, only the part of the magnetic moments of the sample which are distributed in an anisotropic manner on the precession cone is shown

2.2.5.1  Physical Model of Relaxation Processes Fig. 2.2.10  RF pulse in the time and frequency domain. a RF pulse with carrier frequency ωRF and duration tp. b Fourier transformation of the RF pulse. Due to its finite duration, the frequency spectrum of the pulse is not monochromatic, but contains an entire frequency band, which is distributed around the nominal frequency ωRF

In real spin systems, every nucleus is surrounded by other intra- and intermolecular magnetic moments, which are in motion due to rotations, translations, and vibrations of molecules as well as exchange processes. These processes induce an additional fluctuating magnetic field Blok(t) at the position of a given nucleus, which has to be added to the external field. As the movements and exchange processes are random, the fluctuating fields differ in time from nucleus to nucleus—in contrast to the coherent RF field Blok(t) irradiated from the outside. As any other temporal process, the locally fluctuating magnetic fields Blok(t) can be decomposed into its frequency components. Remark: The decomposition of a function into harmonic (i.e., sinusoidal) basis functions is denoted as Fourier analysis, the mathematical operation that gives the intensity (amplitude) of the harmonic basis functions as Fourier transformation. If the given function is periodic with period T, it can be decomposed into a sum of sinus and/or cosine functions with the discrete frequencies ω, 2ω, 3ω . . . (ω = 2π/T ). In contrast, a nonperiodic function has a continuous spectrum of frequencies.

Fig. 2.2.11  Definition of the longitudinal and transverse magnetization. As the macro­scopic magnetization M precesses in the stationary frame around the z-axis, it is beneficial to split it into two components: the rotating transverse magnetiza­tion Mxy and the longitudinal magnetization Mz

The contribution of the different frequency components to the fluctuating local field Blok(t) is described by the spectral density function J(ω). A general feature of this function is that the more rapidly the molecular motion is, the broader the frequency spectrum (Fig. 2.2.13).

2.2  Physical Basics

In order to understand the effect of the fluctuating local magnetic fields Blok(t) on a spin system, the components parallel and perpendicular to B0 have to be discussed separately. Whereas the parallel component exclusively contributes to T2 relaxation, the perpendicular component influences both T1 and T2 relaxation: • The field component perpendicular to the B0 field induces—in analogy to the external RF field B1(t) —transitions between the energy levels (precession cones) of an individual spin. The probability of these transitions depends on the intensity of the frequency component of the fluctuating fields that oscillates at the Larmor frequency ω0: the higher the spectral density J(ω0), the more transitions are induced. As Fig. 2.2.13 shows, J(ω0) assumes a maximum when the limiting frequency ωG of the spectral density function is comparable to the Larmor frequency ω0. The described relaxation process allows the excited spin system to emit and absorb photons of energy ħω0 until the Boltzmann distribution of the energy levels is reached. The energy difference between the excited and the equilibrium state is dissipated to the surrounding medium or “lattice.” Since the change in the occupational numbers of the spin states (precession cones) is related with a change in the macroscopic longitudinal magnetization Mz, the described mechanism contributes to longitudinal relaxation. Moreover, it contributes to T2 relaxation, as the locally induced transitions between the precession cones destroy the phase coherence between those spins which form, as a spin-package, the macroscopic transverse magnetization (cf. Fig. 2.2.12). • The component of the fluctuating field Blok(t) oriented parallel to the z-axis locally modulates the static field B0 at the position of a nucleus and thereby changes the precession frequency ω0 of its nuclear magnetic moment μ. Since the local fluctuations seen by the nuclei are spatially uncorrelated, the precessing magnetic moments within a sample lose their phase coherence, which causes the transverse magnetization to decay (see Fig. 2.2.12). Given the fact that the effect of the high-frequency components of the fluctuating field vanishes when averaged over time, only the quasistatic frequency components, the intensity of which is approximately given by J(ω = 0), have a measurable effect on the transversal magnetization (see Fig. 2.2.13). As no transitions between the energy levels (precession cones) are induced by the described relaxation mechanism, the longitudinal magnetization Mz remains unchanged, which means that the mechanism solely contributes to transversal relaxation. The qualitative discussion of the relaxation mechanisms reveals that their effectiveness depends on two different factors, namely on the magnitude and the temporal characteristics of the field fluctuations. The dependence from the magnitude is utilized when using paramagnetic con-

Fig. 2.2.13  Schematic representation of the density function J(ω) for three substances with a different thermal mobility of the constituting atoms or molecules. a If the atoms or molecules move very slowly (such as in solids), the intensity of high-frequency components is very low. b This is different in fluids. In this case, the atoms or molecules move very rapidly, so that the spectral density function contains high-frequency components to a significant degree. c At a given frequency ω0 the intensity J (ω0) will attain a maximum if the cut-off frequency ωG of the spectral density function approximately corresponds to the given frequency ω0. At low frequencies, J (ω) is nearly independent on the frequency, so that the density of the quasi-static frequency components can be approximated by J (ω = 0)

trast agents (see Sect. 2.6), which possess unpaired electron spins and consequently a magnetic moment. When considering the fact that the magnetic moment of an electron amounts to 658 times the magnetic moment of a proton, one can easily understand why even the slightest amounts of paramagnetic substances can lower the relaxation times considerably. 2.2.5.2  Phenomenological Description of Relaxation Processes For spin systems with a sufficiently high molecular mobility, relaxation processes can be described by exponential functions with the time constant T1 or T2. The longitudinal magnetization increases exponentially toward its equilibrium value Mz = M0, the transverse magnetization decreases exponentially toward Mxy = 0. Figure 2.2.14 shows the exponential relaxation of both magnetization components after excitation of the spin system by a 90° pulse and gives a simple interpretation of the relaxation times T1 and T2: • The longitudinal relaxation time T1 gives the time required for the longitudinal magnetization after a 90° pulse to grow again to 63% of its equilibrium value M0. • The transverse relaxation time T2 gives the time required for the transverse magnetization after a 90° pulse do drop to 37% of its original magnitude.

15

16

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.2.14  Relaxation of the longitudinal and transverse magnetization. After excitation by a 90° pulse, the longitudinal magnetization Mz relaxes toward the equilib­rium magnetization Mz = M0, and the transverse magnetization toward Mxy = 0. The temporal evolution is defined by the relaxation times T1 and T2, respec­tively (e–1 ≅ 0.37)

Fig. 2.2.15  Dephasing of the transverse magnetization. The transverse magnetization Mxy of the sample is split up into several magnetization components, which precess with slightly differing Larmor frequencies around the direction of the B0 field. a Immediately after the 90° pulse, all magnetization components are aligned parallel. b–d Afterward, the components dephase due to their different Larmor frequencies, and thus the macroscopic transverse magnetization decays

2.2  Physical Basics

The process of transverse relaxation can be described intuitively on the macroscopic level. To this end, the transversal magnetization Mxy is split into different magnetization components, or spin packets. Whereas the spins of each spin packet precess with the same Larmor frequency, the spins in different packets slightly differ in their Lamor frequencies. Right after excitation, all components of the magnetization point toward the same direction; shortly afterward, however, some parts precess more quickly than others around the direction of the B0 field. Due to this fact, the components fan out (dephasing), and the resulting transverse magnetization decreases (Fig. 2.2.15). In real MR experiments always macroscopic samples are examined, so that not only the fluctuating local magnetic fields, but also spatial field inhomogeneities of the external field B0, introduced by technical imperfections, contribute to the transverse relaxation. As both effects superpose on one another, the resulting effective relaxation time T2* is always shorter than the real, substancespecific transverse relaxation time T2. 2.2.5.3  Proton Relaxation Times of Biological Tissues Relaxation times in solids and fluids differ markedly (Fig. 2.2.16). Whereas the longitudinal relaxation in solids can take hours or even days, in pure fluids it only takes some seconds. This difference is because the spectral density function J(ω0) at the Larmor frequency is much larger in fluids than it is in solids, in which the low-frequency components dominate (see Fig. 2.2.13). For the same physical reason, the T2 relaxation time in solids usually only amounts to some microseconds, whereas in fluids it is only slightly shorter than the longitudinal relaxation time T1. Soft tissues range, based on their consistency, between solids and pure fluids: with regard to their relaxation be-

Fig. 2.2.16  Relaxation in fluids and solids. The relaxation behavior of a substance depends strongly on the thermal mobility of the constituting atoms and mole­cules. For fluids with a high thermal mobility the relation T1 ≅ T2 holds, for solids T1 >> T2. The relaxation time T1 is minimal, when the cut-off frequency of the spectral density function J(ω) of the substance approximately corre­sponds to the Larmor frequency (see Fig. 2.2.13)

havior, they can in general be treated as viscose fluids. Table 2.2.2 summarizes representative proton relaxation times for different biological tissues. Due to the considerable differences in the tissue relaxation times, it is possible to acquire MR images with an excellent tissue contrast even when the proton densities of the tissues or organs only slightly differ from one another. When interpreting the relaxation times, two aspects have to be taken into account: • The relaxation time T1 of biological tissues strongly depends on the Larmor frequency, whereas the relaxation time T2 is nearly independent of the frequency. When comparing T1 values, one therefore needs to

Table 2.2.2  1H relaxation times of biological tissues at different magnetic flux densities B0 (Bottomley et al. 1984 [2]) Tissue

T2 (ms)

T1 (s) at 0.5 T

T1 (s) at 1.0 T

T1 (s) at 1.5 T

Skeletal muscle

47±13

0.55±0.10

0.73±0.13

0.87±0.16

Heart muscle

57±16

0.58±0.09

0.75±0.12

0.87±0.14

Liver

43±14

0.33±0.07

0.43±0.09

0.50±0.11

Kidney

58±24

0.50±0.13

0.59±0.16

0.65±0.18

Spleen

62±27

0.54±0.10

0.68±0.13

0.78±0.15

Fatty tissuea

84±36

0.21±0.06

0.24±0.07

0.26±0.07

Grey brain matter

101±13

0.66±0.11

0.81±0.14

0.92±0.16

White brain matter

92±22

0.54±0.09

0.68±0.12

0.79±0.13

In fatty tissue, the single component relaxation times have to be considered as rough estimations

a

17

18

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

consider the magnetic flux density B0 at which the T1 measurement was done. • Relaxation processes often consist of multiple components, so that the description by a mono-exponential function is only a rough approximation. The relaxation times given in Table 2.2.2 therefore only represent weighted mean values of an entire spectrum of exponential functions, characterizing the relaxation behavior of protons in different cell and tissue compartments between which the water exchange is slow. However, at the timescale relevant for MRI, relaxation processes of most tissues can be approximated rather well by a single exponential function. An exception is fat-containing tissue (such as subcutaneous fatty tissue or bone marrow), which demands at least two exponential functions to be considered for the parameterization of the relaxation processes observed. 2.2.6  The MR Experiment Figure 2.2.17 shows the general setup of an MR experiment; technical details will be presented in Sect. 2.5. The sample to be examined is located within a very homogeneous static magnetic field B0, which is created either by a permanent magnet or by a (superconducting) coil. The RF field required for the excitation of the spin system is generated by a transmit coil connected to the RF transmit system. This RF coil is positioned in such a way that the radiofrequency field B1(t) is irradiated perpendicular to the B0 field into the sample volume. Remark: Whereas atomic nuclei with a nuclear spin quantum number of I ≥ 1 can interact with both the electric and the magnetic component of the electromagnetic RF field, spin-1/2 nuclei are only affected by the magnetic component B1(t) of the RF field.

Fig. 2.2.17  Principle setup of an MR experiment. The object to be measured is placed within a homogeneous static magnetic field B0. Excitation of the spin system is performed by an RF field B1(t) irradiated perpendicularly to B0 by an RF coil. After

After excitation of the spin system by an RF pulse, the precessing transverse magnetization Mxy in turn induces a weak alternating voltage in a receiver coil, which in general is identical to the transmit coil (Fig. 2.2.18a). The measured voltage is amplified, filtered, digitalized, and fed to the computer of the MR system. The measured MR signal S(t) has the form of a damped oscillation (Fig. 2.2.18b), which is denoted as the free induction decay (FID). The FID signal has the following characteristic features: • It oscillates with the Larmor frequency ω0 of the stimulated nuclei. • It decays in time with the time constant T2*. • Its initial amplitude is proportional to the number N of the excited spins in the sample (N = ρV ∝ M0V; cf. Eq. 2.2.5). If the sample contains nuclei of a certain type whose resonance frequency slightly differs due to intramolecular interactions (see Sect. 2.2.8), the MR signal induced in the receiver coil will consist of several interfering decay curves. However, such a curve is rather complicated to analyze and interpret. Therefore, the detected curve is usually spit up into its frequency components (Fourier analysis, see Sect. 2.2.5.1) and presented as frequency spectrum. Both types of description are merely different representations of the same data, which can be transformed into one another mathematically by a Fourier transformation. Example: Figure 2.2.18b,c illustrates the relation between the description of the MR signal in the time or frequency domain by the example of a substance whose MR spectrum only shows one resonance line.

excitation, the MR signal of the sample is detected by an RF coil and transferred via a receiver channel to the computer of the MR system. (For details, see Sect. 2.5)

2.2  Physical Basics

For quantitative analysis of an MR spectrum, the following features are important:

• The center of the resonance curve is at the Larmor frequency ω0. • The full width Δω at half maximum of the curve is related with the characteristic time constant T2* of the FID by the relation Δω = 2/T2*. • The area under the curve is approximately proportional to the number of excited nuclei in the sample. 2.2.7  Standard Pulse Sequences In an MR experiment, only the RF signal can be determined by measurement, which is induced by the rotating transverse magnetization Mxy in the receiver coil (cf. Sect. 2.2.6). Nevertheless, a large variety of MR experiments can be realized that differ in the way by which the spin system is excited and prepared by means of RF pulses before the signal is acquired. A defined sequence of RF pulses, which is usually repeated several times, is called a pulse sequence. In the following, three “classical” pulse sequences are described that are frequently used for MR experiments (imaging sequences are described in Sect. 2.4): • The saturation recovery sequence • The inversion recovery sequence • The spin-echo sequence. 2.2.7.1  Saturation Recovery Sequence

Fig. 2.2.18  Free induction decay (FID) and frequency spectrum. a After excitation of the spin system by a 90° pulse the magnetization Mxy precesses with the Larmor frequency ω0 around the direction of the B0 field and induces an electric voltage in the receiver coil. b The measured FID signal S(t) has the form of a damped oscillation, the frequency of which is given by the Larmor frequency ω0. The decay of the signal is defined by the time constant T2*. c A Fourier transformation of the FID signal gives the frequency spectrum of the MR signal. The resonance curve has its center at the Larmor frequency ω0; its full width at half maximum (FWHM) is related with the characteristic time constant T2* of the FID by the relation Δω = 2/T2*

The saturation recovery (SR) sequence consists of only a single 90° pulse, which rotates the longitudinal magnetization Mz into the x–y-plane. The FID signal is acquired immediately after the RF excitation of the spin system. After a delay time, the repetition time TR, the sequence is repeated. The SR sequence is described schematically by the pulse scheme (90°–AQ– TR) (AQ = signal acquisition period; Fig. 2.2.19a). If the repetition time TR is long compared to T1, the magnetization M relaxes back to its equilibrium state (see Fig. 2.2.14). In this case, the initial amplitude of the FID, even after repeated excitations, does only depend on the equilibrium magnetization M0 and does not show any T1 dependency. However, if the repetition time TR is shortened to a value that is comparable to T1, the longitudinal magnetization Mz will not fully relax after excitation, and the following 90° pulse will rotate the reduced longitudinal magnetization Mz(TR) = M0[1–exp(–TR/T1)] into the x–y-plane (Fig. 2.2.19b, c). Under the assumption that the transverse magnetization after the repetition time TR has been decreased to zero (TR >> T2*), the following expression is obtained for the initial amplitude SSR of the FID signal: SSR ∝ N (1–e–TR/T1),

(2.2.8)

which exclusively depends on the relaxation time T1 and the number N of the excited spins in the sample.

19

20

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.2.19  Saturation recovery sequence. a Pulse scheme of the SR sequence (AQ: signal acquisition). b The 90° pulse rotates the actual longitudinal magnetization into the x–y-plane. During the repetition time TR, the longitudinal magnetization relaxes toward the equilibrium magnetization M0. The speed of this process is described by the longitudinal relaxation time T1. Note that by the first 90° pulse, the equilibrium magnetization M0 is rotated into the x–y-plane, whereas the subsequent 90° pulses rotate the reduced longitudinal magnetization Mz(TR) = M0[1 – exp(–TR /T1)]. c Temporal evolution of the transverse magneti­zation Mxy in the rotating frame. d Induced MR signal SSR(t)

2.2.7.2  Inversion Recovery Sequence In the inversion recovery (IR) method, the longitudinal magnetization is inverted by a 180° pulse (inversion pulse), which is followed after an inversion time TI by a 90° pulse (readout pulse). Immediately after the 90°

Fig. 2.2.20  Inversion recovery sequence. a Pulse scheme of the IR sequence (AQ: signal acquisition). b Initially, the longitudinal magnetization is inverted by the 180° pulse (inversion pulse), which is followed after an inversion time TI by a 90° pulse (readout pulse), which rotates the existing longitudinal magnetization Mz(TI) into the x–y-plane. After the 90° pulse, the longitudinal magnetization relaxes toward the equilibrium magnetization M0. c Temporal evolution of the transverse magnetization Mxy in the rotating frame. d Induced MR signal SIR(t)

pulse, which rotates the partially relaxed longitudinal magnetization Mz(TI) into the x–y-plane, the FID signal is acquired (Fig. 2.2.20). The IR sequence is described by the pulse scheme (180°– TI – 90° – AQ). The initial amplitude SIR of the FID signal is directly proportional to the longitudinal magnetization immediately before irradiation of the read-out pulse, just as is the case in the SR method. In contrast to

2.2  Physical Basics

the SR sequence, however, the change in the longitudinal magnetization is twice as high and thus—in analogy to Eq. 2.2.8—the following expression is obtained (compare Figs. 2.2.19b and 2.2.20)

2.2.7.3  Spin-Echo Sequence

Remark: If the IR sequence is repeated several times with different inversion times TI, it is possible to sample the temporal course of the longitudinal magnetization step by step, since the initial amplitude of the FID signal is directly proportional to the longitudinal magnetization at time TI (see Fig. 2.2.20). This procedure is applied frequently in order to determine the relaxation time T1 of a sample according to Eq. 2.2.9.

As explained in Sect. 2.2.5.2 the temporal decay of the transverse magnetization Mxy is caused by two effects: fluctuating local magnetic fields and spatial inhomogeneities of the magnetic field B0. The transverse magnetization Mxy therefore relaxes not with the substance-specific relaxation time T2 but rather with the effective time constant T2* (T2* < T2). When determining the relaxation time T2, it is therefore important to compensate the effect of the field inhomogeneities. This can be done, as E. Hahn has already shown in 1950, by using the so-called spinecho (SE) sequence. This sequence utilizes the fact that the dephasing of the transverse magnetization caused by B0 inhomogeneities is reversible since they do not vary in time, whereas the influence of the fluctuating local magnetic fields is irreversible. In order to understand the principle of the SE sequence with the pulse scheme (90° – τ – 180° – τ – AQ; see Fig. 2.2.22a), we initially neglect the influence of the

Fig. 2.2.21  Explanation of the spin-echo experiment in the rotating frame. For the sake of simplicity, the substance-specific transverse relaxation is not been considered in this figure. a The 90° pulse rotates the longitudinal magnetization into the x′–y′plane. b,c In the course of time, the magnetization components, which form together the transverse magnetization Mxy, dephase so that the transverse magnetization decays with the characteristic time constant T2* (see Fig. 2.2.15). d,e Irradiation of the 180° pulse along the x′-axis mirrors the dephased magnetiza-

tion vectors at the x′-axis. As neither the precession direction nor the precession velocity of the magnetization components are altered by the 180° pulse, the components rephase and thus the transverse magnetization increases. The regeneration of the transverse magnetization is called a spin echo. f At the time TE = 2τ, all magnetization components point into the same direction again. Due to the rephrasing effect of the 180° pulse, the amplitude Mz(TE = 2τ) of the spin echo is independent of the static inhomogeneities of the B0 field

SSR ∝ N (1–2e–TI/T1). (2.2.9) The derivation of this relation is based on the assumption that the spin system is in its equilibrium state before it is excited by the inversion pulse. When repeating the IR sequence, one has therefore to make sure that the repetition time TR is markedly longer than the relaxation time T1.

21

22

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

frame, one observes a fanning out of the magnetization components around the y′-axis (Fig. 2.2.21.b,c). If a 180° pulse is applied after a time delay τ along the x′-axis, the magnetization components will be mirrored with respect to this axis (Fig. 2.2.21d). However, the 180° pulse does not change the rotational direction of the magnetization components, but merely inverts the distribution of the components: the faster components now follow the slower ones (Fig. 2.2.21e). After the time t = 2τ, all magnetization components again point to the same direction, and the signal comes to a maximum (Fig. 2.2.21f). The 180° pulse thus induces a rephasing of the dephased transverse magnetization, which causes the MR signal to increase and to generate a spin echo (Fig. 2.2.22). After the spin-echo time TE = 2τ, the echo decays again— as the original FID does—with the time constant T2*. Due to the rephasing effect of the 180° pulse, the spinecho signal SSE(TE) is independent from the inhomogeneities of the static magnetic field: the loss of signal at the time t = TE as compared to the initial signal SSE(0) is determined exclusively via the substance-specific relaxation time T2. If one irradiates a sequence of K 180° pulses at the times τ, 3τ, 5τ, …, (2K–1)τ, one can detect a spin echo in between the subsequent 180° pulses (Fig. 2.2.23). The envelope of the echo signals SSE(2τk) (k = 1, 2, 3, …, K) decays exponentially with the relaxation time T2. SSE ∝ N e–2τk/T2.

(2.2.10)

The major advantage of this multi echo sequence consists in the fact that the T2 decay can very effectively be detected by a single measurement (Fig. 2.2.23). 2.2.8  Influence of the Electron Shell on the Local Magnetic Field Fig. 2.2.22  Spin-echo sequence. a Pulse scheme of the SE sequence (AQ: signal acquisi­tion). b Temporal evolution of the longitudinal magnetization Mz. Note that the 180° pulse at the time t = τ inverts the longitudinal magnetization. c Temporal evolution of the transverse magnetization Mxy in the rotating frame. After excitation of the spin system by the 90° pulse, the transverse magnetization decays with the characteristic time constant T2*. The 180° pulse results in a regeneration of the transverse magnetization denoted as spin-echo. d Induced MR signal SSE(t)

fluctuating local magnetic fields and solely consider the static magnetic field inhomogeneities. Immediately after the 90° pulse, all magnetization components composing the transverse magnetization Mxy point along the y′-axis (Fig. 2.2.21a). Shortly afterward, some components precess faster, others more slowly around the direction of the B0 field, so that the initial phase coherence is lost (see Fig. 2.2.15). When looking at this situation from a rotating

All the considerations so far have been based on the assumption that the external magnetic field B0 created by the RF coil is not altered by the electrons surrounding a nucleus. However, this is not the case as the electrons interact with the applied external magnetic field. In biological tissues in which atoms are covalently bound, two related effects need to be considered, the diamagnetism and the chemical shift. 2.2.8.1  Macroscopic Effect: Diamagnetism Diamagnetism is a general feature of matter and is because electrons attempt to shield the interior of the sample against the external magnetic field. In electrodynamics, this effect is described by Lenz’s law. It states that the current induced in a circuit by the change of a magnetic field is directed in such a way that the secondary magnetic field induced by the electric current weakens the

2.2  Physical Basics

in the electron shell are “frictionless,” which means that an induced electron current remains constant until the external magnetic field changes or until the sample is removed from the magnetic field. The sum of the induced magnetic moments of the electrons per volume is—similar to the nuclear magnetization M—denoted as electron magnetization Me. For averaging, the volume has to be chosen in such a way that, on the one hand, a great number of atoms and molecules is contained, and, on the other hand, that it is small compared to the volume of the sample (for example, 1 µm3 water contains about 3.3 · 1010 water molecules). The magnetization Me thus represents a macroscopic quantity per definitionem.

Fig. 2.2.23  Multi echo sequence. a Pulse scheme (AQ: signal acquisition). b Decay of the echo amplitudes as a function of time. The signal decay is determined exclu­sively by the substance-specific transverse relaxation time T2, whereas the decay and the regeneration of the FID are essentially determined by technically conditioned field inhomogeneities

Remark: Due to practical reasons, distinction is made in electrodynamics between free and bound currents: Free currents are experimentally controllable and are linked to macroscopic circuits, whereas bound currents are linked to atomic and molecular magnetic moments in matter. The field related to free currents is denoted as magnetic field H (unit: ampere/meter), the field created by the total current, i.e. by both the free and the bound current, as magnetic flux density B (unit: Tesla). At every point in space, the vector quantities H, B, and Me are related by B = μ0(H + Me),

(2.2.11)

with Me ≠ 0 only inside the sample. In free space, Eq. 2.2.11 reduces to B = μ0 H. The constant μ0 = 1.257 · 10–6 Vs/Am is known as the magnetic permeability of vacuum.

For most (non-ferromagnetic) substances, the electron magnetization Me is proportional to the magnetic field strength H: Me = χH.

(2.2.12)

Fig. 2.2.24  Lenz’s law. When a circuit is approached to a bar magnet with the magnetic flux density B, a current I is induced in the circuit. This current induces a magnetic dipolar field, which is directed in such a way that it weakens the primary magnetic field. Magnitude and orientation of the dipolar field are described by the magnetic moment μ

The dimensionless proportionality constant χ is called the magnetic susceptibility. For diamagnetic substances, χ is, according to Lenz’s law, always negative and has a very small absolute value (e.g. water: χ = –0.72 · 10–6). When putting a diamagnetic sample into an originally homogeneous magnetic field, a magnetization Me is induced according to Eq. 2.2.12, which itself creates a magnetic field that counters the primary field. Therefore, the field distribution of the magnetic flux density B differs both inside and outside of the sample from the original field distribution.

primary magnetic field (Fig. 2.2.24). If a sample is positioned in an external magnetic field, a current is induced in the electron shell of the atoms and molecules, whose magnetic moment is directed against the external magnetic field, following Lenz’s law. However, in contrast to the electrons within a macroscopic circuit, the electrons

Example: Figure 2.2.25 shows the field distribution of the magnetic flux density B inside and outside of a homogeneously magnetized sphere (χ = constant), which has been brought into an originally homogeneous field B0. Inside the sphere, the magnetic flux density B is given by B = (1+2χ /3) B0. It should be noted, however, that the homogeneously magnetized sphere represents an ideal case in which the B field is homogeneous on

23

24

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

by the macroscopic field B, which has been averaged over a small but microscopically large volume surrounding the nucleus. Denoting the perturbation of the mean field B at the position of a nucleus caused by the surrounding electrons of the molecule by ΔB, we get the relation Blok = B + ΔB.

Fig. 2.2.25  Variation of the magnetic field by a diamagnetic sphere. Distribution of the B field inside and outside a homogeneously magnetized sphere, which was posi­tioned in an originally homogeneous magnetic field. It should be noted, however, that field variations caused by biological tissue are markedly weaker than illustrated here

its inside, whereas in general, the B field is also inhomogeneous inside the object.

The discussion provides three important aspects with respect to MR examinations of humans: • The distribution of the magnetic flux density B in the human body depends on the position, size, form, and magnetic susceptibility of all tissues and organs of the body. • At the interface between tissues with different magnetic susceptibilities, there are local field inhomogeneities. • The distortion of the external magnetic field caused by the body adds to the technical imperfections of the external field B0. In MRI, susceptibility-related inhomogeneities of the static magnetic field inside the body are obviously unavoidable and can result in image artifacts. In spectroscopic examinations, however, this problem can be reduced by acquiring only MR signals from small, morphologically homogeneous tissues regions. Furthermore, one has the possibility to locally adjust the B field by means of external shim coils generating a weak additional magnetic field, so the homogeneity within the examined region fulfils the demands. When speaking of the homogeneity of the B field within a given region of the body, this relates to the average macroscopic field; on the microscopic scale, the magnetic field is always inhomogeneous. 2.2.8.2  Microscopic Effect: Chemical Shift The Larmor frequency of a nucleus is determined by the local magnetic field Blok at the position of the nucleus, not

(2.2.13)

As experimental and theoretical investigations have shown, the small local field perturbation ΔB is proportional to the macroscopic field ΔB = –σB,

(2.2.14)

which yields the following expression for the resonance frequency of the nucleus, considered ω = γ Blok  = γ(1–σ)B.

(2.2.15)

The dimensionless “shielding constant” σ gives the relative resonance frequency shift that is independent of the magnitude of the magnetic field. This shift depends on the distribution of the electrons around the nucleus and thus has different values in different molecules. Remark: The magnitude of the additional field ΔB at the position of a nucleus generally depends on the orientation of the molecule relative to the macroscopic field B. In molecules that rotate rapidly, such as in fluids and soft tissues, the chemical shift anisotropy vanishes, so the quantity σ in Eq. 2.2.14 could be defined as a direction-independent constant. This quantity describes the shielding effect of the electron shell averaged over all spatial directions.

As the absolute value of the frequency shift cannot easily by measured, it is usually determined relative to the resonance frequency ωR of a reference substance. The difference (ω – ωR) of the resonance frequencies is expressed as dimensionless constant

=

−

R 0

⋅10 6 ≅ (

R

− ) ⋅10 6

(2.2.16)

relative to the frequency ω0 = γB0 of the MR system in parts per million (ppm). The chemical shift δ provides information about how the atom with the nucleus under study is bonded in the molecule and thus makes MR spectroscopy a powerful tool for the determination of the structure of molecules as well as for the investigation of biochemical processes. For the 1H nucleus, which is surrounded by only one electron, the chemical shift is about 10 ppm; for atoms with several electrons (e.g., 13C, 19F, and 31P) it can amount to several hundreds of ppm. To resolve these small differences in frequency it is necessary to use a strong and homogeneous static magnetic field (B0 ≥ 1.5 T, ∆B/B0 < 0.1–0.5 ppm depending on the nucleus).

2.2  Physical Basics Example: Figure 2.2.26 shows the 1H spectrum of ethanol (CH3–CH2–OH). Due to the different chemical surroundings of the protons in the hydroxyl, methylene, and methyl group, the spectrum shows three different resonance lines. The ratio of the areas under the resonance lines is 1 : 2 : 3 and thereby corresponds to the number of protons in the three groups.

References 1.

2.

Bottomley PA, Foster TH, Argersinger RE, Pfeifer LM (1984) A review of normal tissue NMR relaxation times and relaxation mechanisms from 1 to 100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Med Phys 11:425–448 Harris RK (1986) Nuclear magnetic resonance spectroscopy. Longman Scientific Technical, Harlow

Suggested Reading 1. 2. 3. 4. 5. 6.

Fig. 2.2.26  Proton MR spectrum of ethanol (CH3–CH2–OH). The three resonances can be assigned to the protons of the hydroxyl, methylene and methyl groups. The ratio of the area under the resonance curves is 1:2:3 and thereby corresponds to the number of protons in the three groups

Abragam A (1986) Principles of nuclear magnetism. Oxford University Press, London Becker ED (1980) High-resolution NMR. Academic, New York Harris RK (1986) Nuclear magnetic resonance spectroscopy. Longman Scientific Technical, Harlow Hauser KH, Kalbitzer KR (1991) NMR in medicine and biology. Springer, Berlin Heidelberg New York Levitt MH (2001) Spin dynamics: basics of nuclear magnetic resonance. Wiley, New York Slichter CP (2006) Principles of magnetic resonance, 3rd edn. Springer, Berlin Heidelberg New York Tokyo

25

26

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

where the proportionality constants Gx, Gy, and Gz describe the magnitude or steepness of the orthogonal gradient fields.

2.3  Image Reconstruction G. Brix If the object to be imaged, such as the human body, is divided into small cuboidal volume elements (i.e., voxels), the task in MR imaging is to distinguish the signal contribu­tions of the voxels to the detected summation signal from one another and to present them in form of sectional images (tomograms). This can be achieved by superimposing the homoge­neous magnetic field B0 by an additional magnetic field with a well-defined dependence on the spatial position, so the Larmor frequency of the MR signal becomes a function of space. 2.3.1  Magnetic Gradient Fields In practice, image reconstruction is achieved almost exclusively by means of magnetic gradient fields. These are three additional magnetic fields B x, B y, and B z, whose field vectors point toward the z-direction and whose field strengths depend linearly on the spatial position x, y, or z, respectively (Fig. 2.3.1). If the z-components of the three magnetic gradient fields are denoted by Bx, By, and Bz, the fields can be expressed as Bx = Gx x, By = Gy y , and Bz = Gz z,

(2.3.1)

Fig. 2.3.1  Gradient fields. For image reconstruction, the homogeneous magnetic field B0 shown in a is superimposed by additional magnetic fields Bx, By, and Bz, so-called gradient fields, the field vector of which points into the z-direction and the magnitude of which (length of the black arrows) depends linearly on the spatial coordinate x, y, and z, respectively. b and c show the field distributions in case the field B0 is superimposed by a gra-

Remark: The magnetic gradient fields are shortly denoted as x-, y-, or z-gradi­ents. What is meant are magnetic fields B x, B y, and B z, the magnitude of which varies linearly along the x-, y-, or zaxis, respectively (see Fig. 2.3.1).

In order to avoid image distortions, the magnitude of the gradients has to be chosen in such a way that the local field variations are markedly greater than the local inhomogeneities of the main magnetic field B0; typical values are between 1 and 50 mT/m. Technically, the gradient fields B x, B y, and B z are produced by three coil systems (gradient coils), which can be operated independ­ently from one another. Example: Assuming a patient diameter in the x-direction of X = 30 cm, a magnetic flux density of the static field of B0 = 1 T, and a gradient strength of Gx = 10 mT/m, the magnetic field B = B0 + Gx x will increase within the patient (–X/2 ≤ x ≤ + X/2) linearly from 0.9985 to 1.0015 T.

In MR imaging, the magnetic gradient fields are used in two different ways: • For selective excitation of the nuclear spins in a partial body region (e.g., a slice). • For position encoding within an excited partial body region (e.g., a slice).

dient field B y and B z, respectively. The open arrows indicate the direction of the field variation, whereas the constants G  y and G z represent the magnitude of the field variation per unit of length. If the z-components of the two magnetic gradient fields are given by B y and B z, then the gradi­ent strengths are defined by G y = ΔB y/Δy and G z = ΔB z/Δz, respectively

2.3  Image Reconstruction

2.3.2  Slice-Selective Excitation An MR signal can principally only be detected from the volume in which the nuclei have been excited before by an RF pulse. This fact is used in planar imaging methods, in order to reduce the primarily 3D reconstruction problem to a 2D one by selectively exciting only nuclei in a thin slice of the body. Remark: Depending on the type of the selectively excited partial body volume, one distinguishes between single-point, line, planar, and volume sampling strate­gies (Fig. 2.3.2). As the intensity of the detected MR signal is proportional to the number of nuclei within the excited volume, the different strategies markedly differ in the time required for the acquisition of qualitatively comparable MR images. Due to the long measurement time, single-point and line scanning techniques have not been successful in clinical practice.

In order to excite a distinct slice of the body selectively, the homogeneous static magnetic field B0 is superimposed with a gradient field (slice-selection gradient) that varies perpendicular to the slice, i.e., for an axial slice a gradient field in longitudinal direction of the body. Due to this superposition, the Larmor frequency ω of the nu-

clei varies along the direc­tion of the gradient. If we consider, for instance, a z-gradient with magnitude Gz, the Larmor frequency is given by (see Fig. 2.3.1c): ω(z) = γ(B0 + Gz z).

(2.3.2)

Consequently, an object slice z1 ≤ z ≤ z2 is characterized by a narrow frequency inter­val γ(B0 + Gz z1) ≤ ω ≤ γ(B0 + Gz z2). If one irradiates an RF pulse, the frequency spectrum of which coincides with this frequency range, only the nuclei within the chosen slice will be excited (Fig. 2.3.3). For the definition of a body slice, this has two implications: • The width d = z2 – z1 of the slice can be varied by changing either the bandwidth of the RF pulse, i.e. the width of the frequency distribution, or the gradient strength Gz. • The position of the slice can be altered by shifting the frequency spectrum of the RF pulse. The practical realization of the concept of slice-selective excitation requires not only the shape of the RF pulse but also the switching mode of the slice-selection gradient to be care­fully optimized:

Fig. 2.3.2  Imaging strategies. a Single-point, b line, c planar, and d volume scanning of a sampling object. In each case the excited partial volumes are marked in gray

Fig. 2.3.3  Principle of slice-selective excitation. The main magnetic field B0 is super­im­posed with a magnetic gradient field Bz = Gzz in z-direction (slice-selection gradi­ent), so the Larmor frequency ω(z) = γ(B0 + Gz z) of the nuclei depends linearly from the spatial coordinate z. The slice z1 ≤ z ≤ z2 within the object is thus unambiguously described by the frequency interval ω(z1) ≤ ω(z) ≤ ω(z2). If one irradiates an RF pulse with a frequency spectrum that corresponds to this frequency interval, only the nuclei within the chosen slice will be excited

27

28

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.3.3  Principle of Spatial Encoding within a Partial Volume: Projections After selective excitation of a partial body region, the MR signal from each voxel within this volume, e.g., a slice, needs to be spatially encoded. This can be achieved by two techniques: frequency and phase encoding of the MR signal. The principle of both encoding techniques will be explained in the following by the example of an axial slice parallel to the x–y-plane. For the sake of simplicity, relaxation effects are neglected in this connection. 2.3.3.1  Frequency Encoding

Fig. 2.3.4  Pulse modulation and gradient refocusing. a In order to obtain an approximately rectangular slice profile, one uses an RF pulse, the envelope of which is not rectangular but modulated in time. b If one dissects a thicker slice of the object into several thin subslices, an optimized RF pulse will deflect the magnetization of each subslice by the same angle from the z-direction, but the magnetization components are dephased after RF excitation, because the Larmor frequencies in the subslices differ. If a 90° pulse with duration tp is used for slice-selective excitation, then the dephasing effect can be compensated in good approximation by inverting the gradient field after excitation for the duration tp/2

• Pulse modulation: As shown in Fig. 2.2.10, the frequency spectrum of a rectangular RF pulse consists of several frequency bands with varying intensities. If such a pulse is used for the RF excitation, the profile of the excited slice is defined insufficiently. In order to obtain a uniform distribution of the transverse magnetization over the slice width, the shape of the selective RF pulse is modulated so that the frequency spectrum becomes as rectangular as possible (“sinc pulse,” see Fig. 2.2.23a). • Compensation gradients: If one imaginatively dissects an object slice into several thin subslices, then the magnetization components of all subslices will be deflected by the same angle from the z-direction when an optimized RF pulse is used for slice-selective excitation. However, the magnetization components will be dephased at the end of the excitation period tp, since the Larmor frequencies of the distinct subslices differ from one another as the slice-selection gradient is switched on. This effect can be compensated by reversing the polarity of the gradient field for a well-defined period after RF excitation (Fig. 2.3.4b).

If the body to be imaged is placed in a homogenous magnetic field with the flux density B0, the magnetization components of all voxels in the excited slice will precess with the same frequency around the direction of the B0 field. Thus, the frequency spectrum consist of only a single resonance line at the Larmor frequency ω0 = γ B0 —it does not contain any spatial information. However, if a magnetic gradient field, e.g., Bx = Gxx, is switched on during the acquisition phase of the MR signal (Fig. 2.3.5),

Fig. 2.3.5  Readout or frequency-encoding gradient. In frequency encoding, the main magnetic field B0 is superimposed with a gradient field (here Bx + Gx x) during the acquisition of the RF signal, so the precession frequency of the transverse magnetization in the selectively excited slice becomes a function of the coordinate x

2.3  Image Reconstruction

the Larmor frequency is related to the position x (see Fig. 2.3.1) by the resonance condition ω(x) = γ(B0 + Gx x).

(2.3.3)

Or in other words, nuclei in parallel strips oriented perpendicular to the direction of the readout (or frequencyencoding) gradient will experience a different magnetic field and thus contribute with different Larmor frequencies ω(x) to the detected MR signal of the excited slice— the spatial infor­mation is encoded in the resonance frequency. In order to determine the contribution of the distinct frequency components to the summation signal, a Fourier transformation of the measured FID has to be performed (cf. Sect. 2.2.6). The intensity I(ω) of the resulting spectrum at the frequency ω is proportional to the number of nuclei precessing with this frequency, i.e., to the number of nuclei that are accord­ing to Eq. 2.3.3 at the position x = (ω – γB0) / γ Gx. The frequency spectrum of the FID signal therefore gives the projection of the spin density distribution in the excited slice onto the direction of the readout gradient (Fig. 2.3.6). Remark: When explaining the concept of frequency encoding, we assumed that there is only one resonance line in the frequency spectrum of the excited body region in the absence of the readout gradient. If this assumption is not fulfilled, i.e., if the spectrum contains several resonance lines due to chemical shift effects described in Sect. 2.2.8.2, then these frequency shifts will be interpreted by the decoding procedure (i.e., the Fourier analysis) of the FID signal as position information. Consequently, the spin density projections of molecules with different chemical shifts are shifted in space against one another. In 1H imaging, the situation is rather easy, as only two dominant proton components contribute to the MR signal of the organism, the protons of the water molecules and those of the CH2 group of fatty acids. As the resonance frequencies of the two components differ by about 3.5 ppm (Fig. 2.3.7), fat- and water-containing structures of the body are slightly shifted against one another in readout direction. This chemical-shift artifact becomes apparent predominantly at the interfaces between fat- and water-containing tissue.

Fig. 2.3.6  Principle of frequency encoding. If the FID signal from a slice is measured in the presence of a gradient field Bx = Gx x (see Fig. 2.3.5), then nuclei in strips oriented perpen­ dicular to the direction of the gradient contribute with different Larmor frequencies ω(x) = γ(B0 = Gx x) to the measured MR signal. The contribution I(ω) of the distinct frequency components to the summation signal can be calculated by a Fourier transformation of the FID signal. As the intensity I(ω) of the resulting spectrum at the frequency ω is, on the one hand, proportional to the number of nuclei precessing with this frequency, and, on the other hand, the spatial information is encoded in the frequency, the Fourier transformation yields the projection of the spin density distribution within the considered object slice on the direction of the readout gradient

From a technical point of view, the FID signal S(t) can only be sampled and stored in discrete steps over a limited period of time tAQ. Consequently, there is only a limited number N = tAQ  /Δt of data points S(Δt), S(2Δt), S(3Δt), …, S(NΔt)

(2.3.4)

that can be used for Fourier transformation. Due to this reason, the spatial sampling inter­val Δx of the spin density projection is limited too. The following relations hold between the number N of data points, the maximum object size X, the spatial sampling interval Δx, the temporal

Fig. 2.3.7  In vivo 1H spectrum of a human thigh at 1.5 T. The two resonance lines can be attributed to protons in water and in the CH2 groups of fatty acids

29

30

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

sampling interval ∆t, and the gradient strength Gx (sampling theorem): Δx = X / N = 2π  / γGx NΔt.

(2.3.5)

Example: If the FID signal is sampled 256 times at ∆t = 30 µs in the presence of a magnetic gradient field of the strength Gx = 1.566 mT/m, then the resolution along the x-axis is ∆x = 1.953 mm, and the maximum object size that can be imaged is X = NΔx = 50 cm.

With frequency encoding, the MR signal is sampled at discrete points in time tn = nΔt (1 ≤ n ≤ N). According to Eq. 2.3.3, the transverse magnetization at the position x precesses under the influence of the readout gradient until the time tn by the angle φn (x) = (γGxx)tn.

(2.3.6)

The spatial information is therefore encoded via the frequency ω(x) in the phase angles φn(x) (1 ≤ n ≤ N). However, the same phase angles can be realized by increasing the gradient strength Gnx = nΔGx in equidistant steps ΔGx at a fixed switch-on time of the gradient. This equivalent approach is called phase encoding. 2.3.3.2  Phase Encoding The concept of phase encoding can easily be realized by applying a magnetic gradient field, e.g., Bx = Gxx, for a fixed time tx before the FID signal is detected (Fig. 2.3.8). Under the effect of this phase-encoding gradient, the magnetization at the position x precesses by the phase angle φn (x) = (γ Gx n tx)x = knx.

(2.3.7)

The parameter kn = γGnx tx = γnΔGx tx is named spatial frequency. After switching off the phase-encoding gradient, the magnetization components of the voxels in the slice precess again with the original, position-independent Larmor frequency ω0 = γB0 around the direction of the B0 field—now, however, with position-dependent phase angles φn (x). This is to say, in phase-encoding, all magnetization components of the excited voxels contribute to the detected MR signal with the same frequency ω0, but with differing phases φn (x). In order to calculate the projection of the spin density distribution in the slice onto the direction of the phase-encoding gradient, the chosen sequence is repeated N times with different spatial frequencies kn = n(γ ΔGx tx) = nΔk (1 ≤ n ≤ N) (Fig. 2.3.9). However, in contrast to frequency encoding, during phase encoding not the entire FID is sampled, but only the MR signal S(kn, t0) at a definitive time t0. After N measurements (sequence cycles), the spin density projection can be calculated by a Fourier transformation of the acquired data set

Fig. 2.3.8  Phase-encoding gradient. In phase encoding, a gradient field (here Bx = Gxx), the magnitude of which is increased in equidistant steps ∆Gx at each sequence cycle, is switched on for a fixed time tx before the FID signal is acquired.

S(Δk, t0), S(2Δk, t0), S(3Δk, t0), …, S(NΔk, t0).

(2.3.8)

Even though both encoding-techniques described are absolutely equivalent from a mathe­matical point of view (see Fig. 2.3.10), they differ considerably with regard to the time needed to acquire the data sets given in Eqs. 2.3.4 and 2.3.8 for the calculation of the spin density projection: whereas frequency encoding only requires one single sequence cycle, phase-encoding needs to repeat the sequence N times. Remark: In practice, this difference in the measurement times renders the phase-encoding technique especially vulnerable to movements (of the patient, blood flow, liquor pulsation, etc.). On the other hand, however, phase encoding does not show any chemical shift artifacts.

2.3.4   Methods of Image Reconstruction in MRI In practical MRI, techniques of image reconstruction have prevailed that merely differ only in the way the aforementioned techniques of selective excitation and spatial encoding are combined.

2.3  Image Reconstruction Fig. 2.3.9  Principle of phase encoding (displayed in the rotating frame). As shown in Fig. 2.3.8, a phase-encoding gradient Gx is switched on for a fixed time tx before the FID signal is acquired (AQ). a–d The sequence is repeated several times at equidistantly increasing gradient strengths. Under the influence of the gradient field Bx = G xx, the magnetization components of the different voxels in the slice precess with different Larmor frequencies. If the gradient is switched off after the time tx, all components rotate again with the original, position-independ­ ent frequency ω0 = γB0 around the direction of the B0 field. However, magnetiza­tion components which precess more quickly during the operating time tx of the gradient field, will maintain their advance compared with the slower ones. This advance is described by the phase angle φ(x) = γGxxtx of the different magnetiza­tion components. The figure shows the dependence of the phase angle φ(x) on the gradient strength Gx and the spatial coor­dinate x schematically for four different gradient strengths (Gx = 0, ∆Gx, 2 ∆Gx, and 3∆Gx) and three adjacent voxels (with the magnetization vectors M0, M1, and M2). As shown in b, the magnetization M1 will rotate at the position x1 under the influence of the gradi­ent field by the phase angle φ (x1) = 45° and the magnetization M2 at the posi­tion x2 by the phase angle φ (x2) = 90°

2.3.4.1  Projection Reconstruction Method This method, which P. Lauterbur used in 1973 to generate the first MR image, is based on a technique of image reconstruction used in computed tomography. Its basic idea is easy to understand: if projections of the spin density distribution of an object slice are available for various viewing angels Φn (1 ≤ n ≤ N), the spin density distribution in the slice can be recon­structed by “smearing back” the (filtered) profiles over the image plane along their viewing directions (Fig. 2.3.11). This approach can be implemented easily by making use of the frequency-encoding technique by repeating the sequence shown in Fig. 2.3.5 several times while rotating step by step the direction of the readout gradient in

the slice plane. In order to reconstruct a planar image of N × N picture elements (pixel), a minimum of N projec­ tions with N data points each is needed. The stepwise rotation of the readout gradient by the angle ΔΦ = 180°/N is performed electronically by a weighted superposition of two orthogonal gradient fields. The projection reconstruction method is easy to understand, but both mathematical description and data processing are rather complex. Furthermore, it carries the disadvantage that magnetic field inhomogeneities and patient movements result in considerable image arti­facts. Due to these reasons, the Fourier techniques described in the following sections are preferred for the reconstruction of MR images.

31

32

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.3.10  Comparison between frequency and phase encoding. With both encoding tech­niques, the transverse magnetizations of all voxels within the excited slice contribute to the detected MR signal; the spatial information is encoded in both cases by the different phase of the magnetization components, which was formed by the influence of a magnetic gradient field (here Bx = Gxx) up to the moment of signal detection. In order to calculate the projection of an object onto the direction of the gradient, the MR signal S has to be measured N times (e.g., N = 128 or 256), with the phase difference between the magnetization components of the different voxels varying in a well-defined way form measurement to measurement. The difference between the two encoding techniques merely consists in the technique with which the data set {S1, S2, . . . , SN} is acquired. a With frequency encoding, the MR signal is sampled at equidistant time steps ∆t in presence of a constant gradient field (in the figure ∆Gx). As the magnetiza­tion components of the voxels within the excited slice steadily dephase under the influence of the gradient field, there is a different phase difference between them at every point in time tn = n∆t (1 ≤ n ≤ N), so the whole data set {S1, S2, . . . , SN} can be detected by a single application of the sequence. The figure shows the first three values of the signal. It should be noted that only the temporal change of the MR signal caused by the gradient field is shown here, whereas the rapid oscillation of the signal with the Larmor frequency

ω0 = γB0 as well as the T2* decay of the signal is neglected. b With phase encoding, a phase-encoding gradient is switched on for a fixed duration tx before the FID signal is acquired. The magnitude of this gradi­ent is increased at each sequence repetition by ∆Gx. During the switch-on period of the gradient field, the magneti­zation components of the different voxels precess with different frequencies so that a phase difference is established between them which is proportional to the magni­tude of the gradient applied (see Fig. 2.3.9). After switching off the gradient, all components rotate again with the original, position-independent frequency ω0 = γB0 around the direction of the B0 field. In the chosen description, one therefore observes an MR signal Sn that is constant over time. In order to acquire the entire data set {S1, S2, …, SN}, the sequence needs to be repeated N times with different gradient strengths Gx = n∆Gx. The figure shows the dependence of the MR signal from the gradient strength schematically for three different gradient strengths (Gx = 0, ∆Gx, 2∆Gx). If the product of the gradi­ent strength and the switch-on time of the gradient up to the time of signal detec­ tion is equal in both encoding tech­niques, the phase difference between the various magnetization components at the time of signal detection is also identical and thus the same MR signal is measured. The product of the two quantities is indicated in the figure by the dark areas

2.3  Image Reconstruction Fig. 2.3.11  Reconstruction by back projection. The figure shows three different projections of two objects in the field of view. If many projections are acquired at different viewing angles, an image can be reconstructed by (filtered) back projection of the profiles. For the measurement of the various projections, the frequency-encoding technique is used, with the readout gradient rotating step by step

2.3.4.2  2D Fourier Method In the planar version of Fourier imaging, just as in projection reconstruction, the spins in a slice are selectively excited by an RF pulse in the first step. Afterwards, however, spatial encoding of the spins in the slice is not done by a successive rotation of a readout gradient, but by a combination of frequency and phase encoding using two orthogonal gradient fields. If we consider an axial slice parallel to the x–y-plane, then these gradients are Gx and Gy (Fig. 2.3.12). The sequence is repeated N times for different values of the phase-encoding gradient Gxn = nΔGx (1 ≤ n ≤ N), with the MR signal being measured M times during each sequence cycle at the times tm = mΔt (1 ≤ m ≤ M) in the presence of the readout gradient Gy. Thus, one obtains a measurement value for each combination (kn, tm) of the parameters kn = γnΔGx tx and tm = mΔt, i.e., a matrix of N × M data points. A 2D Fourier trans­ formation of this data set, the so-called hologram or kspace matrix (see Fig. 2.4.26), yields the MR image of the slice with a resolution of N × M pixels. 2.3.4.3  3D Fourier Method In order to extend the 2D Fourier method to a 3D one, the slice-selection gradient is replaced by a second phaseencoding gradient as shown in Fig. 2.3.13. This means that the RF pulse excites all spins in the sensitive volume of the RF coil and that the spatial information is encoded exclusively by orthogonal gradi­ents—by two phase-encoding gradients and one frequency-encoding gradient. The spatial resolution in the third dimension is defined by the strength of the related phase-encoding gradient and the number K of the phase-encoding steps. Depending

Fig. 2.3.12  Typical pulse and gradient sequence in 2D Fourier imaging. Gz is the slice-selec­tion gradient, Gx the phase-encoding gradient, and Gy the readout gradient

33

34

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

matted, which offers—among others—the possibility to look at an organ or a body structure from various viewing directions. 2.3.4.4  Alternative k-Space Sampling In addition to the described conventional fashion of filling the k-space in Fourier imaging, there are a number of alternative strategies. • Spiral acquisition. As will be described later, the EPI sequence is commonly using an oscillating frequency– encoding gradient. If both, the phase encoding as well as the frequency-encoding gradient are oscillating with increasing gradient amplitudes, the acquired data points will be along a spiral trajectory through k-space. That is why such an acquisition is called spiral EPI. • Radial acquisition. If the direction of the frequencyencoding gradient is rotated as described in Sect. 2.3.4.1, the k-space trajectories will present a star. • Blade, propeller, multivane. These hybrid techniques sample k-space data in blocks (so-called blades) each of which consists of some parallel k-space lines. In order to successively cover the entire k-space, the direction of the blades is rotated with a fixed radial increment. This sampling strategy offers some advantages. Since each blade contains data points close to the center of the k-space, patient movements can, for example, be easily detected and corrected.

Fig. 2.3.13  Typical pulse and gradient sequence in 3D Fourier imaging. Gx and Gy are the phase-encoding gradients and Gz the readout gradient

on the choice of these parameters, the voxels have a cubic or cuboidal shape (isotropic or anisotropic resolution). In order to acquire a 3D k-space matrix with N × M × K independent measurement values, the imaging sequence needs to be repeated N × K times. Example: Using a standard sequence, the repetition time TR is typically of the order of the spin-lattice relaxation time T1, to allow the longitudinal magnetization to relax at least partially before the next RF pulse is irradiated. For instance, if one assumes TR = 500 ms and an image resolution of 256 × 256 × 64 voxels, one obtains a measurement time of 136 min. 3D imaging is thus only feasible with fast imaging sequences (see Sect. 2.4.3).

A 3D Fourier transformation of the acquired 3D k-space matrix yields the 3D image data set of the partial body region excited by the RF pulse. Based on this image data set, multiplanar images in any orientation can be refor-

To reconstruct MR images from alternative k-space trajectories by means of a conventional 2D or 3D Fourier transformation, it is necessary to re-grid the sampled kspace data to a rectangular grid. 2.3.5  Multiple-Slice Technique Data acquisition by 2D imaging techniques can be carried out very efficiently when considering the fact that the time required for slice-selective excitation, spatial encoding, and acquisition of the MR signal is much shorter than the time needed by the spin system to relax at least partially after RF excitation, before it can be excited once again. The long waiting periods can be used to excite—in a temporally shifted manner—adjacent slices and to detect the spatially encoded MR signal from these slices. Thus, MR images from different parallel slices can be acquired simultaneously without prolongation of the total acquisition time (Fig. 2.3.14). Example: Let us consider that 50 ms are needed for excitation, spatial encoding, and data acquisition per sequence cycle and that the sequence is repeated after TR = 1,000 ms. Then MR data from 20 adjacent slices can be acquired simultaneously without prolonging the measurement time.

2.3  Image Reconstruction Fig. 2.3.14  Principle of the multiple-slice technique. In most 2D imaging sequences, the time T required for slice-selective excitation, spatial encoding, and detection of the MR signal is markedly shorter than the repetition time TR. The long waiting periods can be used to subsequently excite spins in parallel slices and to detect the spatially encoded signals from these slices

Fig. 2.3.15  Consideration of the slice profile by the multiple– slice technique. The profile of a slice is generally not rectangular but rather bell shaped. The thickness (TH) of the slice is therefore usually defined by the full-width at half-maximum. In order to prevent overlapping of adjacent slices in the multiple-slice technique, a sufficient gap (G) between the two adjacent slices has to be chosen (G ≥ TH). Often the distance (D = G + TH) between the slices is indicated instead of the gap G. Images from adjacent slices can be detected without overlap by using in a first step a sequence that acquires data from the even slices and in a second step a sequence that acquires data from the odd slices. In both measurements, the gap G should be identical with the slice thickness TH (D = G + TH = 2TH)

However, when using the multiple-slice technique, one has to consider that the distance between slices may not be too small, as the slice profile usually is not rectangular, but bell shaped. In order to avoid repeated excitation of spins in overlapping slice regions, the gap between adjacent slices should correspond approximately to the width of the slice itself. Images from adjacent slices can be obtained in an interleaved manner by applying the sequence twice: in the first measurement, data are acquired from the odd slices and in the second from the even slices (Fig. 2.3.15).

Suggested Reading 1. 2.

3. 4.

Barrett HH, Myers K (2004) Foundations of image science. Wiley, New Jersey Haacke EM, Brown RW, Thompson MR, Venkatesan R (1999) Magnetic resonance imaging: physical principles and sequence design. Wiley, New York Oppelt A (2005) (ed) Imaging systems for medical diagnostics. Publicis MCD, Erlangen Vlaardingerbroek MT, den Boer JA (2004) Magnetic resonance imaging: theory and practice. Springer, Berlin Heidelberg New York

35

36

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4  Image Contrasts and Imaging Sequences G. Brix, H. Kolem, and W.R. Nitz 2.4.1  Image Contrasts 2.4.1.1  Contrast Determinants and Optimization in MRI The main advantage of MR imaging, apart from the flexibility in slice orientation, is the excellent soft-tissue contrast in the reconstructed MR images. It is based on the different relaxation times T1 and T2 of the tissues, which depend on the complex interaction between the hydrogen nuclei and their surroundings. Compared to that, differences in proton densities (PD) are only of minor relevance, at least when considering soft tissues. Remark: The term proton density in MR imaging designates only those hydro­gen nuclei whose magnetization contributes to the detectable image signal. Essentially, this refers to hydrogen nuclei in the ubiquitous water molecules and in the methylene groups of the mobile fatty acids (see Sect. 2.3.3.1 and Fig. 2.3.7). Hydro­gen atoms, which are included in cellular membranes, proteins, or other relatively immobile macromolecular structures, usually do not contribute to the MR signal; their FID signal has already decayed to zero at the time of data acquisition (T2 << TE) (Brix 1990).

Another important contrast factor is the collective flow of the nuclei. The influence of flowing blood on the image signal will be discussed in Sect. 2.7 separately, in the context of MR angiography. Whereas the image contrast of a CT scan only depends on the electron density of the tissues considered (as well as on the tube voltage and beam filtering), the MR signal and thus the character of an MR image is determined by the intrinsic tissue parameter PD, T1, and T2 as well as by the type of the sequence used and by the selected acquisi­tion parameters. This variability offers the opportunity to enhance the image contrast between distinct tissues by cleverly selecting the type of sequence and the corresponding acquisition parame­ters, and thus to optimize the differentiation between these tissue structures. However, the subtle interplay of the many parameters bears the danger of mis­interpretations. In order to prevent these, several MR images are always acquired in clinical routine, with different sequence parameters that are selected in such a way that the tissue contrast of the various images is determined mainly by a single tissue parameter; in this context, one uses the term T1-, T2-, or PD-weighted images. Sometimes, one even goes one step further to calculate “pure” T1, T2, and PD parameter maps on the basis of several MR images that were acquired with different acquisition parameters. The advantage in doing this consists of the fact that the image contrast on the calculated

parameter maps is usually more accentuated than in the weighted images. The calculated tissue parameters can furthermore be used to characterize various normal and pathological tissues. However, experience has shown that a characterization or typing of tissues by means of calculated MR tissue parameters is only possible with reservations (Bottomley et al. 1987; Higer and Bielke 1986; Pfannenstiel et al. 1987). This may be due not only to the insufficient measurement and analysis techniques used, but also to the fact that morphological information of the MR images as well as the clinical expertise of the radiologist have been left aside in many cases. These considerations indicate that each MR practitioners should be aware of the depend­ence of the image contrast on the selected type of imaging sequence as well as on the sequence and tissue parameters in order to fully benefit from the potential of MRI and to avoid misinterpretations. Remark: The term imaging sequence designates the temporal sequence of RF pulses and magnetic gradient fields, which are used to determine the image contrast and for image reconstruction, respectively.

2.4.1.2  Definition of Image Contrast The foregone section has made intuitive use of the term image contrast in order to describe the possibility to distinguish between adjacent tissue structures in an MR image. We will now define this term. If one describes the signal intensities of two adjacent tissues structures A and B with SA and SB, the image contrast between the two tissues can be expressed by the absolute value of the signal difference CAB = |SA – SB|

(2.4.1)

or by the normalized difference

(2.4.2)

Remark: The delineation of a tissue structure depends, of course, also on the signal-to-noise ratio (S/N) as tiny, weakly contrasted structures can be masked by image noise. Some authors therefore proposed to use the contrast-to-noise ratio for evaluating the detectability of a detail. However, the explanatory power of this quantity can hardly be objecti­fied since the contrast–detail detectability strongly depends on the signal detection in the human retina as well as on the signal processing in the central visual system of the observer. In the following, we will therefore use the absolute contrast defined in Eq. 2.4.1. Example: In order to analyze the influence of the tissue and acquisition parame­ters on the image contrast by an example, we will consider in the following the contrast between white and

2.4  Image Contrasts and Imaging Sequences Table 2.4.1  Tissue parameters for gray and white brain matter at 1.5 T Tissue

T1 (ms)

T2 (ms)

ρ (a.u.)

Gray brain matter

972 ± 124

109 ± 17

100 ± 7

White brain matter

599 ± 90

 87 ± 7

 96 ± 5

a.u. arbitrary units

gray brain matter. Representative tissues parameters, which have been measured for a patient collective at 1.5 T, are summarized Table 2.4.1.

2.4.2  Classical Imaging Sequences 2.4.2.1  Spin-Echo Sequence In clinical routine, the spin-echo (SE) sequence is still a frequently applied imaging sequence, due to two reasons: • It is rather insensitive to static field inhomogeneities and other inaccuracies of the MR system. • It allows for the acquisition of T1-, T2-, and PDweighted images by an appropriate choice of the acquisition parameters TR and TE. The basic principle of the SE technique has already been described in Sect. 2.2.7.3. For image reconstruction, in general the 2D Fourier method is used. One of the possible combina­tions of the SE pulse sequence {90° – TE / 2 – 180° – TE / 2 – AQ – TD} (TE, echo time; TD, delay time; AQ, acquisition period), and the gradient sequence is shown in Fig. 2.4.1. The signal intensity SSE of a voxel with the tissue parameters PD, T1, and T2 can be described in good approximation by the equation (TR = TE + TD, repetition time)

(2.4.3)

According to this equation, the signal intensity of a voxel is determined by three independent factors, each of which depends on only one of the three tissue parameters. This leads to a clear dependence of the image signal on the tissue and acquisition parameters: • PD dependence: The signal intensity is directly proportional to the PD of the considered tissue, i.e., to the number of excited nuclei which contribute per unit volume to the signal of the MR image. The influence of the proton density on the image contrast is constant and cannot be varied. • T1 dependence: after RF excitation, the longitudinal magnetization Mz relaxes with the time constant T1 against the equilibrium magnetization M0 (see Fig. 2.2.22); the equilibrium state is reached after a period

Fig. 2.4.1  Pulse and gradient scheme of the SE sequence. Gz sliceselection gradient, Gx phase-encoding gradient, Gy readout gradient. (The effect of the first gradient pulse in readout direction is explained in Fig. 2.4.16)

of about 3T1. Usually, however, the sequence is repeated much earlier, so that the longitudinal magnetiza­tion will be reduced at the beginning of the next sequence cycle compared to the equilibrium magnetization by the T1 factor [1 – exp(–TR / T1)]. Accordingly, the T1 contrast of an SE image can be varied by the choice of the repetition time TR. In Fig. 2.4.2a, the T1 factor is plotted for white and gray brain matter. As this example shows, the T1 contrast reaches a maximum if the repetition time TR is between the T1 relaxation times of the two tissues considered. If TR is markedly longer than the longer T1 time, then the T1 contrast vanishes. • T2 dependence: The influence of the T2 relaxation process on the signal intensity is described by the T2 factor exp (–TE / T2) in the signal equation. For a given T2 time the signal loss is the bigger, the longer the echo time TE becomes. In Fig. 2.4.2b, the T2 factor is plotted for white and gray brain matter versus the echo time TE. The contrast will reach a maximum when the echo time TE ranges between the T2 relaxation times of the two tissues considered. For small TE values (TE << T2), the contrast approximates zero, as the signal intensities in this case are inde­pendent of T2.

37

38

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.2  T1 and T2 dependence of the SE signal intensity. a Plotted is the T1 factor [1 – exp(–TR / T1)] of the SE signal equation as a function of the repetition time TR for the white (WM) and gray (GM) brain matter. As can be seen, the T1 contrast bet­ ween the two tissues approaches 0 for very long as well as for very short repetition times. The highest T1 contrast is obtained for TR ~ 750 ms, i.e. for a repetition time in between the T1 times of the two tissues considered (see Table 2.4.1). b The same considerations hold for the T2 factor exp(–TE /T2). The T2 contrast maximum is at TE ~ 100 ms

Fig. 2.4.3  Influence of the acquisition parameters TR and TE on the contrast behavior of an SE image. The figure shows the interplay of the longitudinal and transversal relaxation for this sequence at the example of white (WM) and gray brain matter (GM) for a fixed repetition time of TR = 1.000 ms. In the left part, the temporal evolution of the longitudinal magnetization Mz during the recovery period (0 ≤ t ≤ TR) is depicted. At t = TR, the partially relaxed longitudinal magnetiza­tion is flipped into

the x–y-plane by the 90° excitation pulse. The T2 relaxation of the resulting transversal magnetization Mxy is plotted in the right part as a function of the echo time TE. As can be seen, there is a reversal behavior of the T1 and T2 contrast. For TE = 83 ms, the contrast is 0, so that the two types of brain matter cannot be differentiated in the relating SE image in spite of differing tissue parameters (see Fig. 2.4.5d). Note that the detected MR signal is directly propor­tional to the transversal magnetization Mxy

2.4  Image Contrasts and Imaging Sequences

In general, adjacent tissues differ in all three tissue para­ meters PD, T1 and T2, so the different factors of the SE signal equation, which can partially compensate one another, need to be considered altogether. This holds even more as the relaxation times are usually positively correlated, i.e., the tissues with longer T1 times usually also have longer T2 times. In order to illustrate this statement, Fig. 2.4.3 shows the course of the longitudinal and transverse magnetization for both white and gray brain matter for a repetition time of TR = 1,000 ms. As can be seen, the transverse magnetization of both substances—and therefore the signal inten­sities (SSE ∝ Mxy)—are identical at an echo time of TE = 83 ms, so the tissues cannot be distinguished on the related SE image in spite of different tissue parameters. Fig. 2.4.4 shows the contrast between white and gray matter as a function of both the repetition and echo time. As expected, there are two regions with a high tissue contrast, which are separated by a low contrast region (cf. Fig. 2.4.3): For TE << T2 and TR ~ T1, the T1 contrast is dominant (T1-weighted image), for TR >> T1 and TE ~ T2, the T2 contrast (T2-weighted image). For TR >> T1 and TE << T2, the contrast vanishes, as the signal intensities in this case are nearly inde­pendent from the relaxation times (PD-weighted image). Example: For TE = 15 ms, the contrast maximum between white and gray brain matter is at TR ~ 640 ms, for TR = 2,400 ms at TE ~110 ms. In order to illustrate the dependence of the image contrast on the acquisition parameters TE and TR, Fig. 2.4.5 shows four SE images of a transversal head section: a T1-weighted image (TR /TE = 600/15), a T2-weighted image (2,400/120), a PDweighted image (2,400/22 ms), and a mixed T1 /T2-weighted image with disappearing contrast between white and gray matter (1,000/85 ms).

Table 2.4.2  Influence of the sequence parameters TR and TE on the contrast of an SE image Weighting

Condition

PD

TR >> T1

TE << T2

T1

TR ≈ T1

TE << T2

T2

TR >> T1

TE ≈ T2

Remark: PD-weighted and T2-weighted SE images can be acquired with one single measurement, if a double echo sequence (e.g., TR/TE1/TE2 = 2,400/22/120) is used (cf. Fig. 2.2.23).

The influence of the acquisition parameters TE and TR on the contrast of an SE image is summarized in Table 2.4.2. 2.4.2.2  Inversion Recovery Sequence In order to adapt the inversion recovery technique explained in Sect. 2.2.7.2 for MR imaging, one has to extend the conventional IR pulse sequence {180° – TI – 90° – AQ – TD} by a 180° pulse {180° – TI – 90° – TE /2 – 180° – TE /2 – AQ – TD} (TI , inversion time; TE, echo time; TD, delay time; AQ , signal acquisition). In this imaging technique, the FID is not acquired directly after the readout pulse, but rather the SE signal, which is created by the additional rephasing 180° pulse. This modification is necessary to integrate the gradient fields required for image reconstruc­tion into the sequence (Fig. 2.4.6).

Fig. 2.4.4  Parameter weighting in SE imaging. The figure shows the image contrast between white and gray brain matter in dependence of the repetition time TR and the echo time TE. There are two regions with a high tissue contrast that are separated by a low contrast region (Fig. 2.4.3): For TE << T2 and TR ~ T1, the T1 contrast is dominant (T1-weighted image), for TR >> T1 and TE ~ T2 the T2 contrast. For TR >> T1 and TE << T2, the contrast is only determined by the differences in the proton densities (PD-weighted image)

39

40

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.5  SE images of a transversal head section with a thickness of 5 mm. a T1-weighted image (TR/TE/NAQ = 600/15/4), b T2-weighted image (2,400/120/1), c PD-weighted image

(2,400/22/1), and d mixed T1/T2-weighted image (1,000/85/1) with disappearing contrast between white and gray matter (cf. Fig. 2.4.3)

The signal intensity SIR of a tissue with the intrinsic parameters PD, T1, and T2 can be described approximately with the expression

pendence, the IR sequence is especially apt for acquiring T1-weighted images: • PD dependence: see Sect. 2.4.2.1 • T1 dependence: The longitudinal magnetization Mz(t) at the time t = 0 is inverted by the 180° pulse of the IR sequence (the inversion pulse), so the longitudinal magnetization will relax strictly monotonously from negative values toward the equilibrium magnetization M0 (inversion recovery). This is described by the T1 factor of the IR signal Eq. 2.4.4, which depends on the

S IR = PD ⋅ [ 1 2 e

TI / T 1 TE / T 2 + e TR /T 1 ] ⋅ e12 3 14442444 3 T factor T1 factor

2

(2.4.4)

The image contrast of an IR image can therefore be varied via three acquisition parameters, at principally: via the inversion time TI, the echo time TE, and the repetition time TR = TI + TE + TD. Due to the pronounced T1 de-

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.6  Pulse and gradient scheme of the IR sequence. Gz sliceselection gradient, Gx phase-encoding gradient, Gy readout gradient. (The effect of the first gradient pulse in readout direction is explained in Fig. 2.4.16)

inversion time TI as well as a on the repetition time TR. In order to optimize the T1 contrast, the inver­sion time TI is usually varied, whereas with the parameter TD, respectively TR is chosen as high as possible (TD >> T1), to allow the recovery of a considerable longitudinal magnetization after RF excitation. The maximum range of values of the T1 factor is between –1 and +1, thus being double the range of values of the SE sequence. There are two types of IR sequences depending on signal interpretation: If only the absolute values of the signal are considered (magnitude recon­struction, IRM), the range of values is limited de facto to the inter­val between 0 and 1, as in the SE sequence. If this mode of data representation is chosen, then the T1 factor will initially decrease to 0 and then converges toward the equilibrium magnetiza­tion M0. Figure 2.4.7 shows the dependence of the T1 factor from the inversion time TI for both possible modes of data representation for white and gray brain matter (TR = 3,000 ms). As this example reveals, the neglect of the sign of the T1 factors in the absolute value representation leads to a destructive T1 contrast behavior in the region between the zeros of both T1 functions

Fig. 2.4.7  T1 dependence on the IR signal intensity. Plotted is the T1 factor [1 – 2 exp(– TI/T1) + exp(– TR / T1)] of the IR signal equation for white (WM) and gray (GM) brain matter as a function of the inversion time TI for a fixed repetition time of TR = 3,000 ms for two different modes of data representation: a by considering and b by neglecting the sign of the longitudinal magnetization Mz. If the last-mentioned mode of data representation is used, there will be a destructive T1 contrast behavior in the region between the zeros of the two tissue curves. The T1 contrast maximum in the considered case is at TI ~ 760 ms, i.e., in between the T1 times of the two tissues considered (see Table 2.4.1)

consid­ered. An IR sequence differentiating between parallel or antiparallel alignment of the longitudinal magnetization at the time of the excitation pulse is called phase sensitive. For evaluation of the tissue contrast, the PD and T2 dependence of the image signal SIR needs to be included in the considerations, too. Figure 2.4.8 demonstrates this with an example. In this figure, the tissue contrast between white and gray brain matter is plotted as a func­tion of the echo time TE for TI = 800 ms and TR = 2,400 ms. For the chosen TI value, there is a reversal behavior of the T1 and T2 contrast, as the relaxation times of the two tissues

41

42

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.8  Influence of the acquisition parameters TI and TE on the contrast behavior of the IR sequence (absolute value representation). The figure shows the interplay of the longitudinal and transversal relaxation for this sequence at the example of white (WM) and gray brain matter (GM) for a fixed inversion time TI = 800 ms and a fixed repetition time TR = 2,400 ms. The left part shows the temporal evolution of the longitudinal magnetization Mz during the inversion phase (0 ≤ t ≤ TI). At t = TI,

the partially relaxed longitudinal magnetization is flipped into the x–y-plane by the 90° excitation pulse. The T2 relaxation of the resulting transversal magnetization Mxy is plotted in the right part as a function of the echo time TE. In the case considered, there is a reversal behavior of the T1 and T2 contrast, so that the contrast between the two brain tissues rapidly reduces with prolonged echo time. Note that the detected MR signal is directly proportional to the transversal magnetization Mxy

exam­ined are positively correlated, i.e., the substance with the longer T1 time also has a longer T2 time. In order to fully grasp the complex interplay between the different tissue and acquisition parameters as a whole, the image contrast between white and gray brain matter is plotted in Fig. 2.4.9 as a function of both the inversion time TI and the echo time TE.

tissue to be suppressed is approximately 0 (Fig. 2.4.7). For TR > 3TI, the corresponding TI value can be estimated by TI = T1 · ln 2.

Example: In the case considered, the T1 contrast assumes a maximum for TE = 20 ms at TI = 620 ms. Figure 2.4.10a shows the corresponding T1-weighted IR image (TR /TI /TE = 2,400/600/20). Compare this image with the T1-weighted SE image shown in Fig. 2.4.5. The two images differ mainly in that the gray matter is not represented in the IR image as the T1 factor of gray matter is approximately 0 for TI = 600 ms (see Fig. 2.4.7).

The influence of the acquisition parameters TE, TI, and TD on the contrast of an IR image is summarized in Table 2.4.3. In order to maximize the T1 contrast (T1-weighted image), the TI time should be between the T1 times of the two tissues considered, and the echo time TE should be chosen as short as possible. As even for the acquisition of T1-weighted images, relatively long repeti­tion times are needed (TD >> T1), the IR sequence requires much more time than the SE sequence. Advantages occur mainly when the image signal of a given tissue structure shall be suppressed, e.g., the retrobulbar fatty tissue for the evaluation of the optic nerve. In this case, the acquisition parameter TI needs to be chosen so that the T1 factor of the

Example: At a magnetic flux density of 1.5 T, the T1 time of fatty tissue is about 260 ms (see Table 2.2.2), so fatty tissue structures are not displayed in the IR image at a TI value of 180 ms (Fig. 2.4.10b). At lower flux densities, the corresponding TI value is somewhat lower. Remark: An IR sequence with a very short TI time is called STIR (short-tau inversion recovery) sequence. If the TI time is selected to suppress the signal from liquor, the sequence is called FLAIR (fluid-attenuated inversion recovery).

Table 2.4.3  Influence of the sequence parameters TE, TI, and TD on the contrast of an IR image (TR = TI + TE + TD) Weighting

Condition

PD

TI >> T1

TD arbitrary

TE << T2

TI << T1

TD >> T1

TE << T2

T1

TI ≈ T1

TD >> T1

TE << T2

T2

TI >> T1

TD arbitrary

TE ≈ T2

TI << T1

TD >> T1

TE ≈ T2

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.9  Parameter weighting in IR imaging (absolute value representation). Plotted is the image contrast between white and gray brain matter in dependence of the inversion time TI and the echo time TE for a fixed repetition time of TR = 2,400 ms. For

TE << T2 and TI ~ T1, the T1 contrast is dominant (T1-weighted image), for TI >> T1 and TE ~ T2, the T2 contrast (T2-weighted image). Note the breakdown of the image contrast at short TI times (see Fig. 2.4.7b)

Fig. 2.4.10  IR images of a transversal head section with a thickness of 5 mm. a T1-weighted image (TR /TI /TE / NAQ = 2,400/600/20/1). Gray brain matter is depicted with very little signal in this image, as the corresponding T1 factor for TI = 600 ms is about 0 (see Fig. 2.4.7). b STIR image

(1,500/180/20/1). In this case, the inversion time TI has been chosen in such a way that the T1 factor of fatty tissues is about 0. Consequently, fatty tissue in the retrobulbar and temple regions is not depicted

43

44

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4.2.3  Limiting Factor: Acquisition Time

• Dynamic imaging studies are limited.

The numerous advantages of conventional imaging techniques are opposed by the long acquisition time as a limiting factor. The time T required for the acquisition of an MR image using the 2D Fourier method (see Sect. 2.3.4.2.) is given by the product of the repetition time TR, the number of the sequence cycles needed for phase encoding NPH, and the number of acquisitions NAQ (or averages):

To overcome these limitations, several methods aiming to shorten the acquisition examination times have been developed. They can be categorized into two groups, depending on whether the repetition time TR or the number of sequence cycles NPH needed for phase encoding is reduced (see Eq. 2.4.5). The two strategies will be discussed in the following sections at the example of some selected imaging sequences. An almost complete overview of the clinically used fast imaging sequences will be provided in Sect. 2.4.6.

T = TR NPH NAQ.

(2.4.5)

• The number NPH of sequence cycles needed for phase encoding is determined by the desired resolution of the image matrix. For the reconstruction of an MR image with a resolution of 256 × 256 pixels, for instance, 256 sequence cycles are required. • The repetition time TR is determined by the selected type of weighting. Typical values are TR = 500–600 ms for T1-weighted and TR ≥ 3,000 ms for T2-weighted or PD-weighted SE images. • In some cases, the imaging sequence is repeated several times (e.g., NAQ = 2 or 4), in order to improve the signal-to-noise ratio (S /N ∝ N AQ ). This is especially valid for T1-weighted SE images, which have a relatively low S/N ratio due to the short repetition time. Example: Based on these considerations, the following representative acquisition times are obtained for SE images: T = 2.1 min for a T1-weighted image (TR = 500 ms, NPH = 256, NAQ = 1) and T = 10.2 min for a T2-weighted and/or a PD-weighted image (TR = 2,400 ms, NPH = 256, NAQ = 1).

By using the multiple slice technique described in Sect. 2.3.5, one can simultaneously acquire MR images from multiple parallel slices within the given acquisition times, but the overall acqui­sition time required for the acquisition of the images will not be reduced. In clinical prac­tice, this basic limitation of conventional imaging sequences leads to the follow­ing problems: • Depending on the clinical question, the time needed for a patient examination ranges between 15 and 45 min. • This demands high cooperation from the patient, as the patient will be asked to remain motionless during the examination in order to assure the comparability of differ­ently weighted MR images. • Critically ill patients may not be examined full scale or might not fit for examina­tion at all. • The image quality is impaired by motion artifacts (such as heart beat, blood flow, breathing, or peristaltic movement). This problem is especially acute in patients with thorax and abdominal diseases, as MR images in general cannot be acquired completely during breath hold, as is the case in CT.

2.4.3  Gradient-Echo Techniques 2.4.3.1  Low–Flip Angle Excitation and Gradient Echoes The long scan times of conventional imaging sequences are due to the fact that the 90° excitation pulse rotates the entire longitudinal magnetization into the x–y-plane, so the pulse sequence can only be repeated when the longitudinal magnetization has been—at least partially—recovered by T1 relaxation processes. To acquire MR images with an acceptable S/N, the sequence repetition time TR has to be of the order of the T1 relaxation time. This basic problem in conventional imaging can be prevented, however, by using an RF pulse with a flip angle of α < 90° to excite the spin system, so that only a part of the longitudi­nal magnetization Mz will be rotated to the x–y-plane, Nevertheless, one obtains a relatively large transverse magnetization. Example: If, for instance, a flip angle of α = 20° is used, then the longitudinal magnetiza­tion Mz will be reduced by 6%, whereas the transverse magnetization Mxy amounts to 34% of the maximum value (Fig. 2.4.11).

In order to discuss the principle of low–flip angle excitation, we will initially neglect the gradient fields needed for spatial encoding and consider the simple sequence shown in Fig. 2.4.12a. It consists of a single RF pulse with a flip angle α < 90° and a spoiler gradient, which destroys the remaining transverse magnetization after the acquisition of the FID. Remark: As an alternative to spoiler gradients, the phase of the RF excitation pulse may be varied with every sequence cycle, in order to prevent the buildup of a steady state for the transverse magnetization (RF spoiling).

If the considered sequence is repeated several times, then the spin system already reaches a dynamic equilibrium after a few sequence cycles. Figure 2.4.13 shows the transient behavior of the longitudinal magnetization in white

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.11  Principle of low–flip angle excitation. In contrast to a conventional 90° excitation, the magnetization M in low–flip angle excitation is only rotated by a flip angle α < 90°, so that the longitudinal magnetization Mz is only slightly reduced by the pulse. Nevertheless, this results in a relatively high transversal magnetization Mxy

7 Fig. 2.4.12  Excitation scheme of the FLASH sequence. a This sequence consists of a single RF pulse with a flip angle α < 90° and a spoiler gradient that destroys the transversal magnetization after data acquisition. b Temporal evolution of the longitudinal magnetization Mz in the steady state. The steady-state magnetization is denoted by M zSS. c Temporal evolution of the transversal magnetization Mxy in the rotating frame

Fig. 2.4.13  Transient behavior of the longitudinal magnetization when using a FLASH sequence (TR = 25ms, T1 = 600 ms). If the sequence shown in Fig. 2.4.12 is repeated several times, then

the longitudinal magnetization reaches a steady-state value M zSS after several repetitions. For α = 90°, the steady state is reached already after the first excitation

45

46

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

brain matter (T1 ≅ 600 ms) for different flip angles α. The equilibrium value of the longitudinal magnetization, the so-called steady-state magneti­zation M zSS, is given by the equation

(2.4.6)

The value of the steady-state longitudinal magnetization depends not only on the flip angle α of the excitation pulse, but also on the repetition time TR and the longitudinal relaxation time T1. It will be smaller when α becomes bigger. For α = 90°, the longitudinal magnetization reaches the steady-state value M zSS = M0 [1–e–TR/T1] after the first excitation, as expected. However, the MR signal is not given by the longitudinal magnetization but by the transverse magnetization Mxy at the time of data acquisition. By using Eq. 2.4.6 the amplitude S of the MR signal can be described by S ∝ Mxy = M zSS  sin α · e–T E / T 2*.

(2.4.7)

Whereas the factor exp (–TE/T2*) describes the decay of the FID signal during the delay time TE, the factor sin α gives the fraction of the steady state magnetization M zSS, which is rotated in the x–y-plane (see Fig. 2.4.12b,c). To illustrate this relation, Fig. 2.4.14 shows the signal intensity S as a function of the ratio TR/T1. From this plot, two important statements can be derived: • As compared with conventional 90° excitation, low– flip angle excitation yields consid­erably higher signal values for short repetition times. • When using low–flip angle excitation, the signal is already independent of T1 for TR < T1. The signal increase realized by low–flip angle excitation in combination with short repetition times is obtained,

however, by omitting the 180° pulse generating a spinecho, as the 180° pulse not only inverts the phase of the transverse magnetization, but also the longitudinal magnetization (see Fig. 2.2.22b). This means that the dephasing of the spins due to static field inhomogeneities as well as to chemical shift effects cannot be compensated, when the concept of low–flip angle excitation is applied (see Sects. 2.4.3.4 and 2.4.3.5). In the following, several advantages and disadvantages of low flip angle excitation are summarized. • Advantages: – Signal increase at short TR times – Shorter TE times, as the 180° pulse is omitted – Reduced energy deposition in the body by choosing an appropriate sequence • Disadvantage: – Image artifacts due to local magnetic field inhomogeneities and chemical shift effects In order to benefit from the advantages of low flip angel excitation for fast MR imaging, the sequence shown in Fig. 2.4.12a needs to be completed with the gradient fields required for slice selective excitation and spatial encoding. The pulse and gradient scheme of a typical 2D fast imaging sequence is shown schematically in Fig. 2.4.15. As is the case in the conventional imaging sequences, there is an additional rephasing gradient in slice-selection direction and a pre-dephasing gradient in readout direction. With this gradient scheme, the dephasing of the transversal magnetization caused by the three magnetic gradient fields is compensated (cf. Sect. 2.3.2), so there will be an echo signal called the gradient echo (GRE). Figure 2.4.16 shows the de- and rephasing behavior of the transverse magnetization for the three gradients in detail. Fast imaging sequences generating an echo by switching a pair of dephasing and rephasing

Fig. 2.4.14  Signal intensity of the FLASH sequence plotted as a function of TR /T1 for four different flip angles α. For short repetition times (TR << T1), the detectable MR signal is markedly higher when RF pulses with low flip angles are used instead of a conventional 90° pulse

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.15  Pulse and gradient scheme of the FLASH sequence. α flip angle of the excitation pulse, Gz slice selection-gradient, Gx phase-encoding gradient, Gy readout gradient

gradients without a rephasing 180° RF pulse as with the spin-echo tech­nique are called gradient-echo sequences (GRE sequences). Remark: This term, however, may not conceal the fact that in conventional imag­ing sequences, there is always a gradientecho created (see Figs. 2.4.1 and 2.4.6). The difference consists only in the fact that the 180° pulse of the SE technique creates an additional spin-echo, which temporally coincides with the gradient echo.

In contrast to the conventional imaging sequences, the nomenclature of the GRE sequences is not unified, but is handled differently by different manufacturers. In the following, the fundamentals of GRE imaging will be discussed in detail at the example of two representative sequences denoted by the acronyms FLASH and trueFISP. The excitation of the spin system and position encoding are identical in both sequences; they differ only in that the transverse magnetization is destroyed after acquisition of the MR signal in the FLASH sequence (spoiled

Fig. 2.4.16  Gradient refocusing of the FLASH sequence. Dephasing of the transversal magnetization caused by the slice selection and the readout gradient is compensated by two additional inverted gradients, so that a gradient-echo occurs. The figure shows the de- and rephasing process of two magnetization components (a,b), which are at different positions and therefore precess under the influence of the gradient-fields with different Larmor frequencies. φx, φy, and φz are the corresponding phase angles

GRE sequence), whereas it is maximized in the trueFISP sequence (refocused GRE sequence). This difference, however, leads to an entirely different contrast behavior.

Table 2.4.4  Recommendation of acquisition parameters for the measurement of weighted FLASH images Weighting

α

TR (ms)

TE (ms)

PD

10–20°

50–500

TE << T2*

T1

40–80°

20–300

TE << T2*

T2*

10–20°

50–500

15–40

47

48

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4.3.2  Example FLASH The spoiled GRE sequence shown in Fig 2.4.15 was introduced in 1985 by A. Haase et al. (1986) under the acronym FLASH (fast low-angle shot); its essential characteristics have already been outlined in the previous section. The signal intensity of a FLASH image can be described by the equation (cf. Sect. 2.4.3.1) S FLASH = PD ⋅

−TR / T 1

(1 − e ) sin α −TE /T 2* ⋅ e123 −TR / T 1 1 −4e 424 cos 1 4α3 T 2* factor

(2.4.8)

T 1 factor

and thus can be varied through the choice of the repetition time TR, the echo time TE , and the flip angle α of the excitation pulse. • PD dependence: see Sect. 2.4.2.1 • T1 dependence: In contrast to the SE sequence, the T1 contrast behavior of the FLASH sequence is not only influenced by the repetition time TR, but also by the choice of the flip angle α. In order to illustrate this, the T1 contrast between white and gray brain matter is plotted in Fig. 2.4.17 as a function of the repetition time TR for different flip angles. As can be seen, the T1 contrast at low flip angles is only different from 0 if very short repetition times are selected. If the flip angle is increased, then the T1 contrast maximum is shifted toward larger TR values and approximates the known T1 contrast behavior of the SE sequence. For a given value of TR/T1, the T1 factor of the signal Eq. 2.4.8 is maximized, when α is given by the so-called Ernst angle αE = arccos(e–TR/T1).

(2.4.9)

However, this does not imply that for this angle the tissue contrast between two structures is at its maximum. In Fig. 2.4.18 the tissue contrast between white

and gray brain matter is plotted for TE << T2*, i.e., exp(–TE  / T2) ≈ 1, as a func­tion of the repetition time TR and the flip angle α. For low flip angles, two contrast regions can be distinguished: for short TR times, the T1 contrast dominates (T1-weighted images), for longer TR times, the PD contrast (PD-weighted image). Example: In order to illustrated the discussed contrast behavior, Fig. 2.4.19 shows a T1-weighted (TR /TE /α = 150/6/60°) and a PD-weighted FLASH image (400/6/20°).

• T2 dependence: In order to acquire a T2*-weighted FLASH image, the influence of T1 relaxa­tion effects need to be minimized. For an SE sequence, this condition can only be fulfilled if a large repetition time TR is chosen (TR >> T1). This is different in FLASH imaging. As Fig. 2.4.18 shows, the recovery due to T1 relaxation can be neglected for much shorter TR values if the flip angle is chosen rather small. The T2* contrast of a FLASH image can be varied through the choice of the echo time TE. When doing this, however, one needs to remember that, due to the rapid T2* decay, T2* weight­ ing happens already at markedly shorter TE times than in SE imaging (see Sect. 2.4.3.4). The influence of the acquisition parameters on the contrast of a FLASH image is summarized in Table 2.4.4. 2.4.3.3  Example trueFISP In 1986, Oppelt et al. introduced a GRE sequence with the acronym FISP (fast imaging with steady precession), which considerably differs in its contrast from the FLASH sequence. This sequence was later renamed to trueFISP (see below). The pulse and gradient scheme of this sequence is shown in Fig. 2.4.20. Instead of the spoiler gra-

Fig. 2.4.17  Influence of the acquisition parameters TR and α on the T1 contrast of a FLASH image. For low excitation angles α, there will only be a considerable T1 contrast (here between white and gray brain matter) when short repetition times are selected. If the flip angle is increased, the T1 contrast max­ imum will shift to a higher TR value

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.18  Parameter weighting in FLASH imaging. Plotted is the contrast between white and gray brain matter as a function of the repetition time TR and the flip angle α for TE << T2*. Two different contrast regions can be distinguished for low flip angles: for short TR times, the T1 contrast dominates (T1-weighted image), for longer TR times, the PD contrast (PD-weighted image). If the flip angle is increased, then the T1 contrast curve will gradually approach the known contrast behavior of the SE sequence (α = 90°). Correspondingly, long repetition times need to be chosen in order to acquire PDweighted or T2-weighted images

Fig. 2.4.19  FLASH images of a transversal head section with a thickness of 5 mm. a T1-weighted image (TR /TE /α/NAQ = 150/6/60°/4) b PD-weighted image (400/6/20°/4)

dient of the FLASH sequence, refocusing gradient pulses are intro­duced in slice-selection direction as well as in the direction of frequency and phase encoding, through which the transverse magnetization is not destroyed after the data acquisition of the MR signal, but rather rephased or refocused (Fig. 2.4.21). Remark: As practice has shown, the trueFISP sequence is very susceptible to inhomogeneities of the static magnetic field, which are rendered visible as disturbing image artifacts. A more

favorable behavior is achieved by omitting the gradient pulses (which have been shaded darkly in Fig. 2.4.20). In this case, only the dephasing of the transverse magnetization caused by the slice selection and phase-encoding gradient is completely compensated. This realization is called FISP sequence.

In the trueFISP sequence, not only the longitudinal, but also the transversal magnetization reaches an equilibrium state after several sequence cycles. As both magnetization components are different from zero at the end

49

50

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.20  Pulse and gradient scheme of the trueFISP sequence. α flip angle of the excitation pulse, Gz slice-selection gradient, Gx phase-encoding gradient, Gy readout gradient. Instead of the spoiler gradient used in the FLASH sequence, there are refocusing gradient pulses in all three gradient directions, so that the transversal magnetization after acquisition of the gradient-echo is not destroyed but recon­structed (see Fig. 2.4.21). In practice, the gradient pulses (marked darkly) here is frequently omitted in order to reduce the susceptibility of the sequence for artifacts leading to a FISP sequence

of a sequence cycle, they will be mixed by the following RF pulse, i.e., a part of the longitudinal magnetization is flipped into the x–y-plane and a part of the transverse magnetization into the z-direction. Consequently, both magnetization compo­nents dependent on T1 and on T2. The T2 dependence increases proportional to the magnitude of the transverse magnetization remaining at the end of the sequence cycle (i.e., with decreasing TR/T2 ratio). Vice versa, this means that the FISP signal for high TR values (TR>>T2) will approximate the FLASH signal. Remark: The difference in the latter case merely consists in the fact, that in the FLASH sequence the transverse magnetization

Fig. 2.4.21  Gradient refocusing of the FISP sequence. Due to the complex gradient switch­ing, the dephasing of the transversal magnetization caused by the three gradients is completely compensated after acquisition of the gradient-echo, so the transversal magnetization is restored before irradiation of the subsequent excita­tion pulse. The figure shows the de- and rephasing process for two magnetization compo­nents (a,b), which are at different positions and therefore precess with different Larmor frequencies under the influence of the gradient fields. φx, φy, and φz are the corresponding phase angles. At the end of the sequence, both magnetiza­tion components are in phase again (φx = φy = φz = 0), independent of their spatial position

is rapidly destroyed by a spoiler gradient after the acquisition of the FID, whereas in the FISP sequence it decays with the time constant T2. Therefore, the FLASH sequence is more useful for the acquisition of T1-weighted and PD-weighted images than is the FISP sequence.

As the discussion has shown, the characteristic signal behavior of the FISP sequence manifests itself only for very short repetition times. For this special situation, the dependence of the signal intensity of the FISP sequence on the tissue parameters T1, T2, and PD can be described

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.22  T1/T2 dependence of the signal intensity of the trueFISP sequence. Plotted is the T1/T2 factor of the trueFISP signal equation as a function of the flip angle α for different T1/T2 ratios and very short repetition times (TR << T2). For a given flip angle α, the signal intensity will be the bigger, the smaller the ratio T1/T2 is

approximately by the expression , (2.4.10) which is independent of the repetition time TR. As this equation shows, the signal intensity of a FISP image for TR<
 T1 / T 2 1   . = arccos   T1 / T 2 + 1 

(2.4.11)

Remark: The considerations are only valid for stationary spins. If the phase coherence and thus the steady-state magnetization are disturbed by flow, as is the case in blood vessels or in liquor spaces, then the signal maximum will be shifted toward lower excitation angles.

In practice, flip angles between 70° and 90° and repetition times between 35 and 50 ms are used for the acquisition of T1/T2-weighted trueFISP images. In these images, tissues with lower T1/T2 ratio (i.e., fatty tissue and liquor) are more signal intensive than are tissues with high T1/T2 ratio (e.g., white and gray brain matter). For the acquisition of T1-weighted and PD-weighted images, the FLASH sequence is better suited than is the trueFISP sequence. Example: Figure 2.4.23 shows a T1/T2-weighted trueFISP image (TR /TE /y = 35/7/80°). The comparison of this image with the T1weighted FLASH image shown in Fig. 2.4.19 reveals the complementary contrast behavior of the two GRE sequences.

2.4.3.4  Influence of Magnetic Field Inhomogeneities Opposite to conventional imaging sequences (i.e., the SE and the IR sequence), the dephasing of spins due to static magnetic field inhomogeneities cannot be compensated in GRE sequences, due to the missing rephasing 180° RF pulse (see Sect. 2.4.3.1). Consequently, the transverse magnetization Mxy does not decay with the tissue-specific transverse relaxation time T2, but with the time constant T2* (cf. Sect. 2.2.5.2 and Fig. 2.2.22). Compared with T2, T2* shortens proportional to an increase in local magnetic field inhomogeneity. If the magnetic field inhomogeneity per voxel is denoted as ∆B, the relation between the three quantities can be expressed as 1 / T2* = 1 / T2 + γΔB / 2.

(2.4.12)

51

52

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Table 2.4.5  Reduction of the transverse relaxation time T2* due to inhomogeneities ∆B of the static magnetic field ∆B/voxel (ppm)

2/γ∆B (ms)

0.02

250

0.2

Fig. 2.4.23  T1/T2-weighted trueFISP image (TR /TE /α/NAQ = 35/7/80°/4) of a transversal head section with a thickness of 5 mm. (Compare this image with the T1-weighted FLASH image shown in Fig. 2.4.19a)

Numerical examples are given in Table 2.4.5. In MR imaging, local field inhomogeneities have essentially three reasons: • Technical imperfections of the main magnetic field. They increase with the distance from the center of the magnet and can cause severe signal losses in the outer zone. When using GRE sequences, one therefore has to remember that the body region to be analyzed should not be too far away from the center of the magnet. Example: In modern MR tomographs, the magnetic field inhomogeneity over a sphere of 50 cm diameter is smaller than 5 ppm. Assuming that the magnetic field changes linearly with distance to the center of the magnet, a magnet field inhomogeneity of about 0.02 ppm per pixel is obtained for a matrix size of 256 × 256 and a field of view of 50 cm (Table 2.4.5).

• Para- or ferromagnetic implants. Such implants (e.g., prostheses or clips) cause strong magnetic field inhomogeneities in the surrounding tissues. As a conse­quence, T2* times of these tissues are drastically reduced which results in signal losses in the vicinity of the implants. In these cases, conventional imaging tech­niques have to be used, which are less susceptible to field inhomogeneities. • Susceptibility effects. As discussed in Sect. 2.2.8, the magnetic field distribution within the human body

25

T2 (ms)

T2* (ms)

100

71

1,000

200

100

20

1,000

24

depends on the position, size, form, and magnetic susceptibility of the various tissues and organs in the body. Whereas the magnetic field is locally homogeneous in larger uniform organs or tissue regions, there are local field inhomogeneities at the interfaces of anatomical structures, which differ in their magnetic susceptibility. They can be strong enough to cause—via a reduction of the T2* times—signal losses in the vicinity of the interface. One can observe this effect in GRE images mainly in the vicinity of air–tissue interfaces, such as in the lung or in paranasal tissues. Even though susceptibility artifacts are unavoidable in GRE imaging, they can be minimized by two techniques: by shortening the echo time TE of the sequence and by reducing the voxel size, especially the slice thickness. By both means the average field inhomogeneity per voxel is reduced. 2.4.3.5  Influence of the Chemical Shift As shown in Fig. 2.3.7, the 1H spectrum of the human organism consists of two dominant peaks, whose resonance frequencies differ due to the chemical shift by about 3.5 ppm. The two resonance peaks are assigned to hydrogen nuclei in water molecules and those in methylene groups of fatty acids. If both components contribute to the MR signal of a tissue, for instance fatty tissue or bone marrow, then a modulation of the signal intensity of these tissues can be observed in GRE imaging in dependence on the echo time TE. This interesting phenomenon is due to the fact that the transverse magnetization of the water component precesses after RF excitation a little faster than the fat component, so both magnetization components are oriented parallel or antiparallel in regular intervals (Fig. 2.4.24). As the detected GRE signal is proportional to the vector sum of the two magnetization components, this periodic behavior manifests in a modulation of the signal intensity: the intensity is at its highest when the two components are oriented parallel; it is at its minimum

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.24  Temporal evolution of the transversal magnetization in fatty tissue. In fat-containing tissues, two pools of hydrogen nuclei contribute to the MR signal, whose resonance frequencies differ by about 3.5 ppm. They are assigned to 1H nuclei in water and in methylene groups of fatty acids. As the transversal magnetization of the water component (M xyw) after RF excita­tion

precesses slightly faster than the methylene component (M xyF), the two magneti­zation vectors are oriented a parallel or b antiparallel in regular intervals. The figure shows this in a rotating frame, in which the transversal magnetization of the fat component does not rotate

Example: At a magnetic flux density of B0 = 1.5 T, the frequency difference between the fat and the water component is about Δω / 2π = 220 Hz. This implies that the two magnetization components are oriented parallel or antiparallel after RF excitation in equidistant time intervals of Δτ = π / Δω = 2.3 ms. Figure 2.4.25 illus­trates the signal modulation at 1.5 T for a realistic example.

2.4.4  Modification of k-Space Sampling 2.4.4.1  Half-Fourier Technique Fig. 2.4.25  Modulation of the GRE signal in fat-containing tissues. At a magnetic flux density of 1.5 T, the difference between the resonance frequencies of the water and methylene protons is about 220Hz, so that the transversal magnetization of the two proton components is oriented parallel or antiparallel every 2.3 ms (see Fig. 2.4.24). This periodic behavior results in a modulation of the measured GRE signal: the signal intensity is at its maximum if the two transversal magnetiza­tions are parallel; it is at its minimum if they are antiparallel

when they are antiparallel. The amplitude of the signal modulation depends on the ratio of the two signal components.

In 2D Fourier imaging, the MR image is reconstructed via a Fourier transformation from the acquired raw data set, the k-space matrix. Even though for the reconstruction of an MR image with N × N pixels a raw data matrix with N × N data points is required, it is sufficient to measure only the first N/2 rows of the raw matrix, when the symmetry of the k-space is taken into account (halfFourier technique) (Feinberg et al. 1986). To illustrate this aspect, Fig. 2.4.26 shows a raw matrix that has been acquired with the SE sequence shown in Fig. 2.4.1. The data points of the k-space matrix are positioned symmetrically to the center of the matrix. Due to this fact, it is sufficient to measure only the data points of the upper half of the k-space and to fill the lower half through (conjugate) point reflection of the measured data in the center of the matrix (Fig. 2.4.27).

53

54

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.26  k-Space matrix. The raw data matrix shown has been acquired with the SE sequence shown in Fig. 2.4.1. As can be seen, the data points are positioned symmetrical to the center of the matrix. This symmetric property of the k-space matrix is utilized in the half-Fourier technique (see Fig. 2.4.27)

Remark: In order to generate this raw data matrix row by row, the sequence has been repeated N times for different values of the phase-encoding gradient Gx = n∆Gx (– N / 2 < n ≤ N / 2) with the SE signal being measured N times in the presence of the readout gradient in equidistant steps ∆t symmetrically to the echo time TE at each sequence cycle (see Fig. 2.4.28a).

With this technique, the number of sequence cycles NPH required for phase encoding can be halved without reducing the spatial resolution, which—according to Eq. 2.4.5—also implies a 50% reduction of the acquisition time. The reduction of acquisition time, however, results in a reduction of the S/N ratio by the factor 2 . Remark: In practice, some more rows of the k-space matrix are acquired, which makes it possible to perform a phase correction. Due to this reason, the reduction of the acquisition time is slightly less than 50%.

2.4.4.2  Fourier Interpolation The measurement time can also be reduced by sampling only the central segment of the k-space matrix. However, Fourier transformation of such a sparsely sampled k-space matrix yields an image that is only coarsely sampled (i.e., has large voxel dimensions) in the phase-encoding direction(s). This may result in severe partial volume effects in case of small tissue structures, such as vessels in

Fig. 2.4.27  Half-Fourier technique. a In this technique, only the data points of one half of the raw matrix are measured. b The missing points are generated through (conju­gate) point reflection of the measured data in the center of the matrix (see Fig. 2.4.26). c After this preparatory step, the MR image can be reconstructed as usual by a 2D Fourier transformation of the k-space matrix

MR angiography. The nominal resolution of the reconstructed image can be improved when the “missing” lines of the k-space are “zero filled” before Fourier transformation. This approach, denoted as Fourier interpolation, increases the number of pixels in the reconstructed image and thus the image resolution, but does not, of course, improve the physical resolution of the imaging process. Nevertheless, this trick improves the appearance of the reconstructed images and reduces partial volume effects. Fourier interpolation is routinely used in 2D imaging as, for example, to display a 512 × 512 image matrix reconstructed from a k-space matrix with only 256 × 256

2.4  Image Contrasts and Imaging Sequences

2.4.4.3  Parallel Imaging

Fig. 2.4.28  Possible sampling schemes for the k-space matrix. a In the conven­tional 2D Fourier technique, only one row of the k-space matrix is acquired per sequence cycle so that N sequence cycles are required to completely construct the raw data matrix. b In the single-shot echo planar technique, all rows of the k-space matrix are detected in one single sequence cycle. The figure shows one possi­ble sampling strategy. c In segmented k-space sampling, several rows of the k-space matrix are acquired per sequence cycle. The figure indicates the rows of the raw matrix that are acquired in one sequence cycle in gray

measurement points. The concept can be extended to 3D imaging, in particular to compensate for the sparse sampling of the k-space matrix in the direction of the second phase-encoding gradient (slab direction). Remark: A special application is the VIBE (volume interpolated breath-hold examination) technique, which is being used to reconstruct images from a sparsely sampled k-space matrix acquired with a fast 3D imaging sequence within a single breath hold.

What happens, when the number of phase-encoding steps NPH—and thus the image acquisition time—is halved, while the sampling interval of the phase-encoding gradient, here ∆Gx, is doubled? According to the sampling theorem discussed in Sect. 2.3.3.1, the field of view FOVx that can be imaged in x-direction without reconstruction errors is inversely proportional to ∆Gx. Therefore, the FOVx is also halved by the modified sampling strategy. In contrast, the spatial resolution ∆x = X/N is not changed. The modification discussed has thus—apart from a reduction of the S/N ratio due to the reduced number of measurements—no effect on the reconstructed MR image, as long as the object size X in direction of the phase-encoding gradient is less than FOVx /2. However, if the object size X is greater, then this will result in aliasing artifacts, i.e., the portions of the object outside of the reduced field of view get mapped to an incorrect location inside this field of view. The reason is that at these positions the phase angles ϕ induced by the phase-encoding gradient are greater than 2π or 360°) and thus cannot be distinguished from phase angles in the interval 0–2π (see Fig. 2.3.9) that encode positions inside the reduced field of view. The missing information, however, can be obtained by using two adjacent coil elements in an RF receive array with non-overlapping sensitivity profiles, each of which detects the MR signals from only one half of the full field of view. This strategy is called parallel imaging. Obviously, the number of phaseencoding steps NPH—and thus the acquisition time—can be further decreased if receive arrays with more than two coil elements are used (see Sect. 2.5). Parallel imaging offers the major advantage that it can be combined with any existing imaging sequence (e.g., SE, GRE, and echo planar imaging sequences) and that the contrast behavior of the sequence used is not changed. When implementing parallel imaging techniques, one has to deal with the problem that the sensitivity profiles of the coil elements are not rectangular but rather overlap. To avoid image artifacts arising from this fact, sophisticated image recon­struction techniques have been developed. They can be classified into two major groups, image-based reconstruction (e.g., SENSE, or sensitivity encoding) and k-space–based reconstruction techniques (e.g., GRAPPA, or generalized autocalibrating partially parallel acquisitions). 2.4.4.4  Segmented k-Space Sampling The basic idea of the echo planar imaging (EPI) technique, described by P. Mansfield in 1976 (Mansfield et al. 1976), aims at generating a series of spin or gradient-echoes in a short period of time after excitation of the spin system by an RF pulse, which are differently phase encoded by an appropriate gradient-switching scheme. In this way,

55

56

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

all rows of the k-space matrix can be acquired within one single sequence cycle. The various published versions of the EPI technique only differ in the way the phase-encoding gradients are switched, i.e., in what order the data points of the raw data matrix are sampled. One possible sampling scheme is shown in Fig. 2.4.28b. The limiting factor evidently is the speed with which the transverse magnetization decays after RF excitation. If the acquisition time is markedly longer than the decay constant (T2 or T2*, respectively), the different rows of the k-space matrix are differently T2 or T2* weighted, depending on the sampling scheme. This means that distinct spatial frequencies are over- and others are underweighted. In order to prevent this problem, the entire data set needs to be acquired within a time interval, which is comparable to the decay constant of the trans­verse magnetization. The EPI technique therefore challenges the gradient system. The mentioned problems in EPI can be avoided by sampling only few rows of the k-space matrix during each sequence cycle. If NE echoes are generated per excitation, then the number of sequence repetitions necessary to completely fill the raw data matrix and, in consequence, the acquisition time can be reduced by the factor 1/NE as compared with the conventional 2D Fourier technique. In contrast to the EPI technique, the demands on the gradient system for this so-called segmented k-space sampling strategy are much lower. 2.4.5  Preparation Techniques In MR imaging, there are two major pools of hydrogen nuclei that contribute to the image signal: the 1H nuclei of freely moveable water molecules and those of the CH2 groups of mobile fatty acids (see Fig. 2.3.7). In contrast, the MR signal from hydrogen nuclei which are bound in relatively immobile macro-molecular structures (such as proteins, cell membranes, etc.) cannot be detected due to their very short T2 times (cf. Sect. 2.2.5.3). Nevertheless, they affect indirectly the magnetization of the “free” 1 H nuclei in water and thereby also modify the image signal. Based on these facts, different methods have been developed, which offer the possibility to modify the magnetization detectable in MR imaging by preparation pulses. 2.4.5.1  Fat Saturation As already discussed in Sect. 2.3.3.1, the chemical shift of about 3.5 ppm between 1H nuclei in water and fatty acids becomes evident in the fact that fat- and water-containing tissue structures of the body are slightly shifted against one another. This image artifact, which is especially pronounced in EPI, can be avoided by reducing the contribution of the fatty acids to the MR signal.

One possibility to reduce the signal from fatty acids has already been discussed in relation with the STIR sequence (see Sect. 2.4.2.2), in which the magnetization is prepared (here inverted) by an inversion pulse. If the inversion time is chosen so that the longitudinal magnetization of the fatty acids is almost zero at this point in time, only the water pool will contribute to the STIR image (see Fig. 2.4.10b). A drawback of this technique is, however, that the inversion time is determined by the T1 time of the 1H nuclei in fatty acids, and thus the T1 contrast of the images cannot be chosen freely. This problem can be avoided by utilizing not the specific relaxation properties of fatty tissues (see Table 2.2.2) for fat suppres­ sion, but the chemical shift between 1H nuclei in water and in fatty acids. The latter strategy offers the opportunity to selectively saturate the magnetization of the fatty acids by frequency-selective RF pulses, so-called CHESS (chemical shift selective) pulses, and to use the remaining magnetization of the water pool for image generation. In the simplest case, the CHESS pulse is a frequency-selective 90° pulse, which flips the longitudinal magnetization of the CH2 groups into the x–y-plane, where it will be dephased by a spoiler gradient (Frahm et al. 1985; Haase et al. 1985). However, this places high demands on the homogeneity of the magnetic field within the patient; in order to separate the two resonance lines, the field homogeneity must be better than 3 ppm. In practice, this condition can only be rendered if the homogeneity of the magnetic field is adjusted by a special shim procedure before the examination of the patient. In order to avoid this procedure, optimized CHESS pulses for fat suppression have been developed (e.g., binominal pulses). 2.4.5.2  Magnetization Transfer In biological tissues, there are two pools of hydrogen nuclei that differ markedly regarding their biophysical properties: 1H nuclei in water molecules with high mobility (1Hf, or free water protons), and 1H nuclei in macromolecules with reduced mobility (1Hr). Whereas the T1 relaxation times of the two pools do not differ markedly (T1 > 100 ms), the T2 times differ strongly (see Fig. 2.2.16). The T2 times of the 1Hf nuclei are generally higher than 40 ms, whereas for 1Hr nuclei they are less than 100 µs due to the strong dephasing effect of neighboring spins. The different T2 times are mirrored in the 1H spectrum (Fig. 2.4.29). As the width ∆ω of a resonance line is inversely proportional to the T2 time (see Sect. 2.2.6), the 1Hf pool has a line width of a few Hertz, whereas the spectral width of the 1Hr nuclei is more than 10 kHz. It is crucial that the two 1H pools interact due to intermolecular processes (spin–spin interaction) and/or chemical exchange processes (Wolff and Balaban 1989). Due to this reason, any change in the magnetization in one pool results in an alteration of magnetization of the other pool. This effect is called magnetization transfer (MT).

2.4  Image Contrasts and Imaging Sequences

signal. In the simplest case, the frequency spectrum of the preparation pulse is defined by a rectangular function below and/or above the resonance frequency of the 1Hf pool, as is shown in Fig. 2.4.29. When doing this, the offset frequency has to be chosen big enough, so that local variations of the resonance frequencies of the 1Hf nuclei due to inhomogeneities of the static magnetic field and differences in tissue susceptibilities do not lead to a direct effect on the 1Hf magnetization. MT preparation pulses are often used in MR angiography to increase the blood-tissue contrast. This interesting application is based on the fact that the MT effect reduces the 1Hf magnetization of stationary tissue, whereas the magnetization of the flowing blood is not affected. Fig. 2.4.29  Schematic depiction of a 1H spectrum of biological tissues. Apart from the reso­nance line of 1H nuclei in free water (1Hf), with a low spectral line width (< 20 Hz), there is a broad underground due to 1H nuclei in macromolecules with reduced mobility (1Hr), the MR signal of which cannot be detected directly because of their short T2 times. Note that the spectral widths are not depicted in true scale. The frequency spectrum of a MT preparation pulse is marked in gray

To utilize this effect for MR imaging, the magnetization of the 1Hr pool is saturated by frequency-selective preparation pulses (saturation transfer). Due to the MT effect, this leads to a significant reduction of the MR signal of 1Hf nuclei and thereby to a reduction of the image

2.4.6  Sequence Families Imaging in magnetic resonance is based on spin warp imaging but is commonly referred to as Fourier imaging. The main underlying principle is the use of magnetic field gradients to prepare the slice-selective excitation and to phase and frequency encode the signal that is induced by the rotating transverse magnetization. The motivation of continued sequence development is fuelled by the aim to improve the tissue distinction and the shortening of measurement time. In recent years, a great number of sequences have been developed (see Table 2.4.6), each of which are utilized in routine clinical applications. The following paragraph provides a systematic overview of the sequence families.

Table 2.4.6  The most important MRI sequences Acronym

Phrase

Explanation

Synonyms

CISS

Constructive interference in steady state

Data merging of two 3D trueFISP sequences with and without RF phase alternation to eliminate destructive interference pattern

PC-3D-FIESTA

DESS

Double-echo steady state

Image data merging of FISP and PSIF appearing in adjacent acquisition windows

DRIVE

TSE with a –90° RF pulse at the end of an echo train in order to increase the signal for tissue with a long T2 relaxation time

RESTORE, FRFSE

EPI

Echo planar imaging

Single-shot technique using only one excitation, followed by multiple phase-encoded gradient-echoes to fill the k-space

FFE

Fast field echo

GRE with low–flip angle excitation and rephasing in the direction of phase encoding after data acquisition of a single Fourier line

FISP, GRASS

FIESTA

Fast imaging–employing steady-state acquisition

GRE with low–flip angle excitation and rephasing in all directions after data acquisition of a single Fourier line

trueFISP, bFFE

57

58

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Table 2.4.6  (continued) The most important MRI sequences Acronym

Phrase

Explanation

Synonyms

FISP

Fast imaging with steady-state precession

GRE with low–flip angle excitation and rephasing in the direction of phase encoding after data acquisition of a single Fourier line

FFE, GRASS

bFFE

Balanced FFE

GRE with low–flip angle excitation and rephasing in all directions after data acquisition of a single Fourier line

trueFISP, FIESTA

FLASH

Fast low-angle shot

GRE with low–flip angle excitation and “spoiling” after data acquisition of a single Fourier line

T1-FFE, SPGR

FLAIR

Fluid-attenuated inversion recovery

Liquor suppressed imaging protocol using an IRM sequence (long inversion time)

–

FRFSE

Fast-recovery fast spin echo

FSE with a –90° RF pulse at the end of an echo train in order to increase the signal for tissue with a long T2 relaxation time

DRIVE, RESTORE

FSE

Fast spin echo

SE using multiple phase–encoded echoes for faster filling of k-space

TSE, RARE

FSPGR

Fast spoiled GRASS

Spoiled GRASS with inversion or saturation pulse preceding the whole measurement in order to establish T1 weighting or for the nulling of the signal of a specific tissue

TFL, TFE

GRASE

Gradient and spin echo

FSE with multiple phase–encoded gradient-echoes within a SE envelope

TGSE

GRASS

Gradient-recalled acquisition in the steady state

GRE with low–flip angle excitation and rephasing in the direction of phase encoding after data acquisition of a single Fourier line

FISP, FFE

HASTE

Half-Fourier singleshot turbo spin echo

TSE utilizing the half-Fourier technique

IR

Inversion recovery

SE with preceding RF inversion pulse

MP-RAGE

Magnetization prepared rapid gradient echo

3D TFL version

T1-FFE

T1 fast field echo

GRE with low–flip angle excitation and “spoiling” after data acquisition of a single Fourier line

FLASH, SPGR

PSIF

Backward-running FISP

Acquiring an RF refocused signal using a time reversed FISP sequence

T2-FFE, SSFP

RARE

Rapid acquisition with relaxation enhancement

SE using multiple phase–encoded echoes for faster filling of k-space

TSE, FSE

RESTORE

Fast-recovery fast spin echo

TSE with a –90° RF pulse at the end of an echo train in order to increase the signal for tissue with a long T2 relaxation time

DRIVE, FRFSE

SE

Spin echo

90–180° sequence

SPGR

Spoiled gradient-recalled acquisition in the steady state

GRE with low–flip angle excitation and “spoiling” after data acquisition of a single Fourier line

STEAM

Stimulated echo acquisition mode

Sequence using three 90° RF pulses

FLASH, T1-FFE

2.4  Image Contrasts and Imaging Sequences Table 2.4.6  (continued) The most important MRI sequences Acronym

Phrase

Explanation

STIR

Short TI inversion recovery

TIRM sequence using a short inversion time suitable for suppressing the signal from fat

TGSE

Turbo gradient spin-echo

TSE with multiple phase–encoded gradient-echoes within an SE envelope

GRASE

TIR

Turbo inversion recovery

TSE with preceding inversion pulse (phase sensitive)

IR-FSE

TIRM

Turbo inversion recovery magnitude

TSE with preceding inversion pulse and utilizing only the magnitude of the signal (phase insensitive)

trueFISP

True fast imaging with steady-state precession

GRE with low–flip angle excitation and rephasing in all directions after data acquisition of a single Fourier line

FIESTA, bFFE

TSE

Turbo spin echo

SE using multiple phase–encoded echoes for faster filling of k-space

FSE, RARE

TFE

Turbo field echo

T1-FFE with inversion or saturation pulse preceding the whole measurement in order to establish T1 weighting or for the nulling of the signal of a specific tissue

TFL, FSPGR

TFL

Turbo fast low-angle shot

FLASH with inversion or saturation pulse preceding the whole measurement in order to establish T1 weighting or for the nulling of the signal of a specific tissue

TFE, FSPGR

VIBE

Volume-interpolated breath-hold examination

3D GRE with low–flip angle excitation and “spoiling” after data acquisition of a single Fourier line and Fourier interpolation in the direction of partition encoding

2.4.6.1  Sequence Classification A first sequence classification can be performed in assigning the type of sequence in either a spin-echo or a gradient-echo group. The main difference between SE and GRE is the influence of susceptibility gradients on image contrast. In general, in GRE imaging susceptibility gradients lead to a faster decay of the signal, whereas in SE imaging dephasing mechanisms that are fixed in location and consistent over time are refocused by the 180° refocusing RF pulse. SE image contrast depends on the tissue specific transversal relaxation time T2, whereas GRE image contrast is a function of the transversal relaxation time T2*. Some GRE techniques utilize the excitation pulse also as a refocusing pulse, causing spin-echo components to contribute to the image contrast. Within the SE and the GRE group, the contrast can be manipulated by preparing the longitudinal magnetization prior to starting the imaging sequence or prior to the measurement of a Fourier line. In multi-echo imaging, the transverse magnetization is refocused and reutilized after the collection of a Fourier line, omitting the necessity of a further excitation for the collection of another Fourier line. This method is applicable within the SE group as well as the GRE group. Again, a preparation of the magnetization is generating another

Synonyms

sequence family. Using only one excitation and multiple phase–encoded echoes to acquire all required k-space lines without a further excitation is called a single-shot technique. Figure 2.4.30 shows an overview scheme that provides one possible sequence-classification. Within the SE sequences, there are: • The conventional SE sequence (see Sect. 2.4.2.1) • The IR sequences (see Sect. 2.4.2.1) • The multi-echo sequences (e.g., TSE, FSE) • The multi-echo sequences with preparation of the magnetization (e.g., TIR, TIRM, IR-FSE, RESTORE, DRIVE, FRFSE) • The single-shot techniques (e.g., SS-FSE, or using a Half-Fourier technique [see Sect. 2.4.4.1] like in HASTE) • The single-shot techniques with preparation of the magnetization (e.g., HASTIRM) Multi-echo sequences use several phase–encoded echoes in order to fill the k-space (see Sect. 2.4.4.2). They are called fast spin-echoes (FSE), turbo spin-echoes (TSE), and RARE (rapid acquisition with relaxation enhancement). The combination of the basic TSE technique and an inversion pulse for the preparation of the longitudinal magnetization is called TRIM (turbo inversion recovery magnitude) or TIR (turbo inversion recovery). If one

59

60

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.4.30  The sequence family. The shown sequence table provides an overview of the sequence classification

combines the Half-Fourier method with a TSE sequence to a degree at which only a single excitation pulse suffices to fill the raw-data matrix with following spin-echoes, one applies the so-called HASTE technique (half-Fourier single-shot turbo spin-echo). The mix of spin-echoes with gradient-echoes or, more precisely the acquisition of gradient-echoes within an SE envelope leads to the TGSE (turbo gradient-echo sequence) sequence, also called GRASE (gradient and spin-echo). As expected, with the introduction of gradient-echoes within a multi-echo spin-echo sequence, the contrast behavior is also T2* related. This sequence is also called a hybrid. Similar to the SE sequences, the GRE sequences can be grouped into: • Conventional GRE sequences (e.g., FLASH, FISP, trueFISP, DESS, CISS, PSIF) • GRE sequences with preparation of the magnetization (e.g., turboFLASH, MP-RAGE) • Multi-echo GRE sequences (e.g., MEDIC, segmented EPI) • Multi-echo GRE sequences with preparation of the magnetization (e.g., segmented DW-SE-EPI) • Single-shot GRE sequences (EPI) • Single-shot GRE sequences with preparation of the magnetization (e.g., DW-SE-EPI) As indicated above, conventional GRE sequences can be further divided into: • SSI group (steady-state incoherent), which only aims at a steady state in the longitudinal magnetization (e.g., FLASH, SPGR, T1-FFE) and • The SSC group (steady-state coherent), during which the steady state of the transversal magnetization

equally contributes to the signal (e.g., FISP, trueFISP, GRASS, FIESTA, FFE, bFFE). Acronyms of the SSI group are FLASH (fast low-angle shot), SPGR (spoiled gradient-recalled acquisition in the steady state), and T1-FFE (T1-fast field echo). In the SSC group there are trueFISP (fast imaging with steady precession), GRASS (gradient-recalled acquisition in the steady state), and FFE (fast field echo). Within the SSC group, there is a slow transition toward spin-echoes, as excitation pulses do not only excite, but also refocus various echo paths of a remaining or refocused transverse magnetization. The extreme form is PSIF (a backward-running FISP), also named SSFP (General Electric) or T2-FFE (Philips). In these techniques, the excitation pulse of the following measurement operates as a refocusing pulse for the transverse magnetization of the previous excitation. The contrast is T2 weighted, as the effective echo time amounts to almost two repetition times. A combination of FISP echo and PSIF echo is called DESS (double-echo steady state), and having the FISP and PSIF echo coincide in time will result in a CISS (constructive interference steady state) or a trueFISP sequence. The same preparation of the magnetization utilized for SE techniques can be applied to GRE techniques. With a very rapid GRE sequence (RAGE, or rapid acquired gradient-echoes), with the aim of measuring as fast as possible, the TR is set to a minimum and consequently so is TE, and the excitation angle is set to an optimum (Ernst angle) in order to generate as much signal as possible. To reestablish a T1 weighting, a saturation or inversion pulse is applied, but not prior each Fourier line as in SE imaging, but at the beginning of the whole

2.4  Image Contrasts and Imaging Sequences

measurement. Those techniques are called turboFLASH, FSPGR (fast spoiled gradient-recalled acquisition in the steady state), TFE (turbo field echo) or, placing the inversion within the partition loop of a 3D sequence, MPRAGE (magnetization prepared rapid acquired gradientechoes). As is the case in fast SE sequences, GRE sequences also can make use of multi-echo acquisitions. MEDIC (multiecho data image combination) uses multiple echoes for averaging, thus improving SNR and T2* contrast. The classical form of a single-shot GRE technique, during which the raw data matrix is filled after a single excitation with several phase-encoded gradient-echoes, is called EPI (echo planar imaging). Simply collecting the free induction decay with multiple phase–encoded gradient-echoes is called FID-EPI. Placing the gradient-echoes beneath an SE envelope is called SE-EPI. The most common magnetization prepared single shot gradient-echo technique is the diffusion–weighted spinecho echo planar imaging sequence (DW-SE-EPI). 2.4.6.2  Fast Spin-Echo Techniques The idea of using multiple phase-encoded spin-echoes to fill the k-space more rapidly as compared with conventional imaging has surfaced as early as 1986, with the acronym RARE—rapid acquisition with relaxation enhancement (Fig. 2.4.31). The number of applied echoes is directly proportional to the potential reduction in measurement time. The overall image contrast is dominated by the weighting of those Fourier lines acquired in the center of k-space (effective echo time). The qualities of the early images were not close to the quality of conventional T2-weighted SE imaging. In the course of hard- and software developments, Mulkern and Melki “re-discovered” multi-echo SE imaging during a search for a fast T2-localizer, creating the acronym FSE (fast spin-echo). Siemens and Philips use the acronym TSE for turbo spin-echo. Since the higher spatial frequencies, the “outer” k-space lines, are usually acquired using late echoes, early concern has been that small objects might be missed. Fortu-

nately, the time saving achieved with the use of multiple echoes has been utilized to improve the contrast by selecting longer repetition times and to improve the spatial resolution by increasing the matrix size. Both measures have more than compensated the effect of an under-representation of high spatial frequencies within the k-space matrix. T2-weighted TSE has replaced conventional spinecho imaging in all clinical applications. The acquired phase-encoded echo train can also be used to create PDweighted and T2-weighted images similar to conventional dual-echo spin-echo imaging. The use of phase-encoded echoes for the k-space of the PD-weighted image as well as the k-space of the T2-weighted image is customary and this procedure is called shared echo. T1-weighted TSE imaging is also an option for some applications, although additional echoes will increase (unwanted) T2-weighting. T1-weighted TSE imaging is rarely applied to the central nervous system as the use of additional echoes prolongs the time needed for a single slice and the number of necessary slices may not fit into the desired TR. For the genitourinary system (uterus, cervix, bladder, etc.) about three echoes are used to improve SNR or to reduce measurement time. In areas where the amount of T1-weighting is less of an issue, e.g. T1weighted imaging of the cervical and the lumbar spine for degenerative disease, TSE is usually used with an echo train length (ETL) of five echoes. The same protocol is applied for enhanced and unenhanced studies of suspected vertebral metastases. T1-weighted TSE imaging for the abdomen is not an issue, since the restriction of the measurement time down to a breath hold period is suggesting T1-weighted GRE imaging. The remaining point of concern in comparing TSE imaging with conventional SE imaging is the reduced sensitivity to susceptibility artifacts. Hemorrhagic lesions appear less suspicious on TSE imaging as compared with conventional SE imaging. The Fourier transformation assumes a consistent signal contribution for all Fourier lines. Any violation of this assumption will lead to over- or under-representation of spatial frequencies, with a correlated image blurring. Although TSE violates this assumption in using multiple phase–encoded echoes to

R

Fig. 2.4.31  Structure of the TSE   sequence. Excitation, refocusing,   frequency, and phase encoding are   done as in the conventional SE sequence. The dephasing done for the purpose of spatial encoding is rephased prior to generating another spin-echo using a 180° RF pulse followed by another phase-encoding step

61

62

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

fill the k-space, where the signal amplitude of the echoes diminishes following T2 decay, there are several parameter that can be utilized to minimize the artifacts related to the T2 decay–related k-space asymmetry: • First, especially for T2-weighted imaging, the signal amplitudes for the late and closely spaced echoes can be approximated as being constant. TSE sequences are mainly used for the acquisition of T2-weighted images. As lesions usually have a long T2 relaxation time, the signal loss caused by the T2 decay does not play a major role during data acquisition. • Second, the matrix dimensions used in TSE imaging are usually higher as compared with conventional SE imaging. This will significantly minimize the risk of missing small objects. • Third, in T2-weighted TSE imaging, the repetition time is increased considerably, which leads to a remarkable improvement in contrast, again reducing the potential risk of missing small objects due to k-space asymmetry. In a typical TSE protocol, 13–15 echoes per excitation are used for imaging, implying a theoretical shortening of measurement time of the factor 13–15. In practice, the shortening is about the factor 2–6. Longer repetition times are selected for improved PD and T2 weighting, and larger image matrices are used for improved spatial resolution, diminishing the potential shortening of the measurement time when using multiple phase–encoded echoes. As the mentioned influence on the space encod-

ing is only present in the direction of the phase encoding, the effect can be demonstrated by exchanging of the frequency-and phase-encoding gradients. Figure 2.4.32 shows this in the example of the cauda equine. Apart from use for high-resolution images, TSE sequences are also applied in cardiology, as shown in Fig. 2.4.33. With the help of TSE sequences, T2-weighted images of the beating heart can be acquired based on the breath-hold method. Other acronyms for equal or similar techniques are FSE (fast spin-echo [Mulkern et al. 1990]) and RARE (rapid acquisition with relaxation enhancement [Henning et al. 1986]). The 3D encoding is another application of the TSE sequence that would lead to intolerable measurement times in conventional SE (see Fig. 2.4.36). Here, the slice-selection gradient is provided with a second phase-encoding gradient, as has been discussed in Sect. 2.3.4.3. Thus, thinner slices than in the 2D technique can be obtained. Furthermore, there are no slice gaps and the problem of “crosstalk,” which is found in multiple-slice techniques can be avoided (see Sect. 2.3.5). Fast spin-echo imaging demonstrates two essential differences in imaging appearance as compared with conventional spin-echo imaging: Fat appears bright, and there is a reduced sensitivity to hemorrhagic lesions. The bright appearance of fat is related to the J-coupling. The so-called J-coupling (see Sect. 2.2.9) of the carbon-bound protons provides a slow dephasing of transversal magnetization in conventional SE imaging, in spite of the refocusing pulse. If the refocusing pulses follow shortly after

Fig. 2.4.32  Image of the cauda equina, acquired with a TSE sequence with a the phase-encoding direction from left to right, and b the phase-encoding direction from head to foot. The longitudinal structures of the relatively thin nerves will be better

visible, if the frequency-encoding direction is perpendicular to the nerve fiber. The resolution in frequency-encoding direction is not influenced as much by the T2 decay as by the resolution in phase-encoding direction

2.4  Image Contrasts and Imaging Sequences

to diffusion in between excitation, refocusing, and data acquisition. Reducing this diffusion time by rapidly succeeding refocusing pulses will also reduce the related artifacts, thus making TSE imaging less sensitive for hemorrhagic lesions. 2.4.6.2.1  Fast Spin-Echo Techniques with Inversion to Suppress the Signal from Fat

Fig. 2.4.33  T2-weighted image of the heart in short-axis perspective. Duration of data acquisition 20 s (with ECG triggering at breath holding) (TSE 900/57, slice thickness = 8 mm)

one another, as is the case in TSE imaging, the J-coupling will be overcome; the dephasing will be suppressed. Consequently, fat tissue appears brighter in the TSE image than it does in a conventional image. If desired fat saturation or fat suppression (see Sect. 2.4.5.1) can be utilized to suppress this appearance. The susceptibility-related artifact of hemorrhagic lesions in spin-echo imaging is due

The TSE sequence, like the conventional SE technique, can be used with an inversion pulse for preparation of the longitudinal magnetization. Thus, it becomes possible to yield a suppression of the fat signal, based on the short relaxation time of fat (see Fig. 2.4.10b). Relaxation dependent fat suppression using an inversion pulse prior to the fast spin-echo train is routinely used to demonstrate bone infarctions and bone marrow abnormalities like bone marrow edema, e.g., in sickle cell anemia. This fat suppression scheme is also used in genitourinary applications, where the high signal intensity of fat may obscure contrast-enhanced tumor spread. Since only the fat suppression is desired, the used inversion recovery technique is not phase sensitive, only the magnitude of the longitudinal magnetization is used. The acronym used in this case is TIRM (turbo inversion recovery with magnitude consideration). The structure of a turbo inversion recovery (TIR) sequence is presented in Fig. 2.4.34.

Fig. 2.4.34  Structure of the turbo inversion recovery (TIR) sequence

63

64

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4.6.2.2  Fast Spin-Echo Techniques with Inversion to Suppress the Signal from Fluid The reduction in measurement time due to the utilization of multiple phase-encoded spin-echoes permits the use of inversion times in the order of 2 s, keeping the measurement time acceptable. An inversion time of 1.9 s will provide a relaxation dependent suppression of the cerebral spinal fluid (CSF) signal (Fig. 2.4.35). The utilization

of a long inversion time is called fluid-attenuated inversion recovery, or FLAIR. In combination with TSE imaging (the structure of a 3D TSE sequence is presented in Fig. 2.4.36), the technique is called turboFLAIR or simply TIRM. Since CSF has the longest T1 relaxation time, the longitudinal magnetization within all other tissues will be aligned parallel to the main magnetic field, and it is not necessary to have a phase sensitive IR method for this application. The attenuated CSF signal allows a better differentiation of periventricular lesions and has demonstrated a superior sensitivity for focal white matter changes in the supratentorial brain, whereas posterior fossa located lesions can be missed. The turboFLAIR method apparently allows the identification of hyperacute subarachnoid hemorrhage with MR, precluding the need for an additional CT. 2.4.6.2.3   Fast Spin-Echo Techniques with Inversion to Improve T1 Contrast

Fig. 2.4.35  Axial head image measured with a TIR sequence. The cerebrospinal fluid (CSF) does not contribute any signal, as the inversion time is selected for the CSF not to have a longitudinal component at the time of excitation (TIRM 8,000/2,200/105, slice thickness = 6 mm)

R

The time-consuming IR method has been used in the past for studying the development of white matter tracts in developmental pediatrics. This technique has been replaced using an inversion pulse prior to a spin-echo train of a TSE sequence. The selected inversion time (~350 ms) allows a better delineation of small differences in T1 relaxation times, e.g., for the documentation of the development of the pediatric brain. The improved tissue characterization between gray matter and white matter tracts allows, e.g., the demonstration of mesial temporal sclerosis and visualizing hippocampal atrophy. For this application a so-called phase-sensitive inversion recovery is required to differentiate nuclear magnetization aligned parallel to the magnetic field as compared with antiparallel alignment at the time of excitation. The according acronym is TIR.

Fig. 2.4.36  Structure of the 3D TSE sequence (additional phase-encoding gradients in direction of slice selection)

2.4  Image Contrasts and Imaging Sequences

2.4.6.2.4  Driven Equilibrium or Restore Techniques The residual transverse magnetization after measuring a single Fourier line, or, in case of multi-echo imaging, after measuring the “package” of Fourier lines, is usually spoiled. A later-introduced concept refocuses the transverse magnetization at the end of the echo train and uses an RF pulse to “restore” the residual transverse magnetization back to the longitudinal direction. The method “improves” the recovery of the longitudinal magnetization for tissue with long relaxation times, allowing a further shortening of the repetition time without loss of contrast. The technique is called RESTORE (SIEMENS), fast recovery fast spin-echo FRFSE (GE) and DRIVE (Philips). It does not make a difference whether the magnetization is prepared after the measurement of a Fourier line or at the very beginning of a new excitation cycle. For this reason it is justified to list RESTORE as a turbo spin-echo scheme with preparation of the longitudinal magnetization. 2.4.6.2.5  Dark-Blood Preparation Scheme Multi-echo spin-echo imaging has the potential to acquire T1- and/or T2-weighted spin-echo imaging of the beating heart within a breath hold. The only obstacle that needs to be addressed is the significant flow artifacts caused by the flowing blood. The introduction of dark-blood preparation scheme finally revolutionized cardiac MR imaging. With this preparation scheme, it is now possible to acquire T1- and T2-weighted images of the beating heart within a breath hold, without any flow artifacts. The magnetization of the whole imaging volume is inverted nonselectively, followed by a selective reinversion of the slice. This is done at end diastole, with the detection of the QRS complex. During the waiting period to follow, most of the reinverted blood will be washed out of the slice, being replaced by the inverted blood—and the spin-echo train acquired again toward end diastole will show “black” blood. A double inversion pulse will even allow not only the black-blood preparation, but also the suppression of fat signal, which will be helpful in characterizing fatty infiltration of the myocardium in arrhythmogenic right ventricular dysplasia (ARD). 2.4.6.2.6  Single-Shot Spin-Echo Imaging Single shot, per definition, refers to a single excitation pulse and the use of multiple phase–encoded echoes to fill the required Fourier lines. The original RARE has been published as a single-shot spin-echo technique. Other acronyms found in the literature are SS FSE for

single-shot fast spin-echo (GE) or SS TSE of single-shot turbo spin-echo (Philips). The combination with a halfFourier technique allows a further reduction in measurement time and has been named HASTE: half-Fourier acquired single-shot turbo spin-echo. As elaborated on earlier, the first and the last data point of a Fourier line are characterized by the transverse magnetization of adjacent voxel pointing into opposite direction. The same situation is found for the first and last Fourier line within k-space, considering the transverse magnetization within adjacent voxel aligned in the direction of phase encoding. k-Space is symmetrical. Although this hermitian symmetry is ideal and reality is slightly different, it has been claimed, that the deviation from the ideal situation are only of coarse nature and that a few (e.g., eight) Fourier lines measured beyond the center of k-space should be sufficient to correct for this insufficiency. As an example for a 128*256 matrix, a single-shot spin-echo technique using the half Fourier approach would use 128 /2 + 8 = 72 phase-encoded echoes to fill the k-space. The measurement time using this acquisition method is about a second per slice. The high numbers of echoes suggest that this technique is only useful for T2-weighted imaging and the blurring effect due to signal variation in k-space as a result of T2 decay will be prohibitive for high resolution studies. Nevertheless, it is an alternative, even in the brain, for a fast T2-weighted study for patients who are not able or willing to cooperate. Since it is the perfect technique to visualize fluid-filled cavities, HASTE is used, e.g., for MRCP (magnetic resonance cholangiopancreatography). A typical result of this sequence is shown in Fig. 2.4.37. Progress in hardware development and the correlated improvement in image quality, together with the pioneering research within this field have recently led to an impressive increase in HASTE utilization for obstetric imaging. Although sonography remains the imaging technique of choice for prenatal assessment, the complementary role of MR imaging is getting more and more important in the early evaluation of brain development of the unborn child or even in the early detection of complications within the fetal circulatory system. 2.4.6.3  Gradient-Echo Techniques The search for shorter measurement times for faster imaging led to the group of gradient-echoes (GRE) in 1983. An MR signal can be detected immediately after the excitation pulse. That signal is the free induction decay (FID). In addition to the spin–spin interaction causing the T2 relaxation, other dephasing mechanism will contribute to the image contrast, dephasing mechanisms that are based on differences in precessional frequencies due to magnetic field variations across a voxel. The main sources of local magnetic field variations are differences in tis-

65

66

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4.6.3.1  Spoiled Techniques

Fig. 2.4.37  Axial abdominal image acquired with HASTE. Every image is produced with a single excitation, followed by a train of phase-encoded spin echoes. The scan time of the image was 570 ms; the slice thickness is 8 mm

sue-specific susceptibility values. Since these dephasing mechanisms are fixed in location and are constant over time, they are refocused using a 180° RF refocusing pulse in SE imaging. Omitting this pulse will lead to a contribution of these dephasing mechanisms to the image contrast. The observed tissue specific relaxation parameter is then called T2* rather than T2. 1 1 1 = + T 2* T 2 T 2'

with T2´ being both machine and sample dependent. Although the missing 180° RF refocusing pulse will cause a rapid dephasing of the transverse magnetization and with that a rapidly dephasing MR signal, the echo time can be reduced as well and so can the repetition time. The shorter echo time will allow, in most cases, a detection of a signal despite the rapid dephasing of the transverse magnetization. Since the echo is now formed by using a bipolar gradient pulse in the direction of frequency encoding, these techniques are called GRE. With shorter repetition times an extension of phase encoding for the direction of slice or slab selection can be considered, and 3D imaging becomes feasible. The short excitation pulses used in common GRE imaging will result in a less perfect slice profile as compared with the slice profile achieved with a 90–180° combination of longer RF pulses as typically utilized in SE imaging. As a result, there will be significant contributions of the low angle–excited outer regions of a slice, explaining the basic difference in contrast between a GRE image as compared with an SE image, even if a 90° excitation angle is utilized.

Similar to the spin-echo sequence acquisition scheme, there is residual transverse magnetization left at the end of the acquisition of one Fourier line—and similar to the spoiling of the transverse magnetization at the end of the measurement in SE imaging, the same process can be applied to GRE imaging as well. Spoiling can be done with a gradient pulse, distributing the transverse magnetization evenly, so that the next excitation pulse will not generate a stimulated echo. Or the phases of the excitation pulses can be randomized in order to avoid the buildup of a steady state for the transverse magnetization (RF spoiling). Spoiled gradient-echo imaging has been introduced as fast low angle shot (FLASH), T1-weighted fast field echo (T1-FFE), or spoiled gradient recalled acquisition in the steady state (SPGR). FLASH imaging allows multislice imaging in measurement times short enough to allow breath hold acquisitions. Since the contrast mainly depends on the T1 relaxation time, FLASH images are usually called T1 weighted. In clinical routine, FLASH sequences have been introduced for diagnosing cartilage lesions (Fig. 2.4.38), for abdominal breath-hold T1weighted imaging (Fig. 2.4.39), and in dynamic contrast enhanced studies. As has been discussed in Sect. 2.4.3.2, not only the amplitude of the signal can be controlled, but also the basic contrast behavior can be influenced. For instance, when using an extremely small excitation angle and moderate repetition times, one can minimize the influence of the T1 relaxation time (see Fig. 2.4.17). Thus, one can obtain

Fig. 2.4.38  Sagittal 2D image of an ankle using a FLASH sequence (FLASH 770/11/60° SD = 3 mm, 512 × 512 matrix)

2.4  Image Contrasts and Imaging Sequences

Fig. 2.4.39  Axial T1-weighted image of the abdomen acquired with a FLASH sequence during a breath hold (2D FLASH 174/4/80°, 23 slices)

FISP. The original implementation and publication of FISP uses a gradient refocusing in phase encoding as well as in frequency-encoding direction and slice-select direction. As this sequence was susceptible to artifacts at the time, the implemented and released FISP sequence, still used today, is only refocused in the direction of phase encoding, and no refocusing in readout and slice-selection direction. Such a sequence has been called ROAST (resonant offset acquired steady state [Haacke et al. 1991]). For the FISP sequence, the phase encoding is reversed after the acquisition of the Fourier line, undoing the dephasing that was applied for spatial encoding. This approach will lead to a steady state not only for the longitudinal magnetization, but also for the transverse magnetization—for tissue with long T2 relaxation times. Differences in FISP contrast as compared with FLASH applications will only be visible for short repetition times; large excitation angles and will only enhance signal within tissue with long T2 relaxation times. General Electric introduced this technique as gradient-recalled acquisition in the steady state (GRASS). Philips is using the fast field echo (FFE). 2.4.6.3.3  Balanced Techniques

Fig. 2.4.40  Axial image of the cervical spine (MR myelography), acquired with a FLASH sequence with relatively small excitation angle at a magnetic field strength of 0.2 T (2D FLASH 1,200/50/40°, slice thickness = 4 mm)

If one rephases at the end of the measurement of a Fourier line, all parts of the transverse magnetization that have been dephased for spatial encoding and if one compensates in advance for the dephasing to be expected while the slice selection gradient is switched on, one obtains the trueFISP sequence (Fig. 2.4.41). This technique combines the advantages of the FISP sequence and the PSIF sequence, with further echo paths contributing to the overall signal. A clinical application of this sequence is shown in Fig. 2.4.42. This original approach of refocusing all transverse magnetization at the end of the mea-

RF

PD-weighted or T2-weighted images in spite of short repe­ tition times. This effect is used in MR myelography (see Fig. 2.4.40). 2.4.6.3.2  Refocused Techniques The alternative to spoiling the residual transverse magnetization after the end of the Fourier line acquisition is to rephase what has been dephased for spatial encoding. This was introduced as fast imaging with steady precession (FISP) (Oppelt et al. 1986), later to be called true-

Fig. 2.4.41  Structure of the trueFISP sequence. The sequence seems to be “balanced” due to a symmetry in time. All components of the transverse magnetization are refocused at the end of the measurement, leading to a steady state

67

68

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

surement of one Fourier line will not only cause a steady state for the transverse magnetization, but also the next excitation pulse will also operate as a refocusing pulse. The excitation pulse will not only convert longitudinal magnetization to transverse magnetization, but will also generate a spin-echo. The sequence seems to be symmetric, balanced. The challenge is to get all the generated echoes to have one phase, otherwise the echoes will destructively interfere, causing band-like artifacts. The po-

sitions of these bands also depend on the starting phase of the RF pulse. Adding another acquisition with a phase shifted RF will lead to a technique called constructive interference steady state technique, (CISS), or phasecycled fast imaging–employing steady-state acquisition (PC-FIESTA, the acronym used by GE). Since CISS contains spin-echo components, the technique is even useful in regions with significant susceptibility gradients, e.g., nerve imaging at the base of the skull. Since this technique is a fast technique with hyperintense appearance of fluid filled cavities, it is primarily applied to study abnormalities of the internal auditory canal. The originally published FISP is the trueFISP, where all the dephasing is reversed and even the slice selection gradient is preparing the dephasing to be expected during the first half of the next excitation pulse. The trueFISP technique is a fast gradient-echo sequence with spin-echo contributions leading to hyperintense appearance of all tissues with long T2 relaxation times. The technique is primarily used

Fig. 2.4.42  Sagittal T2-weighted image of the head, measured with a trueFISP sequence (7/3/80°, 512 × 512, 4 acquisitions, overall scan time = 13 s)

Fig. 2.4.44  Image of the cochlea using a maximum intensity projection (MIP) applied to an image stack, produced by a 3D PSIF sequence (3D PSIF 17/7/80°, slice thickness = 0.5 mm)

Fig. 2.4.43  Structure of the PSIF   sequence

2.4  Image Contrasts and Imaging Sequences Fig. 2.4.45  Structure of the DESS sequence. FISP and PSIF signals are acquired in adjacent acquisition windows and later added prior to image reconstruction. DESS is used as a 3D technique

in fast cardiac imaging, for cine snapshots of the beating heart. General Electric is using the acronym FIESTA for the same technique. Philips is using bFFE (balanced fast field echo) as the acronym for his technique. 2.4.6.3.4  The Gradient Echo That is a Spin Echo The previously mentioned spin-echo component of a balanced technique can be isolated and can be used to generate an image. The PSIF sequence shown in Fig. 2.4.43 appears to be violating the causality at first: a FISP sequence running backward. The signal inducing transverse magnetization is produced with the first excitation at the end of the first cycle, refocused with the second excitation at the end of the second cycle, and inducing a signal at the beginning of the third cycle. The effective echo time therefore amounts to almost two repetition times. The resulting images consequently show a remarkable T2 weighting. (Note that in this case, it is a spin-echo and not a gradient-echo.) The PSIF sequence is insensitive to susceptibility gradients. In contrast to CISS, the PSIF is very sensitive to flow and motion, thus it is not applied for IAC imaging but rather used as an adjunct to demonstrate abnormal CSF flow pattern. General Electric is calling this technique simply steady state free precession (SFFP), while Philips is using the acronym T2-FFE. Imaging of the cochlea (Fig. 2.4.44) is no longer performed with PSIF but rather with CISS, due to the intrinsic flow insensitivity of the latter. When combining a FISP image with a PSIF image, one obtains an image with a T2* weighting via the GRE signal and a T2 weighting via the

Fig. 2.4.46  Sagittal image of a knee acquired with a DESS sequence using a special excitation pulse, which only excites the water protons (water excitation) (DESS 26/9 or 45/40°)

SE signal. Such a sequence is called double echo steady state (DESS, Fig. 2.4.45). The DESS sequence is routinely used in orthopedic imaging (Fig. 2.4.46). It links the advantages of the FISP sequence with the additional signal enhancement of the PSIF sequence for tissues with long T2 (e.g., edema and joint effusion).

69

70

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.4.6.3.5  Gradient-echo Single-Echo Imaging with Preparation of the Magnetization In theory, all the magnetization preparation schemes previously applied for the SE group can also be applied for the GRE sequences. However, it has to be kept in mind that GRE sequences are usually using shorter TE and shorter TR than SE imaging does, and therefore some preparation schemes should be slightly altered. As, for an example, the fat saturation scheme: The time necessary for a spectral saturation pulse followed by a gradient spoiler will add up to the slice-loop time for an otherwise short TR GRE sequence. A feasible modification is to skip the fat saturation for a few slices—this is referred to as “quick fat sat.” Although a slice-dependent recovery of the fat signal is observed, the compromise is in general acceptable, since the fat signal stays low and more slices can be measured per TR. Composite Spectral Saturation Pulses: to Suppress Signal from Adipose Tissue, or to Excite just Water The quality of a spectral saturation pulse depends on the overall homogeneity within the imaging volume. In addition, the spectral fat saturating RF pulse is very close to the water resonance, causing a loss in overall SNR. For a nonselective excitation, it is theoretically possible to also simply excite either fat or water using the tissue specific Larmor frequency. In practice, such an approach is very prone to artifacts due to imperfect field homogeneity within the volume of interest. Better results in water excitation or fat excitation have been achieved with binomial pulses (1-1, 1-2-1, or 1-3-3-1). The mechanism of e.g., a 1-2-1 RF pulse is described as follows, leading finally to a 90° RF excitation pulse for just water. After an initial 22.5° RF pulse, there will be a waiting period, allowing the magnetization within fat to fall behind the magnetization of water. At the point of opposite position of the magnetizations, a 45° excitation angle will than move the magnetization within water to a 67.5° position with respect to the longitudinal direction, whereas the magnetization within fat will be flipped back to the 22.5° position. After the previously mentioned waiting period, another 22.5° excitation pulse will accomplish the 90° excitation for water, while the magnetization of fat will be restored to the longitudinal position, not contributing to the MR signal. Another advantage of these binomial pulses is that they can be either executed using nonselective RF pulses or selective RF pulses. In the latter case, they are called spatial spectral frequency or simply composite pulses. Inversion Pulse prior to the Measurement to Improve T1 Contrast Short TR, short TE GRE imaging, utilizing the Ernst angle leads to PD-weighted images rather than T1-weighted images. In SE imaging, the T1 contrast is improved by placing an inversion pulse prior to the acquisition of the

Fig. 2.4.47  Structure of the TFL sequence. After an inversion pulse and an inversion time, the small-angle excitation is repeated several times until the raw data matrix is filled

Fourier line. This approach is not feasible in GRE imaging, since the inversion time would be much larger than the commonly used repetition times. In fast GRE imaging, an inversion pulse is used prior to the whole imaging sequence (Fig. 2.4.47). That concept has been introduced as turboFLASH (snapshotFLASH (Haase et al. 1989), fast SPGR (FSPGR) or turbo Field Echo (TFE)). The minor drawback is that the longitudinal magnetization, and consequently the generated transverse magnetization, will change throughout the measurement. The resulting violation of k-space symmetry will cause an under- or over-representation of some spatial frequencies producing a slight image blurring, typical for turboFLASH imaging. When using this method, one has to consider the following three facts: 1 The longitudinal component of the macroscopic magnetization will recover after the inversion pulse with the T1 relaxation time. This relaxation process also takes place during data acquisition. The various measured Fourier lines will have different T1 weightings. The image contrast is dominated by the T1 weighting of the Fourier line measured at the center of k-space. 2 The recovery of the longitudinal magnetization is influenced by the excitation angles of the rapid GRE data

2.4  Image Contrasts and Imaging Sequences

acquisition. In order to minimize this influence and to obtain a maximum effect of the preparation-pulse on the image contrast, the rapid GRE acquisition needs to be executed with small excitation angles. 3 Every Fourier line is measured with a different phase encoding gradient and contains the spatial information of the object in direction of phase encoding. The k-space symmetry is significantly violated due to the change in signal contribution for each spatial frequency measured as a consequence of T1 relaxation during sampling. As a result, the images appear to be blurred, with imprecise edges and coarse signal oscillations parallel to the edges. With the introduction of short TR gradient-echo acquisition schemes, 3D imaging became feasible. The application of an inversion pulse prior to a 3D acquisition scheme is not very promising, since the preparation of the longitudinal magnetization would vastly diminish during the relatively long measurement time and the significant number of low angle excitation pulses. A feasible alternative is to repeat the preparation of the longitudinal magnetization in either the partition-encoding loop, or the phase-encoding loop. Although the timesavings would be larger for placing the inversion pulse prior to the longer phase encoding loop, fortunately at the time it was only possible to place the inversion pulse prior to the partition-encoding loop. Fortunately, because the previously described turboFLASH-artifact based on the overand under-representation of k-space lines is now omitted. The phase-encoding gradient is prepared; the inversion pulse is set followed by a rapid execution of the partitionencoding loop, during which the amount of longitudinal magnetization will change according to the course of the T1 relaxation (recovery influenced by the low-angle excitation pulses). After this, the next phase-encoding line is prepared, the inversion pulse set, and again the whole partition loop executed. The amount of signal within each partition is identical for all phase-encoding steps and the k-space is again symmetric, resulting in artifact-free images. This technique has been introduced as magnetization prepared rapid acquired gradient-echoes (MP-RAGE, Mulger and Brookeman 1990) (Fig. 2.4.48). Figure 2.4.49 shows as a typical application of the MPRAGE sequence showing the medial–sagittal T1-weighted slice out of 64 of an examination covering the entire skull in less than 5 min. The sequence had some promise to replace the conventional T1-weighted spin-echo imaging of the brain, since it allows the gapless coverage of the whole brain in less than 6 min. But, it is a gradient-echo sequence. Susceptibility gradients especially at the base of the skull will cause geometric distorted representation of the anatomy or even signal voids. Another disturbing effect is the appearance of contrast enhancement in active lesions. Due to the commonly “squishy” content of lesions, the appear-

Fig. 2.4.48  Structure of the MP-RAGE sequence. Similar to the TFL sequence there is an inversion of the longitudinal magnetization prior to the partition-encoding loop

Fig. 2.4.49  Sagittal T1-weighted-head study, acquired with a MP RAGE sequence (12/5/15°, 64 slices, partition thickness = 2 mm, no slice gaps, scan time = 5 min)

71

72

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

ance is usually iso- to hypointense in T1-weighted imaging. MP-RAGE allows better control over T1 weighting, potentially causing the lesion to be more hypointense as compared with SE imaging. In conjunction with contrast uptake, lesions will show up hyperintense on T1-weighted SE imaging. They may or may not show up hyperintense on MP-RAGE imaging. The appearance has been reported to be inconsistent, likely to be due to the better T1weighting (a hypointense lesion may show up isointense after contrast uptake). Inversion Pulse prior to the Measurement for Tissue-Selective Signal Suppression The rapid acquisition of a gradient-echo or steady state sequence following an inversion is sometimes referred to as single-shot technique. This is not quite correct since as many low angle excitation pulses are applied as Fourier lines are needed to fill the k-space. But the singleshot nomenclature allows a differentiation compared to the segmented, or multi-phase, imaging of the beating heart. Similar to the STIR approach used in fat signal– suppressed imaging, the inversion pulse prior to a singleshot technique enables the nulling of signal for a specific tissue (depending on the T1 relaxation time). The turboFLASH technique is used to study the firstpass of a contrast bolus through the cardiac chamber, showing a delayed enhancement in perfusion restricted ischemic myocardium. The inversion time is adjusted, so that normal myocardium will give no signal. In the early phase normal myocardium will be perfused with the T1-shortening contrast agent, whereas the perfusion restricted ischemic myocardium will remain hypointense. The same method of tissue signal nulling can be applied to the trueFISP. This technique has been used to demonstrate the late enhancement of infarcted myocardium. An advantage for the trueFISP versus turboFLASH is that the additional signal contributions due to refocusing and balancing (spin-echo components), allowing a higher bandwidth acquisition correlated with a shorter TE, a shorter TR, and therefore a shorter measurement time (~450 ms). In addition, the trueFISP has a significant lower sensitivity to flow and motion artifacts as compared to the turboFLASH, leading to (almost) artifact-free images. Both methods are currently evaluated regarding their value in characterizing myocardial viability. 2.4.6.4  Single-Shot Gradient-echo Imaging The single-shot gradient-echo imaging is echo planar imaging (EPI). Similar to TSE imaging, EPI makes use of several phase-encoded echoes to fill the raw data matrix (Fig. 2.4.50). There are multiple ways to acquire the data. A single excitation can be utilized, followed by multiple phase–encoded gradient-echoes with a small, constant

RF

Fig. 2.4.50  Classical FID-EPI sequence. After an excitation pulse, multiple GREs are generated using an oscillating frequency–encoding gradient. In this example the phase encoding is achieved with a low-amplitude, constant phase-encoding gradient throughout the measurement

RF

Fig. 2.4.51  SE-EPI sequence. After an excitation and refocusing-pulse, multiple GREs are generated using an oscillating frequency–encoding gradient. The phase encoding is achieved in this sequence using small gradient pulses (blips) during ramping of the frequency-encoding gradient

phase-encoding gradient activated during readout period. Such a technique would be called FID-EPI, since signal sampling is done during free induction decay. Another variant is placing the gradient-echoes under an SE envelope (Fig. 2.4.51). In this case the central k-space contains a T2 contrast, in opposition to the T2* contrast of the FID-EPI version. The SE-EPI sequence shows a lower sensitivity for susceptibility gradients. Fig. 2.4.51 shows a SE-EPI with an alternative approach to phase encoding. In this example, phase encoding is done using gradient “blips” during the ramping time of the frequency-encoding gradient. Such a technique is called blipped EPI. Figure 2.4.52 shows a blipped EPI T2-weighted image of the head.

2.4  Image Contrasts and Imaging Sequences

in any phase encoding, and sometimes it is a byproduct of another desired functionality, e.g. the frequency encoding. To rephase or refocus the dephased transverse magnetization, a magnetic field gradient of opposite polarity can be used prior to the frequency-encoding gradient. But, this will only work if the transverse magnetization does not change the position in the meantime, as is the case for diffusion. If the transverse magnetization changed positions, then the phase history will be different as compared with stationary tissue at that new location, and the rephasing will be insufficient. Insufficient rephasing will result in a reduced signal. The signal drop is characterized by S ≈ e–b D,

Fig. 2.4.52  Axial T2-weighted-head-image, acquired with an SE-EPI sequence (TE 70, measurement time per slice = 180 ms, slice thickness = 3 mm). There is only one excitation per slice. No repeated excitations are necessary. All data are acquired using multiple phase–encoded gradient echoes

Addressing a different way of k-space sampling, both, the frequency-encoding” gradient and the “phase-encoding” gradient may oscillate, causing a spiral trajectory through k-space. Such a method is known as spiral EPI. The quotation marks are used to indicate that the magnetic field gradients do no longer have the apparent meaning of frequency and phase encoding. The high sensitivity of EPI to local field inhomogeneities is utilized in (brain) perfusion imaging and for monitoring the oxygen level to identify cortical activation in BOLD (blood oxygenation level–dependent) imaging. In spite of many limitations, the EPI sequences have attained high clinical potential in functional imaging and in perfusion studies.

(2.4.13)

with b being a system or method specific parameter, and D being the diffusion coefficient for the tissue. The method specific parameter b b = γ 2 · G 2 · δ2 · (Δ – δ / 3)

(2.4.14)

is a function of the gradient amplitude G used, the duration δ for each amplitude and the temporal distance ∆ between the two gradient amplitudes. ∆ is also called the diffusion time. A typical value for b = 1,000 s/mm2. A sequence illustration is given in Fig. 2.4.53. The result of the application of such a technique to a patient with an acute infarction is shown in Fig. 2.4.54. Diffusion-weighted imaging allows an evaluation of the extent of cerebral ischemia in a period where possible interventions could limit or prevent further brain injury. The diffusion anisotropy potentially measured with this method allows the mapping of neuronal connectivity and offers an exciting perspective to brain research.

RF

2.4.6.5  Single-Shot Gradient-echo Imaging with Preparation of Magnetization (Diffusion-Weighted Imaging) The preparation of the longitudinal magnetization is not only possible with the previously described multi-shot techniques, but also with the single-shot version. Single shot, per definition, means one excitation pulse and multiple phase–encoded gradient-echoes for sampling of all Fourier lines. The primary preparation scheme for singleshot gradient-echo imaging is diffusion weighting. Any magnetic field gradient in the presence of a transverse magnetization will cause the Larmor frequency to be a function of location. Sometimes this effect is desired, as

Fig. 2.4.53  Diffusion-weighted SE-EPI sequence. Diffusion weighting is provided using high gradient pulses placed symmetrically around the RF refocusing pulse. This method of diffusion weighting is called Stejskal-Tanner approach

73

74

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.4.54  The left image shows an axial head image acquired with an EPI sequence without diffusion weighting. The right image shows the result of the same technique with diffusion weighting. The infarcted region is only visible with diffusion weighting. (This example was provided by Dr. R. Edelmann, Beth Israel Hospital, Boston, Mass.)

2.4.6.6  Hybrid Techniques The turbo gradient spin-echo sequence (TGSE), also called gradient and spin-echo (GRASE) is a combination of multiple gradient-echoes that are acquired within multiple SE envelopes of a TSE sequence, as shown in Fig. 2.4.55. This method holds several advantages in comparison to the “simple” TSE sequence: The use of several phase-encoded gradient-echoes has the potential of further shortening measurement time. Figure 2.4.56 shows a transversal T2-weighted head image with a matrix size of 1,024, which has been measured with a TGSE sequence in 2.56 min. Another advantage is the fact that with the use of several gradient-echoes per spin-echo envelope, the gap between the refocusing pulses widens. Therefore, the J-coupling remains intact. Fat appears darker, and the contrast approaches the contrast of conventional SE sequences. Further, the enhanced sensitivity toward the susceptibility gradients that has been introduced with the gradient-echoes allows for a better depiction of blooddecay products, similar to conventional SE-imaging.

Fig. 2.4.56  Axial T2-weighted head image acquired with a TGSE sequence (matrix 1024 × 1024, TGSE 7,000 /115, slice thickness = 3 mm, measurement time = 2.56 min)

References 1.

2.

3.

4.

5.

Bottomley PA, Hardy CJ, Argersinger RE et al (1987) A review of 1H nuclear magnetic resonance relaxation in pathology: are T1 and T2 diagnostic? Med Phys 14:1–37 Brix G, Schad LR, Lorenz WJ (1990) Evaluation of proton density by magnetic resonance imaging: phantom experiments and analysis of multiple component proton transverse relaxation. Phys Med Biol 35:53–66 Feinberg DA, Hale JD, Watts JC et al (1986) Halving MR imaging time by conjugation: demonstration at 3.5 kG. Radiology 161:527–531 Frahm J, Haase A, Hänicke W, Matthaei D, Bomsdorf H, Helzel T (1985) chemical shift selective MR imaging using a whole-body magnet. Radiology 156:441–444 Haacke EM, Wielopolski PA, Tkach JA (1991) A comprehensive technical review of short TR, fast, magnetic resonance imaging. Rev Magn Res Med 3:53

RF

Fig. 2.4.55  Structure of the TGSE sequence. Several phase encoded GRE are produced within the spin-echo envelope. (In this example, three gradient-echoes are acquired within a spin-echo envelope)

2.4  Image Contrasts and Imaging Sequences 6.

7.

8.

9.

10.

11. 12.

13.

14.

15.

16.

Haase A (1990) Snapshot FLASH MRI. Application to T1, T2, and chemical-shift imaging. Magn Reson Med 13:77–89 Haase A, Frahm J, Hänicke W, Matthaei D (1985) 1H-NMR chemical shift selective (CHESS) imaging. Phys Med Biol 30:341–344 Haase A, Frahm J, Matthaei D et al (1986) FLASH imaging: rapid NMR imaging using low flip angle pulses. J Magn Reson 67:258–266 Haase A, Matthaei W, Bartkowski R, Duhmke E, Leibfritz D (1989) Inversion recovery snapshot FLASH MR imaging. J Comput Assist Tomogr 13:1036 Henning J, Nauerth A, Friedburg H (1986) RARE-imaging: a fast imaging method for clinical MR. Magn Reson Med 3:823–833 Mansfield P, Mosley AA, Baines T (1976) Fast scan proton density imaging by NMR. J Phys 9:271–278 Mugler JP, Brookeman JR (1990) Three-dimensional magnetization-prepared rapid gradientecho imaging (3D MP RAGE). Magn Res Med 15:152 Mulkern RV, Wong STS, Winalski C, Jolesz FA (1990) Contrast manipulation and artifact assessment of 2D and 3D RARE sequences. Magn Reson Imag 8:557–566 Oppelt A, Graumann R, Barfuß H, Fischer H, Hartl W, Schajor W (1986) FISP: eine neue schnelle Pulssequenz für die Kernspintomographie. Electromedica 54:15–18 Pfannenstiel P, Just M, Higer HP et al (1987) Erste klinische Ergebnisse der Gewebecharakterisierung durch T1, T2 und Protonendichte bei der Kernspintomographie. RoFo 146:591–596 Wolff S, Balaban R (1989) Magnetization transfer via cross relaxation. Magn Reson Med 10:135–144

Suggested Reading 1.

Bruder H, Fischer H, Graumann R, Deimling M (1988) A new steady-state sequence for simultaneous acquisition of two MR images with clearly different contrast. Magn Reson Med 7:35 2. Haacke EM, Brown RW, Thompson MR, Venkatesan R (1999) Magnetic resonance imaging: physical principles and sequence design. Wiley, New York 3. Heidemann RM, Özsarlak Ö, Parizel PM et al (2003) A brief review of parallel magnetic resonance imaging. Eur Radiology 13:2323–2337 4. Higer HP, Bielke G (1986) Gewebecharakterisierung mit T1, T2 und Protonendichte: Traum und Wirklichkeit. ROFO 144/5:597–605 5. Kiefer B, Grässner J, Hausmann R (1994) Image acquisition in a second with half-Fourier acquired single shot turbo spin echo. JMRI 4:86 6. Mansfield P, Pykett IL (1978) Biological and medical imaging by NMR. J Magn Reson 29:355–373 [see also German patent no. 2755956 C2 from 15 December 1977, Government Patent Office, Germany) 7. Margosian P, Schmitt F, Purdy D (1986) Faster MR imaging: imaging with half the data. Health Care Instrum 1:195 8. Oppelt A (ed) (2005) Imaging systems for medical diagnostics. Publicis MCD, Erlangen 9. Oshio K, Feinberg DA (1991) Magn Reson Med 20:344 10. Pfannenstiel P, Just M, Higer HP et al. (1987) Erste klinische Ergebnisse der Gewebecharakterisierung durch T1, T2 und Protonendichte bei der Kernspintomographie [in German]. RoFo 146/5:591–596 11. Vlaardingerbroek MT, den Boer JA (2004) Magnetic resonance imaging: theory and practice. Springer, Berlin Heidelberg New York

75

76

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.5  Technical Components M. Bock Clinical MR systems come in various types and shapes; however, the fundamental components of a clinical MR tomograph are essentially the same. These are: • The magnet: The magnet creates a static and homogeneous magnetic field B0, which is needed to establish a longitudinal magnetization. • The gradients: The gradient coils generate additional linearly ascending magnetic fields that can be switched on and off. The gradient fields allow assigning a spatial location to the received MR signals (spatial en­co­ding). For an image acquisition three independent gradient systems in x-, y-, and z-direction are required. • The radio frequency (RF) system: To rotate the longitudinal magnetization from its equilibrium orientation along B0 into the transverse plane, an oscillating magnetic field B1 is required. This RF field is generated by a transmitter and coupled into the patient via an antenna, the RF coil. Radio frequency coils are also used to receive the weak induced MR signals from the patient, which are then amplified and digitized. • The computer system: Measurement setup and image post-processing are performed by (distributed) com­ pu­ters that are controlled by a host computer. At this host computer, new measurements are planned and started and the reconstructed images are stored and analyzed. A schematic of the components of a clinical MR system is shown in Fig. 2.5.1; more detailed descriptions can be found in the works of Oppelt (2005), Vlaardingerbroek et al. (2002), and Chen and Hoult (1989). In the upper part, a cross-section through a superconducting magnet can be seen, with field-generating magnet windings embedded in a cryostatic tank. Closer to the patient, the gradient coil and the whole-body RF coil are located outside the cryostat. The magnet is surrounded by an electrically conducting cabin (Faraday cage) which is needed to optimally detect the weak MR signals, without RF background from other RF sources (e.g., radio transmitters). In the lower part, the computing architecture and the hardware control cabinets are shown. A hardware computer controls the gradient amplifier, the RF transmitter, and the receiver. The received and digitized MR signals are passed on to an image-reconstruction computer, which finally transfers the reconstructed image data sets to the host computer for display and storage. 2.5.1  Magnet To generate the main magnetic field three different types of magnets can be utilized: permanent magnets, resistive

magnets, and superconducting magnets (Oppelt 2005; Vlaardingerbroek 2002; Chen 1989). The choice of an individual magnet type is determined by the re­quire­ments on the magnetic field. Important characteristics are the field strength B0, spatial field homogeneity, temporal field stability, patient accessibility, as well as construction and servicing costs. As outlined in Sect. 2.2 a high magnetic induction is desirable as the MR signal S is approximately pro­ portional to B0², and the signal-to-noise ratio (SNR) increases approximately linearly with B0. It is thus expected that with increasing field strength the measurement time can be substantially decreased. The field strength is limited however for the following reasons: • For tissues in typical magnetic fields of 0.5 T and higher, the longitudinal relaxation time T1 increases with field strength. If the same pulse sequence with identical measurement parameters (TR, TE, etc.) would be used at low field and high field, the T1 contrast would be less pronounced in the high-field image, since image contrast typically depends on the ratio of TR over T1. To achieve a similar T1 contrast with a conventional SE or GRE pulse sequence, TR (and thus the total measurement time) needs to be increased. • The resonance frequency ω0 increases linearly with field strength according to ω0 = γ B0. At higher frequencies the wavelength of the RF waves are of the order of or even smaller than the dimensions of the objects to be imaged. Under these circumstances, standing waves can be created in the human body, which manifests in areas of higher RF fields (hot spots) and neighboring areas of reduced RF intensity. These unwanted RF inhomogeneities are difficult to control, as they are dependent on the geometry and the electric properties of the imaged object. • The power that is deposited in the tissue during RF excitation rises quadratically with ω0 (and thus with B0). To ensure patient safety at all times during the imaging procedure the specific absorption rate (SAR), i.e., the amount of RF power deposited per kilogram of body weight, is monitored and limited by the MR system. With increasing field strength, the RF power generated by a pulse sequence increases, and thus the flip angle needs to be lowered to stay within the guidelines of SAR monitoring. Since most pulse sequences require certain flip angles (e.g., a 90–180° pulse pair for an SE), the RF pulses need to be lengthened at higher field strength to reduce the RF power per pulse. Additionally, the time-averaged power can be lowered by increasing the TR. • At field strengths above 1 T, only superconducting magnets are used for whole-body imaging systems. These magnets become very heavy and expensive. A typical 1.5-T MR magnet weighs about 6 t, whereas a 3-T magnet already has a weight of about 12 t. Shielding of the stray fields, which is, e.g., necessary to avoid

2.5  Technical Components

Fig. 2.5.1  Schematic of a clinical MR system. The main magnet is located in an RF cabin to suppress external RF signals. The superconducting, actively shielded solenoid magnet shown here consists of a cryotank (1) that is filled with liquid helium. The cryotank also houses the primary magnet coils (4) together with the shielding coils (2) that create the magnetic field. The cryotank is embedded in a vacuum tank (3). In a separate tubular structure in the magnet bore, the gradient coil (5) and the RF body coil (6) are mounted. An MR measurement is initiated by

the user from the host computer. The timing of the sequence is monitored by the hardware computer, which controls (among others) the RF transmitter, RF receiver, and the gradient system. During the measurement, the RF pulses generated by the transmitter are applied (typically) via the integrated body coil, whereas signal reception is done with multiple receive coils. The digitized MR signals are reconstructed at the image-reconstruction computer, which finally sends the image data to the host for further post-processing and storage

77

78

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

interference with cardiac pacemakers, becomes increasingly difficult. • The absolute differences in resonance frequency between chemical substances increase with field strength. This effect is beneficial for high-resolution spectroscopy, as high field strengths allow separating the indi­vi­dual resonance lines. During imaging, however, a substantially increased chemical shift artifact (i.e., a geo­metric shift of the fatty tissues versus the water-containing tissues) is seen, which can only be com­pensated using higher readout bandwidths. • Differences in magnetic susceptibility between neighboring tissues create a static field gradient at the tissue boundaries. The strength of these unwanted intrinsic field gradients scales with B0. Therefore, increasingly higher imaging gradients are required at higher field strength to encode the imaging signal without geometric distortion; however, gradient strengths are technically limited. On the other hand, in neurofunctional MRI (fMRI) the increased sensitivity at higher field strengths is utilized to visualize those brain areas where local susceptibility differences in the blood are modulated during task performance. In the clinical environment, magnetic field strengths between 0.2 and 3 T are common. Low-field MR systems (B0 < 1 T) are often used for orthopedic or interventional MRI, where the access to the patient during the imaging procedure is important. High field strengths between 1 T and 3 T are used for all other diagnostic imaging applications. Recently, whole body MR systems with field strengths up to 9.4 T have been realized (Robitaille et al. 1999). With these systems, in particular neurofunctional studies, high-resolution imaging and spectroscopy as well as non-proton imaging (e.g., for molecular imaging) are planned, since these applications are expected to profit most from the high static magnetic field. In the following, the three different types of magnet are described that are used to create the static magnetic field. 2.5.1.1  Permanent Magnets Permanent MRI magnets are typically constructed of the magnetic material NdBFe. Permanent magnet materials are characterized by the hysteresis curve, which describes the non-linear response of the material to an external magnetic field. If an external field is slowly increased, the magnetization of the material will also increase until all magnetic domains in the material are aligned—at this point the magnet is saturated, and no further amplification of the external field is possible. If the external field is then switched off a constant, non-vanishing magnetic field remains in the material because some of the

Fig. 2.5.2  C-arc–type permanent magnet 0.35-T MR system (Magnetom C!, Siemens) with a weight of 16 t. The vertical magnetic field is created by two pole shoes that have a free opening of 38.5 cm for imaging

domains remain aligned. Permanent magnets offer very high remanence field strength. Permanent magnets require nearly no maintenance because they provide the magnetic field directly without any electrical components. Permanent magnets often use a design with two poles, which are either above or below (Fig. 2.5.2) or at the sides of the imaging volume. Within this volume the magnetic field lines should be as parallel as possible (high field homogeneity), which is achieved by shaping of the pole shoes. Due to their construction, the magnetic field is typically orthogonal to the patient axis, whereas high-field superconductors use solenoid magnets with a parallel field orientation. Magnetic field lines are always closed; therefore, an iron yoke is used in permanent magnets to guide the magnetic flux between the pole shoes. With increasing field strength permanent magnets become very heavy (10 t and more), and the high price of the material NdBFe becomes a limiting factor. Additionally, to achieve high temporal field stability the material requires a constant room temperature, which should not vary by more than 1 K. For these reasons, permanent magnets are typically used only for field strengths below 0.3 T. 2.5.1.2  Resistive Magnets If an electrical current is flowing through a conductor, a magnetic field is created perpendicular to the flow direc-

2.5  Technical Components

tion that is proportional to the current amplitude. Unfortunately, in conventional conductors (e.g., copper wire) the electric resistance converts most of the electric energy into unwanted thermal energy and not into a magnetic field. Therefore, a permanent current supply is required to maintain the magnetic field and to compensate for the Ohmic losses in the wire. Additionally, to dissipate the thermal energy resistive magnets need permanent water cooling as their power consumption reaches several 100 kW. Resistive magnets use iron yokes to amplify and guide the magnetic field created by the electric currents. The iron yoke is surrounded by the current-bearing wires so that the field lines stay within the iron. In the simplest form, the closed iron yoke has a gap at the imaging location and the magnet takes the form of a C-arc, which can also be rotated by 90° to provide a good access for the patient (Fig. 2.5.3). Other magnet designs use two or four iron posts that connect the pole shoes. The magnetic field of a resistive magnet is typically not as homogeneous as that of a superconducting magnet of the same size. To achieve high field homogeneity within the imaging field-of-view, the diameter of the pole shoes should not be less than 2.5 times the desired diameter of the imaging volume (DFOV), and the pole separation should be more than 1.5 DFOV. At a typical pole separation of 45 cm, the imaging volume would thus have a diam­eter of 30 cm, and the pole shoe diameter amounts to 75 cm. Resistive magnets are susceptible to field variations caused by instabilities of the electric power supply. To minimize this effect the magnetic field of the magnet can be stabilized using an independent method to measure the field strength (e.g., electron spin resonance). The difference between actual and desired field strength is then used to regulate the current in the magnet in a closed feedback loop. 2.5.1.3  Superconducting Magnets To create magnetic fields of more than 0.3 T with a bore size of 60 cm or more, today, typically superconducting magnets are utilized (Fig. 2.5.4). In principle, these magnets operate in a similar fashion as resistive magnets without iron yoke—superconducting magnets also generate their magnetic field by wire loops that carry a current. Instead of copper wire, superconductors use special metallic alloys such as niobium–titanium (NbTi). The alloys completely lose their electric resistance below a certain transition temperature that is characteristic for the material; this effect is called superconductivity. The transition temperature itself is a function of the magnetic field, so that lower temperatures are required when a current is flowing through the wire. Unfortunately, an upper

Fig. 2.5.3  Iron-frame electromagnet 0.6-T MR system (UprightTM MRI, FONAR) with a horizontal magnetic field. This special construction of the MR systems allows for imaging in both upright and lying positions. This flexibility is especially advantageous for MR imaging of the musculoskeletal system

limit for the current density in the wire exists, which is also a function of the temperature and the magnetic field. To maintain the required low temperatures cooling with liquid helium is typically necessary (T < –270°C). The imaging volume of the MR system is typically kept at room temperature (T = 20°C), whereas the sur­ rounding super­conducting wires require temperatures near the absolute zero (–273.15°C). To maintain this enormous temperature gra­dient, the field-generating superconducting coils are encased in an isolating tank, the cryostat. The cryostat is a non-magnetic steel structure that contains radiation shields to prevent heat diffusion, heat conduction, and heat transport. If this isolation is not working properly and the wire is locally warming up over the transition temperature, then this section of the wire will become normally conducting, and the energy stored in the current will be dissipated as heat. The heat will then be transported to adjacent sections of the wire, which will also lose their superconductivity. This very rapid process is called a quench. When the magnet wire is heating up the liquid helium will evaporate, and the cryo-

79

80

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.5.4  Conventional 1.5-T superconducting MR magnet (Magnetom Symphony, Siemens) with a magnet length of 160 cm and a free open–inner bore diameter of 60 cm. The system is equipped with an in-room monitor (left), which allows controlling the MR system from within the RF cabin

stat is exposed to an enormous pressure. To prevent the cryostat from ex­ploding, a so-called quench tube is connected to superconducting magnets with helium cooling, which safely guides the cold helium vapor out of the magnet room. Recently also ceramic superconducting materials on the basis of niobium–tin (Nb3Sn) alloys have been used to make superconducting wires. The brittle Nb3Sn alloys show a higher transition temperature (–263°C) and thus do not necessarily require liquid helium cooling. If the cryostat is equipped with a good thermal vacuum isolation, a conventional cooling system (e.g., Gifford-McMahon cooler) can be used to maintain the temperature. This technology has been realized both in a dual-magnet system (General Electric SP, B0 = 0.5 T) and a low-field open MR system (Toshiba OPART, B0 = 0.35 T). Because a helium-filled cryostat requires more space than does a system without helium, these magnets can be installed in smaller areas than can comparable magnets with helium. In the recent years, several MR magnets have been equipped with helium liquefiers to regain the evaporated helium gas in the magnet. Once filled with helium these so-called zero boil-off magnets can operate in principle without any additional helium filling. Magnets without helium liquefiers require replenishment of the helium at intervals between several months and 1–2 years, depending on the quality of the cryostat and the usage of the MR system. The most widespread form of a superconducting magnet is the solenoid, where the windings of the super­ conducting wire form loops around the horizontal bore of the cylindrical magnet. At a typical inner bore diameter of 60 cm for clinical MRI systems, solenoid magnets can create very homogeneous magnetic fields with varia-

tions of only a few parts per million (ppm). Because the relatively bulky magnet structure limits access to the patient, shorter magnets of 1.5 m length with wider diameters of 70 cm have been designed (Siemens Magnetom Espree, B0 = 1.5 T) (Fig. 2.5.5). In these magnets, obese patients can be imaged more conveniently, claustrophobic patients feel more at ease and some MR-guided percutaneous interventions might become feasible. Another variant of the solenoid is a dual magnet MR system consisting of two collinear short solenoid magnets (General Electric SP, B0 = 0.5 T) — here the imaging area is located in between the two magnets and even intra-operative MR imaging is possible. Recently, also two-pole systems with a magnet design similar to low-field resistive magnets have become, which offer a good patient access in combination with higher field strengths (Fig. 2.5.6). Outside a superconducting magnet, the field strength his falling off with the inverse third power of the distance (1/r ³) so that the stray fields can extend far outside the MR room. Magnetic fields in commonly accessible areas must not exceed 0.5 mT, because higher fields can affect pace makers and other active electric devices (Fig. 2.5.7). For this reason, two shielding technologies have been utilized to reduce the magnetic fringe fields. With passive shielding ferromagnetic materials such as steel are mounted near the magnet. This shielding technique confines the field lines to the interior of the shielding material, and the stray fields are reduced. Unfortunately, the amount of shielding material rapidly increases with increasing magnetic field, and between 400 t and 600 t of steel are required to shield a 7-T magnet (Schmitt et al. 1998). With active shielding a second set of wire loops is integrated in the cryostat of the magnet. The shielding

2.5  Technical Components Fig. 2.5.5  Clinical 1.5-T superconducting MR system (Magnetom Espree, Siemens) with a very short magnet (125-cm length) and an open-bore diameter of 70 cm. The additional 10 cm in bore diameter over conventional MR system with solenoid magnets and the shorter magnet length offer a better access to the patient, so that, e.g., percutaneous interventions can be performed in this magnet structure

Fig. 2.5.6  Superconducting 1-T MR system (Panorama 1 T, Philips) with a two-pole open configuration and a vertical magnetic field

coils create a magnetic field in the opposite direction of the imaging field so that the stray field falls off more rapidly. The shielding coils have a larger diameter than do the field-generating primary coils. Thus, the desired magnetic field within the magnet can be maintained by increasing the current in both coil systems. Additionally, the shielding coils and the primary coils repel each other

(Lorentz forces), which requires a magnet design with more stable coil formers. The attractive forces acting on paramagnetic or ferromagnetic objects near such an actively shielded magnet are significantly higher than near an unshielded magnet; device compatibility and safety should thus always be specified with regard to the investigated magnet type.

81

82

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.5.7  Magnetic field line plots for an actively shielded 1.5-T magnet (Magnetom Symphony, Siemens). In the horizontal plane (top) the 0.5-mT (5 Gs) line stays within the RF cabin or the adjacent walls, so that access needs to be restricted

to the RF cabin (controlled area) only. In vertical direction (bottom) the 0.5-mT line extends partly into the lower floor (here: the basement), and warning signs need to be mounted to avoid involuntary exposure to the static magnetic field

2.5  Technical Components

To localize the MR signals emitted by the imaging object, a linearly increasing magnetic field, the gradient G, is superimposed on the static magnetic field B0. The gradient fields are created by gradient coils that are located between the magnet and the imaging volume (Schmitt et al. 1998). For each spatial direction (x, y, and z) a separate gradient coil is required, and angulated gradient fields are realized by linear superposition of the physical gradient fields. In a cylindrical bore superconducting magnet, the gradient coils are mounted on a cylindrical structure, which is often made of epoxy resin. This gradient tube reduces the available space in the cryostat from typically 90 cm, without gradient coils, to 60 cm, with gradient coils. The functional principle of a gradient system is best illustrated by a setup of two coaxial wire loops with a radius a that are separated by a distance d (Fig. 2.5.8). If the two coils both carry the same current, however, in counterpropagating directions, their respective magnetic fields cancel at the iso-center of the setup. At distances not too far from iso-center the magnetic field will increase linearly, which is exactly the desired behavior of a gradient field. To achieve this linear gradient field the condition d = 3a must be met (Maxwell coil pair) (Jin 1999). In commercially available gradient system, much more complicated wiring paths are utilized, which are optimized using the so-called target field approach (Turner 1993). This often results in wire patterns that, when plotted on a sheet of paper, resemble fingerprints (fingerprint design). Nevertheless, a common feature of all gradient systems is the absence of current at the cen-

tral plane, which allows separating the gradient coils, e.g., for C-arc–type magnets. The quality of the gradient system is characterized by several parameters: the maximum gradient strength Gmax, the slew rate smax, the homogeneity, the duty cycle, the type of shielding, and gradient pulse stability and pre­ cision. Today, clinical MR systems have maximum gradient strengths of up to Gmax = 40 mT/m at bore diameters of 60 cm. Even higher gradient strengths of 80 mT/m and more can be realized when so-called gradient inserts with smaller diameters are used (e.g., for head imaging). The maximum gradient strength is limited by the capa­bi­li­ties of the power supply of the gradient system—modern gradient systems use power supplies that can deliver voltages up to 2,000 V and currents up to 500 A. Another limiting factor for Gmax is gradient heating: With in­crea­sing current through the gradient coil, the windings heat up to levels at which the gradient could be destroyed. Therefore, to remove the heat from the gradient tube, pipes are integrated in the gradient coils for water-cooling. The maximum slew rate smax is the ratio of Gmax over the shortest time required to switch on the gradient (rise time). When the current in the gradient coil is increased during gradient switching, according to Lenz’s law the coil will produce a current, which opposes the change. Thus, it counteracts the switching process, and the rise time cannot be made infinitely short. During MR imaging, however, it is desirable to have very short rise times (i.e., high slew rates), as these times only prolong the imaging process. Clinical MRI systems have slew rates between 10 mT/m/ms and 200 mT/m/ms. If the gradient coil is connected to a capacitance via a fast switch,

Fig. 2.5.8  Magnetic field of two wire loops (radius a, distance d) that are carrying a current I with opposite polarity. Over a certain range between the loops, the magnetic field increases lin-

early with the position z (dashed line). With realistic values of a = 30 cm, d = 52 cm, and I = 200 A, a gradient amplitude of 18 mT/m is achieved

2.5.2  Gradients

83

84

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

very short rise times can be achieved, as the inductance of the gradient coil and the capacitance form a resonance circuit. Such a resonant gradient system has the disadvantage that the characteristic frequency of the resonance circuit determines the possible rise times. Additionally, gradients can only be switched on after the capacitances have been charged. Resonant gradient systems have nevertheless been successfully applied to EPI studies, in which the sinusoidal gradient waveforms are beneficial, and multistage resonant systems have been utilized to appro­ximate the trapezoidal gradient waveforms (Harvey 1994). When the gradient is switched on, the maximal rate of field change is observed at the ends of the gradient coil (i.e., FOVmax/2): dB/dt = smax FOVmax/2. A changing magnetic field induces currents in electrically conducting struc­tures in its vicinity—outside the gradient coil this structure is given by the cryostat, and on the inside the patient can act as a conductor. To avoid these parasitic currents (eddy currents) in the cryostat, which in turn create magnetic fields counteracting the gradients, often a second outer gradient structure is integrated in the gradient tube. The inner and outer gradient coils are designed so that their combined gradient field vanishes everywhere outside the gradient coil, whereas the desired gradient amplitudes are realized on the inside. This technique is called active shielding, and is conceptionally similar to the active shielding of superconducting magnets (Mansfield and Chapman 1986; Harvey 1994). Gradient-induced currents in the human body pose a more severe problem, as these currents can potentially lead to painful peripheral nerve stimulation or, at higher amplitudes, to cardiac stimulation (Mansfield and Harvey 1994; Schaefer 1998; Liu et al. 2003). These physiologic effects are not only dependent on the amplitude, but also on the frequency of field change. For clinical MR systems, different theoretical models have been established to determine the threshold for peripheral nerve stimulation. To make the best use of the available gradient system some fast pulse sequences (e.g., for contrast-enhanced MRA or EPI) operate very close to these threshold values. As individuals are more or less susceptible to peripheral nerve stimulation, for some patients the individual threshold might be exceeded, and they experience a tickling sensation during fast MR imaging. This physiologic effect currently prohibits the use of stronger gradient systems. Since the field change is lower at shorter distances from iso-center, peripheral nerve stimulation can be avoided if shorter gradient systems are used. Unfortunately, a shorter gradient system only covers a limited FOV, and the anatomical coverage is compromised. To overcome this limitation a combined gradient system with a shorter, more powerful inner coil and a longer, less intense outer coil has been proposed (twin gradients) (Harvey 1999). Such a system can be used, e.g., to rapidly image the beat-

ing heart with the small coil, or to acquire image data from the surrounding anatomy at lower frame rates. When the gradient system is mounted in the MR magnet, strong mechanical forces act on the gradient tube, which are proportional to the gradient current. These forces are generated by the interaction of the gradient field with the static magnetic field and thus increase with B0. The permanent gradient switching creates time-varying forces that lead to acoustic noise. Several techniques have been proposed to reduce noise generation, which in some cases can exceed dangerous sound pressure levels of 110 dB. The wire paths in the coil can be designed in such a way that the forces are locally balanced, the gradient tube can be mechanically stabilized, the gradients can be integrated in a vacuum chamber to prevent sound propagation in air, or the gradient system can be mounted externally to reduce acoustic coupling to the cryostat (Pianissimo gradient, Toshiba). Another possibility to reduce acoustic noise is to limit the slew rates in the pulse sequences to lower values than technically possible; in some pulse sequences (e.g., spin-echo sequences), this does not significantly affect the pulse sequence performance, but severely increase patient comfort. 2.5.3  Shim Shimming is a procedure to make the static magnetic field in the MR system as homogeneous as possible. In­ho­ mogeneities of the magnetic field that are caused during the manufacturing of the magnet structure can be compensated with small magnetic plates (passive shim). After a localized measurement of the initial magnetic field, the position of the plates is calculated, and the plates are placed in the magnet. This procedure is repeated until the desired homogeneity of the field is achieved (e.g., 0.5 ppm in a sphere of radius 15 cm). During MR imaging, objects are present in the static magnetic field that distort the homogeneous static field. Field distortion is caused by susceptibility differences at the tissue interfaces and is thus specific for each patient. To at least locally compensate these field distortions, adjustable magnetic fields are required (active shim). If the field distortion is linear in space, then the gradient coils can be used for compensation. For higher-order field variations, additional shim coils are required. Typically, shim coils up to fifth order are present in an MR system. Higher-order shimming is particularly important for MR spectroscopy, where the field homogeneity directly affects the spectral line width. To optimize the shim currents, an interactive measurement process (the shim) is started after the patient is positioned in the magnet. During active shimming the field homogeneity is measured (e.g., using localized MR spectroscopy or a field mapping technique), and the currents are then adjusted to improve the field homogeneity (Webb and Macovski 1991).

2.5  Technical Components

2.5.4  Radiofrequency System The radiofrequency (RF) system of an MR scanner is used to both create the transverse magnetization via resonant excitation and to acquire the MR signals (Oppelt 2005; Vlaardingerbroek et al. 2002; Chen and Hoult 1989). The RF system consists of a transmit chain and a receive chain. In the following, the details of the RF system are described. 2.5.4.1  RF Cabin The MR signals, which are acquired by the RF coils of the MR system, are typically very low. To optimally detect these low signals, any other electromagnetic signals (e.g., radio waves) must be suppressed. Therefore, the MR system is placed in a radiofrequency cabin (also called a Faraday cage), which dampens RF signals at the resonance frequency by typically 100 dB and more. In low-field MR tomographs, the RF screening is sometimes realized as a wire mesh that is integrated in the MR system. This has the advantage that RF-emitting equipment such as television screens can be placed very close to the MR unit. At larger magnet dimensions, these local screens are often not suitable. Here, the whole MR room is designed as an RF cabin, and the screening material is integrated into the walls, doors, and windows. For screening often copper sheets are used, which are glued to the wall panels, or the cabin consists completely of steel plates. To be able to transmit signals to and receive signals from the RF cabin, openings are integrated in the cabin. In general, one distinguishes between so-called filter plates, which contain electronic filters and open waveguides. Waveguides are realized as open tubes with a certain length-to-diameter ratio, which is dependent on the wave­length of the RF frequency. Waveguides are used to deliver anesthesia gases to the RF cabin and to guide the quench tube out of the shielded room. 2.5.4.2  Transmitter At the beginning of the transmit chain the RF transmitter is found, which consists of a synthesizer with highfrequency stability and an RF power amplifier. The lowpower synthesizer oscillates at the Larmor frequency. Its output signal is modulated by a digitally controlled pulse shaper to form the RF pulse, which is then amplified by the power amplifier. For typical clinical MR systems, the transmitter needs to provide peak power output at the Larmor frequency of 10 kW and more. Besides high peak power, the RF transmitter should also allow for a high time-averaged power output, as several pulse sequences such as fast spin-echo sequences require RF pulses at

short repetition times. The RF power is then transferred into the RF cabin via a shielded cable, and is delivered to the transmit RF coil. To guarantee a safe operation of the transmitter and to limit the RF power to values below the regulatory constraints for the specific absorption rates (SAR), directional couplers are integrated in the transmission line. These couplers measure the RF power sent to the RF coil as well as the reflected power. High power reflection is an indicator of a malfunctioning of the connected coil, which could endanger the patient. If the reflected power exceeds a given threshold (e.g., 20% of the forward power), then the RF amplifier could be damaged by the reflected RF power and the transmitter is switched off. 2.5.4.3  RF Coils To couple the RF power of the RF transmitter to the human body an RF antenna is required, the so-called RF coil. Before MR imaging starts, the coil is tuned to the resonance frequency of the MR system (RF tuning). Simul­ taneously, the properties of the connecting circuitry are dynamically changed to match the resistance of the coil with the imaging object (loaded coil) to the resistance of the transmit cable (RF matching). Once the coil is tuned and matched, the transmitter is adjusted. During this procedure, the MR system determines the transmitter voltage required to create a certain flip angle. For a given reference RF pulse shape Sref (t), the transmitter voltage Uref is varied until the desired flip angle αref (e.g., 90°) is realized. During the subsequent imaging experiments, use is made of the fact that the flip angle is linearly proportional to the (known) integral over the RF pulse shape, so that the required voltages can be computed from the reference values by linear scaling. Radiofrequency coils are categorized into transmit (Tx) coils, receive (Rx) coils, and transmit/receive (TxRx) coils. Tx coils are only used to expose the imaging object to an RF B1 field during RF excitation, whereas Rx coils detect the weak echo signal emitted from the human body—only if a coil performs both tasks, is it called a TxRx coil. A typical example of a TxRx coil is the body coil integrated into most super­con­duc­ting MR systems; however, in some modern MR systems, it is used as a Tx coil only due to its suboptimal receive characteristics. Rx-only RF coils are the typical local coils found in MR systems that possess a (global) body coil, and local TxRx coils are used in all other MR systems without a body coil (ultra-high field, dedi­cated interventional systems, openconfiguration low field). During signal reception, the oscillating magnetization in the human body induces a voltage in the RF coil. For an optimal detection of this weak signal, the RF coil should be placed as close to the imaging volume as possible. For this reason, optimized imaging coils exist for

85

86

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

nearly any part of the human body. The largest coil of an MR system is typically the body coil (if present), which is often integrated in the magnet cover. To image the head or the knee, smaller volume resonators are used, where the imaging volume is in the interior of the RF coil (Fig. 2.5.9). Flexible coils exist, that can be wrapped around the imaging volume (e.g., the shoulder). Small circular surface coils are used to image structures close to the body surface (e.g., eyes). Unfortunately, the sensitivity of these coils is rapidly decreasing with distance from the coil center, so that they are not suitable for imaging experiments, where a larger volume needs to be covered. During RF transmission, Rx coils need to be deactivated, because a tuned and matched Rx coil would ideally absorb the transmitted RF power, and a significant amount of the RF energy would be deposited in the coil. To avoid any electronic damages, the coil is actively de-

Fig. 2.5.9  Various receive coils on the patient table of a clinical 1.5-T MR system (Magnetom Avanto, Siemens) with 32 receive channels and the possibility to connect a total of 76 coil elements. The head coil with 12 coil elements is com­bined with a neck coil (4 elements), and the remaining parts of the anatomy are imaged with multiple flexible anterior phased-array coils (2 × 3 elements) and the corresponding posterior coils, which are integrated in the patient table. For smaller imaging volumes dedicated surface coils (flexible coil, open loop coil, small loop coil) can be used, which share a common amplifier interface

tuned during RF transmission; this is often accomplished by fast electronic switches (e.g., PIN diodes), which connect a dedicated detuning circuitry. To combine the high sensitivity of small surface coils with the volume coverage of a large volume resonator, the concept of the so-called phased-array coils has been introduced (Roemer et al. 1990). A phased-array coil consists of several small coil elements, which are directly connected to individual receiver channels of the MR system. The separate reconstruction of the coil elements is technically demanding, because a full set of receiver electronics (amplifiers, analog-to-digital converters) as well as an individual image reconstruction are required for each coil element. The signals of the individual coil elements are finally combined using a sum-of-squares algorithm, which yields a noise–optimal signal combination. Under certain conditions when SNR can be sacrificed, also a suboptimal image reconstruction can be achieved by a direct combination of the coil element signals, which reduces the number of receive channels and shortens the image reconstruction time. To be able to manually adjust SNR versus reconstruction overhead, special electronic mixing circuits have been introduced which allow combining, e.g., three coil elements into a primary, a secondary, and a tertiary signal (Total Imaging Matrix Tim, Siemens). In a phased-array coil, the coil elements are positioned in such a way that an induced voltage in one element does not couple to the adjacent element—this can be achieved by an overlapping arrangement of the coil elements (geometric decoupling). Phased-array coils with up to 128 elements have been realized; however, typically the number of elements ranges between 4 and 32. Today, MRI systems with 32 independent receiver channels are available, at which up to 76 coil elements can be positioned simultaneously. The individual coil elements can be selected manually or automatically to achieve the highest possible SNR for a given imaging location. Phased-array coils are not only required to achieve a high SNR. The individual coil elements can also be used to partially encode the spatial location in the image; this procedure is called parallel imaging. The simplest version of parallel imaging uses two adjacent coil elements with non-overlapping sensitivities. If one wants to image the full FOV covered by both coils only FOV/2 needs to be encoded, since each coil element is sensitive over this distance only. If the phase-encoding direction is chosen in this direction, the phase FOV can be reduced by a factor of 2, which in turn halves the total acquisition time. In practice, the sensitivity profiles of the coil elements overlap and more sophisticated techniques such as SMASH Sodickson and Manning 1997) or SENSE (Dumooulin et al. 1993) are required to reconstruct the image. Nevertheless, in parallel imaging the intrinsic spatial encoding present in the different locations of the imaging coils is exploited to reduce the number of phase encoding steps.

2.5  Technical Components

Because the phase encoding direction is different for different slice orientations, the optimal phased-array coil for parallel imaging offers coil elements with separated sensitivity profiles in all directions. For MR spectroscopy and non-proton imaging, RF coils with resonance frequencies for the respective nuclei are required. These non-proton coils can also incorporate a coil at the proton resonance frequency to acquire proton images without the need for patient repositioning. Double-resonant coils are also important in situations when both frequencies are used at the same time as, e.g., in decoupling experiments. For interventional MRI, dedicated tracking coils have been developed that are attached to the interventional devices (e.g., catheters or needles). The signal from these coils can be used for high-resolution imaging (e.g., of the vessel wall), but it is often only utilized to determine the position of the device (Doumoulin et al. 1993). In these tracking experiments, the signal of the coil is encoded in a single direction using a non-selective RF excitation, and the position of the coil in this direction is extracted after a one-dimensional Fourier transform. 2.5.4.4  Receiver The MR signal received by the imaging coil is a weak, analog, high-frequency electric signal. To perform an image reconstruction or a spectral analysis, this signal must be amplified, digitized, and demodulated. The signal amplification is typically performed very close to the imaging coil to avoid signal interference from other signal sources. If the RF coil is a TxRx coil, then the signal passes a transmit–receive switch that separates the transmit from the receive path. The amplified analog signal still contains the high-frequency component of the Larmor frequency. To remove this unwanted frequency component, the signal is sent to a demodulator, which receives the information about the current Larmor frequency from the synthesizer of the transmitter. After demodulation, the MR signal contains only the low-frequency information imposed by the gradients. Finally, the analog voltage is converted into a digital signal using an analog-to-digital converter (ADC). Over the recent years the conversion into a digital signal has increasingly been performed at an earlier stage in the receiver chain (e.g., before demodulation), and all subsequent steps were carried out in the digital domain. At the end of the receiver chain, the digital signal is then handed over to the image reconstruction computer. 2.5.5  Computer System The computing system of an MR tomograph is typically realized by a system of distributed computers that are

connected by a local high-speed network. The requirements for the computing system are manifold: For the user of the system it should provide an intuitive interface for measurement control, image processing, archiving, and printing. During sequence execution, the computers should control the hardware (i.e., gradients, RF, ADCs, patient monitoring, etc.) in real time. Additionally, the computing system must reconstruct and visualize the incoming MR data. Since a single computer cannot perform all of these tasks at the same time, typically three computers are used in an MR system: the host computer for interaction with the user, the hardware control computer for real-time sequence control, and the image reconstruction computer for high-speed data reconstruction. 2.5.5.1  Host Computer The host computer provides the interface between the user and the MR system. Through the MR user interface, the whole MR system can be controlled, MR measurements can be started, and the patient monitoring is visualized. At the host computer, the incoming images are sorted into an internal database for viewing, post-processing, and archiving. The internal database stores and sorts the images by patients, studies, and series. The database is often connected to the picture archiving and communication system (PACS) of the hospital, from where it retrieves the patient information to maintain a unique patient registry. A 256 × 256 MR image typically requires about 130 KB of storage space, and for each patient investigation between 100 and 1,000 images are acquired. On an average working day between 10 and 30 patients can be examined. The data of all of these patients need to be stored in the database so that a storage volume of about 4 GB per day should be provided. With increasing matrix sizes and image acquisition rates, these numbers can easily be multiplied by factors of 10 and more. The host computer is also used to transfer the acquired data to archiving media such as magneto-optical disks (MOD), tapes, compact disks (CD), digital versatile disks (DVD), or external computer archives (typically, the PACS). Data transfer is increasingly accomplished using the image standard DICOM (digital imaging and communications in medicine), which regulates not only the image data format, but also the transfer protocols. It is due to this imaging standard that images can be exchanged between systems from different vendors and can be shared between different modalities. For post-processing, typically different software packages are integrated. In MR spectroscopy, software packages for spectral post-processing are available to calculate, e.g., peak integrals automatically. For MR diffusion measurements, the apparent diffusion coefficient can be

87

88

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

mapped. With flow-evaluation software, the flow velocities and flow volumes can be assessed. To visualize threedimensional data sets often multi-planar re­for­matting tools or projection techniques such as the maximum intensity projection (MIP) are used. All of these software packages retrieve the image data from the integrated image database, into which the calculated images are finally stored. Dedicated computer monitors are connected to the host computer for image visualization, which fulfill the special requirements for diagnostic imaging equipment. In addition, these screens must not be susceptible to distortions due to the magnetic field; for this reason liquid crystal monitors based on the thin-film transistor (TFT) technology are increasingly used. For interventional MR, shielded monitors for in-room image display have been designed, where the monitor is shielded against electromagnetic interference. These monitors can be used within the Faraday cage of the MR system without interfering with the image acquisition. 2.5.5.2  Hardware-Control Computer The control of the imaging hardware (i.e., the gradients in x, y, and z, the RF sub-system, the receiver, and the patient-monitoring system) requires a computer with a real-time operating system. Compared with conventional operating systems where the instructions are processed in an order and at a time that are influenced by many factors, a real-time operating system ensures that operations are executed on an exactly defined time scale. This real-time execution is necessary to maintain, e.g., the phase coherence during spin-echo MRI or to ensure that a given steady state is established during balanced SSFP imaging. During sequence execution, the different instructions for the hardware are typically sent by the control program to digital signal processors (DSP) that control the individual units. Thus, new instructions can be prepared by the control program, whereas the actual execution is controlled close to the individual hardware. To ensure that enough hardware instructions are available, many time steps are computed in advance during sequence execution. For real-time pulse sequences, this advance calculation needs to be minimized to be able to interactively change sequence parameters such as the slice position (controlled by the RF frequency) or orientation (controlled by the gradient rotation matrix). In real-time sequences, the information about the current imaging parameters is thus retrieved not only once at the beginning of the scan, but continuously during the whole imaging experiment.

2.5.5.3  Image-Reconstruction Computer The reconstruction of the data arriving at the ADCs is performed by the image-reconstruction computer. To estimate the amount of data this computer needs to process the following estimate can be used: During high-speed data acquisition about 256 raw data points (i.e., 256 × 16 bytes) arrive per imaging coil at time intervals of TR = 2 ms, so that with 10 Rx coils a data rate of 20 MB/s results. These incoming data need to be re­arranged, corrected, Fourier transformed, combined, and geometrically distorted before the final image is sent to the host computer. To perform this task today multiprocessor CPUs are used to perform some of these tasks in parallel. In particular, the image reconstruction for multiple coils lends itself naturally to parallelization, since each of the coils is independent of the other. Additionally, some manufacturers are including simple post-processing steps into the standard image re­con­ struction. Since the reconstruction computer does not provide a direct user interface, these reconstruction steps need to be designed in such a way that no user interaction is necessary. This is the case for the calculation of activation maps in fMRI, for MIP calculations under standard views in MR angiography, or for the calculation of the arrival time of a contrast agent bolus in perfusion studies. At the end of the image reconstruction, the image data are transferred to the host computer via the internal computer network. 2.5.6  Patient Monitoring Special MR imaging techniques require additional MR components that are not necessarily available at any MR scanner. These components often monitor certain physiologic signals such as the electrical activity of the heart (electrocardiogram, ECG) or breathing motion (Fig. 2.5.10). Typically, the measured physiologic signals are not used to assess the health status of the patient but to synchronise the image acquisition with the organ motion, since heart and breathing motion can cause significant artifacts during abdominal imaging. Synchronization of the image acquisition is performed either with prospective or retrospective gating. With prospective gating (or triggering), the imaging is started with the arrival of a certain physiologic signal (e.g., the R wave in the ECG). Therefore, the physiologic signal is post-processed (e.g., thresholding and low-pass filtering) to create a trigger signal when the physiologic condition is present. With retrospective gating, the measurement is not interrupted, but data are acquired continuously, and for each measured data set, the physiologic state is stored with the data (e.g., the time duration after the last R wave). During image reconstruction, the measured data

2.5  Technical Components Fig. 2.5.10  Whole-body imaging with array coils covering the patient from head to toe (Exelart VantageTM, Toshiba). Since not all coils are in the imaging volume of the MR system at the same time, a lower number of receiver channels (here: 32) are sufficient for signal reception

are sorted in such a way that images are formed from data with similar physiologic signals (e.g., diastolic measurements). The advantage of the retrospective over the prospective data acquisition is the continuous measurement without gaps that could lead to artifacts in steady state pulse sequences. The post-processing effort of retrospectively acquired data is higher because data need to be analyzed and sorted before image reconstruction. Additionally, on average more data need to be acquired as compared to prospective triggering to ensure that for each physiologic condition at least one data set is present (over-sampling). 2.5.6.1  Electrocardiogram To measure the ECG in the MR system MR-compatible electrodes made of silver–silver chloride (Ag/AgCl) are used. The measurement of the ECG in an MR system is difficult, because the switching of the gradients can induce voltages in the ECG cables that completely mask the ECG signal. This effect can be minimized, if short and loopless ECG cables are utilized. Short ECG cables are additionally advantageous since long cables with a loose contact to the skin can be the cause for patient burns that are induced by the interaction with the RF field during RF excitation (Kugel et al. 2003). To reduce this potential danger to a minimum, ECG systems have been developed that amplify the ECG signal close to the electrodes, and which transmit the ECG signal to the MR system either via optical cables (Felblinger et al. 1994) or as an RF signal at a frequency different from the Larmor frequency. With this technology, ECG signals can be acquired even during echo planar imag-

ing when gradients are permanently switched on and off (Ives et al. 1993). It should be noted that the ECG signal in the MR system significantly differs from the signal outside the magnet. The electrically conducting blood is flowing at different velocities in the cardiac cycle. Within the magnetic field the blood flow induces velocity-dependent electric fields (Hall effect) across the blood vessels, which in turn change the electric potentials measured at the ECG electrodes. Typically, the T wave of the ECG is augmented, an effect that is more pronounced at higher field strengths (Kangarlu and Robitaille 2000). For this reason, the ECG acquired in the MR system should not be regarded as of diagnostic quality. 2.5.6.2  Pulse Oximetry Pulse oximeters measure the absorption of a red and an infrared light beam that is sent through perfused tissue (e.g., a finger). The absorption is proportional to the oxygen content, so that devices can determine the partial oxygen pressure (pO2). Additionally, the pulsation of the blood leads to a pulse-related variation of the transmitted light signal, which is used in the MR systems to derive a pulse-related trigger signal (Shellock et al. 1992). Since the pulse wave arrives at the periphery with a significant delay after the onset of systole, it is difficult to use the pO2 signal for triggering in systolic MR imaging. Pulse oximeters consist solely of non-magnetic and non-conduction optical elements, so that they are not susceptible to any interference with the gradient or RF activity.

89

90

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.5.6.3  Breathing Synchronization

2.5.6.4  Electroencephalogram

To detect breathing motion, several mechanical devices such as breathing belts or cushions have been introduced. Essentially, all these systems are air filled and change their internal pressure as a function of the breathing cycle when they are attached to the thorax of the patient. The pressure is continuously monitored and is used as an indicator for breathing status. As with the pulse oximeters, these systems are also free of any electrically conducting elements, so that no RF heating is expected. However, in clinical practice breathing triggering can pose a problem in long-lasting acquisitions since patients start to relax over time, and the initial breathing pattern is not reproduced. An alternative approach to the measurement of the breathing cycle is offered by the MR itself: If a single image line is excited in head–foot direction through the thorax (using, e.g., a 90° and 180° slice that intersect along the desired line), then the signal of this line has high contrast at the liver–lung interface. This diaphragm position can be detected automatically and can be used to extract the relative position in the breathing cycle. This technique is called a navigator echo (Ehman and Felmlee 1989), since an additional echo for navigation needs to be inserted into the pulse sequence. Similar approaches using lowresolution two- or three-dimensional imaging can be used to correct for patient motion in long-lasting image acquisitions such as fMRI (Welch et al. 2002). Here, the change in position is determined and used to realign the imaging slices (prospective motion correction).

For neurofunctional studies, electroencephalogram (EEG) systems have been developed that can be operated in the MR tomography (Muri et al. 1998). Compared with the ECG, the voltages induced during brain activity are about 100 times smaller in EEG recording, which poses a significant detection problem (Goldman et al. 2000; Sijbersa et al. 2000). Blood pulsation, patient motion, as well as induced voltages during gradient and RF activity can cause spurious signals in the EEG leads, which obscure the true EEG signal. To remove the imaging-related artifacts, dynamic filtering can be used, which removes all signal contributions associated with the basic frequencies of the MR system. 2.5.7  Summary A large variety of MR systems with different magnet types, coil configurations, and gradient sets is currently available for diagnostic and interventional MR imaging. To choose from these systems, the desired imaging applications as well as economic factors need to be considered: A small hospital might with few MR patients might want to use a low-field permanent magnet system with low maintenance cost, whereas a university hospital with a diverse patient clientele and high patient throughput should better offer a high-field MR system with state-ofthe-art gradient systems.

Fig. 2.5.11  Physiologic monitoring and triggering units. The three electrodes of the ECG system as well as the tube of the breathing sensor are connected with a transceiver that transmits both signals to the patient monitoring unit of the MR system. The optical pulse sensor is attached to the finger, and the signals are guided via optical fibers to the detection unit. For increased patient safety the ECG system must be used together with a holder system (not shown here), which provides additional distance between the ECG leads and the patient body

2.5  Technical Components

References 1. 2.

3.

4.

5.

6.

7.

8.

9. 10.

11.

12.

13.

14. 15.

Chen CN, Hoult DI (1989) Biomedical magnetic resonance technology. Adam Hilger, Bristol Dumoulin CL, Souza SP, Darrow RD (1993) Real-time position monitoring of invasive devices using magnetic resonance. Magn Reson Med 29:411–415 Ehman RL, Felmlee JP (1989) Adaptive technique for highdefinition MR imaging of moving structures. Radiology 173:255–263 Felblinger J, Lehmann C, Boesch C (1994) Electrocardiogram acquisition during MR examinations for patient monitoring and sequence triggering. Magn Reson Med 32:523–529 Goldman RI, Stern JM, Engel J Jr, Cohen MS (2000) Acquiring simultaneous EEG and functional MRI. Clin Neurophysiol 111:1974–1980 Harvey PR (1999) The modular (twin) gradient coil—high resolution, high contrast, diffusion weighted EPI at 1.0 Tesla. MAGMA 8:43–47 Harvey PR, Mansfield P (1994) Resonant trapezoidal gradient generation for use in echo planar imaging. Magn Reson Imaging 12:93–100 Ives JR, Warach S, Schmitt F, Edelman RR, Schomer DL (1993) Monitoring the patient’s EEG during echo planar MRI. Electroencephalogr Clin Neurophysiol 87:417–420 Jin J (1999) Electromagnetic analysis and design in magnetic resonance imaging. CRC Press, Boca Raton Kangarlu A, Robitaille PML (2000) Biological effects and health implications in magnetic resonance imaging. Concepts Magn Reson 12:321–359 Kugel H, Bremer C, Puschel M, Fischbach R, Lenzen H, Tombach B, Van Aken H, Heindel W (2003) Hazardous situation in the MR bore: induction in ECG leads causes fire. Eur Radiol 13:690–694 Liu F, Zhao H, Crozier S (2003) On the induced electric field gradients in the human body for magnetic stimulation by gradient coils in MRI. IEEE Trans Biomed Eng 50:804–15 Mansfield P, Chapman B (1986) Active magnetic screening of coils for static and time-dependent magnetic field generated in NMR imaging. J Phys E: Sci Instrum 19:540–545 Mansfield P, Harvey PR (1993) Limits to neural stimulation in echo planar imaging. Magn Reson Med 29:746–758 Mispelter J, Lupu M, Briguet A (2006) NMR probeheads for biophysical and biomedical experiments: theoretical principles and practical guidelines. World Scientific, London

16. Muri RM, Felblinger J, Rosler KM, Jung B, Hess CW, Boesch C (1998) Recording of electrical brain activity in a magnetic resonance environment: distorting effects of the static magnetic field. Magn Reson Med 39:18–22 17. Oppelt A (ed) (2005) Imaging systems for medical diagnostics: fundamentals, technical solutions and applications for systems applying ionizing radiation, nuclear magnetic resonance and ultrasound, 2nd edn. Publicis, New York 18. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42:952–962 19. Robitaille PM, Warner R, Jagadeesh J, Abduljalil AM, Kangarlu A, Burgess RE, Yu Y, Yang L, Zhu H, Jiang Z, Bailey RE, Chung W, Somawiharja Y, Feynan P, Rayner DL (1999) Design and assembly of an 8 Tesla whole-body MR scanner. J Comput Assist Tomogr 23:808–820 20. Roemer PB, Edelstein WA, Hayes CE, Souza SP, Mueller OM (1990) The NMR phased array. Magn Reson Med 16:192–225 21. Schaefer DJ (1998) Safety aspects of switched gradient fields. Magn Reson Imaging Clin N Am 6:731–748 22. Schmitt F, Stehling MK, Turner R (1998) Echo planar imaging: theory, technique and application. Springer, Berlin Heidelberg New York 23. Shellock FG, Myers SM, Kimble KJ (1992) Monitoring heart rate and oxygen saturation with a fiber-optic pulse oximeter during MR imaging. AJR Am J Roentgenol 158:663–664 24. Sijbersa J, Van Audekerke J, Verhoye M, Van der Linden A, Van Dyck D (2000) Reduction of ECG and gradient related artefacts in simultaneously recorded human EEG/MRI data. Magn Reson Imaging 18:881–886 25. Sodickson DK, Manning WJ (1997) Simultaneous acquisition of spatial harmonics (SMASH): Fast imaging with radiofrequency coil arrays. Magn Reson Med 38:591–603 26. Turner R (1993) Gradient coil design: a review of methods. Magn Reson Imaging 11:903–920 27. Vlaardingerbroek MT, den Boer JA, Luiten A (2002) Magnetic resonance imaging: theory and practice, 2nd rev. edn. Springer, Berlin Heidelberg New York 28. Webb P, Macovski A (1991) Rapid, fully automatic, arbitrary-volume in vivo shimming. Magn Reson Med 20:113–122 29. Welch EB, Manduca A, Grimm RC, Ward HA, Jack CR Jr (2002) Spherical navigator echoes for full 3D rigid body motion measurement in MRI. Magn Reson Med 47:32–41

91

92

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.6  Contrast Agents A. Huppertz and C.J. Zech During the pioneer period of MR imaging, expectations were that the high inherent contrast in MR imaging makes the use of contrast agents superfluous. However, increasing use of the modality in the clinical setting has revealed that a number of diagnostic questions require the application of a contrast agent. Similar to other imaging modalities, the use of contrast agents in MR imaging aims at increasing sensitivity and specificity and, thereby, the diagnostic accuracy. 2.6.1  Physicochemical Properties of MR Contrast Agents The main contrast parameters in MR imaging are proton density, relaxation times, and magnetic susceptibility (ability of a material or substance to become magnetized by an external magnetic field). MR imaging contrast agents focus upon relaxation time and susceptibility changes. Most of them are either para- or superparamagnetic. The most efficient elements for use as MR imaging contrast agents are gadolinium (Gd), manganese (Mn), dysprosium (Dy), and iron (Fe). The magnetic field produced by an electron is much stronger than that produced by a proton. However, in most substances the electrons are paired, resulting in a weak net magnetic field. Gd with its seven unpaired electrons possesses the highest ability to alter the relaxation time of adjacent protons (relaxivity). For MR contrast agents, differentiation between positive and negative agents has to be made. Paramagnetic contrast agents Gd and Mn have a similar effect on T1 and T2 and are classified as positive agents. Since the T1 of tissues is much higher than the T2, the predominant effect of these contrast agents at low concentrations is that of T1 shortening. Thus, tissues that take up Gd- or Mn-based agents become bright in T1-weighted sequences. On the other hand, negative-contrast agents influence signal intensity by shortening T2 and T2*. Superparamagnetic agents belong to this group and produce local magnetic field inhomogeneities of the local magnetic field. T2 is reduced due to the diffusion of water through these field gradients. Magnetite, Fe3O4, is such a paramagnetic particle. Coated with inert material (e.g., dextranes, starch), it can be used for oral or intravenous applications. In addition to the classification in positive or negative agents, MR contrast agents can be differentiated according to their target tissue. The targeting of an agent is determined by the pharmaceutical profile of the substance. In the clinical environment, we differentiate currently three classes of agents:

• Unspecific extracellular fluid space agents • Blood-pool and intravascular agents • Targeted and organ-specific agents Unspecific extracellular fluid space agents. Low-molecular-weight paramagnetic contrast agents distribute into the intravascular and extracellular fluid space (ECF) of the body. Their contrast effect is caused by the central metal ion. All approved ECF agents contain a Gd ion, which contains seven unpaired electrons. Because Gd itself is toxic, the ion is bound in highly stable complexes. The different complexes and the physicochemical properties of all clinically used agents are listed in Table 2.6.1. The agents are not metabolized and are excreted in unchanged form via the kidneys. Bound, they form low-molecular-weight, water-soluble contrast agents. Gadopentetate dimeglumine (Magnevist, Bayer Schering Pharma, Berlin, Germany) and gadoterate meglumine (Dotarem, Laboratoires Guerbet, Aulnay-Sous-Bois, France) are ionic high-osmolality agents, whereas gadodiamide (Omniscan, GE Healthcare, Buckinghamshire, UK) and gadoteriol (ProHance, Bracco Imaging, Milan, Italy) are non-ionic low-osmolality agents. Due to the low total amount of contrast agent usually applied in MR imaging, no difference in tolerance between both classes could be demonstrated (Oudkerk et al. 1995; Shellock 1999). An estimated 50% of ECF agents (as for example in gadopentetate dimeglumine, size 590 Da) is cleared from the vascular space into the extravascular compartment on the initial passage through the capillaries. Two agents in the group of ECF agents have to be mentioned separately. Gadopentate dimeglumine (MultiHance, Bracco Imaging) is an agent with a weak protein binding (about 10%) in human plasma. The bound fraction of the agent has a higher relaxivity than does the unbound fraction. In sum, the relaxivity of gadopentate dimeglumine is 50% higher as compared with gadopentetate dimeglumine at 1.5 T/37°C in plasma. The effect of higher relaxivity is highest at low field strengths (Table 2.6.1). The concentration of the contrast agent is 0.5 mol/l. Gadopentate dimeglumine was primarily developed as a liver-specific MR imaging agent, and is currently approved both in the indication detection of focal liver lesions and in MR angiography. Most of the injected dose of gadopentate is excreted unchanged in urine within 24 h, although a fraction corresponding to 0.6–4.0% of the injected dose is eliminated through the bile and recovered in the feces (Spinazzi et al. 1999). The second particular ECF agent, gadobutrol (Gadovist, Bayer Schering Pharma) is approved in a higher concentration (1 M) than all other available MR imaging contrast agents. In addition, gadobutrol has a higher relaxivity than most extracellular 0.5 M contrast agents on the market (Table 2.6.1). The higher concentration has revealed to be particularly useful for MR perfusion studies and MR angiography (Tombach et al. 2003).

2.6  Contrast Agents Table 2.6.1  Physicochemical properties of Gd-based MR ECF contrast media INN code

Gadopentetate dimeglumine

Gadodiamide

Gadoterate meglumine

Gadoteriol

Gadobutrol

Gadopentate dimeglumine

Tradename

Magnevist®

Omniscan™

Dotarem

Prohance™

Gadovist®

Multihance™

Manufacturer

Bayer Schering Pharma AG, Germany

GE Bioscience Healthcare, UK

Laboratoires Guerbet, France

Bracco Imaging, Milan, Italy

Bayer Schering Pharma AG, Germany

Bracco, Italy

Concentration (mol/l) Osmolality   (mOsmol/kg H2O at 37°C)

0.5 1,960

0.5 780

0.5 1,350

0.5 630

1 1,603

0.5 1,970

Viscosity (mPA s at 37°C)

2.9

1.9

2.0

1.3

4.96

5.30

Relaxivity r1 in plasma 40°C (l/mmol–1s–1 at 0.47 T)

3.8 ± 0.2

4.4 ± 0.3

4.3 ± 0.2

4.8 ± 0.3

6.1 ± 0.4

9.2 ± 0.6

Relaxivity r1 in plasma at 37°C (l/mmol–1s–1 at 1.5 T)

4.1 ± 0.3

4.3 ± 0.3

3.6 ± 0.2

4.1 ± 0.3

5.2 ± 0.3

6.3 ± 0.4

Relaxivity r1 in plasma at 37°C (l/mmol–1s–1 at 3 T)

3.7 ± 0.2

4.0 ± 0.2

3.5 ± 0.2

3.7 ± 0.2

5.0 ± 0.3

5.5 ± 0.3

Blood-pool and intravascular agents. Blood-pool agents stay within the intravascular space with no or only slow physiologic extravasation. The agents can be used for firstpass imaging and delayed blood-pool phase imaging. The prolonged imaging window allows more favorable image resolution and signal-to-noise ratio. The absence of early extravasation also improves the contrast-to-noise ratio. The pharmacokinetic properties of blood-pool agents are expected to be well suited to MR angiography and coronary angiography, perfusion imaging, and permeability imaging (detection of ischemia and tumor grading). Currently, three types of blood-pool agents are being developed: 1 Gd compounds with a strong but reversible affinity to human proteins such as albumin 2 Macromolecular-bound Gd complexes 3 Ultra small or very small super-paramagnetic particles of iron oxide (USPIO and VSOP) There are important differences between the three groups regarding pharmacokinetics in the body, i.e., distribution and elimination.

Gd compounds with a strong but reversible affinity to human proteins such as albumin exhibit prolonged plasma elimination half-life and increased relaxivity. The elimination is done by glomerular filtration of its unbound fraction. Given that there is equilibrium between the bound and unbound fraction in the presence of albumin, the excreted molecules are immediately substituted due to dissociation of agent from the agent–albumin complex. Two agents with affinities to albumin were developed and tested in clinical trials: gadofosveset (Vasovist®, Bayer Schering Pharma) 80–96% bound in human plasma (Lauffer et al. 1998) and gadocoletic acid (B22956/1, Bracco Imaging), with a protein binding of approximately 95% in humans (Cavagna et al. 2002; La Noce et al. 2002). Currently, gadofosveset is the only blood-pool agent approved (for MRA in Europe). All other contrast agents with blood-pool characteristics are in clinical or in earlier phase development. Gadofosveset is a stable Gd diethylenetriaminepentaacetic acid (Gd-DTPA) chelate substituted with a diphenylcyclohexylphosphate group. The mean plasma concentration at 1, 4 and 24 h after the

93

94

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

bolus injection of 0.03 mmol/kg body weight dose were 56%, respectively 41% and 14% of the concentration reached 3 min after injection. The mean half-life of the distribution phase (t1/2α) was 0.48 ± 0.11 h. Relative to the reported clearance values of the non–protein-bound MRI contrast agents, the clearance values of gadofosveset are markedly slower. Gadofosveset is provided in a concentration of 0.25 mol/l and a dose of 0.03 mmol/kg body weight is recommended for MRA (Perrault et al. 2003). As a further benefit, Gd compounds with a strong, but reversible affinity to human proteins provide a long-lasting blood-pool effect even when small amounts of the substance leak out of the vasculature. The blood-pool effect persists because albumin remains highly concentrated in plasma while it shows a two- to three-times lower concentration in the extravascular space. Thus, even when Vasovist® leaks from the vasculature, the receptor-induced magnetization enhancement (RIME) effect within the vascular spaces ensures that the signal enhancement in the blood dominates the MRI contrast. In rabbits, enhancement with gadofosveset persisted at relatively constant levels from two minutes to up to 1 h, whereas the enhancement of ECF had virtually disappeared within 60 min (Lauffer et al. 1998). The second blood-pool agent with binding to human serum albumin, gadocoletic acid has been tested in coronary MRA (Paetsch et al. 2006). Compared with gadofosveset, the slightly higher percentage of bounded agent may result in a lower percentage of extravasation and a further decreased elimination period. Macromolecular Gd-based blood-pool agents are large molecules with sizes between 30 and 90 kDa. They are eliminated rapidly by glomerular filtration. Due to their large size, they do not extravasate into the interstitial space. The two agents used in clinical trials were Gadomer (Schering) and P 792 (Laboratoires Guerbet, Aulnay-Sous-Bois). Gadomer contains multiple Gd molecules (24 Gd atoms, Mr 35,000). P792 is a monodisperse monogadolinated macromolecular compound with Mr 6.47 kDa, based on a gadoterate meglumine core (Port et al. 2001). Four hydrophilic arms account for its intravascular properties. In a preclinical study, P792 allowed acquisition of high-quality MR angiograms. Image quality was rated as superior for P792 in the postbolus phase images compared with ECF agents. The intravascular properties lead to an excellent signal in the vasculature with limited background enhancement (Ruehm et al. 2002). The first clinical use of USPIO was done in specific parenchymal organ imaging due to the incorporation of USPIO/SPIO into cells of the reticuloendothelial system of the liver, bone marrow, spleen, or lymphatic tissue. These particles produce a strong augmentation of the local magnetic field. Predominant shortening of T2 and

T2* produces a loss of signal intensity on MR images. The agents that have been developed as blood-pool agents provide different characteristics with a predominating T1 effect and a prolonged intravascular residence time due to the small size of the particles. NC 100150 (Clariscan, GE Healthcare) was the first USPIO tested for MRA (Taylor et al. 1999: Weishaupt et al. 2000). It is a strictly intravascular agent with an oxidized starch coating and has an approximate diameter of 20 nm. The half-life is 45–100 min, and it has shown to reduce blood T1 to below 100 ms (Wagenseil 1999). Another iron oxide particle MR contrast agent in the phase of clinical development is VSOP-C184 (Ferupharm, Teltow, Germany). It is classified as a VSOP with a core diameter of 4 nm and a total diameter of 8.6 nm. VSOPC 184 is coated with citrate. The relaxivities in water at 0.94 T are (T1) 20.1 and (T2) 37.1 l/[mmol*s]. The plasma elimination half-life at 0.045 mmol Fe/kg was 21.3 ± 5.5 minutes in rats and 36.1 ± 4.2 minutes in pigs, resulting in a T1 relaxation time of plasma of <100 ms for 30 min in pigs (Wagner et al. 2002). Qualitative evaluation of image quality, contrast, and delineation of vessels showed that the results obtained with VSOP-C184 at doses of 0.025 and 0.035 mmol Fe/ kg was similar to those of gadopentetate dimeglumine at 0.1 and 0.2 mmol Gd/kg. VSOP-C184 is suitable for firstpass MRA and thus, in addition to its blood-pool characteristics, allows for selective visualization of the arteries without interfering venous signal (Schorr et al. 2004). Another USPIO is SH U 555 C (Supravist, Bayer Schering Pharma), an optimized formulation of carboxydextran-coated ferucarbotran (Resovist; Bayer Schering Pharma), which was formerly identified as SH U 555 A, with respect to T1-weighted MR imaging. SH U 555 C has a mean core particle size of about 3–5 nm and a mean hydrodynamic diameter of about 20 nm in an aqueous environment. Relaxivity measurements yielded an r1 of 22 s–1 (mmol/l)–1 and an r2 of 45 s–1 (mmol/l)–1 at 40°C and 20 MHz in water (Reimer et al. 2004). 2.6.2  Dependency of Contrast Agents from the Magnetic Field Strength The efficacy of MRI contrast agents is not just determined by their pharmacokinetic properties (distribution and time dependence of their concentration in the area of interest), but also by their magnetic properties, described by their T1 and T2 relaxivities. For all commercially available MRI contrast agents, relaxivities are published and listed in the respective package inserts. However, the most commonly used field strength for relaxation measurements (0.47 T) is different from the currently most frequently used field strength of clinical MRI instruments (1–3 T). Rohrer et al. evaluated in a well-conducted and standardized phantom measurement study the T1 and T2 relaxivities of all currently commercially available MR

2.6  Contrast Agents

contrast agents in water and in blood plasma at 0.47, 1.5, 3, and 4.7 T, as well as in whole blood at 1.5 T (Rohrer et al. 2005). They quantified significant dependencies of relaxivities on the field strength and solvents (Table 2.6.2). Protein binding leads to both increased field strength and solvent dependencies and hence to significantly altered T1 relaxivity values at higher magnetic field strengths. 2.6.3  Safety of MR Contrast Agents MR contrast agents are in clinical use since 1988, and a wide experience is reported. Severe or acute reactions after single intravenous injection of Gd-based ECF agents are rare. In two large multiple-year surveys including, respectively, 21,000 and more than 9,000 examinations, an incidence of acute adverse reactions between 0.17 and 0.48% were reported (Li et al. 2006: Murphy et al. 1996). The severity of these adverse reactions was classified as mild (75–96%), moderate (2–20%), and severe (2–5%). Typical nonallergic adverse reactions include nausea, headache, taste perversion, or vomiting, and typical reactions resembling allergy include hives, diffuse erythema, skin irritation, or respiratory symptoms. The incidence of severe anaphylactoid reaction is very low and was reported to be between 0.0003 and 0.01% in the literature (De Ridder et al. 2001; Li et al. 2006: Murphy et al. 1996). The reported life-threatening reactions resembling allergy were severe chest tightness, respiratory distress, and periorbital edema. Known risk factors for the development of adverse reactions are prior adverse reactions to iodinated contrast media, prior reactions to a Gd-based contrast agent, asthma, and history of drug/food allergy. Concerning liver-specific contrast media, a higher percentage of associated adverse reactions were reported for mangafodipir trisodium (7–17%) and ferumoxides (15%) (Runge 2000). The recently approved bolus-injectable agent ferucarbotran (Resovist, Bayer Schering Pharma) has proven a better tolerance profile during the clinical development compared to ferumoxides. Even bolus injections caused no cardiovascular side effects, lumbar back pain, or clinically relevant laboratory changes (Reimer and Balzer 2003). For the two approved Gdbased agents gadopentate dimeglumine and gadoxetic acid, far fewer patients have been examined to date. According to the results of the clinical trials conducted for the approval of both agents, they are comparable to Gdbased ECF agents in terms of safety (Bluemke et al. 2005; Halavaara et al. 2006; Huppertz et al. 2004). Post-marketing surveillance of gadopentate dimeglumine reporting approximately 100,000 doses revealed an overall adverse event incidence of <0.03%, with serious adverse eventss reported for <0.005% of patients (Kirchin et al. 2001). In the class of blood-pool agents, only gadofosveset (Bayer Schering Pharma) has been approved recently

in some European countries. The tolerance of the agent must be estimated based on the clinical trials. Based on these data, gadofosveset is well tolerated, and the incidence and profile of undesired side effects is very similar to ECF agents (Goyen et al. 2005; Petersein et al. 2000; Rapp et al. 2005). Magnetic resonance contrast agents, particularly the Gd-based agents, are extremely safe (Niendorf et al. 1994) and lack in the usually applied diagnostic dosage the nephrotoxicity associated with iodinated contrast media. Nevertheless, health care personnel should be aware of the (extremely uncommon) potential for severe anaphylactoid reactions in association with the use of MR contrast media and be prepared should complications arise. Nephrogenic systemic fibrosis (NSF) is a rare disease occurring in renal insufficiency that only has been described since 1997. In 2006, a first report about a potential relationship with intravenous administration of Gdbased MR contrast medium gadodiamide was published (US Food and Drug Administration 2007). NSF appears to occur in patients with kidney failure, along with high levels of acid in body fluids (a condition known as metabolic acidosis) that is common in patients with kidney failure. The disease is characterized by skin changes that mimic progressive systemic sclerosis with a predilection for peripheral extremity involvement that can extend to the torso. However, unlike scleroderma, NSF spares the face and lacks the serologic markers of scleroderma. NSF may also result in fibrosis, or scarring, of body organs. Diagnosis of NSF is done by looking at a sample of skin under a microscope. The risk of NSF in patients with advanced renal insufficiency does not suggest being the same for all Gdbased contrast agents, because distinct physicochemical properties affect their stabilities and thus the release of free Gd ions (Bundesinstitut für Arzneimittel und Medizinprodukte [Federal Institute for Drugs and Medical Devices] 2007). Some Gd-based contrast media are more likely than are others to release free Gd3+ through a process called transmetallation, with endogenous ions from the body (Thomsen et al. 2006). These agents have the largest amount of excess chelate. Gadodiamide and gadoversetamide differ from other Gd-based contrast media because of an excess of chelate and is more likely to release free Gd3+ as compared with other agents. Cyclic molecules offer better protection and binding to Gd3+, compared with linear molecules (Thomsen et al. 2006). The non-linear, non-ionic chelates gadodiamide and gadoversetamide seem to be associated with the highest risk of NSF (Broome et al. 2007; Sadowski et al. 2007). The recommendations to prevent development of NSF are nonspecific (US Food and Drug Administration 2007): • Gd-containing contrast agents, especially at high doses, should be used only if clearly necessary in patients with advanced kidney failure (those currently

95

2.9 ± 0.2 2.9  ±  0.2

1 M ECF 0.5 M, low protein binding

Prohance

Dotarem

Omniscan

Optimark

Gadovist

Multihance

Gadoterate meglumine

Gadodiamide

Gadoversetamide

Gadobutrol

Gadopentate dimeglumine

Endorem®

Supravist

Ferumoxide

Ferucarbotran

NA not applicable

SPIO

NA

Gadomer

USPIO

Macromolecular Gd

Vasovist

Gadofosveset

0.25 M, large-protein binding

Primovist®

Gadoxetic acid

13.2 ± 0.7

4.7 ± 0.3

17.3 ± 0.9

5.2 ± 0.3

4.7 ± 0.2

4.0 ± 0.2

3.3 ± 0.2

3.8 ± 0.2

3.3 ± 0.2

3.3 ± 0.2

Gadoteridol

0.5 M ECF

Magnevist

44 ± 3

41 ± 2

22 ± 1

5.9 ± 0.6

5.1 ± 0.6

4.3 ± 0.5

3.9 ± 0.8

4.2 ± 0.7

3.6 ± 0.6

3.2 ± 0.7

3.2 ± 0.7

3.9 ± 1.1

7.3 ± 0.4

4.1 ± 0.3

13 ± 0.7

5.3 ± 0.3

4.3 ± 0.3

4.0 ± 0.3

3.2 ± 0.3

3.6 ± 0.3

3.2 ± 0.3

2.8 ± 0.2

2.8 ± 0.2

3.1 ± 0.3

57 ± 3

93 ± 6

23 ± 1

6.1 ± 0.4

5.5 ± 0.3

4.7 ± 0.3

3.9 ± 0.3

4.5 ± 0.3

3.8 ± 0.3

3.3 ± 0.3

3.4 ± 0.3

3.7 ± 0.3

10.7 ± 0.6

4.5 ± 0.3

16 ± 1

19 ± 1

6.9 ± 0.4

6.3 ± 0.3

5.2 ± 0.3

4.7 ± 0.3

4.3 ± 0.3

3.6 ± 0.2

4.1 ± 0.4

4.1 ± 0.2

r1

r1

r1

r2

1.5 T

3T

1.5 T r2

Plasma

Water

Gadopentetate dimeglumine

Class

Trade name

INN code

Table 2.6.2  r1 and r2 relaxivities of MR contrast media in water and plasma at 37°C at 1.5 and 3 T

38 ± 2

33 ± 2

19 ± 1

34 ± 2

8.7 ± 0.9

8.7 ± 0.9

6.1 ± 0.9

5.2 ± 0.9

5.2 ± 1

4.3 ± 0.9

5.0 ± 0.8

4.6 ± 0.8

r2

5.6 ± 0.3

2.7 ± 0.2

13 ± 1

9.9 ± 0.5

6.2 ± 0.3

5.5 ± 0.3

5 ± 0.3

4.5 ± 0.3

4 ± 0.2

3.5 ± 0.2

3.7 ± 0.2

3.7 ± 0.2

r1

3T

95 ± 9

45 ± 3

25 ± 2

60 ± 4

11 ± 1

11 ± 1

7.1 ± 0.9

5.9 ± 0.9

5.6 ± 0.9

4.9 ± 0.9

5.7 ± 0.9

5.2 ± 0.9

r2

96 2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.6  Contrast Agents

requiring dialysis or with a glomerular filtration rate (GFR) = 15 ml/min or less). • It may be prudent to institute prompt dialysis in patients with advanced kidney dysfunction who receive a Gd contrast MRA. Although there are no data to determine the utility of dialysis to prevent or treat NSF in patients with decreased kidney function. 2.6.4  Value of Contrast Agents in Clinical Practice 2.6.4.1  Contrast Agents in Neuroimaging The use of contrast agents in neuroimaging is an accepted standard for the assessment of pathological processes, which utilizes the extravasation of contrast agents through a compromised blood–brain or blood–spinal cord barrier. Compared with contrast-enhanced CT, MR imaging with Gd-based contrast agents is far more sensitive and depicts even subtle disruptions of the blood–brain barrier that are caused by a variety of noxious agents as, for example neoplastic or inflammatory processes and ischemic stress. Moreover, MR contrast agents are increasingly used to evaluate brain perfusion in clinical practice for a variety of applications, including tumor characterization, stroke, and dementia. The contrast-enhanced brain perfusion MR examination is based on a magnetic susceptibility contrast phenomenon that occurs owing to the T2 and T2* relaxation effects of rapidly intravenous bolus–injected contrast agents. The contrast agents in current use are the standard ECF Gd chelates (Table 2.6.1). These extracellular agents show no appreciable differences in their enhancement properties and biologic behavior (Akeson et al. 1995; Brugieres et al. 1994; Grossman et al. 2000; Oudkerk et al. 1995; Valk et al. 1993; Yuh et al. 1991). They equilibrate rapidly between the intra- and extracellular spaces of soft tissues and enter central nervous system lesions only at sites of damaged blood–brain barrier. The standard dose for MR imaging of the central nervous system is 0.1 mmol/kg body weight; however, it has been shown that a higher dose of Gd chelate–based contrast agents may help reveal more subtle disease states of the blood– brain barrier regardless whether caused by tumors or by inflammatory lesions (Bastianello et al. 1998; Haustein et al. 2003; Yuh et al. 1994). This raises the question, in how far Gd contrast agents with a higher concentration as for example gadobutrol or agents with a higher relaxivity as for example gadopentate dimeglumine help to increase the sensitivity and accuracy to detect lesions as compared to standard Gd chelates. For gadobutrol, no comparative studies to standard Gd chelates exist up to now; however, based on smaller cohorts it can be assumed that the higher amount of Gd, which can be achieved by the higher Gd concentration, is of value for lesion detection and characterization (Vogl

et al. 1995). Moreover, based on animal experiments the amount of Gd in gliomas was higher after injection of gadobutrol in comparison to gadopentetate dimeglumine although identical doses of Gd per kilogram body weight were injected for both contrast agents (Le Duc et al. 2004). Gadopentate dimeglumine proved significantly superior tumor enhancement of intraaxial enhancing primary and secondary brain tumors at a dosage of 0.1 mmol/kg body weight as compared with the same dosage of gadopentetate dimeglumine (Knopp et al. 2004). Similar results were also obtained in comparison of gadopentate dimeglumine with other contrast agents as well as in special populations as for example in pediatric patients (Colismo et al. 2001, 2004, 2005). The increased contrast enhancement resulted also in an increased number of detected brain metastases. Dynamic susceptibility-weighted (DSC) contrast agent-enhanced MR imaging is increasingly used for the assessment of cerebral perfusion in many different clinical settings, such as ischemic stroke (Parsons et al. 2001), neurovascular diseases (Doerfler et al. 2001), brain tumors (Essig et al. 2004), and neurodegenerative disorders (Bozzao et al. 2001). Unlike MR angiography, which depicts the blood flow within larger vessels, perfusion-weighted MR techniques are sensitive to perfusion on the level of the capillaries. The technique is based on the intravenous injection of a T2*-relaxing contrast agent and subsequent bolus tracking using a fast susceptibility-weighted imaging sequence. After converting voxel signal into concentration values, parametric maps of regional cerebral blood volume (rCBV) and blood flow (rCBF) can be calculated by unfolding tissue concentration curves and the concentration curve of the feeding artery. The contrast agents used for dynamic susceptibility-weighted MR perfusion are usually standard Gd chelates; however, the dosages of Gd per kilogram of body weight as well as the value of higher concentrated agents have been widely discussed. During the first pass of the Gd chelate, the high intravascular concentration of Gd causes the T2* effects, which can be measured by rapid imaging techniques. The length and the peak concentration of the bolus seem to have influence on the resulting measured signal with a highly concentrated small bolus of contrast agent being advantageous for MR brain perfusion imaging (Essig et al. 2002; Heiland et al. 2001). In between the standard Gd chelates, no notably different behavior of the available agents has been published up to now. The recommended dose for DSC perfusion MRI is in the range of 0.15–0.30 mmol/kg body weight, with most authors preferring a value of 0.2, because the volume of the bolus gets too high when higher dosages are applied (Bruening et al. 2000). Therefore, the use of higher concentrated contrast agents or agents with higher relaxivity are also interesting for cerebral perfusion MRI. Again, studies were able to demonstrate the value of the 1 M gabobutrol and gadopentate dimeglumine. Tombach

97

98

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

et al. (2003) showed that 1 M gadobutrol resulted in a significantly improved quality of the perfusion examination in comparison to 0.5 M gadobutrol at the same dosage of 0.3 mmol/kg body weight. The results were explained by the sharper, more concentrated bolus, which could be achieved due to the smaller injection volume. Essig et al. directly compared 1 M gadobutrol and 0.5 M gadopentate dimeglumine at 1.5 T with a similar dosage of 0.1 mmol/ kg body weight and found no significant differences between the two agents (Essig 2006). The benefit of a double dose of 0.2 mmol was observed only as a trend; however, it was not considered to be of clinical relevance. Similar results in a comparison between the two agents were recently obtained also on a 3-T system (Thilmann et al. 2005). For both agents sufficient high-quality perfusion examinations can be achieved with an acceptable injection volume, which is helpful for their clinical application in daily practice and can be considered superior to standard Gd chelates (Essig et al. 2004; Thiman et al. 2005; Tombach et al. 2003). In a limited number of proof-of-concept studies, USPIO were also used in neuroimaging (Corot et al. 2004; Manniger et al. 2005). The long blood-circulating time and the progressive macrophage uptake in inflammatory tissues of USPIOs are two properties of major importance for pathologic tissue characterization. In the human carotid artery USPIO, accumulation in activated macrophages induced a focal drop in signal intensity compared with unenhanced MRI. The USPIO signal alterations observed in ischemic areas of stroke patients is probably related to the visualization of inflammatory macrophage recruitment into human brain infarction, since animal experiments in such models demonstrated the internalization of USPIO into the macrophages localized in these areas. In brain tumors, USPIO particles that do not pass the ruptured blood– brain barrier at early times post injection can be used to assess tumoral microvascular heterogeneity. Twenty-four hours after injection, when the cellular phase of USPIO takes place, the USPIO tumoral contrast enhancement was higher in high-grade than in low-grade tumors. Several experimental studies and a pilot multiple sclerosis clinical trial in 10 patients have shown that USPIO contrast agents can reveal the presence of inflammatory lesions related to multiple sclerosis. The enhancement with USPIO does not completely overlap with the Gd-chelate enhancement. 2.6.4.2  Contrast Agents in MR Angiography During the last few years magnetic resonance angiography (MRA) has been established as a non-invasive alternative to conventional X-ray angiography in the diagnosis of arteriosclerotic and other vascular diseases (Meany et al. 1997; Meany 1999). With the exception of imaging

intracerebral vessels (Gibbs et al. 2005; Ozsarlak et al. 2004), contrast-enhanced techniques have revealed superiority over non–contrast-enhanced techniques as the time-of-flight (TOF-MRA) or phase-contrast (PC MRA) technique (Sharafuddin et al. 2002). The main advantages over unenhanced techniques are the possibilities to acquire larger volumes, allowing, e.g., demonstration of the carotid artery from its origin to the intracranial portion, shorter acquisition times, and reduced sensibility to flow artifacts. Contrast-enhanced MR angiography can be performed during the first-pass of a contrast agent, preferably in breath-hold technique, after rapid bolus injection or during steady-state conditions after injection of vascular specific blood-pool agents. Most experiences were reported for first-pass MRA after injection of ECF contrast agents. The demands on the agent are a high influence on the signal intensity on blood after injection and the possibility of fast and compact bolus injection. The most commonly applied group of contrast agents are 0.5 molar ECF. In the last years, two novel ECF agents with innovative properties were used for MRA. The first one, the 0.5 M contrast agent gadopentate dimeglumine offers a higher T1 relaxivity. In studies in which gadopentate dimeglumine is compared at equal dose with other Gd-based MR contrast agents without relevant protein binding in plasma, gadopentate dimeglumine has consistently shown significantly better quantitative and qualitative performance (Goyen and Debatin 2003). Even at lower doses compared with gadopentetate dimeglumine injected at a dose of 0.2 mmol/kg body weight, the greater relaxivity of gadopentate dimeglumine provides higher intravascular signal and signal-to-noise ratio (Pediconi et al. 2003). Thus, gadopentate dimeglumine can be considered to have a very favorable risk–benefit ratio for MRA. The second one, gadobutrol is available in 1 M concentration. In combination with a higher relaxivity compared to other ECF agents, the agent has revealed in quantitative evaluations a significant increase in signalto-noise and contrast-to-noise ratios in comparison to gadopentetate dimeglumine in pelvic MRA and in whole body MRA (Goyen et al. 2001, 2003). Better delineation of arterial morphology was reported especially for small vessels, but no statistically significant difference in image quality could be seen. Two different options for injection have been described: reduction of the injection rate by 50% compared to injection protocols using 0.5 M ECF (equimolar dosing) or reduction of the injection time by 50%. The equimolar dosing mainly exploits the higher relaxivity potential of gadobutrol. In this case, the injection duration is identical to a corresponding protocol using a 0.5M contrast agent a similar bolus geometry, and contrast delivery in the ROI is obtained (e.g., in a 70-kg-weighing patient 7 ml of gadobutrol are injected at 1 ml/s compared with 14 ml of gadopentetate dimeglumine injected

2.6  Contrast Agents

at 2 ml/s). Hence, well-known protocols can be adopted with good results. The second option keeps the injection speed unchanged in comparison to the 0.5 M agent protocol, resulting in shortening of the initial bolus duration by a factor of two (Fink et al. 2004). The philosophy is to use a very compact, high-relaxivity bolus and to fully exploit the potential of 1 M gadobutrol. This approach is particularly recommended in conjunction with very fast acquisition techniques, e.g., time-resolved (often referred to as 4D) MRA. Although the effective bolus geometry in the respective ROI is broadened, dependent on individual physiology and mainly influenced by the lung passage, this approach requires higher demands on precise bolus timing and is recommended to users with advanced MRA experience and ultrafast imaging equipment. In addition, a further approach was reported by reducing the amount of contrast agent by a factor of two in abdominal MRA (Vosshenrich et al. 2003). The injection speed was kept constant in comparison to a 0.5 M agent protocol, resulting in very short total bolus duration. Vosshenrich et al. used an amount of 0.1 mmol/kg body weight. They compared the examinations qualitatively and quantitatively to exams acquired after injection of gadopentetate dimeglumine (0.2 mmol/kg), and concluded that for MRA of the hepatic arteries and the portal veins, gadobutrol can be used at half the dosage as recommended for a standard 0.5 M contrast agent. The concept of contrast-enhanced MRA based on ECF agents has some limitations. The primary problem is the rapid extravasation of the contrast agents limiting acquisition time and therefore spatial resolution as well as contrast-to-noise-ratio. To improve spatial resolution it is necessary to prolong imaging time. Intravascular contrast agents are able to overcome the restrictions of

spatial resolution. The longer acquisition period can be used to decrease voxel size, to repeat measurements, or to trigger acquisitions by ECG and/or respiratory gating. The second limitation of currently used MRA is the quantification of artery stenoses, which still seems to be inferior to invasive catheter angiography. The cause is the inferior spatial resolution in MRA using ECF agents, with which the increase of spatial resolution is limited by the acquisition time during first-pass (arterial phase). With intravascular contrast agents, a longer data acquisition during the distribution phase is possible. The spatial resolution can be increased on a similar level compared with catheter angiography and, therefore, the accuracy of stenosis quantification is significantly increased. Optimally, a blood-pool agent permits a long acquisition window including first-pass MRA as well as the possibility of separate imaging of arteries and veins by timing the injection and data acquisition. Gadofosveset, the first MR blood-pool agent approved for clinical use, permits both a high-resolution approach with a long acquisition windows and first-pass contrast-enhanced MRA (Fig. 2.6.1). The approval was based on the data of clinical trials in all different types of arterial vessels including high-flow vessels with large diameter (e.g., the pelvic arteries), low-flow vessels (e.g., foot arteries), and high-flow vessels with a small diameter (e.g., the renal arteries).

Fig. 2.6.1  Whole-body MRA of a healthy volunteer after bolus injection of gadofosveset (0.025 mmol/kg body weight). Firstpass and steady-state acquisition acquired immediately (a) and 10 min (b) after injection of the contrast agent. T1-weighted 3D gradient recalled echo sequence (TR/TE/α 3.1/1.1/25, spatial

resolution 1.6 × 1 × 1.5 mm). First-pass imaging depicts exclusively the arteries. Steady-state imaging shows an enhancement of both arteries and veins. Due to the higher concentration of the contrast agent during first-pass imaging the absolute level of enhancement is higher (a)

2.6.4.3  Contrast Agents for Soft-Tissue Lesions ECF contrast agents are widely used in MR imaging of soft-tissue lesions. The enhancement in either inflammatory or neoplastic lesions makes their use inevitable for the detection and characterization of soft tissue lesions.

99

100

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Relevant anatomical sites that in the daily clinical practice are subject to MR imaging are the female breast and the soft tissue related to the musculoskeletal system. For the female breast, MR imaging with extracellular contrast agents (MR mammography) is nowadays widely used for the detection and for the characterization of unclear breast tumors Morris et al. 2002). The histopathological basis of the different enhancement patterns in breast masses is not yet fully understood; however, it is well known that angiogenesis with the formation of new vessels, is an important aspect (Knopp et al. 1999). The amount of angiogenesis and contrast agent extravasation is considered different for several benign and malignant lesions; however, the visible phenomenon of different enhancement is usually too small to be analyzed only visually. The discrete changes of contrast-agent enhancement are usually evaluated by using a semiquantitative evaluation with region-of-interest measurements at different time-points (Kuhl et al. 2005). The thereby achieved enhancement kinetics, as represented by the time–signal intensity curves, differ significantly for benign and malignant enhancing lesions, and are used as an aid in differential diagnosis. Usually four to six measurements with an interval of 1–2 min are applied in the daily clinical practice (Kuhl et al. 1999; Pediconi et al. 2005). A recently published study showed that the temporal resolution for the assessment of time–signal intensity curves is not as critical as the spatial resolution; therefore, the recommendations for the dynamic postcontrast MR imaging tend toward a 2-min interval with a high spatial resolution (e.g., full 512 imaging matrix) (Kuhl et al. 2005). A more detailed evaluation of perfusion parameters needs, however, a very high temporal resolution in a range of 3–5 s. First results for the differentiation of unclear breast tumors in an investigational setting are very promising; however, due to the high temporal resolution only single slices can be measured, which is not feasible for daily practice (Brix et al. 2004). Usually standard Gd chelates at a dose of 0.1 mmol/ kg body weight are used for contrastenhanced MR mammography. First results indicate that the use of the high-relaxivity MR contrast agent gadopentate dimeglumine in the same dosage can achieve a superior detection and identification of malignant breast lesions at MR imaging as compared with gadopentetate dimeglumine. However, up to now gadopentate dimeglumine is not officially approved for this indication. There are also first approaches to perform MR mammography with blood-pool contrast agents. A major limitation of ECF is that they extravasate nonselectively from the vasculature into the interstitium of both normal and pathological tissues in the breast. It is hypothesized that the degree of microvascular endothelial disruption inherent to cancer vessels with the resulting extravasation of macromolecular contrast agents may predict tumor aggressiveness and tumor grade more accurately that with standard Gd chelates (Daldrup-Link and Brasch 2003;

Daldrup-Link et al. 2003). First results with USPIO have shown an improved characterization of unclear breast tumors at the expense of tumor enhancement, which is important for tumor detection. An interesting approach is also the use of small molecular Gd chelates, which bind reversibly to plasma proteins as for example gadofosveset. This might allow for a sensitivity and specificity due to the presence of small and large molecules (Daldrup-Link et al. 2003). The assessment of microvascular changes in experimental breast tumors seem not to be reliably depicted with theses agents in contrast to the macromolecular albumin-Gd-(DTPA)30 (Daldrup-Link et al. 2003; Turetscheck et al. 2001). However, clinical experience on breast tumors does not exist now. Although potential diagnostic applications have been investigated with various sized albumin-Gd-DTPA, this contrast agent is considered a poor candidate for development as a clinical drug due to slow and incomplete elimination and a potentially immunologic toxicity (Daldrup-Link and Brasch 2003). For soft-tissue or bone lesions in the musculoskeletal system, the application of extracellular Gd-contrast agents has become a clinical standard for characterization, staging of the local extent, biopsy planning, and the therapy monitoring (Verstraete and Lang 2000). The basic principle of contrast-enhanced imaging is as described above the distribution of the Gd chelates in the intravascular space, showing enhancement in tumors with dense vascularity and neoangiogenesis as well as distribution into the extracellular space. For these clinical standard applications, there seem to be no relevant differences in the diagnostic performance between the different extracellular Gd chelates, similar to neuroimaging. The role of Gd-enhanced MRI for exact tissue characterization is still very limited. A differential diagnosis in between different sarcomas, nerve sheath tumors, or other mesenchymal tumors is not possible based on the contrast agent behavior up to know. The differentiation between benign and malignant tumors is also often very limited, even with tools like dynamic time-resolved contrast-enhanced MRI (Verstraete and Lange 2000). Nevertheless, surrogate parameters for angiogenesis like histological tumor-vessel density can be correlated with this method (van Dijke et al. 1996). One major limitation is the extravasation of standard Gd chelates through the intact endothelium so that pathological extravasation in tumor vessels disrupted endothelium cannot be separated from the physiological distribution. Therefore, the experimental studies mainly focus on contrast agents which show no or only minor physiological extravasation. Different studies—mainly in the animal-experimental stage—were able to show that characterization in benign and malignant tumors, evaluation of angiogenesis, and even tumor grading is feasible with blood-pool contrast agents (Daldrup et al. 1998; Kobayashi et al. 2001; Preda et al. 2004a, b). There have been promising results with albumin-Gd-(DTPA); however, as mentioned above

2.6  Contrast Agents

this agent is unlikely to be available for diagnostic use in humans (Daldrup et al. 1998; Daldrup-Link and Brasch 2003). Similar to breast tumors, USPIO have also been utilized for the evaluation of perfusion and for the characterization of soft-tissue tumors in the past (Bentzen et al. 2005). 2.6.4.4  Hepatobiliary Imaging The basic group of contrast agents for hepatobiliary imaging is the group of ECF Gd-based contrast agents. However, there are also tissue-specific contrast agents available, which allow for an increased detection and characterization of focal and diffuse liver disease. Liver-specific contrast agents can be divided into two groups: On the one hand, there are iron-oxide particles (SPIO, or superparamagnetic particles of iron oxide), which are targeted to the reticuloendothelial system (RES) to the so-called Kupffer cells. These agents cause a signal decrease in T2/T2*-weighted sequences by inducing local inhomogeneities of the magnetic field. On the other hand, there is the group of hepatobiliary contrast agents, which are targeted directly to the hepatocyte and are excreted via the bile. These agents cause signal increase in T1-weigthed sequences by shortening of the T1 relaxation time. In Europe there are five different liver specific contrast agents available on the market (Table 2.6.3). The basic principle behind SPIO is the fact, that there are usually no Kupffer cells in malignant liver tumors, in contrast to the normal liver parenchyma and to solid benign liver lesions. Therefore, in the liver specific phase, which starts for ferucarbotran after about ten min and for ferumoxide after about 30 min, high contrast is produced in between malignant liver lesions and normal liver parenchyma. Due to the signal loss in normal liver parenchyma the malignant lesions are contrasted as hyperintense lesions in T2*-weighted and T2-weighted sequences against the dark liver parenchyma. The first SPIO on the market in Europe has been ferumoxide (Endorem®, Guerbet, Aulnay Sous Bois, France). Since 2001 in most

European countries and in Asia the bolus-injectable ferucarbotran (Resovist®, Bayer Schering Pharma AG, Berlin, Germany) is available. With regard to the basic principle of imaging there is no difference between the two agents; however, direct comparative studies have not been performed so far. The most striking advantage of ferucarbotran is the better workflow due to the possibility to inject ferucarbotran as a bolus. Bolus-applicability is possible for ferucarbotran due to the different particle sizes and the coating of the particles; this is also responsible for the fewer rate of side effects (especially fewer events of severe back pain), which are encountered with ferucarbotran. In earlier clinical trials the effects of SPIO particles were evaluated almost exclusively on T2-weighted FSE and T2*-weighted GRE sequences, whereas usually not much attention was paid to the T1-effects. However, the effect of SPIO particles on proton relaxation is not confined to T2 and T2*. They also influence T1 relaxivity with increased signal intensity on T1-weighted GRE sequences at low concentrations (Chambon et al. 1993). This gave raise to the hope, that with the bolus-injectable ferucarbotran vascularity of focal liver lesions could be depicted; however, investigations have been shown, that the ferucarbotran-enhanced early dynamic examination with T1-weighted sequences does not permit to evaluate lesion vascularity, since (with exception of the cotton-wool paddling of hemangioma) the expected enhancement pattern cannot be seen with reliability (Zech et al. 2005). With regard to the T2 /T2* effects there might be differences between both agents, which could be related to the different average particle size of both agents (approximately 150 nm for ferumoxide and 60 nm for ferucarbotran). With help of SPIO-enhanced MR an accurate liver lesion detection can be achieved. There have been several studies comparing SPIO to CT during arterial portography (CTAP), which has been considered as best practice and reference standard. These studies showed detection rates of more than 90% (Ba-Ssalamah et al. 2000; Vogl et al. 2003). In comparison to CTAP this detection rate

Table 2.6.3  Liver-specific MR imaging contrast agents Product name

Generic Name

Group

Manufacturer

Endorem®  FerridexTM

Ferumoxide

SPIO/RES specific

Guerbet, Aulnay Sous Bois, France Berlex Laboratories, NY

Resovist®

Ferucarbotran

SPIO/RES specific

Bayer Schering Pharma AG, Berlin, Germany

Teslascan®

Mangafodipir trisodium

Hepatobiliary

GE Healthcare Biosciences, Buckinghamshire, UK

MultiHance®

Gadopentate dimeglumine

Hepatobiliary

Bracco Imaging, Milano, Italy

Primovist®

Gadoxetic acid

Hepatobiliary

Bayer Schering Pharma AG, Berlin, Germany

101

102

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

was comparable; moreover, SPIO-enhanced MR is more specific than CTAP, in which false positive lesions are encountered frequently. The above cited references investigated SPIO-enhance MR in a mixed collective of patients; publications focusing on the cirrhotic liver showed, that in these patients the combination of SPIO and extracellular Gd-contrast agents have to be considered as the gold-standard for lesion detection (Ward et al. 2000). With regard to lesion characterization SPIO particles can be of help for the differential diagnosis of focal liver lesions based on the cellular composition and function of the different lesions (or rather based on different Kuppfer cell density and function). When the same MR sequence is acquired pre-contrast and after a definite time interval, then the signal loss in normal liver parenchyma and in different focal liver lesions can be quantitatively evaluated. This is helpful for the differentiation of benign and malignant lesions; when a threshold of 25% signal loss is chosen, than lesions with less signal loss are of a malignant nature with over 90% sensitivity and specificity (Namkung and Zech et al. 2007). However, the sequence must have the same parameters (including the same acceleration in case parallel imaging is used), since application of parallel imaging makes systematic changes in the spread of image-noise (Zech et al. 2004). The second important group is the group of hepatobiliary contrast agents. The basic principle behind this group of contrast agents is the specific uptake directly into the hepatocyte. Since the agents all shorten the T1 relaxation times, they cause a signal increase in normal liver parenchyma and in solid benign lesions, whereas in malignant lesions like metastases no specific uptake can be seen. These lesions contrast as hypointense lesions against the bright liver parenchyma. Approved agents in Europe are the manganese-based agent mangafodipir trisodium (Teslascan®, GE Healthcare), and the Gd-based agents gadopentate dimeglumine and gadoxetic acid (Primovist®, Bayer Schering Pharma). Mangafodipir has the drawback that it must not be administered as a bolus, but only as a short infusion; therefore, dynamic studies are not possible with mangafodipir. However, the liver specificity is high and the high uptake in normal liver parenchyma enables imaging of, e.g., metastases with high contrast to the surrounding liver parenchyma. Gadopentate dimeglumine and gadoxetic acid are injectable as boluses. With both contrast agents, a valid early dynamic examination is feasible, allowing differentiation of lesions with regard to their hyper- or hypovascularity (Huppertz et al. 2005; Petersein et al. 2000b). Due to the lower liver specificity of gadopentate dimeglumine, the imaging time-point of the liver-specific phase starts about 40 min after injection; whereas gadoxetic acid allows for imaging at 20 min after injection. This can be of value with regard to the workflow in the MR department. Similar to the situation at SPIO agents, direct comparative studies between the agents have not been

published yet; therefore, the following remarks again hold true for all hepatobiliary contrast agents. However, with regard to lesion characterization, only gadoxetic acid has the official approval to be used for this indication. All three hepatobiliary agents are approved for lesion detection. In comparison to SPIO agents, the potential advantage of hepatobilary agents is the fact that T1-weighted sequences usually can be performed with less acquisition time, less artifacts, and substantial higher spatial resolution. This holds true especially for T1-weighted 3D-GRE sequences derived from MR angiography sequences as, for example, VIBE, or volumetric interpolated breathhold examination (Siemens Medical Solutions, Erlangen, Germany), or LAVA, for liver acquisition with volume acceleration (GE Healthcare). In how far these high-resolution sequences with a slice thickness of usually below 3 mm allow further increasing the detection of small (<1 cm) malignant lesions has to be investigated in the future. The present date for the hepatobiliary agents was acquired mostly with conventional 2D-GRE sequences and a slice thickness between 6 and 8 mm; however even in this setting the detection of lesions <1 cm was improved in comparison to baseline MRI and spiral CT (Bartolozzi et al. 2004; Gehl et al. 2001; Huppertz et al. 2004; Peterseing et al. 2000b). An earlier trial showed slight superiority of SPIO-enhanced MRI versus hepatobiliary MRI in detection of liver metastases; however, the potential advantaged of modern T1-weighted 3D GRE sequences were not available for this evaluation (Del Frate et al. 2002). A recent evaluation showed comparable detection rates between these two contrast agent groups (Kim et al. 2005). For the diagnosis of solid benign liver lesions (as focal nodular hyperplasia [FNH] and hepatocellular adenoma), the basis is still the extracellular contrast agent behavior with flush-like, mostly homogenous arterial hyper­vascularization and fast, but only faint washout, being in portovenous and equilibrium phase mostly slightly hyperintense (and not hypointense in contrast to the strong washout in malignant lesions). Contrast agents used for patients with suspected solid benign lesions must enable that this information can be acquired. Therefore, SPIO agents or mangafodipir alone are not sufficient for this indication; however, especially SPIO can contribute to the diagnosis of these lesions in combination with extracellular contrast agents, which have to show the abovementioned enhancement pattern. With SPIO agents solid benign lesions typically show liver-specific uptake of the substance in the range of normal liver parenchyma, thereby allowing the differentiation from malignant lesions as for example hepatocellular carcinoma (HCC). With regard to the differential diagnosis between FNH and adenoma, results in a limited number of patients indicated that the quantification of iron uptake can be helpful for this issue, because in our cohort, adenoma showed stronger iron uptake in comparison to FNH, with only minimal overlap of the percentage sig-

2.6  Contrast Agents

nal intensity loss (PSIL) measured in a T2-weighted FSE sequence with fat-sat (Namkung and Zech et al. 2007). With the hepatobiliary contrast agents gadopentate dimeglumine and gadoxetic acid, diagnosis of FNH and adenoma is also possible on the one hand based on the extracellular contrast phenomena, on the other based on the liver specific uptake of the agents into these lesions (Grazioli et al. 2001, 2005; Huppertz et al. 2005). There is also valid data indicating that the differentiation between FNH and adenoma is feasible with hepatobiliary contrast agents. Therefore, with the bolus-injectable agents of this class a time- and presumably cost-effective diagnosis can be achieved (Fig. 2.6.2). In patients with extrahepatic malignoma, confirmation or ruling out of liver metastasis is often crucial for the therapeutic management. Moreover, the exact staging of metastatic disease of the liver is getting more and more important, since sophisticated, stage-adapted therapeutic regimens with different options from atypical liver resection over local ablative minimally invasive

treatment (e.g., radiofrequency ablation) up to extended liver resection exist. Therefore, contrast agents used for this indication have to provide an excellent detection rate for focal liver lesions; however, the characterization of these lesions is also important, especially the differential diagnosis of small cystic metastases and small benign cysts or atypical hemangioma. After injection of extracellular contrast agents, the vascularity can be differentiated in hypo- und hypervascular metastases. Hypovascular metastases appear as hypointense lesions in the portovenous phase, whereas hypervascular metastases appear as hyperintense lesions in the arterial-dominant phase. In contrast to the enhancement pattern of benign lesions, the nodular “cotton-wool”-like paddling in hemangioma or the homogeneous enhancement of FNH or adenoma metastases typically show a heterogeneous, ring-like enhancement with strong washout in the portovenous and equilibrium phase, resulting in first hyper-, and then hypointense lesions. However, in very small lesions the morphology of vascularization between the different en-

Fig. 2.6.2  MR images of a 27-year-old male patient with a formerly unclear liver lesion. The primary MR examination (upper row) with gadopentetate dimeglumine and ferucarbotran shows a lesion (arrow) with strong arterial enhancement in the Gd-enhanced T1-weighted 3D-GRE sequence (a) (TR, 5.02 ms; TE, 1.77 ms; flip angle 15°) and ongoing washout in portovenous and equilibrium phase (not shown). The T2-weighted FSE sequence with fat-saturation 10 min after application of ferucarbotran (b) depicts the lesion as nearly liver isointense. The calculated percentage iron uptake compared to the pre-contrast T2-weighted sequence was about 40%, showing the benignity of the lesion. Based on these imaging features and the lobulated margins as

well as the central scar, the diagnosis of a FNH was made. The follow-up study (lower row) was performed with gadoxetic acid after a single bolus injection. In the T1-weighted 3D-GRE sequence (same parameters as above) in the arterial phase (c), the same enhancement characteristics as the in prior study can be delineated. In the delayed T1-weighted 2D-GRE sequence (d), the presence of hepatocytes is proven due to the liver-specific enhancement. Note the excellent delineation of the central scar in the delayed images. In contrast to the upper row, the followup study gave information about vascularity and tissue composition, with a single contrast agent injection only

103

104

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

tities is getting more and more similar, so that the present or missing liver-specific uptake is an additional criterion for differentiation. Several publications have shown that detection of metastasis is feasible with the highest accuracy with help of liver-specific contrast agents, regardless if SPIO or hepatobiliary. According to the literature for detection of liver metastases, all liver-specific agents can be used with a very high diagnostic reliability and superiority to merely extracellular MRI or spiral CT (Bartolozzi et al. 2004; Ba-Ssalamah et al. 2000; Del Frate et al. 2002; Gehl et al. 2001; Huppertz et al. 2004; Kim et al. 2005; Petersein et al. 2000b; Vogl et al. 2003). A difficult situation is liver imaging in cirrhotic liver. It is known that extracellular agents are helpful for detection and characterization of HCC nodules, the sensitivity is with T1-weighted 3D-GRE sequences 76 and 75% specificity (Burrel et al. 2003). Moreover, for diagnosis a HCC according to the accepted guidelines, hypervascularity has to be demonstrated. Therefore, a valid early dynamic phase is a mandatory part for imaging in the cirrhotic liver. Because regenerative nodules, which can be found frequently in the cirrhotic liver, also can show hyper­vascularity differentiation between these nodules and HCC nodules, it is a crucial issue for the management of patients suffering from liver cirrhosis. This is the reason that liver-specific contrast agents play an important role for imaging of the cirrhotic liver. It has been demonstrated that HCC shows no relevant uptake of SPIO particles in contrast to benign regenerative nodules (Bhartia et al. 2003; Imai et al. 2000; Ward et al. 2000; Namkung and Zech et al. 2007). Since in the cirrhotic liver fibrotic areas are present frequently, SPIO alone are not sufficient to evaluate the cirrhotic liver. A reasonable approach for the diagnosis of HCC based on imaging alone is the correlation of hypervascularity and missing or at least decreased iron uptake (Bhartia et al. 2003; Ward et al. 2000). However, there is also an indefinite area of overlapping phenomena between dysplastic nodules and well-differentiated HCC, which is the reason for false negative findings—meaning well-differentiated HCC with substantial iron uptake (Imai et al. 2000). With regard to lesion detection, the availability of high-resolution MR sequences gives advantages for Gd-enhanced arterial-phase imaging alone (Kwak et al. 2004) or in combination with SPIO as double contrast (Ward et al. 2000). Imaging with hepatobiliary contrast agents is considered as inferior in comparison to SPIO agents in the cirrhotic liver, mainly due to substantial overlapping in the liver-specific uptake between well-differentiated HCC and regenerative nodules. 2.6.4.5  Lymph Node Imaging The assessment of lymph node affections by extranodal tissue, i.e., lymph node metastasis, is currently based on

morphologic parameters including, lymph node size, shape, irregular border, and signal intensity inhomogeneities (Brown et al. 2003; Zerhouni et al. 1996). For all parameters, no clear cut-off values or cut-off characteristics can be defined. The definition of a cut-off value in individual studies is the result of finding a compromise between sensitivity and specificity (e.g., larger values around 10 mm give a high specificity but low sensitivity, whereas the reverse of low specificity and high sensitivity is observed when smaller diameter <10 mm are defined). The use of unspecific Gd-based extracellular contrast agents has not revealed to overcome this limitation. Lymphotropic MR contrast agents were, therefore, developed to increase the diagnostic accuracy of positive lymph node involvement. Currently, none of these agents is approved for clinical use, and the experience with different formulations is limited to clinical studies. The most frequently used agents are USPIOs. They are administered intravenously and, as a result of their small diameter and their electrical neutrality, pass the first lymphatic bar­ riers, i.e., the liver and the spleen. In the lymphatic nodes, they are phagocytosed by local macrophages. In healthy lymphatic tissue, the local concentration of iron oxides is resulting in a significant decrease of T2 and T2* relaxation, resulting in a marked decrease of signal in T2- and T2*-weighted sequences. In contrast, metastatic tissue replacing the lymphatic tissue shows no relevant uptake of USPIO, and no relevant change in signal intensity can be observed. Gradient-recalled echo T2-weighted sequences are considered the most accurate to detect the signal loss in nonmetastatic nodes. The application of USPIOs offers not only the possibility to differentiate between tumorfree, reactive (Koh et al. 2004) and tumor-positive lymph nodes, but enables to depict micrometastasis in case sequences when high spatial resolutions are used (Harisinghani et al. 2003). One representative of the group of lymph node–specific USPIO, ferumoxtran-10 (Sinerem® Guerbet, Paris, France) is infused after dilution. The recommended dose is 2.6 mg Fe/kg body weight. The optimal time-point for postcontrast imaging is 24–36 h after application. During their clinical development, USPIOs have shown to be effective in staging lymph nodes of patients with various primary malignancies (Deserno et al. 2004; Jager et al. 1996; Michel et al. 2002; Nguyen et al. 1999). The usual way for diagnosis is to perform an initial precontrast scan and to compare the images with postcontrast images acquired 24–36 h after infusion of USPIOs for signal changes between both time-points. The type, onset, and intensity of adverse events after application of ferumoxtran-10 was evaluated in phase III studies and seems to be similar to those related to infusion of ferumoxides (Anzai et al. 2003). In patients with esophageal or gastric cancer, USPIOs revealed a sensitivity of 100% and specificity between 92.6 and 95.4% (diagnostic accuracy between 94.8 and

2.6  Contrast Agents

96.2%) for diagnosis of metastatic nodes (Nishimura et al. 2006; Tatsumi et al. 2006). In patients with carcinomas of the upper aerodigestive tract, application of ferumoxtran-10 has shown to increase the sensitivity from 64 to 94% while maintaining a specificity of 78.9%, compared with precontrast imaging (Curvo-Semedo et al. 2006). In patients with rectum cancer, USPIOs have shown wellpredictable signal characteristics in normal and reactive lymph nodes, and were able to differentiate the latter from malignant lymph nodes (Koh et al. 2004). Dissimilarly, Keller et al. studied females with uterine carcinoma and were able to show high specificity, but a low sensitivity for metastatic lymph nodes; mainly micro-metastases around 5 mm diameter were missed. A possible way to further improve the diagnostic accuracy for detection of small positive lymph nodes could be the use of 3-T high magnetic field strength scanner resulting in a lower spatial resolution (Heesakkers et al. 2006). Different results were published concerning the necessity of both pre- and postcontrast images. Whereas the majority of clinical publications using USPIO for lymph node imaging used both pre-and postcontrast images and Stets et al. (2002) were able to statistically prove the advantage of pre- and postcontrast studies, Harisinghani et al. (2003) were showed that on ferumoxtran-10–enhanced MR lymphangiography, contrast-enhanced images alone may be sufficient for lymph node characterization. However, a certain level of interpretation experience seems to be required before contrast-enhanced images can be used alone. Both USPIOs (Rogers et al. 1998) and SPIOs (Maza et al. 2006) can alternatively be administered with subcutaneous or submucosal injection. This application route is able to identify sentinel lymph nodes and lymphatic drainage patterns (Fig. 2.6.3). Additionally, high diagnostic accuracy of interstitial MR lymphography using blood-pool Gd-based agents has been described (Herborn et al. 2002, 2003).Using different macromolecular agents or Gd-based agents with high protein binding in animal models, Herborn et al. were able to show that the differentiation of tumor-bearing lymph nodes from reactive inflammatory and normal nodes based on a contrast uptake pattern assessed qualitatively as well as quantitatively is possible. In difference to the intravenous administration, subcutaneous injection gives the possibility to acquire the MR images as early as some min after application. The use of Gd-based non-lymphotropic blood-pool agents induced a relatively short and inhomogeneous lymph node enhancement (Misselwitz et al. 2004). With the aim to become more specific, a new generation of lymphotropic T1 contrast agents was developed and tested in animal models after subcutaneous injection. These perfluorinated Gd chelates were able to visualize fine lymphatic vasculature, even the thoracic duct in animal models (Staatz et al. 2001).

Fig. 2.6.3  Lymphatic drainage of a mucosal melanoma in the left nasal cavity to a lymph node located in the left submandibular region. MR images (upper row) and SPECT images (lower row). a Concordant alignment of hot spots caused by the skin markers on the SPECT images with the vitamin E caps on the MR images (yellow arrows). b Accurate sentinel lymph node localization (blue arrow) after subcutaneous injection of ferucarbotran (mg Fe/kg body weight). T2*-weighted 2D gradient–recalled echo sequence TR /TE /a 997/15/90. Homogeneous signal intensity decrease in the depicted lymph node (arrow) indicates normal lymphatic tissue and, thereby, metastatic involvement can be ruled out (Maza et al. 2006)

2.6.4.6  Gastrointestinal Imaging Bowel MR contrast agents are generally classified as either positive (bright lumen) or negative (dark lumen) agents. In addition to enteral contrast agents especially approved for MR imaging, several existing pharmaceutical agents, such as methyl cellulose, mannitol, and polyethylene glycol preparations, licensed for other enteric application than MRI, have also been exploited.

105

106

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

The specific demands for enteral contrast media include: • Opacification of intraluminal space for better delineation of bowel structures and clear differentiation to extraintestinal structures • Suppression of bowel signal in non-enteral indications (MRCP and MR urography) • No interaction with enteral mucosa for clear delineation of bowel wall • Distension of bowel for visualization of the entire wall and lumen • High contrast between bowel lumen and bowel mucosa • Excellent safety profile and no contrast absorption trough the enteral mucosa For enteral and rectal application a special formulation of gadopentetate dimeglumine was developed (Magnevist enteral, Schering). The agent contains a total of 15 g mannitol/l to prevent the absorption of the fluid simultaneously introduced in the GI tract, thus allowing homogeneous filling, distension, and constant Gd concentration during the entire examination period. In the early development, an increase in diagnostic accuracy in examinations of the pancreas in the diagnostics of abdominal lymphoma and in pelvic MR imaging was shown (Claussen et al. 1988). Negative agents provide desirable contrast to those pathologic processes that are signal intense. They have been shown to improve the quality of images obtained by techniques such as MR cholangiopancreaticography (MRCP) and MR urography by eliminating unwanted signal from fluid-containing adjacent bowel loops, thus allowing better visualization of the pancreatic/biliary ducts and the urinary tract. An alternative to oral SPIOs was described by using ordinary pineapple juice. It was demonstrated that pineapple juice decreased T2 signal intensity on a standard MRCP sequence to a similar degree than a commercially available negative contrast agent (ferumoxsil) (Riordan et al. 2004). Oral SPIO preparations usually contain larger particles than injectable agents do. In Europe, two SPIO preparations with the INN code ferumoxsil are approved for oral use: Lumirem (Laboratoires Guerbet, France) with a particle size of 300 nm, and (oral magnetic particles) Abdoscan (GE Healthcare) with a particle size of 350 nm. They are coated with a non-biodegradable and insoluble matrix (siloxane for Lumirem and polystyrene for Abdoscan), and suspended in viscosity-increasing agents (usually based on ordinary food additives, such as starch and cellulose). These preparations can prevent the ingested iron from being absorbed, particles from aggregating, and improve homogeneous contrast distribution throughout the bowel. If SPIO particle aggregating occurs, magnetic susceptibility artifact may result, especially when high magnetic field strength and gradientecho pulse sequence are used (Wang et al. 2001).

Lumirem is composed of crystals of approximately 10 nm; the hydrodynamic diameter is approximately 300 nm (Debatin and Patak 1999). The recommended concentration is 1.5–3.9 mmol Fe/l. Oral SPIO are administered over 30–60 min, with a volume of 900 ml for contrast enhancement of the whole abdomen, and 400 ml for imaging of the upper abdomen. Oral SPIO suspensions are well tolerated by the patients (Haldemann et al. 1995); the iron is not absorbed and the intestinal mucosal membrane is not irritated. Combination of Gd-enhanced T1-weighted sequences and T2-weighted sequences after oral contrast with SPIO has revealed highest accuracy in the evaluation of Crohn’s disease (Maccioni et al. 2006). Furthermore, it has been shown that MRI with negative superparamagnetic oral contrast is comparable to endoscopy in the assessment of ulcerative colitis. In difference to patients with Crohn’s disease, the double-contrast imaging does not provide more information than single oral contrast (De Ridder et al. 2001). In MRCP, negative oral contrast agents can be given before the examination to provide non-superimposed visualization of the bile and pancreatic ducts. There is no negative influence of the oral contrast agents on the diameter of the ducts (Petersein et al. 2000a). 2.6.4.7  Cardiac Imaging In cardiac MRI, contrast agents are obligatory for the assessment of myocardial perfusion, for the evaluation of enhancement of cardiac masses, and for the evaluation of myocardial viability. In addition, contrast agents are frequently used when MR angiography of the coronary arteries is performed. 2.6.4.7.1  Myocardial Perfusion Myocardial perfusion imaging is a promising and rapidly increasing field in cardiac MR imaging. In comparison to radionuclide techniques, MR imaging has several advantages, including higher spatial resolution, no radiation exposure, and no attenuation problem related to anatomical limitations. The examination is performed after rapid intravenous administration (e.g., 3–5 ml/sec) of a contrast agent and evaluation of the first-pass transit of the agent through the myocardium. With the use of fast scan techniques, perfusion imaging can be performed as a multislice technique, with imaging of three to five slice levels per heartbeat, possibly allowing coverage of the entire ventricle. From a series of images, signal intensity–time curves are derived from regions of interest in the myocardial tissue for generation of parametric images. The majority of data are published using ECF agents. In clinical practice, most investigators are using fast T1-weighted imaging and bolus injection of doses of 0.025 up to 0.05 mmol/kg body weight (Edelman 2004).

2.6  Contrast Agents

For the evaluation, both quantitative and qualitative approaches can be used. In case of ECF agents, generally semiquantitative assessments are applied. To quantify myocardial perfusion a calculation was published by Wilke et al. on the basis of first-pass data acquired after fast bolus injection of 0.025 mmol/kg body weight of ECF agents (Wilke et al. 1997). In practice, Gd concentrations between 0.2 and 1.2 mmol/l result in a linear progression of the MR signal compared with the concentration of the agent itself. Above this dose, the maximal relative increase in signal intensity begins to saturate (Schwitter et al. 1997). When the evaluation is performed by visual evaluation, a higher dose of 0.1–0.2 mmol/kg should be preferred to reach better myocardial enhancement and image quality. Good correlation between the perfusion reserve with MR imaging and the coronary flow reserve with Doppler ultrasonography could be proven. Blood-pool agents have the potential to be applied for quantitative measurements also because their volume of distribution is limited to the intravascular space. A requirement for quantitative perfusion measurements is that the relation between the measured signal intensity on the MR images and the contrast agent concentration in the blood is known (Brasch 1991). There are two major differences between first-pass curves obtained from blood-pool agents and extracellular contrast agents. First, blood-pool contrast agents reach a lower tissue signal because their volume of distribution is limited to the intravascular space (Wendland et al. 1997; Wilke et al. 1995). Second, there is a better return to baseline for blood-pool contrast agents. The wash-in kinetics and the signal intensity in the myocardial tissue depend on the concentration of the contrast agent, the coronary flow rate, diffusion of the contrast agent into the interstitium, relative tissue volume fractions, bolus duration, and recirculation effects (Burstein et al. 1991). Absolute quantification of myocardial perfusion has been performed in animal models using NC100150. A high absolute quantification correlation was found between MRI and contrast-enhanced ultrasound (Johansson et al. 2002). 2.6.4.7.2  Myocardial Viability Delayed enhancement allows direct visualization of necrotic or scarred tissue and is an easy and robust method to assess myocardial viability. By measuring the transmural extent of late enhancement, a prognosis toward the degree of functional recovery of cardiac tissue may be possible. Although several studies have been aimed at describing the mechanisms of late enhancement, these could not be fully explained up to now. The extent of late enhancement possibly depends on the time-point after injection as well as the time-point after myocardial infarction. Relevant publications about delayed enhance-

ment are reporting data after the administration of 0.5 M ECF agents in a dose range of 0.1–0.2 mmol/kg. ECF agents are probably more efficient in assessing the cellular integrity when they are distributed homogenously through damaged myocardium (Wendland et al. 1997), but homogenous distribution is not always the case, as in microvascular obstruction (Kroft and de Roos 1995). Differences exist between the distribution patterns of extracellular and blood-pool agents, and hypo-enhanced cores may be observed earlier using blood-pool agents (Schwitter et al. 1997). The sensitivity of blood-pool agents for myocardial infarction and, therefore, their potential value for the evaluation of myocardial viability, is unknown. A different strategy to determine myocardial viability is the use of necrosis-specific MR contrast agents. Gadophrin-2 and -3 (Schering) have shown to possess a marked and specific affinity for necrotic tissue components and showed persistent enhancement in necrotic tissue (40 min to 12 h in myocardial infarction). In preclinical studies, the agent has not proven superiority in estimation of infarcts compared with ECF agents, and the further development of the agent was, therefore, not continued (Barkhausen et al. 2002). 2.6.4.7.3  Coronary MRA To depict coronary arteries in MRI, both unenhanced and contrast-enhanced techniques are used. A frequently performed contrast-enhanced examination strategy with acquisition of multiple 3D slaps in breath-hold depicting each coronary artery separately was primarily described by Wielopolski et al. (1998). For 3D coronary MR angiography, however, the contrast between blood and myocardium in relation to the inflow of unsaturated protons is reduced. Thus, the use of an intravascular contrast agent may be particularly convenient due to the T1 relaxation time reduction in blood. The application of ECF agents has proven to be most effective for breath‑hold acquisitions. However, the concentration of ECF agents declines rapidly as they extravasate into the interstitial space, thereby reducing the contrast between blood and myocardium. Newly developed strategies use high-resolution free-breathing MR sequences for coronary MRA. In this situation, ECF agents are less beneficial due to the relatively long acquisition time of these free breathing 3D sequences. This problem can be solved by the use of intravascular contrast agents. An additional benefit from application of a blood-pool agent is a longer acquisition window, which may be used to further increase both the signal-to-noise ratio and/or image resolution (Nassenstein et al. 2006). In an animal model use of the macromolecular Gd-based agent P792 with a free-breathing technique allowed more distal visualization of the coronary arteries than did an ECF agent or non-enhanced MR images (Dirksen et al. 2003).

107

108

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Akeson P, Jonsson E, Haugen I, Holtas S (1995) Contrastenhanced MRI of the central nervous system: comparison between gadodiamide injection and Gd-DTPA. Neuroradiology 37:229–233 Anzai Y, Piccoli CW, Outwater EK, Stanford W, Bluemke DA, Nurenberg P, Saini S, Maravilla KR, Feldman DE, Schmiedl UP, Brunberg JA, Francis IR, Harms SE, Som PM, Tempany CM (2003) Evaluation of neck and body metastases to nodes with ferumoxtran 10-enhanced MR imaging: phase III safety and efficacy study. Radiology 228:777–788 Barkhausen J, Ebert W, Debatin JF, Weinmann HJ (2002) Imaging of myocardial infarction: comparison of magnevist and gadophrin-3 in rabbits. J Am Coll Cardiol 39:1392–1398 Bartolozzi C, Donati F, Cioni D et al (2004) Detection of colorectal liver metastases: a prospective multicenter trial comparing unenhanced MRI, MnDPDP-enhanced MRI, and spiral CT. Eur Radiol 14:14–20 Ba-Ssalamah A, Heinz-Peer G, Schima W et al (2000) Detection of focal hepatic lesions: comparison of unenhanced and SHU 555 A-enhanced MR imaging versus biphasic helical CTAP. J Magn Reson Imaging 11:665–672 Bastianello S, Gasperini C, Paolillo A et al (1998) Sensitivity of enhanced MR in multiple sclerosis: effects of contrast dose and magnetization transfer contrast. AJNR 19:1863–1867 Bellin MF, Lebleu L, Meric JB (2003) Evaluation of retroperitoneal and pelvic lymph node metastases with MRI and MR lymphangiography. Abdom Imaging 28:155–163 Bentzen L, Vestergaard-Poulsen P, Nielsen T et al (2005) Intravascular contrast agent-enhanced MRI measuring contrast clearance and tumor blood volume and the effects of vascular modifiers in an experimental tumor. Int J Radiat Oncol Biol Phys 61:1208–1215 Bhartia B, Ward J, Guthrie JA, Robinson PJ (2003) Hepatocellular carcinoma in cirrhotic livers: double-contrast thin-section MR imaging with pathologic correlation of explanted tissue. AJR Am J Roentgenol 180:577–584 Bluemke DA, Sahani D, Amendola M, Balzer T, Breuer J et al (2005) Efficacy and safety of MR imaging with liver-specific contrast agent: US multicenter phase III study. Radiology 237:89–98 Bozzao A, Floris R, Baviera ME, Apruzzese A, Simonetti G (2001) Diffusion and perfusion MR imaging in cases of Alzheimer’s disease: correlations with cortical atrophy and lesion load. AJNR Am J Neuroradiol 22:1030–1036 Brasch RC (1991) Rationale and applications for macromolecular Gd-based contrast agents. Magn Reson Med 22:282–287 Brix G, Kiessling F, Lucht R et al (2004) Microcirculation and microvasculature in breast tumors: pharmacokinetic analysis of dynamic MR image series. Magn Reson Med 52:420–429

14. Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA (2007) Gadodiamide-associated nephrogenic systemic fibrosis: why radiologists should be concerned. AJR Am J Roentgenol 188:586–592 15. Brown G, Richards CJ, Bourne MW, Newcombe RG, Radcliffe AG, Dallimore NS, Williams GT (2003) Morphologic predictors of lymph node status in rectal cancer with use of high-spatial-resolution MR imaging with histopathologic comparison. Radiology 227:371–377 16. Bruening R, Berchtenbreiter C, Holzknecht N, Essig M et al (2000) Effects of three different doses of a bolus injection of gadodiamide: assessment of regional cerebral blood volume maps in a blinded reader study. AJNR Am J Neuroradiol 21:1603–1610 17. Brugieres P, Gaston A, Degryse HR et al (1994) Randomized double blind trial of the safety and efficacy of two Gd complexes (Gd-DTPA and Gd-DOTA). Neuroradiology 36:27–30 18. Bundesinstitut für Arzneimittel und Medizinprodukte (Federal Institute for Drugs and Medical Devices) (2007) Public assessment reports increased risk of nephrogenic fibrosing dermopathy/nephrogenic systemic fibrosis and Gd-containing MRI contrast agents. http://www.bfarm.de. Cited 21 March 2007 19. Burrel M, Llovet JM, Ayuso C et al (2003) MRI angiography is superior to helical CT for detection of HCC prior to liver transplantation: an explant correlation. Hepatology 38:1034–1042 20. Burstein D, Taratuta E, Manning WJ (1991) Factors in myocardial “perfusion” imaging with ultrafast MRI and Gd-DTPA administration. Magn Reson Med 20:299–305 21. Cavagna FM, Lorusso V, Anelli PL, Maggioni F, de Haen C (2002) Preclinical profile and clinical potential of gadocoletic acid trisodium salt (B22956/1), a new intravascular contrast medium for MRI. Acad Radiol 9(Suppl 2): S491–S494 22. Chambon C, Clement O, Blanche RL et al (1993) Superparamagnetic iron oxides as positive MR contrast agents: in vitro and in vivo evidence. Magn Reson Imaging 11:509–519 23. Claussen C, Kornmesser W, Laniado M, Kaminsky S, Hamm B, Felix R (1988) Oral contrast media for magnetic resonance tomography of the abdomen. III. Initial patient research with Gd-DTPA. RoFo 148:663–689 24. Colosimo C, Ruscalleda J, Korves M et al (2001) Detection of intracranial metastases: a multicenter, intrapatient comparison of gadopentate dimeglumine-enhanced MRI with routinely used contrast agents at equal dosage. Invest Radiol 36:72–81 25. Colosimo C, Knopp MV, Barreau X et al (2004) A comparison of Gd-BOPTA and Gd-DOTA for contrast-enhanced MRI of intracranial tumors. Neuroradiology 46:655–665

2.6  Contrast Agents 26. Colosimo C, Demaerel P, Tortori-Donati P et al (2005) Comparison of gadopentate dimeglumine (Gd-BOPTA) with gadopentetate dimeglumine (Gd-DTPA) for enhanced MR imaging of brain and spine tumors in children. Pediatr Radiol 35:501–510 27. Corot C, Petry KG, Trivedi R, Saleh A, Jonkmanns C, Le Bas JF, Blezer E, Rausch M, Brochet B, Foster-Gareau P, Baleriaux D, Gaillard S, Dousset V (2004) Macrophage imaging in central nervous system and in carotid atherosclerotic plaque using ultrasmall superparamagnetic iron oxide in magnetic resonance imaging. Invest Radiol 39:619–625 28. Curvo-Semedo L, Diniz M, Migueis J, Juliao MJ, Martins P, Pinto A, Caseiro-Alves F (2006) USPIO-enhanced magnetic resonance imaging for nodal staging in patients with head and neck cancer. J Magn Reson Imaging 24:128–131 29. Daldrup-Link HE, Brasch RC (2003) Macromolecular contrast agents for MR mammography: current status. Eur Radiol 13:354–365 30. Daldrup H, Shames DM, Wendland M et al (1998) Correlation of dynamic contrast-enhanced magnetic resonance imaging with histologic tumor grade: comparison of macromolecular and small-molecular contrast media. Pediatr Radiol 28:67–78 31. Daldrup-Link HE, Rydland J, Helbich TH et al (2003) Quantification of breast tumor microvascular permeability with feruglose-enhanced MR imaging: initial phase II multicenter trial. Radiology 229:885–892 32. D’Arienzo A, Scaglione G, Vicinanza G, Manguso F, Bennato R, Belfiore G, Imbriaco M, Mazzacca G (2000) Magnetic resonance imaging with ferumoxil, a negative superparamagnetic oral contrast agent, in the evaluation of ulcerative colitis. Am J Gastroenterol 95:720–724 33. De Ridder F, De Maeseneer M, Stadnik T, Luypaert R, Osteaux M (2001) Severe adverse reactions with contrast agents for magnetic resonance: clinical experience in 30,000 MR examinations. JBR-BTR 84:150–152 34. Debatin JF, Patak MA (1999) MRI of the small and large bowel. Eur Radiol 9:1523–1534 35. Del Frate C, Bazzocchi M, Mortele KJ et al (2002) Detection of liver metastases: comparison of gadopentate dimeglumine-enhanced and ferumoxides-enhanced MR imaging examinations. Radiology 225:766–772 36. Deserno WM, Harisinghani MG, Taupitz M et al (2004) Urinary bladder cancer: preoperative nodal staging with ferumoxtran-10-enhanced MR imaging. Radiology 233:449–456 37. Dijke C van, Brasch R, Roberts T et al (1996) Mammary carcinoma model: correlation of macromolecular contrastenhanced MR imaging characterization of tumor microvasculature and histologic capillary density. Radiology 198:813–818 38. Dirksen MS, Lamb HJ, Kunz P, Robert P, Corot C, de Roos A (2003) Improved MR coronary angiography with use of a new rapid clearance blood pool contrast agent in pigs. Radiology 227:802–808

39. Doerfler A, Eckstein HH, Eichbaum M et al (2001) Perfusion-weighted magnetic resonance imaging in patients with carotid artery disease before and after carotid endarterectomy. J Vasc Surg 34:587–593 40. Edelman RR (2004) Contrast-enhanced MR imaging of the heart: overview of the literature. Radiology 232:653–668 41. Essig M, Wenz F, Scholdei R et al (2002) Effect of contrast media dosage and contrast media extravasation on dynamic susceptibility-contrast enhanced dynamic echo planar imaging of cerebral gliomas. Acta Radiol 43:354–359 42. Essig M, Waschkies M, Wenz F, Debus J, Hentrich HR, Knopp MV (2003) Assessment of brain metastases by means of dynamic susceptibility contrast enhanced MRI – initial results. Radiology 228:193–199 43. Essig M, Giesel F, Le-Huu M, Stieltjes B, von Tengg H, Weber MA (2004) Perfusion MRI in CNS disease: current concepts. Neuroradiology 46(Suppl 2):S201–S207 44. Essig M, Lodemann KP, Le-Huu M, Bruning R, Kirchin M, Reith W (2006) Intraindividual comparison of gadopentate dimeglumine and gadobutrol for cerebral magnetic resonance perfusion imaging at 1.5 T. Invest Radiol 41:256–263 45. Fink C, Bock M, Kiessling F, Lichy MP, Krissak R, Zuna I, Schmahl A, Delorme S, Kauczor HU (2004) Time-resolved contrast-enhanced three-dimensional pulmonary MR-angiography: 1 M gadobutrol vs. 0.5 M gadopentetate dimeglumine. J Magn Reson Imaging 19:202–208 46. Gehl HB, Bourne M, Grazioli L, Moller A, Lodemann KP (2001) Off-site evaluation of liver lesion detection by GdBOPTA-enhanced MR imaging. Eur Radiol 11:187–192 47. Gibbs GF, Huston J 3rd, Bernstein MA, Riederer SJ, Brown RD Jr (2005) 3.0-Tesla MR angiography of intracranial aneurysms: comparison of time-of-flight and contrast-enhanced techniques. J Magn Reson Imaging 21:97–102 48. Goyen M, Debatin JF (2003) Gadopentate dimeglumine (MultiHance) for magnetic resonance angiography: review of the literature. Eur Radiol 13(Suppl 3):N19–N27 49. Goyen M, Lauenstein TC, Herborn CU, Debatin JF, Bosk S, Ruehm SG (2001) 0.5 M Gd chelate (Magnevist) versus 1.0 M Gd chelate (Gadovist): dose-independent effect on image quality of pelvic three-dimensional MR-angiography. J Magn Reson Imaging 14:602–607 50. Goyen M, Herborn CU, Vogt FM, Kroger K, Verhagen R, Yang F, Bosk S, Debatin JF, Ruehm SG (2003) Using a 1 M Gd-chelate (gadobutrol) for total-body three-dimensional MR angiography: Preliminary experience. J Magn Reson Imaging 17:565-571 51. Goyen M, Edelman M, Perreault P, O‘Riordan E, Bertoni H, Taylor J, Siragusa D, Sharafuddin M, Mohler ER 3rd, Breger R, Yucel EK, Shamsi K, Weisskoff RM (2005) MR angiography of aortoiliac occlusive disease: a phase III study of the safety and effectiveness of the blood-pool contrast agent MS-325. Radiology 236:825–833 52. Grazioli L, Morana G, Federle MP et al (2001) Focal nodular hyperplasia: morphologic and functional information from MR imaging with gadopentate dimeglumine. Radiology 221:731–739

109

110

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 53. Grazioli L, Morana G, Kirchin MA, Schneider G (2005) Accurate differentiation of focal nodular hyperplasia from hepatic adenoma at gadopentate dimeglumine-enhanced MR imaging: prospective study. Radiology 236:166–177 54. Grossman RI, Rubin DI, Hunter G et al (2000) Magnetic resonance imaging in patients with central nervous system pathology: a comparison of OptiMARK (GdDTPABMEA)and Magnevist (Gd-DTPA). Invest Radiol 35:412–419 55. Haen C, Anelli PL, Lorusso V, Morisetti A, Maggioni F, Zheng J, Uggeri F, Cavagna FM (2006) Gadocoletic acid trisodium salt (b22956/1): a new blood pool magnetic resonance contrast agent with application in coronary angiography. Invest Radiol 41:279–291 56. Halavaara J, Breuer J, Ayuso C, Balzer T, Bellin MF et al (2006) Liver tumor characterization: comparison between liver-specific gadoxetic acid disodium-enhanced MRI and biphasic CT—a multicenter trial. J Comput Assist Tomogr 30:345–354 57. Haldemann Heusler RC, Wight E, Marincek B (1995) Oral superparamagnetic contrast agent (ferumoxsil): tolerance and efficacy in MR imaging of gynecologic diseases. J Magn Reson Imaging 5:385–391 58. Harisinghani MG, Barentsz J, Hahn PF et al (2003) Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 348:2491–2499 59. Haustein J, Laniado M, Niendorf HP et al (1993) Tripledose versus standard-dose gadopentetate dimeglumine: a randomized study in 199 patients. Radiology 186:855–860 60. Heesakkers RA, Futterer JJ, Hovels AM, van den Bosch HC, Scheenen TW, Hoogeveen YL, Barentsz JO (2006) Prostate cancer evaluated with ferumoxtran-10-enhanced T2*-weighted MR Imaging at 1.5 and 3.0 T: early experience. Radiology 239:481–487 61. Heiland S, Reith W, Forsting M, Sartor K (2001) How do concentration and dosage of the contrast agent affect the signal change in perfusion-weighted magnetic resonance imaging? A computer simulation. Magn Reson Imaging 19:813–820 62. Herborn CU, Lauenstein TC, Vogt FM, Lauffer RB, Debatin JF, Ruehm SG (2002) Interstitial MR lymphography with MS-325: characterization of normal and tumor-invaded lymph nodes in a rabbit model. AJR Am J Roentgenol 179:1567–1572 63. Herborn CU, Vogt FM, Lauenstein TC, Goyen M, Dirsch O, Corot C, Debatin JF, Ruehm SG (2003) Assessment of normal, inflammatory, and tumor-bearing lymph nodes with contrast-enhanced interstitial magnetic resonance lymphography: preliminary results in rabbits. J Magn Reson Imaging 18:328–335 64. Huppertz A, Balzer T, Blakeborough A et al (2004) Improved detection of focal liver lesions at MR imaging: multicenter comparison of gadoxetic acid-enhanced MR images with intraoperative findings. Radiology 230:266–275

65. Huppertz A, Haraida S, Kraus A, Zech CJ et al (2005) Enhancement of focal liver lesions at gadoxetic acid-enhanced MR imaging: correlation with histopathologic findings and spiral CT—initial observations. Radiology 234:468–478 66. Imai Y, Murakami T, Yoshida S et al (2000) Superparamagnetic iron oxide-enhanced magnetic resonance images of hepatocellular carcinoma: correlation with histological grading. Hepatology 32:205–212 67. Jager GJ, Barentsz JO, Oosterhof GO, Witjes JA, Ruijs SJ (1996) Pelvic adenopathy in prostatic and urinary bladder carcinoma: MR imaging with a three-dimensional TI-weighted magnetization-prepared-rapid gradient-echo sequence. AJR Am J Roentgenol 167:1503–1507 68. Johansson LO, Bjerner T, Bjornerud A, Ahlstrom H, Tarlo KS, Lorenz CH (2002) Utility of NC100150 injection in cardiac MRI. Acad Radiol 9(Suppl 1):S79-S81 69. Kim YK, Lee JM, Kim CS, Chung GH, Kim CY, Kim IH (2005) Detection of liver metastases: gadopentate dimeglumine-enhanced three-dimensional dynamic phases and one-h delayed phase MR imaging versus superparamagnetic iron oxide-enhanced MR imaging. Eur Radiol 15:220–228 70. Kirchin MA, Pirovano G, Venetianer C, Spinazzi A (2001) Safety assessment of gadopentate dimeglumine (MultiHance): extended clinical experience from phase I studies to post-marketing surveillance. J Magn Reson Imaging 14:281–294 71. Knopp MV, Weiss E, Sinn HP et al (1999) Pathophysiologic basis of contrast enhancement in breast tumors. J Magn Reson Imag 10:260–266 72. Knopp MV, Runge VM, Essig M et al (2004) Primary and secondary brain tumors at MR imaging: bicentric intraindividual crossover comparison of gadopentate dimeglumine and gadopentetate dimeglumine. Radiology 230:55–64 73. Kobayashi H, Sato N, Kawamoto S et al (2001) 3D MR angiography of intratumoral vasculature using a novel macromolecular MR contrast agent. Magn Reson Med 46:579–585 74. Koh DM, Brown G, Temple L et al (2004) Rectal cancer: mesorectal lymph nodes at MR imaging with USPIO versus histopathologic findings – initial observations. Radiology 231:91–99 75. Kroft LJ, de Roos A (1995) Blood pool contrast agents for cardiovascular MR Imaging. J Magn Reson Imaging 10:395–403 76. Kuhl CK, Mielcareck P, Klaschik S et al (1999) Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 211:101–110 77. Kuhl CK, Schild HH, Morakkabati N (2005) Dynamic bilateral contrast-enhanced MR imaging of the breast: tradeoff between spatial and temporal resolution. Radiology 236:789–800

2.6  Contrast Agents 78. Kwak HS, Lee JM, Kim CS (2004) Preoperative detection of hepatocellular carcinoma: comparison of combined contrast-enhanced MR imaging and combined CT during arterial portography and CT hepatic arteriography. Eur Radiol 14:447–457 79. La Noce A, Stoelben S, Scheffler K, Hennig J, Lenz HM, La Ferla R, Lorusso V, Maggioni F, Cavagna F (2002) B22956/1, a new intravascular contrast agent for MRI: first administration to humans—preliminary results. Acad Radiol 9(Suppl 2):S404–S406 80. Lauffer RB, Parmelee DJ, Dunham SU, Ouellet HS, Dolan RP, Witte S, McMurry TJ, Walovitch RC (1998) MS-325: albumin-targeted contrast agent for MR angiography. Radiology 207:529–538 81. Le Duc G, Corde S, Charvet AM et al (2004) In vivo measurement of Gd concentration in a rat glioma model by monochromatic quantitative computed tomography: comparison between gadopentetate dimeglumine and gadobutrol. Invest Radiol 39:385–393 82. Li A, Wong CS, Wong MK, Lee CM, Au Yeung MC (2006) Acute adverse reactions to magnetic resonance contrast media—Gd chelates. Br J Radiol 79:368–371 83. Maccioni F, Bruni A, Viscido A, Colaiacomo MC, Cocco A, Montesani C, Caprilli R, Marini M (2006) MR imaging in patients with Crohn disease: value of T2- versus T1weighted Gd-enhanced MR sequences with use of an oral superparamagnetic contrast agent. Radiology 238:517–530 84. Manninger SP, Muldoon LL, Nesbit G, Murillo T, Jacobs PM, Neuwelt EA (2005) An exploratory study of ferumoxtran-10 nanoparticles as a blood–brain barrier imaging agent targeting phagocytic cells in CNS inflammatory lesions. AJNR Am J Neuroradiol 26:2290–2300 85. Maza S, Taupitz M, Taymorian K, Winzer KJ, Rückert J, Paschen C, Räber G, Schneider S, Trefzer U, Munz DL (2006) Multimodal fusion imaging ensemble for targeted sentinel lymph node management: initial results of an innovative promising approach for anatomically difficult lymphatic drainage in different tumor entities. Eur J Nucl Med Mol Imaging 34:378–383 86. Meaney JF (1999) Non-invasive evaluation of the visceral arteries with magnetic resonance angiography. Eur Radiol 9:1267–1276 87. Meaney JF, Weg JG, Chenevert TL, Stafford-Johnson D, Hamilton BH, Prince MR (1997) Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 336:1422–1427 88. Michel SC, Keller TM, Frohlich JM, Fink D, Caduff R, Seifert B, Marincek B, Kubik-Huch RA (2002) Preoperative breast cancer staging: MR imaging of the axilla with ultrasmall superparamagnetic iron oxide enhancement. Radiology 225:527–536 89. Misselwitz B, Platzek J, Weinmann HJ (2004) Early MR lymphography with gadofluorine M in rabbits. Radiology 231:682–688 90. Morris EA. Breast cancer imaging with MRI (2002) Radiol Clin North Am 40:443–466

91. Murphy KJ, Brunberg JA, Cohan RH (1996) Adverse reactions to Gd contrast media: a review of 36 cases. AJR Am J Roentgenol 167:847–849 92. Namkung S, Zech CJ, Helmberger T, Reiser MF, Schönberg SO (2007) Superparamagnetic iron oxide (SPIO)-enhanced liver MR Imaging with ferucarbotran: efficacy for characterization of focal liver lesions. J Magn Reson Imaging 25:755–765 93. Nassenstein K, Waltering KU, Eggebrecht H, Schlosser T, Hunold P, Barkhausen J (2006) [MR coronary angiography with MS-325, a blood pool contrast agent: comparison of an inversion recovery steady-state free precession with an inversion recovery fast low angle shot sequence in volunteers]. RoFo 178:508–514 94. Nguyen BC, Stanford W, Thompson BH, Rossi NP, Kernstine KH, Kern JA, Robinson RA, Amorosa JK, Mammone JF, Outwater EK (1999) Multicenter clinical trial of ultrasmall superparamagnetic iron oxide in the evaluation of mediastinal lymph nodes in patients with primary lung carcinoma. J Magn Reson Imaging 10:468–473 95. Niendorf HP, Alhassan A, Geens VR, Clauss W (1994) Safety review of gadopentetate dimeglumine. Extended clinical experience after more than five million applications. Invest Radiol 29(Suppl 2):S179–S182 96. Nishimura H, Tanigawa N, Hiramatsu M, Tatsumi Y, Matsuki M, Narabayashi I (2006) Preoperative esophageal cancer staging: magnetic resonance imaging of lymph node with ferumoxtran-10, an ultrasmall superparamagnetic iron oxide. J Am Coll Surg 202:604–611 97. Oudkerk M, Sijens PE, van Beek EJ, Kuijpers TJ (1995) Safety and efficacy of Dotarem (Gd-DOTA) versus Magnevist (Gd-DTPA) in magnetic resonance imaging of the central nervous system. Invest Radiol 30:75–78 98. Ozsarlak O, Van Goethem JW, Maes M, Parizel PM (2004) MR angiography of the intracranial vessels: technical aspects and clinical applications. Neuroradiology 46:955–972 99. Paetsch I, Jahnke C, Barkhausen J, Spuentrup E, Cavagna F, Schnackenburg B, Huber M, Stuber M, Fleck E, Nagel E (2006) Detection of coronary stenoses with contrast enhanced, three-dimensional free breathing coronary MR angiography using the Gd-based intravascular contrast agent gadocoletic acid (B-22956). J Cardiovasc Magn Reson 8:509–516 100. Parsons MW, Yang Q, Barber PA et al (2001) Perfusion magnetic resonance imaging maps in hyperacute stroke: relative cerebral blood flow most accurately identifies tissue destined to infarct. Stroke 32:1581–1587 101. Pediconi F, Fraioli F, Catalano C, Napoli A, Danti M, Francone M, Venditti F, Nardis P, Passariello R (2003) Gadopentate dimeglumine (Gd-DTPA) vs. gadopentetate dimeglumine (Gd-BOPTA) for contrast-enhanced magnetic resonance angiography (MRA): improvement in intravascular signal intensity and contrast to noise ratio. Radiol Med (Torino) 106:87–93

111

112

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 102. Pediconi F, Catalano C, Occhiato R et al (2005) Breast lesion detection and characterization at contrast-enhanced MR mammography: gadopentate dimeglumine versus gadopentetate dimeglumine. Radiology 237:45–56 103. Perreault P, Edelman MA, Baum RA, Yucel EK, Weisskoff RM, Shamsi K, Mohler ER III (2003) MR angiography with gadofosveset trisodium for peripheral vascular disease: phase II trial. Radiology 229:811–820 104. Petersein J, Reisinger W, Mutze S, Hamm B (2000a) Value of negative oral contrast media in MR cholangiopancreatography (MRCP). RoFo 172:55–60 105. Petersein J, Spinazzi A, Giovagnoni A et al (2000b) Focal liver lesions: evaluation of the efficacy of gadopentate dimeglumine in MR imaging—a multicenter phase III clinical study. Radiology 215:727–736 106. Port M, Corot C, Rousseaux O, Raynal I, Devoldere L, Idee JM, Dencausse A, Le Greneur S, Simonot C, Meyer D (2001) P792: a rapid clearance blood pool agent for magnetic resonance imaging: preliminary results. MAGMA 12:121–127 107. Preda A, Novikov V, Moglich M et al (2004a) MRI monitoring of Avastin antiangiogenesis therapy using B22956/1, a new blood pool contrast agent, in an experimental model of human cancer. J Magn Reson Imaging 20:865–873 108. Preda A, Wielopolski PA, Ten Hagen TL et al (2004b) Dynamic contrast-enhanced MRI using macromolecular contrast media for monitoring the response to isolated limb perfusion in experimental soft-tissue sarcomas. MAGMA 17:296–302 109. Prince MR, Narasimham DL, Stanley JC et al (1995) Breath-hold Gd-enhanced MR angiography of the abdominal aorta and its major branches. Radiology 197:785–792 110. Rapp JH, Wolff SD, Quinn SF, Soto JA, Meranze SG, Muluk S, Blebea J, Johnson SP, Rofsky NM, Duerinckx A, Foster GS, Kent KC, Moneta G, Middlebrook MR, Narra VR, Toombs BD, Pollak J, Yucel EK, Shamsi K, Weisskoff RM (2005) Aortoiliac occlusive disease in patients with known or suspected peripheral vascular disease: safety and efficacy of gadofosveset-enhanced MR angiography—multicenter comparative phase III study. Radiology 236:71–78 111. Reimer P, Balzer T (2003) Ferucarbotran (Resovist): a new clinically approved RES-specific contrast agent for contrast-enhanced MRI of the liver: properties, clinical development, and applications. Eur Radiol 13:1266–1276 112. Reimer P, Bremer C, Allkemper T, Engelhardt M, Mahler M, Ebert W, Tombach B (2004) Myocardial perfusion and MR angiography of chest with SH U 555 C: results of placebo-controlled clinical phase I study. Radiology 231:474–481 113. Riordan RD, Khonsari M, Jeffries J, Maskell GF, Cook PG (2004) Pineapple juice as a negative oral contrast agent in magnetic resonance cholangiopancreatography: a preliminary evaluation. Br J Radiol 77:991–999

114. Rogers JM, Jung CW, Lewis J, Groman EV /91998) Use of USPIO-induced magnetic susceptibility artifacts to identify sentinel lymph nodes and lymphatic drainage patterns. I. Dependence of artifact size with subcutaneous Combidex dose in rats. Magn Reson Imaging 16:917–23 115. Rohrer M, Bauer H, Mintorovitch J, Requardt M, Weinmann HJ (2005) Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol 40:715–724 116. Ruehm SG, Christina H, Violas X, Corot C, Debatin JF (2002) MR angiography with a new rapid-clearance blood pool agent: Initial experience in rabbits. Magn Reson Med 48:844–851 117. Runge VM (2000) Safety of approved MR contrast media for intravenous injection. J Magn Reson Imaging 12:205–213 118. Sadowski EA, Bennett LK, Chan MR, Wentland AL, Garrett AL, Garrett RW, Djamali A (2007) Nephrogenic systemic fibrosis: risk factors and incidence estimation. Radiology 243:148–157 119. Schnorr J, Wagner S, Abramjuk C, Wojner I, Schink T, Kroencke TJ, Schellenberger E, Hamm B, Pilgrimm H, Taupitz M (2004) Comparison of the iron oxide-based blood-pool contrast medium VSOP-C184 with gadopentetate dimeglumine for first-pass magnetic resonance angiography of the aorta and renal arteries in pigs. Invest Radiol 39:546–553 120. Schwitter J, Saeed M, Wendland MF et al (1997) Influence of severity of myocardial injury on distribution of macromolecules: extravascular versus intravascular Gd-based magnetic resonance contrast agents. J Am Coll Cardiol 30:1086–1094 121. Sharafuddin MJ, Stolpen AH, Sun S, Leusner CR, Safvi AA, Hoballah JJ, Sharp WJ, Corson JD (2002) High-resolution multiphase contrast-enhanced three-dimensional MR angiography compared with two-dimensional time-of-flight MR angiography for the identification of pedal vessels. J Vasc Interv Radiol 13:695–702 122. Shellock FG (1999) Safety of MRI contrast agents. JMRI 10:477–484 123. Spinazzi A, Lorusso V, Pirovano G, Kirchin M (1999) Safety, tolerance, biodistribution, and MR imaging enhancement of the liver with gadopentate dimeglumine: results of clinical pharmacologic and pilot imaging studies in nonpatient and patient volunteers. Acad Radiol 6:282–291 124. Staatz G, Nolte-Ernsting CC, Adam GB, Grosskortenhaus S, Misselwitz B, Bucker A, Gunther RW (2001) Interstitial T1-weighted MR lymphography: lipophilic perfluorinated Gd chelates in pigs. Radiology 220:129–134 125. Stets C, Brandt S, Wallis F, Buchmann J, Gilbert FJ, Heywang-Kobrunner SH (2002) Axillary lymph node metastases: a statistical analysis of various parameters in MRI with USPIO. J Magn Reson Imaging 16:60–68

2.6  Contrast Agents 126. Tatsumi Y, Tanigawa N, Nishimura H, Nomura E, Mabuchi H, Matsuki M, Narabayashi I (2006) Preoperative diagnosis of lymph node metastases in gastric cancer by magnetic resonance imaging with ferumoxtran-10. Gastric Cancer 9:120–128 127. Taylor AM, Panting JR, Keegan J, Gatehouse PD, Amin D, Jhooti P, Yang GZ, McGill S, Burman ED, Francis JM, Firmin DN, Pennell DJ (1999) Safety and preliminary findings with the intravascular contrast agent NC100150 injection for MR coronary angiography. J Magn Reson Imaging 9:220–227 128. Thilmann O, Larsson EM, Bjorkman-Burtscher IM, Stahlberg F, Wirestam R (2005) Comparison of contrast agents with high molarity and with weak protein binding in cerebral perfusion imaging at 3 T. J Magn Reson Imaging 22:597–604 129. Thomsen HS, Morcos SK, Dawson P (2006) Is there a causal relation between the administration of Gd based contrast media and the development of nephrogenic systemic fibrosis (NSF)? Clin Radiol 61:905–906 130. Tombach B, Benner T, Reimer P et al (2003) Do highly concentrated Gd chelates improve MR brain perfusion imaging? Intraindividually controlled randomized crossover concentration comparison study of 0.5 versus 1.0 mol/l gadobutrol. Radiology 226:880–888 131. Turetschek K, Floyd E, Helbich T et al (2001) MRI assessment of microvascular characteristics in experimental breast tumors using a new blood pool contrast agent (MS325) with correlations to histopathology. J Magn Reson Imaging 14:237–242 132. US Food and Drug Administration (2006) Public health advisory: Gd-containing contrast agents for magnetic resonance imaging (MRI): Omniscan, OptiMARK, Magnevist, ProHance, and MultiHance. http://www.fda.gov/cder/ drug/advisory/Gd_agents.htm. Cited 21 March 2007 133. Valk J, Algra PR, Hazenberg CJ, Slooff WB, Slavand MG (1993) A double-blind, comparative study of gadodiamide injection and gadopentetate dimeglumine in MRI of the central nervous system. Neuroradiology 35:173–177 134. Verstraete KL, Lang P (2000) Bone and soft tissue tumors: the role of contrast agents for MR imaging. Eur J Radiol 34:229–246 135. Vogl TJ, Friebe CE, Balzer T et al (1995) Diagnosis of cerebral metastasis with standard dose gadobutrol vs. a high dose protocol. Intraindividual evaluation of a phase II high dose study. Radiologe 35:508–516 136. Vogl TJ, Schwarz W, Blume S et al (2003) Preoperative evaluation of malignant liver tumors: comparison of unenhanced and SPIO (Resovist)-enhanced MR imaging with biphasic CTAP and intraoperative US. Eur Radiol 13:262–272 137. Vosshenrich R, Engeroff B, Obenauer S, Grabbe E (2003) Kontrastmittel-gestützte 3D-Angiographie des arteriellen und portalvenösen Gefäßsystems der Leber mit unterschiedlicher KM-Konzentration. RoFo 175:1239–1243

138. Wagenseil JE, Johansson LO, Lorenz CH (1999) Characterization of t1 relaxation and blood-myocardial contrast enhancement of NC100150 injection in cardiac MRI. J Magn Reson Imaging 10:784–789 139. Wagner S, Schnorr J, Pilgrimm H, Hamm B, Taupitz M (2002) Monomer-coated very small superparamagnetic iron oxide particles as contrast medium for magnetic resonance imaging: preclinical in vivo characterization. Invest Radiol 37:167–177 140. Wang YX, Hussain SM, Krestin GP (2001) Superparamagnetic iron oxide contrast agents: physicochemical characteristics and applications in MR imaging. Eur Radiol 11:2319–31 141. Ward J, Guthrie JA, Scott DJ et al (2000) Hepatocellular carcinoma in the cirrhotic liver: double-contrast MR imaging for diagnosis. Radiology 216:154–162 142. Weishaupt D, Ruhm SG, Binkert CA, Schmidt M, Patak MA, Steybe F, McGill S, Debatin JF (2000) Equilibriumphase MR angiography of the aortoiliac and renal arteries using a blood pool contrast agent. AJR Am J Roentgenol 175:189–195 143. Wendland MF, Saeed M, Lauerman K et al (1999) Alterations in T1 of normal and reperfused infarcted myocardium after Gd-BOPTA versus Gd-DTPA on inversion recovery EPI. Magn Reson Med 37:448–456 144. Wielopolski PA, van Geuns RJ, de Feyter PJ, Oudkerk M (1998) Breath-hold coronary MR angiography with volume-targeted imaging. Radiology 209:209–219 145. Wilke N, Kroll K, Merkle H et al (1995) Regional myocardial blood volume and flow: first-pass MR imaging with polylysine-Gd-DTPA. J Magn Reson Imaging 5:227–237 146. Wilke N, Jerosch-Herold M, Wang Y et al (1997) Myocardial perfusion reserve: assessment with multisection, quantitative, first-pass MR imaging. Radiology 204:373–384 147. Yuh WT, Fisher DJ, Engelken JD et al (1991) MR evaluation of CNS tumors: dose comparison study with gadopentetate dimeglumine and gadoteridol. Radiology 180:485–491 148. Yuh WT, Nguyen HD, Tali ET et al (1994) Delineation of gliomas with various doses of MR contrast material. AJNR Am J Neuroradiol 15:983–989 149. Zech CJ, Herrmann KA, Huber A et al (2004) High-resolution MR-imaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Reson Imaging 20:443–450 150. Zech CJ, Namkung S, Helmberger T, Reiser MF, Schönberg SO (2005) Efficacy of ferucarbotran-enhanced early dynamic MR Imaging with T1-weighted sequences for characterization of focal liver lesions. Eur Radiol 15(Suppl 3):37 151. Zerhouni EA, Rutter C, Hamilton SR et al (1996) CT and MR imaging in the staging of colorectal carcinoma: report of the Radiology Diagnostic Oncology Group II. Radiology 200:443–451

113

114

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.7  Flow Phenomena and MR Angiographic Techniques M. Bock 2.7.1  Introduction “Blut ist ein ganz besonderer Saft.” [Blood is a very special juice.] J. W. Goethe (1749–1832), Faust One of the strengths of MRI is the ability to visualize soft tissues with different image contrasts. Additionally, various two and three-dimensional MR imaging techniques for morphologic and functional examinations exist. Among the functional techniques the visualization and measurement of blood flow is of particular interest, since nearly all physiologic processes rely on an adequate blood supply. As with many other MR imaging techniques, the sensitivity of MRI to blood flow was first observed in artifacts visible near larger blood vessels. To suppress these artifacts new imaging methods have been investigated. In a further refinement of these techniques the artifact (here: the blood flow) has been made the primary source of the imaging technique; thus, in search for new methods of flow artifact suppression the blood flow itself became the contrast-generating element. The delineation of the vascular tree with MRI, MR angiography (MRA), is such a development: in T1-weighted 3D gradient-echo data, it was observed that the blood vessel signal of the margin partitions was significantly higher than at the center of the image stack. Furthermore, signal voids were seen in regions of turbulent flow, and in blood vessels with pulsating flow, ghost images of the vessel were visible in phase encoding direction. In the following, the underlying physical phenomena of these artifacts will be discussed, as they form the basis for time-of-flight MRA, phase-contrast MRA, and MR flow measurements. Since the pioneering work of Prince (1994), many MR angiographies are acquired using contrast-enhan­ ced ac­qui­sition techniques. In contrast-enhanced MRA, the signal difference between the bright blood vessel and the dark surrounding tissue is induced by a reduction of the blood’s T1 relaxation time. Again, this technique has evolved from an unwanted vascular signal artifact in spin-echo images acquired after contrast agent injection into a major MR application. With the development of new contrast agents with a longer half-life in the vascular system, the so-called intravascular contrast agents, contrast-enhanced MRA has been developed even further. In this section, techniques for MRA with either intravascular or extracellular contrast agents will be presented.

2.7.2  MR Properties of Blood The visualization of the blood vessels with MRI relies particularly on the specific properties of blood. Blood consists nearly entirely of liquids, so that blood has a very high spin density and thus yields a strong MR signal. The T1 time of blood is long compared to that of other tissues (e.g., 1 200 ms at 1.5 T) (Gomori et al. 1987), and it depends on its oxygenation state. Long T1 values are a disadvantage in T1-weighted acquisition strategies as the signal decreases with increasing T1. This disadvantage can be converted into an advantage if the blood signal needs to be suppressed (black-blood angiography) for visualization of the vessel walls. Furthermore, using an inversion (or saturation) recovery technique, the prepared magnetization can be tracked for a longer time, as the preparation persists much longer than in other tissues; this is the basis of arterial spin labeling techniques. Typical T2 values are of the order of 150–200 ms (at 1.5 T). This T2 value is long enough to provide a high signal in the blood vessels using dedicated T2-weighted image acquisition strategies. With conventional T2weighted spin-echo techniques, MR angiographies are difficult to acquire since the motion of the blood needs to be compensated; nevertheless, T2-weighted MRA pulse sequences for imaging of the peripheral vasculature have been reported (Miyazaki et al. 2000). Another approach is the use of balanced SSFP pulse sequences, where the contrast is dependent on the ratio of T1/T2—these fast pulse sequences have found a widespread use in the visualization of the cardiac system. In addition to the relaxation times, blood velocity is an important parameter. In healthy arterial vessels velocity values between 100 cm/s (e.g., in the aortic arch) and 30 cm/s (e.g., in the intracranial vessels) are common, whereas much lower values are found in the venous vasculature. High blood flow velocities lead to a pronounced inflow of fresh, unsaturated magnetization into an imaging slice, which increases the signal in the blood vessels— this is the well-known time-of-flight contrast. Furthermore, the velocity in the arterial system is not constant but changes as a function of time in the cardiac cycle. This pulsatility can be exploited to separate arterial from venous vessels, if image data are acquired with cardiac synchronization. All of the presented MRA techniques rely on these properties of the blood; some exploit only one of them, whereas others use a combination of them to increase the vascular contrast even further. 2.7.3  Time-of-Flight MRA In any MR pulse sequence, the magnetization in a measurement slice is exposed to a series of radio frequency (RF) pulses. If the magnetization does not move out of

2.7  Flow Phenomena and MR Angiographic Techniques Fig. 2.7.1  Transient longitudinal magnetization, which is subjected to a series of excitation pulses (30°) at a repetition time of 30 ms after entering the readout slice at t = 0 during TOF MRA. The longer the blood spins remain in the slice the more they are saturated, and a differentiation between blood and surrounding tissue becomes difficult

the measurement slice (e.g., in static tissue) it approaches a so-called steady state which, for a spoiled gradient-echo pulse sequence, depends on the flip angle, the repetition time TR, and the relaxation times T1 and T2. The steady state magnetization is smaller than the magnetization at the beginning of the experiment—it is partially saturated. Fresh, unsaturated blood flowing into the imaging slice is carrying the full magnetization and thus generates a significantly higher MR signal (Fig. 2.7.1); this is known as time-of-flight (TOF) contrast (Anderson and Lee 1993; Potchen et al. 1993). A major disadvantage of TOF MRA is the sensitivity to blood signal saturation: the longer the inflowing blood remains in the measurement slice, the more its signal is saturated. In situations where the blood vessel is oriented over a long distance parallel to the imaging slice (or 3D slab), the inflowing magnetization is progressively saturated. Thus, blood appears bright near the entry site but is seen less intense with increasing distance from this position. To maintain the TOF contrast over the whole imaging volume TOF MRA should therefore be per­for­med as a 2D acquisition with thin slices or, if a 3D acquisition technique is preferred, in the arterial vascu­lature where high-flow velocities result in fewer saturation pulses for the arterial blood. The inflow effect can be maximized if the measurement slice is oriented perpendicular to the blood vessel. This is often possible for the straight arterial vessels (e.g., the carotids in the head), but can be difficult for extended vascular territories with tortuous vessels. In 3D acquisitions of larger vessel structures, the saturation effect can be partially com­pensated, if the flip angle is increased from the entry side of the slab to the exit side (Fig. 2.7.2). Thus, the saturation effect is less pronounced during entry, and the magnetization is still visible when it enters smaller

vessels that are far away from the entry side. Often, an RF pulse with a linearly increasing flip angle is utilized (tilted optimized non-saturating excitation, or TONE [Nagele et al. 1995]). For an optimal vessel contrast the blood flow velocity, the repetition time, the mean flip angle, and the slope of the RF pulse profile are important parameters. In 2D TOF MRA, a very strong TOF contrast can be achieved, if the slice thickness D is chosen such that the magnetization flowing with a velocity v is completely replaced during one TR interval, i.e. D ≤ TR · v. At a typi-

Fig. 2.7.2  3D TOF MRA data set of the intracranial vasculature in lateral (top) and axial (bottom) maximum intensity projection. To minimize saturation, a TONE RF pulse was used for excitation, and the signal from static brain tissue was additionally suppressed using magnetization transfer pulses

115

116

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.7.3  Motion artifacts in lateral views of multi-slice 2D TOF MRA data sets of the arterial vasculature in the neck. Due to swallowing, the blood vessel can move from one acquisition to the next, and the edges appear with discontinuities in the lateral views of the maximum intensity projection

cal TR of 10 ms and a blood flow velocity of 40 cm/s, the slice thickness should thus not be larger than 4 mm. With these small slice thicknesses, the data acquisition in larger vascular territories such as the legs is very timecon­suming, and patient movements cannot be excluded during the several minutes of scan time. Patient movements lead to artificial vessels shifts between the imaging slices, which are particularly observed in orthogonal data reformats—these artifacts can mimic pathologies such as stenoses and thus significantly reduce the diag­nostic quality of the data sets (Fig. 2.7.3). 3D TOF MRA is advantageous over sequential 2D TOF MRA because an isotropic spatial resolution in all direc­tions can be achieved. To reduce the saturation effects in 3D TOF MRA not only one thick, but also several thinner 3D slabs are acquired consecutively. Thus, the saturation effects are smaller for the individual slabs and a stron­ger TOF contrast is seen. Unfortunately, the flip angle in fast 3D acquisitions is not constant over the slab, but is declining towards its margins. This inhomogeneous excitation results in a higher signal for the stationary tissue at the slab margin, providing an inhomogeneous signal background in lateral views of the data. Combined with a higher TOF contrast at the entry side compared with the exit side, spatially varying signal intensity is seen

in lateral views of the whole data set (venetian blind artifact). To reduce this artifact, overlapping 3D slabs are acquired (multiple overlapping thin slab acquisition, or MOTSA [Parker et al. 1991]), and the marginal slices of each slab are removed; however, this results in an increased total scan time. To increase the contrast between the blood vessels and the surrounding tissue in TOF MRA, often magnetization transfer pulses are included in the pulse sequences (Edelman et al. 1992). Using off-resonant RF pulses, the magnetization transfer contrast (MTC) selectively saturates those tissues where macromolecules are present. For brain tissue, these additional RF pulses can reduce the signal from background tissue by 40% and more, which increases the conspicuity espe­cially of the smaller blood vessels. The use of magnetization transfer pulses however increases the minimally achievable TR and, thus, the total acquisition time. Additionally, through the integration of the MTC pulses more RF power is applied to the patient, so that the regulatory power limits for the specific absorption rate (SAR) might be exceeded, an effect that is more pronounced at higher field strengths. Never­theless, MTC is often included in intracranial TOF MRA protocols where longer TRs can be an advantage, as long TRs additionally reduce the saturation effect. The flow velocity in arteries is typically not constant but varies over the cardiac cycle. Thus, the TOF contrast is a function of time, so that for image acquisition times that are longer than one cardiac cycle, a signal variation during k-space sampling is present. This periodic signal variation results in phantom images of the blood vessels in phase encoding direction after image reconstruction: the so-called pulsation artifacts or ghost images (Haacke and Patrick 1986; Wood and Henkelman 1985). To avoid pulsation artifacts, the image acquisition can be synchronized with the cardiac cycle using ECG triggering, which typically prolongs the total acquisition time, as only part of the measurement time is used for data acquisition. Another option to reduce pulsation artifacts is to saturate the inflowing blood in a slice upstream of the imaging slice. Therefore, a slice-selective RF excitation is applied in a (typically parallel) saturation slice, so that the magnetization of the inflowing blood is significantly reduced. Spatial presaturation avoids pulsation artifacts; however, the interior of the blood vessel now has a negative contrast, and the positive TOF contrast is gone. Another important ingredient of a TOF MRA pulse sequence is flow compensation (cf. paragraphs on flow measurements, below): The movement of the spins causes an additional velocity-dependent phase shift that is seen in TOF MRA data sets without flow compensation as a displacement. If multiple velocities are present as in turbulent flow, the different phases can cause signal cancellation (intra-voxel dephasing) that manifests, e.g., as a signal void behind a stenosis (Saloner et al. 1996). With special compensation gradients the velocity-depen-

2.7  Flow Phenomena and MR Angiographic Techniques

dent phase shifts can be reduced; however, this typically prolongs the echo time TE. TOF MRA is susceptible to several artifacts and is strongly dependent on a sufficient inflow velocity of un­ sa­tu­rated blood. Therefore, 3D TOF MRA techniques are typically only used in the head, where the arterial flow velocities are high and enough time is available for imaging. For abdominal studies, TOF techniques are of minor interest, because long measurement times are not possible due to respiratory motion. 2.7.4  Arterial Spin Labeling In conventional TOF MRA, the difference in longitudinal magnetization between the saturated stationary tissue and the unsaturated inflowing blood is exploited to create a positive contrast between blood and tissue. With arterial spin-labeling techniques, a similar approach is taken to the visualization of the inflowing blood; however, here only a certain fraction of the inflowing blood is tagged (or labeled) and subsequently visualized, whereas in TOF MRA all inflowing material is detected (Detre et al. 1994). Spin-labeling pulse sequences consist typically of a labeling section, during which an RF pulse is applied to the spins upstream of the imaging slice (Fig. 2.7.4). For labeling often adiabatic inversion pulses are used, which are less susceptible to motion during the inversion and that allow inverting the magnetization even in RF coils with a limited transmit homogeneity (e.g., a transmit/receive head coil). After an (often variable) inflow delay time TI, during which the labeled blood is flowing into the vascular target structure, the signal in the imaging slice is acquired. For signal reception different image-acquisition strategies can be employed such as segmented spoiled gradient-echo (FLASH), fast spin-echo (RARE, HASTE), or even echo planar imaging (EPI). Note: This image data set contains both the signal from the labeled blood and the static background tissue. In a second acquisition, the entire pulse sequence is repeated without labeling of the blood, and a second image data set is acquired. To selectively visualize only the labeled blood the two data sets are subtracted; since the signal intensity of the blood differs in both acquisitions, a non-vanishing blood signal is seen, whereas the signal contribution from static tissue cancels. If the phase of the second image data set is shifted by 180° compared to the first, labeled data set, the images can be added (the minus sign is provided by the phase), and the technique is called signal targeting with alternating radiofrequencies (STAR) (Edelman et al. 1994). In clinical MRI systems, arterial spin labeling (ASL) is typically implemented with the described labeling pulses, which are applied only once per data readout;

Fig. 2.7.4  Concept of arterial spin labeling: magnetization is prepared (e.g., using a slice-selective inversion pulse) in a section of the artery (red). After an inflow delay of several hundred milliseconds, the magnetization has reached the imaging slice (green), and an image is acquired. The procedure is repeated without preparation, and the two data sets are subtracted to remove the signal background from static tissue

this approach is also termed pulsed arterial spin labeling (PASL). Another method for ASL uses a small transmit coil for the labeling pulse, which continuously applies an RF pulse to the arterial vessel, and thus achieves a much higher degree of inversion. Unfortunately, these continuous ASL techniques often cannot be used in a clinical MR system due to the regulatory constraints for the maximum RF power applied to the patient (SAR limits). ASL techniques are typically applied to study perfusion in the brain and other organs, where the inflow delay is chosen long enough for the labeled blood to have reached the capillary bed (Golay et al. 2004). Unfortunately, the labeling in the blood does not persist for much longer than one T1 time (i.e., 1–2 s at 1.5 T), and the signal differences are generally very small (2–5% of the total signal), which makes perfusion measurement ASL a time-consuming procedure. Another application to ASL is the time-resolved visualization of blood flow, e.g., in intracranial malformations (Essig et al. 1996), where saturation effects limit the diagnostic quality of conventional 3D TOF MRAs. Here, dynamic ASL data sets are acquired at a series of inflow delays to visualize the transit of the labeled bolus through the nidus of the malformation, and, more importantly, the arrival of the blood in the draining venous vessels,

117

118

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.7.5  TOF MRA (top) and timeresolved dynamic MRA with arterial spin labeling (bottom) of an intracranial arteriovenous malformation. In the TOF MRA, the nidus of the malformation is clearly seen, but the draining vein can hardly be identified because the inflowing blood is already completely saturated when it arrives in this part of the AVM. In the three ASL images acquired 100, 600, and 1 200 ms after signal preparation, the filling of the nidus and the drainage through the vein is clearly visible

that cannot be seen on the TOF data sets (Fig. 2.7.5). In addition to a morphologic representation of these blood vessels, transit-time measu­re­ment of the blood becomes feasible, which could be used as an indicator, e.g., for an increase in vascular resistan­ce after a radiation therapy. 2.7.5  Native-Blood Contrast MR angiographies can also be acquired using the special contrast properties of blood. In balanced steady state free precession pulse sequences (bSSFP, trueFISP, FIESTA), an image contrast is created that depends on the ratio of the relaxation times T1 and T2 (Oppelt et al. 1986). For blood, this ratio is high, and thus the interior of the blood vessels are shown with higher signal intensity than the surrounding tissue. Unfortunately, other liquid-filled spaces such as the ventricles also appear with a bright signal, so that conventional MRA post-processing strategies such as the maximum intensity projection cannot be used to visualize the vascular tree (Fig. 2.7.6). Despite their short repetition times and balanced gradient schemes, these pulse sequences are susceptible to flow artifacts caused by intra-voxel dephasing, which can be compensated using flow-compensation gradients (Storey et al. 2004; Bieri and Scheffler 2005). Another problem with balanced steady state pulse sequences is the susceptibility to off-resonance artifacts: Since both transverse and longitudinal magnetizations contribute to the MR signal, perfect phase coherence must be main­

tained within one TR to establish the desired contrast. In off-resonant regions, this phase coherence is perturbed, and a contrast variation is seen in the form of dark bands. The banding artifacts can be reduced using a repetition time that is shorter than the inverse of the off-resonance frequency, i.e., for a 200-Hz off-resonance, the TR should be shorter than 5 ms. Off-resonance frequencies scale with field strength, so that banding artifacts become an increasing problem at higher field strengths. Nevertheless, fast balanced SSFP pulse sequences are increasingly used in MRA studies of the heart and the neighboring vessels in combination with ECG triggering to visualize the vascular anatomy and to assess. 2.7.6  Black-Blood MRA In conventional spin-echo images, one often observes that the interior of the blood vessels is darker than the surrounding tissue. This so-called black-blood contrast is caused by an incomplete signal refocusing of the 180° pulse. Compared with TOF MRA with gradient-echo sequences, where the inflow of blood causes signal am­ pli­fication, spin-echo sequences attenuate the signal from flowing blood because spins leave the imaging slice between the 90° excitation pulse and the 180° refocusing pulse, and thus do not contribute to the MR signal. Therefore, blood signal attenuation can be increased with a longer spacing between the two RF pulses, i.e., with longer echo times TE. To further suppress signals from slowly flowing blood near the vessel walls, often addi­

2.7  Flow Phenomena and MR Angiographic Techniques

Fig. 2.7.6  Balanced SSFP image (left) and contrast-enhanced MRA (right) of the heart and the abdominal aorta of a patient. In this image blood is seen with high signal intensity; however, the surrounding tissue also appears with a strong MR signal. In the ascending aorta, signal voids are seen which are caused by turbulent flow, and banding artifacts are visible in the subcuta-

Fig. 2.7.7  ECG-triggered dark blood image of the heart acquired with a single-shot fast spin-echo technique (HASTE). The blood signal both in the heart and in the cross-section of the descending aorta is completely suppressed

neous fat. Nevertheless, balanced SSFP sequences provide good angiographic overview images in very short acquisition times, without the need for contrast agent injection. In the con­trastenhanced data acquisition, a better background suppression is possible, and the projection image of the 3D data set clearly delineates the aorta and the adjacent vessels

tional strong gradients are introduced in the black-blood pulse sequences, which cause an increased intra-voxel dephasing and thus suppress the signal (Lin et al. 1993). A different technique for blood signal suppression makes use of an inversion recovery blood signal preparation (Edelman et al. 1991): Similar to arterial spin labeling a non-selective 180° inversion pulse is applied; however, the signal in the ima­ging slice is reinverted by a subsequent slice-selective inversion pulse. With this preparation ,the magnetization of the blood (and of all other tissues) outside the imaging slice is selectively inverted. After a delay-time that is chosen to achieve a zero crossing of the longitudinal magnetization of the inverted blood, an image is acquired. If the blood has been completely exchanged during the delay, then the signal of the labeled blood is nulled and only the static tissue is visible. This technique is often used in combination with cardiac triggering to visualize, e.g., the myocardium (Fig. 2.7.7). In cardiac black-blood applications, both techniques are combined, which is possible, because data are acquired during diastole when the heart is nearly at rest, whereas the signal preparation is applied during systole.

119

120

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.7.7  Velocity-Dependent Phase

2.7.7.1  Flow Measurements

The TOF contrast relies on the increase in signal amplitude due to the inflow of unsaturated magnetization. In addition to elevated signal amplitude, the spin movement can also create a change in the phase of the MR signal. If a gradient is turned on and blood moves along the gradient direction (here: the x-direction), the phase ϕ(t) of the MR signal is given by:

If the two phase images are directly subtracted, the result is a phase difference image that is linearly dependent on the spin velocity: this is the basis of an MR flow meas­urement (Bryant et al. 1984). Since phase data are only unambiguous in the angular range of ±180°, so is the velocity information in an MR flow measurement. To avoid artifacts due to multiple rotations of the spin phase (so-called wrap around artifacts), the first moment needs to be chosen such that the maximum velocity in the image creates a phase shift of 180°. In general, this velocity is set via the so-called velocity encoding, or VENC, parameter in the pulse sequence. Higher VENC values require weaker en­co­ding gradients, which can be realized in shorter echo times. Despite an inadequate choice of the VENC value, MR flow measurements are susceptible to phase noise, which is present in regions of low signal amplitude. If the SNR is 1 or less, the phase in the image is nearly uniformly distributed between –180° and +180°; under these conditions a meaningful flow measurement is not possible. Unfortunately, phase noise is also often present near blood vessels (e.g., in the air-filled spaces of the lung, close to the pulmonary vessels). Here, the meas­ urement of velocity values requires a very careful placement of the regions of interest (ROIs) to avoid systematic errors from included noise pixels. In a conventional flow measurement, velocity encoding is typically performed in slice-selection direction only, because the orthogonal placement of the flow measurement slice induces a high TOF signal in the cross-section of the vessel lumen. Additionally, a parallel orientation of the velocity encoding direction with the image plane makes the image acquisition susceptible to systematic errors due to displacement, as, e.g., the readout gradient are used for both spatial encoding and velocity measurement simultaneously. Flow measurements in arterial vessels are often performed with cardiac synchronization to account for the pulsatility of the blood flow. Cardiac synchronization can be performed by prospective ECG triggering or retrospective ECG gating. With prospective triggering, data acquisition is started by a trigger signal, which is generated by the ECG electronics during the QRS complex of the ECG (Fig. 2.7.8). After data have been acquired for a certain number of cardiac phases, the measurement sequence is stopped until a new cardiac trigger signal is detected. In retrospective gating, image data are continuously acquired, and the time between the last trigger and the current data set is stored. Later, data are resorted into predefined time intervals (bins) in the cardiac cycle, and the images are reconstructed. Prospective triggering is less time-consuming during image reconstruction and is very precise in the delineation of the cardiac activity; however, a temporal gap at the end of the cardiac cycle is required and thus flow measurements at late diastole are difficult.



(2.7.1)

Here, the motion x(t) of the magnetization is expressed as a Taylor series, and only the constant term (i.e., the initial position x0) and the linear term (i.e., the velocity v0) are considered. The two integrals M0 and M1 solely depend on the gradient timing and are called the zeroth and first moment of G(t). The next higher order term is proportional to the acceleration of the spins; however, the proportionality constant M2 only becomes large if long time scales are considered; thus, the estimation of the spin phase from the zeroth and first moments is justified for gradient-echo sequences with short echo times. If the gradient timing is modified such that the first moment is zero, the gradients are called flow compensated. Flow compensation is important ingredient in many MR pulse sequences: If a range of velocities is present in a single voxel, then the MR signal amplitude is attenuated due to the incoherent addition of the signals. With flow compensation, the individual signals all have the same phase, and the signals of the different velocities add up coherently. Flow compensation is especially important in regions of high velocity gradients as, e.g., turbulent jets or in highly angulated vessels. In general, both M0 and M1 will be non-vanishing, and the phase of the signal will become proportional to the local spin velocity. Unfortunately, many other factors such as off-resonance, field inhomogeneity, or chemical shift also affect the spin phase, so that a direct velocity measurement is not possible with a single MR experiment alone. To create an MR image that is dependent on the local velocity, a minimum of two image acquisitions are required. In the first, velocity-sensitized acquisition a gradient timing is used with a carefully selected, nonvanishing first gradient moment (the zeroth moment is defining the spatial encoding, i.e., the k-space trajectory). In a second, flow-compensated acquisition, a gradient timing is chosen that cancels M1.

2.7  Flow Phenomena and MR Angiographic Techniques

Fig. 2.7.8  ECG-triggered MR velocity measurement showing a typical velocity time curve in the aorta (left). From the phasedifference images, the mean velocity in the aorta are calculated after a region of interest is defined around the cross section of

the blood vessel. (right) The enlarged region around the aorta and the vena cava shows the aortic pulsating flow (dark, round region) and the nearly constant flow (bright region) in the opposite direction

Prospective gating uses continuous image acquisition, and the magnetization steady state is always maintained. Unfortunately, more data need to be acquired than with prospective triggering to ensure a sufficient coverage of the cardiac cycle, and a temporal blurring due to the interpolation is seen in the velocity data.

the imaging slice is twice the VENC value, a phase shift of 360° (or 0°, which cannot be distinguished) is created. Under these conditions, the velocity-encoded and the velocity-compensated acquisition have the same phase, and no PC MRA signal would be observable. As blood does not flow with a constant velocity and velocity values can be reduced by pathologies (aneurysms) or increased (stenoses), the optimum choice of VENC value is often difficult. Because PC MRA is more time-consuming, is susceptible to artifacts, and suffers from the same signal saturation as TOF MRA, it is rarely used in clinical routine.

2.7.7.2  Phase-Contrast MRA When the complex image data of the two acquisitions are subtracted instead of the phases, and the magnitude of the difference is displayed, a so-called phase-contrast MRA image is created (Dumoulin 1995). This PC MRA image is not only dependent on the velocity of the spins, but also on the signal amplitude in both acquisitions; thus, every PC MRA data set always has an overlaid TOF contrast (Fig. 2.7.9). An advantage of PC MRA is the fact that signal background of the surrounding stationary tissue is almost completely suppressed, and vessels can be traced further into the vascular periphery than with TOF MRA, with comparable measurement parameters. In PC MRA, often not only the velocity in one spatial direction is encoded, but in all three directions. Since separate velocity-encoded acquisitions have to be performed for each direction, the measurement time of a PC MRA is two- to fourfold longer than that of a TOF MRA. A careful selection of the VENC value is especially important in PC MRA. If, e.g., the maximum velocity in

2.7.8  Contrast-Enhanced MRA With TOF and PC MRA techniques, the blood motion is used to create a signal difference between the vessel lumen and the surrounding tissue, whereas contrastenhanced MRA utilizes the reduction of the longitudinal relaxation time T1 after administration of a contrast agent. When a contrast agent is injected, the T1 of blood is shortened from T1blood = 1.2 s (for B0 = 1.5 T) to less than 100 ms during the first bolus passage (first pass). The relaxation rate R1 (i.e., 1/ T1) is a function of the local contrast agent concentration:

(2.7.2)

121

122

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.7.9  Phase-contrast images encoding flow in head–foot (top) and left–right (center) direction, and phase contrast MRA image (bottom). In the flow images, a velocity-sensitive and a velocity-compensated data set are subtracted, whereas the PC MRA image is generated by complex subtraction of the respective signal amplitudes. Note, that the PC MRA image has nearly no background signal from static tissue

The proportionality constant between contrast agent concentration C and the change in relaxation rate is called the relaxivity r1. The relaxivity is different for each contrast agent – typical values range from 4 to 10 mmol–1s–1. In general, high relaxivities are desirable because lower contrast-agent concentrations are needed to achieve the same change in image contrast. To enhance the signal in the contrast agent bolus and to suppress the signal background from static tissue, heavily T1-weighted spoiled gradient-echo sequences (FLASH) with very short repetition times (TR < 5 ms) and high flip angles (α = 20°–50°) are used (Fig. 2.7.10). The use of short TRs is advantageous because very short acquisition times of only a few seconds can be achieved even for the acquisition of a complete 3D data set. These short acquisition times are needed, because the contrast agent is progressively diluted during the passage, which reduces the vessel-to-background contrast. Short acquisition times are also favorable because MRA data sets can thus be acquired in a single breath hold; for this reason, contrast-enhanced MRA techniques are especially suited for abdominal applications (Prince 1996; Sodickson and Manning 1997). To ensure isotropic visualization of the vascular territories, typically 3D techniques are used for data acquisition. Conventional 3D techniques have measurement times of several minutes, so that even with short repetition times larger parts of the k-space data are acquired after the contrast agent concentration has fallen to levels where only a weak signal enhancement is observable. For this reason, the measurement times are reduced using partial k-space sampling, parallel imaging, and view sharing between subsequent 3D data sets (Sodickson and Manning 1997; Wilson et al. 2004; Goyen 2006). In general, contrast agents can be categorized into extracellular agents that can leave the blood stream and intravascular agents that are specifically designed to remain in the vascular system. Historically, the first ap­pro­ved MR contrast agent was Gd-DTPA (gadopentate dimeglumine, Magnevist, Schering, Germany), an extra­cellular agent, which has the paramagnetic Gd3+ ion as the central atom in an open-chain ionic complex (chelate). Over the years, several similar extracellular agents such as Gd-BT-DO3A (Gadovist, Schering), Gd-DOTA (Dotarem, Guerbet, France), Gd-BMA (Omniscan, GE Healthcare), Gd-HP-DO3A (ProHance, Bracco Imaging, Italy), and Gd-BOPTA (MultiHance, Bracco, Italy) have been approved for clinical use, which only slightly differ in the stability of the Gd chelates, pharmacokinetic properties, and safety profiles. In general, the most recently appro­ved contrast agents have higher relaxivities and thus allow acquiring MRA data sets with higher contrast at the same dose or with similar contrast at lower dose. Only recently, the first intravascular contrast agent gadofosveset trisodium (Vasovist, Schering) has been approved for clinical use in Europe (Goyen 2006). This

2.7  Flow Phenomena and MR Angiographic Techniques Fig. 2.7.10  Signal intensity as a function of T1-contrast in a spoiled gradient-echo pulse sequence (FLASH). At high flip angles and short repetition times the signal from tissue (T1 > 300 ms at 1.5 T) is nearly completely saturated, whereas a high intraluminal signal is seen due to the high concentrations of the contrast agent

molecule has a diphenylcyclohexyl group, which is co­va­ lently bound to a Gd complex, which creates a reversible, non-covalent binding of the molecule to serum albumin that significantly prolongs the half-life of the agent in blood to about 16 h. After injection of the agent, at first a more rapid decline of concentration is observed, because the fraction of the contrast agent bound to albumin is dependent on the contrast agent concentration; thus, a steady-state concentration is established after the unbound fraction is renally excreted. Both extracellular and intravascular agents can be imaged during the first pass of the contrast agent, when a high vessel-to-background contrast is present, whereas intravascular contrast agents additionally allow angiographic imaging during the subsequent steady state. 2.7.8.1  First-Pass Studies The T1 shortening is dependent on the contrast agent concentration, which is getting smaller already a few seconds after infusion of the contrast agent, as the contrast agent bolus in the blood is in­creasingly diluted and, for the extracellular agents, the contrast agent is extravasculized. Therefore, contrast-enhanced MRA techniques usually use pulse sequences with very short acquisition times (TA < 30 s). The short passage time of the contrast agent bolus of a few seconds requires that imaging be precisely syn­ chronized with the contrast agent infusion. The transit time of the bolus from the point of injection (usually a vein in the arm) through to the vascular target structure (e.g., the renal arteries) varies significantly with the heart rate and cardiac output and can be difficult to predict. Therefore, various synchronization and acquisition tech­

niques have been proposed for a reliable MRA data acquisition: • Automatic Start An automatic technique to start the 3D MRA data acquisition (SmartPrep™, General Electric) uses a fast pulse sequence before the 3D MRA, which continuously acquires the signal in the vascular target region (Prince et al. 1997). After ad­mini­stration of the contrast agent, this signal exceeds a certain signal threshold, and the 3D MRA acquisition is automatically started. If the signal threshold is selected too low, image noise can mimic a bolus arrival and the mea­s­urement is triggered too early, whereas a too high value of the threshold can lead to an omission of the data acquisition. • Test Bolus With the test bolus technique, a small bolus of a few milliliters of the contrast agent is infused, and the passage of the bolus is imaged near the target vessel with a fast time-resolved 2D MR measurement (Earls et al. 1997). The trigger delay TD for the subsequent 3D measurement is then calculated from the transit time of the bolus TT and the acquisition time of the 3D MRA TA as: TD = TT – necho × TA. Here, necho denotes the fraction of TA before the center of k-space is acquired (Fig. 2.7.11). • MR Fluoroscopy As in the automatic start of the sequence, fluoroscopic previews (CareBolus™, Siemens Medical Solutions) ima­ge the contrast bolus during its passage; however, here fast 2D sequences are used with real-time image re­con­struc­tion and display Riederer et al. 1988; Wilman et al. 1997). Once the bolus has reached the target region, the operator of the MR scanner ma­nually switches to the predefined 3D MRA pulse sequence, which is then executed with minimal time delay.

123

124

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.7.11  Test bolus measurement in the heart (blue) and the aorta (red) of a patient. For an optimal visualization of the aortic arch, a time delay of 16 s is required between contrast agent injection and the acquisition of the central k-space lines

• Multiphase MRA Time-resolved MRA has been increasingly used to completely avoid manual or automated synchronization. Multiphasic acquisitions consecutively acquire 3D MRA data sets during the bolus passage so that the opti­mal vessel contrast is obtained in at least one of the data sets. Various methods of mea­s­ure­ment acceleration are combined to ensure adequate temporal resolution; these include parallel imaging, asymmetric k-space readout, and temporal data interpolation (Korosec et al. 1996; Fink et al. 2005). Never­the­less, timeresolved MRA data sets are usually of a lower spatial resolution than are optimally acquired MRA data with bolus synchronization (Fig. 2.7.12). Artifacts arise, if the 3D MRA data acquisition is not perfectly synchronized with the bolus passage. The appearance of these artifacts depends on the relative timing of the k-space acquisition and the concentration-time curve of the contrast agent. If the bolus arrives too late in the target vessel, then the center of the k-space has already been sampled and large structures, such as the interior of the blood vessel, appear dark, whereas fine structures, such as vessel margins, have a high signal if the bolus arrives during sampling of the k-space periphery. If the data acquisition is started too late, then the bolus has already reached the target vessel and the contrast has partly disappeared – thus, the signal is significantly reduced compared with an optimally synchronized data acquisition (Maki et al. 1996; Wilman et al. 2001; Svensson et al. 1999). Another disadvantage of suboptimal bolus timing is the fact that the bolus may have passed from the arterial

to the venous system in some vascular regions (e.g., the extremities) so that both veins and arteries are seen in the images. This venous contamination makes the interpretation of the image data difficult in cases where arterial and venous vessels are parallel to each other. The variation in contrast agent concentration over time also results in a reduction in the achievable image resolution (Fain et al. 1999). The spatial resolution of an MR image sampled with Cartesian data acquisition is always uni­quely defined by the measured number of k-space lines. With increasing number of k-space lines (i.e., larger k-space coverage), finer image details are encoded in the image. This so-called Nyquist scanning theorem only applies if the signal intensity is constant during data acquisition. Even with perfect synchronization with the contrast agent bolus, the contrast agent concentration is only optimal during acquisition of the central k-space lines. Later on, the concentration is reduced and the peripheral k-space regions are acquired with significantly reduced signal intensity (Fig. 2.7.13). This different weighting of the k-space regions results in a reduction in spatial resolution (blurring), which is mathematically described by the point-spread function, PSF. The PSF is the image of a point object – for linear imaging systems, it is used to describe the imperfections of the image acquisition system. The PSF depends on the acquisition time, the contrast agent dynamics, and the measurement parameters of the pulse sequence. The deviation from an ideal PSF is particularly visible in those spatial directions that are acquired with the lowest sampling velocity. In conventional 3D acquisition, this is either the phase encoding direction or the partition encoding direction. To eliminate this asymmetry and to

2.7  Flow Phenomena and MR Angiographic Techniques

Fig. 2.7.12  Multiphase MRA of the lung vasculature in a patient with a patent ductus arteriosus. The irregular flow pattern of the early enhancing descending aorta is clearly identified in the time series of 3D MRA data sets. To achieve a high temporal

resolution of 2.3 s, temporal interpolation techniques (TRICKS) were applied. In the later phases, additionally, a qualitative assessment of lung perfusion is possible

evenly distribute the blurring in both spatial directions, elliptical scanning of the phase and partition encoding steps has been proposed (Bampton et al. 1992; Wilman and Riederer 1996; Wilman et al. 1996), where the encoding steps are acquired along an elliptical path, starting from the center of the k-space. Using the signal–time curve of a test bolus, the signal variation during data acquisition can be avoided. Therefore, an injection scheme is calculated for the signal–time curve using linear system theory, where the injection rate is modulated such that there is a constant contrast agent concentration (i.e., an ideal PSF) in the target region throughout the data acquisition. This technique requires a programmable contrast-agent injector and additional computations, and the constant concentration can only be achieved in a limited target volume. Another option for reducing the intensity changes is to induce a blood flow stasis for a brief period after contrast agent inflow. This can be achieved in the peripheral blood vessels, without endangering the patient using an inflatable cuff, which temporarily blocks the blood flow during data acquisition; this technique has been used successfully applied to contrast-enhanced MRA studies of the hand (Zhang et al. 2004) and the legs (Zhang et al. 2004; Vogt et al. 2004). In addition to reducing T1, an MR contrast agent always also reduces T2 (and T2*). The reduction in T2* can lead to a significant signal reduction in the MRA image at high contrast agent concentrations; often the venous vessel through which the bolus is infused are seen dark in the MRA data sets (Albert et al. 1993). To avoid these ar-

tifacts, the contrast agent concentration or the echo time TE can be reduced. Besides reducing T2*, the contrast agent also causes a concentration-dependent resonance frequency shift. Radial or spiral k-space data acquisitions espe­cial­ly susceptible to these frequency shifts that cause blurring artifacts, which can be compensated using dedi­ cated off-resonance correction algorithms. Contrast-enhanced MRA studies are also susceptible to artifacts known from TOF MRA. In particular, pulsation artifacts are visible in contrast-enhanced measurements (Al-Kwifi et al. 2004). Intra-voxel dephasing is observed, although the effect is much lower due to the shorter TEs used here. To keep the acquisition time short, generally flow com­pen­sation is not integrated into contrast-enhanced MRA pulse sequences, because the additional gradients significantly prolong the measurement time. Contrast-enhanced MRA offers substantial advantages over TOF or PC MRA, because the saturation effects seen in TOF MRA are almost completely avoided. Thus, extended vascular structures for example in the abdomen or the extremities can be visualized with a few slices oriented parallel to the vessel. The short acquisition times of contrast-enhanced MRA allow breath-held acquisitions (TA < 30 s), which significantly reduces motion artifacts (Maki et al. 1997). The dynamic information of multiphase 3D MRA contains information about vascular anatomy, flow direction (e.g., in aortal aneurysms), tissue perfusion (e.g., in the kidney), and vascular anomalies, which might not be visible on a single MRA data set.

125

126

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.7.13  (Top) Simulated signal–time curve during a firstpass MRA (solid line) and a study with an intravascular contrast agent (dotted line). At the center of the acquisition window, when the central k-space lines are acquired, a maximal signal is seen in the first-pass study, whereas a constant signal is present in

with the intravascular contrast agent. (Bottom) The cor­respond point spread functions show that during a first-pass study objects appear blurred in phase-encoding direction, whereas the theoretical resolution of the data acquisition can be achieved when the concentration is constant

2.7  Flow Phenomena and MR Angiographic Techniques Fig. 2.7.14  Surface rendering (left) and MIP (right) visualization of an abdominal MRA with an intravascular contrast agent. The three-dimensional character of the data set is better captured with the surface display, whereas the finer details are better visualized on the MIP. The presence of venous signal is making the interpretation of the data more difficult; however, a significantly higher spatial resolution can be achieved with intravascular contrast agents

2.7.8.2  Intravascular Contrast Agents Using intravascular agents, the contrast agent concentration in the blood is maintained over time spans of minutes to hours. In general, the same pulse sequences can be applied for MRA with intravascular contrast agents as with the extracellular contrast agents, as they share the same contrast mechanism; however, as the concentration of intravascular contrast agent in blood attains an equilibrium state after a few re-circulations (typically 20–40 s), MRA data sets can also be acquired over longer acquisition times. This prolonged acquisition window can be used to increase the image resolution, because an ideal PSF can be achieved and, thus, no blurring should be present (van Bemmel et al. 2003a; Grist et al. 1998). Because data acquisition does not have to be synchronized with the contrast agent bolus, data acquisition can be started once the contrast agent concentration has reached equilibrium. The contrast agent injection does not need to be performed with a contrast agent pump, but can be infused manually via a venous access port even prior to the MR examination. With intravascular contrast agents, the acquisition time TA is not limited by the transit time of the bolus, and data sets can be acquired over much longer acquisition periods. With longer acquisition times, special trigger and gating techniques (ECG triggering, respiratory gating, navigator echoes [Ahlstrom et al. 1999]) to suppress motion artifacts are required. If image data are acquired in the equilibrium phase, venous overlay is a fundamental problem in MRA with intravascular contrast agents (Fig. 2.7.14). Venous contamination particularly makes those images hard to interpret that are calculated through a projection technique such as the maximum intensity projection (MIP), because here, the depth information is lost. A separation of arterial and venous vessels is possible with the help of dedicated post-processing software. Therefore, a region in an arterial vessel is identified and a region-growing algorithm is used to find all connected regions. Unfortu-

nately, when arteries and veins are in close proximity the algo­rithm may artifactually connect arterial to venous vessels. With equilibrium-phase MRA, data sets of intra­ vascular contrast agents, these artifacts are easier to avoid than in first-pass studies because of the higher spatial resolution. Nevertheless, for direct arteriovenous connections or shunts, a manual correction of the segmen­tation is always required (van Bemmel et al. 2003b; Svensson et al. 2002). The use of intravascular contrast agent is not limited to the equilibrium phase, but can be combined with a first-pass study during the initial contrast agent injection to obtain both the dynamics of the contrast agent passage as well as the vascular morphology (Grist et al. 1998). Additionally, the dynamic information can be utilized to separate arteries from veins in the high-resolution equilibrium-phase 3D MRA data sets (Bock et al. 2000). Although the long half-life of the intravascular contrast agent is advantageous for intraluminal studies, it can become a problem if a dynamic study has to be repeated. With extracellular contrast agents, this is possible within a few minutes, whereas up to several hours have to be waited after a study with an intravascular contrast agent. In practice, intravascular contrast agents are still advantageous for MRA, as they allow combining high spatial resolution in the equilibrium phase with dynamic information during first passage. Additionally, these contrast agents can be used to quantify perfusion (Prasad et al. 1999) or to delineate vessels during MR-guided intra­vas­ cu­lar pro­cedures (Wacker et al. 2002; Martin et al. 2003). 2.7.9  Summary Various techniques for MR angiography and MR flow measurements exist that make use of the different physical properties of blood: flow, pulsation, or signal variation following administration of contrast agent. TOF MRA is often used in anatomical regions where a high inflow is

127

128

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

present, and long measurement times can be tolerated. Phase-contrast flow measurements provide a quantitative assessment of blood flow when combined with cardiac triggering. Contrast-enhanced studies are favorable in abdominal regions and the periphery, where saturation effects are a limiting factor for TOF MRA. Intravascular contrast agents further extend the capabilities of contrastenhanced MRA because high-resolution data sets can be acquired over an extended time. References 1.

Ahlstrom KH, Johansson LO, Rodenburg JB, Ragnarsson AS, Akeson P, Borseth A (1999) Pulmonary MR angio­ graphy with ultrasmall superparamagnetic iron oxide particles as a blood pool agent and a navigator echo for respiratory gating: pilot study. Radiology 211:865–869 2. Albert MS, Huang W, Lee JH, Patlak CS, Springer CS Jr (1993) Susceptibility changes following bolus injections. Magn Reson Med 29:700–708 3. Al-Kwifi O, Kim JK, Stainsby J, Huang Y, Sussman MS, Farb RI, Wright GA (2004) Pulsatile motion effects on 3D magnetic resonance angiography: implications for evaluating caro­tid artery stenoses. Magn Reson Med 52:605–611 4. Anderson CM, Lee RE (1993) Time-of-flight techniques. Pulse sequences and clinical protocols. Magn Reson Imaging Clin N Am 1:217–227 5. Bampton AEH, Riederer SJ, Korin HW (1992) Centric phase-encoding order in three-di­men­sional MP-RAGE sequences: application to abdominal imaging. J Magn Reson Imaging 2:327–334 6. Bemmel CM van, Spreeuwers LJ, Viergever MA, Niessen WJ (2003a) Level-set-based artery-vein separation in blood pool agent CE-MR angiograms. IEEE Trans Med Imaging 22:1224–1234 7. Bemmel CM van, Wink O, Verdonck B, Viergever MA, Niessen WJ (2003b) Blood pool con­trast-enhanced MRA: improved arterial visualization in the steady state. IEEE Trans Med Imaging 22:645–652 8. Bieri O, Scheffler K (2005) Flow compensation in balanced SSFP sequences. Magn Reson Med 54:901–907 9. Bock M, Schönberg SO, Flomer F, Schad LR (2000) Separation of arteries and veins in 3D MR angiography using correlation analysis. Magn Reson Med 43:481–487 10. Bryant DJ, Payne JA, Firmin DN, Longmore DB (1984) Measurement of flow with NMR imaging using a gradient pulse and phase difference technique. J Comput Assist Tomogr 8:588–593 11. Detre JA, Zhang W, Roberts DA, Silva AC, Williams DS, Grandis DJ, Koretsky AP, Leigh JS (1994) Tissue spe­cific perfusion imaging using arterial spin labeling. NMR Biomed 7:75–82 12. Dumoulin CL (1995) Phase contrast MR angiography techniques. Magn Reson Imaging Clin N Am 3:399–411

13. Earls JP, Rofsky NM, DeCorato DR, Krinsky GA, Weinreb JC (1997) Hepatic arterial-phase dynamic gadolinium-enhanced MR imaging: optimization with a test examination and a power injector. Radiology 202:268–273 14. Edelman RR, Chien D, Kim D (1991) Fast selective black blood MR imaging. Radiology 181:655–660 15. Edelman RR, Ahn SS, Chien D, Li W, Goldmann A, Mantello M, Kramer J, Kleefield J (1992) Improved time-offlight MR angiography of the brain with magnetization transfer contrast. Radiology 184:395–399 16. Edelman RR, Siewert B, Adamis M, Gaa J, Laub G, Wielopolski P (1994) Signal targeting with alternating radio­ frequency (STAR) sequences: application to MR angiography. Magn Reson Med 31:233–238 17. Essig M, Engenhart R, Knopp MV, Bock M, Scharf J, Debus J, Wenz F, Hawighorst H, Schad LR, van Kaick G (1996) Cerebral arteriovenous malformations: improved nidus demarcation by means of dynamic tagging MR-angio­graphy. Magn Reson Imaging 14:227–233 18. Fain SB, Riederer SJ, Bernstein MA, Huston J III (1999) Theoretical limits of spatial resolution in elliptical-centric contrast-enhanced 3D-MRA. Magn Reson Med 42:1106–1116 19. Fink C, Ley S, Kroeker R, Requardt M, Kauczor HU, Bock M (2005) Time-resolved contrast-enhanced three-dimensional magnetic resonance angiography of the chest: combination of parallel imaging with view sharing (TREAT). Invest Radiol 40:40–48 20. Golay X, Hendrikse J, Lim TC (2004) Perfusion imaging using arterial spin labeling. Top Magn Reson Imaging 15:10–27 21. Gomori JM, Grossman RI, Yu-Ip C, Asakura T (1987) NMR relaxation times of blood: dependence on field strength, oxidation state, and cell integrity. J Comput Assist Tomogr 11:684–690 22. Goyen M (2006) (ed) MR angiography with Vasovist®. ABW Wissenschaftsverlag, Berlin 23. Grist TM, Korosec FR, Peters DC, Witte S, Walovitch RC, Dolan RP, Bridson WE, Yucel EK, Mistretta CA (1998) Steady-state and dynamic MR angiography with MS-325: initial ex­pe­rience in humans. Radiology 207:539–544 24. Haacke EM, Patrick JL (1986) Reducing motion artifacts in two-dimensional fourier trans­form imaging. Magn Reson Imaging 4:359–376 25. Korosec FR, Frayne R, Grist TM, Mistretta CA (1996) Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 36:345–351 26. Lin W, Haacke EM, Edelman RR (1993) Black blood angiography. In: Potchen EJ, Haacke EM, Siebert JE, Gottschalk A (eds) Magnetic resonance angiography, concepts and applications. Mosby, St. Louis pp 160–172 27. Maki JH, Prince MR, Londy FJ, Chenevert TL (1996) The effects of time varying intravascular signal intensity and kspace acquisition order on three-dimensional MR angiography image quality. J Magn Reson Imaging 6:642–651

2.7  Flow Phenomena and MR Angiographic Techniques 28. Maki JH, Prince MR, Chenevert TL (1997) The effects of incomplete breath-holding on 3D MR image quality. J Magn Reson Imaging 7:1132–1139 29. Martin AJ, Weber OM, Saeed M, Roberts TP (2003) Steady-state imaging for visualization of endovascular interventions. Magn Reson Med 50:434–438 30. Miyazaki M, Sugiura S, Tateishi F, Wada H, Kassai Y, Abe H (2000) Non-contrast-enhanced MR angiography using 3D ECG-synchronized half-Fourier fast spin-echo. J Magn Reson Imaging 12:776–783 31. Nagele T, Klose U, Grodd W, Nusslin F, Voigt K (1995) Nonlinear excitation profiles for three-dimensional inflow MR angiography. J Magn Reson Imaging 5:416–420 32. Oppelt A, Grauman R, Barfuss H, Fischer H, Hartl W, Schajor W (1986) FISP—a new fast MRI sequence. Electro­ medica 54:15–19 33. Parker DL, Yuan C, Blatter DD (1991) MR angiography by multiple thin slab 3D acquisition. Magn Reson Med 17:434–451 34. Potchen EJ, Haacke EM, Siebert JE (1993) Magnetic resonance angiography. Mosby, St. Louis 35. Prasad PV, Cannillo J, Chavez DR, Pinchasin ES, Dolan RP, Walovitch R, Edelman RR (1999) First-pass renal perfusion imaging using MS-325, an albumin-targeted MRI contrast agent. Invest Radiol 34:566–571 36. Prince MR (1994) Gadolinium-enhanced MR aortography. Radiology 191:155–164 37. Prince MR (1996) Body MR angiography with gadolinium contrast agents. Magn Reson Imaging Clin N Am 4:11–24 38. Prince MR, Chenevert TL, Foo TK, Londy FJ, Ward JS, Maki JH (1997) Contrast-enhanced abdominal MR angiography: optimization of imaging delay time by automating the de­tection of contrast material arrival in the aorta. Radiology 203:109–114 39. Prince MR, Grist TM, Debatin JF (2003) 3D contrast MR angiography. Springer, Berlin Heidelberg New York 40. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42:952–962 41. Riederer SJ, Tasciyan T, Farzaneh F (1988) MR flouroscopy: technical feasibility. Magn Reson Med 8:1–15 42. Saloner D, van Tyen R, Dillon WP, Jou LD, Berger SA (1996) Central intraluminal saturation stripe on MR an­gio­ grams of curved vessels: simulation, phantom, and clinical analysis. Radiology 198:733–739 43. Sodickson DK, Manning W (1997) Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 38:591–603

44. Storey P, Li W, Chen Q, Edelman RR (2004) Flow artifacts in steady-state free precession cine imaging. Magn Reson Med 51:115–122 45. Svensson J, Petersson JS, Stahlberg F, Larsson EM, Leander P, Olsson LE (1999) Image artifacts due to a time-varying contrast medium concentration in 3D contrast-enhanced MRA. J Magn Reson Imaging 10:919–928 46. Svensson J, Leander P, Maki JH, Stahlberg F, Olsson LE (2002) Separation of arteries and veins using flow-induced phase effects in contrast-enhanced MRA of the lower extremities. Magn Reson Imaging 20:49–57 47. Vogt FM, Ajaj W, Hunold P, Herborn CU, Quick HH, Debatin JF, Ruehm SG (2004) Venous compression at highspatial-resolution three-dimensional MR angiography of peripheral arteries. Radiology 233:913–920 48. Wacker FK, Wendt M, Ebert W, Hillenbrandt C, Wolf KJ, Lewin JS (2002) Use of a blood-pool contrast agent for MR-guided vascular procedures: feasibility of ultrasmall super­para­magnetic iron oxide particles. Acad Radiol 9:1251–1254 49. Wentz K, Fröhlich J, von Weymarn C, Patak M, Jenelten R, Zollikofer C (2003) High-resolution magnetic re­so­nance angiography of hands with timed arterial compression (tac-MRA). Lancet 361:49–50 50. Wilman AH, Riederer SJ (1996) Improved centric phase encoding orders for three-dimensional magnetization-prepared MR angiography. Magn Reson Med 36:384–392 51. Wilman AH, Riederer SJ, Breen JF et al (1996) Elliptical spiral phase encoding order: an optimal, field-of-viewdependent ordering scheme for breath-hold contrast-enhanced 3D MR angiography. Radiology 201:328–329 52. Wilman AH, Riederer SJ, King BF, Debbins JP, Rossman PJ, Ehman RL (1997) Fluoroscopi­cal­ly triggered contrast-enhanced three-dimensional MR angiography with elliptical centric view order: application to the renal arteries. Radiology 205:137–146 53. Wilman AH, Yep TC, Al-Kwifi O (2001) Quantitative evaluation of nonrepetitive phase-enco­ding orders for firstpass, 3D contrast-enhanced MR angiography. Magn Reson Med 46:541–547 54. Wilson GJ, Hoogeveen RM, Willinek WA, Muthupillai R, Maki JH (2004) Parallel imaging in MR angiography. Top Magn Reson Imaging 15:169–185 55. Wood ML, Henkelman RM (1985) MR image artifacts from periodic motion. Med Phys 12:143–151 56. Zhang HL, Ho BY, Chao M, Kent KC, Bush HL, Faries PL, Benvenisty AI, Prince MR (2004) Decreased venous contamination on 3D gadolinium-enhanced bolus chase peripheral MR angiography using thigh com­pres­sion. Am J Roentgenol 183:1041–1047

129

130

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging O. Dietrich 2.8.1  Introduction Diffusion in the context of diffusion-weighted MRI or diffusion tensor imaging (DTI) refers to the stochastic thermal motion of molecules or atoms in fluids and gases, a phenomenon also known as Brownian motion. This motion depends on the size, the temperature, and in particular on the microscopic environment of the examined molecules. Diffusion measurements can therefore be used to derive information about the microstructure of tissue. In MRI, stochastic molecular motion can be observed as signal attenuation. This was first recognized in the NMR spin-echo experiment by E.L. Hahn in 1950 (Hahn 1950), long before the invention of actual magnetic resonance imaging. A number of more sophisticated experiments were described in following years that allowed the quantitative measurement of the diffusion coefficient (Carr and Purcell 1954; Torrey 1956; Woessner 1961). Of particular importance is the pulsed-gradient spin-echo (PGSE) technique proposed by Stejskal and Tanner in 1965 (Stejskal and Tanner 1965) that is described in detail in Sect. 2.8.3.1. Diffusion as an imaging-contrast mechanism was first incorporated in MRI pulse sequences in 1985 (Taylor and Bushell 1985; Merboldt et al. 1985) and applied in vivo in 1986 (LeBihan et al. 1986). Its great potential for clinical MRI became evident in around 1990, when diffusionweighted images were recognized to be extremely valuable for the early detection of stroke (Moseley et al. 1990a,b; Chien et al. 1990, 1992). Areas of focal cerebral ischemia appear hyperintense in diffusion-weighted images only minutes after the onset of symptoms (see also Chap. 3, Sect. 3.4). Having thus been pushed into publicity, diffusion-weighted imaging was evaluated in many other applications such as the characterization of brain tumors (Tien et al. 1994; Sugahara et al. 1999; Okamoto et al. 2000) and of multiple sclerosis lesions (Cercignani et al. 2000; Filippi and Inglese 2001), but none of these reached the clinical significance of stroke diagnosis. Mainly due to limitations of image quality, there are considerably fewer publications about diffusion-weighted imaging outside the central nervous system. Examples of these are studies with the purpose of differentiating osteoporotic and malignant vertebral compression fractures (Baur et al. 1998, 2003; Herneth et al. 2002) or benign and malignant lesions of the liver (Moteki et al. 2002; Taouli et al. 2003) and the kidneys (Cova et al. 2004). Molecular diffusion is a three-dimensional process and is—depending on the tissue microstructure—in general anisotropic, i.e., the extent of molecular motion depends on spatial orientation. A physical quantity called

the diffusion tensor is required to fully describe anisotropic diffusion. MRI techniques to measure the diffusion tensor have been introduced in the 1990s (Basser et al. 1994; Pierpaoli et al. 1996) and gained considerably more popularity when tracking algorithms were proposed for three-dimensional reconstruction of white matter fiber tracts (Mori et al. 1999; Conturo et al. 1999). Today, diffusion tensor imaging is a valuable research tool with applications, e.g., in neurodevelopment (Neil 2002; Snook et al. 2005), neuropsychiatry (Taber et al. 2002, Moseley et al. 2002, Sullivan and Pfefferbaum 2003), or aging (Moseley 2002, Sullivan and Pfefferbaum 2003, Sullivan et al. 2006). 2.8.2  Physics of Diffusion 2.8.2.1  Brownian Molecular Motion All molecules in fluids or gases perform microscopic random motions. This motion is called molecular diffusion or Brownian motion after Robert Brown (1773–1858), who observed a minute motion of plant pollens floating in water in 1827 (Brown 1866). These pollens were constantly hit by fast-moving water molecules, resulting in a visible irregular motion of the much larger particles. Due to Brownian motion, a tracer such as a droplet of ink given into water will diffuse into its surroundings, resulting in spatially and temporally varying tracer concentrations, until the ink is diluted homogeneously in the water. However, Brownian molecular motion does not require concentration gradients, but occurs also in fluids consisting of only a single kind of molecule. The molecules of any arbitrary droplet of water within a larger water reservoir will stochastically disperse into their surroundings; this process is called diffusion or, to emphasize that the observed molecules do not diffuse into an external medium, self-diffusion. It should be noted that diffusion always refers to a stochastic and not directed motion and is strictly to be distinguished from any kind of directional flow of a liquid. The molecules in fluids or gases perform random motions due to their thermal kinetic energy, Ekin, which is proportional to the temperature, T: Ekin = –23 kT (k = 1.38 × 10–23 J/K is the Boltzmann constant). This energy corresponds to a mean velocity v = 2 E kin m for a molecule of mass m; in the case of water at room temperature (T = 300 K), the mean velocity is about 650 m/s. Due to frequent collisions with other particles, however, mol-

 Strictly speaking, we calculate the square root of the mean value of squared velocity, i.e., mean (ν�) , which is slightly different (by a factor of 3π / 8 ≈ 1.085 , i.e., by 8.5%) from the actual mean value of the velocity due to the asymmetry of the Maxwell distribution.

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging

Fig. 2.8.1  Simulated diffusion path of a single molecule. a Random-walk simulation after N = 2,000, 10,000, 50,000, and 150,000 simulated steps. b Dependence of diffusion distance, s, on diffusion time, t

Fig. 2.8.2  Gaussian probability distribution of individual molecular diffusion distances, d, after diffusion times of t = 50 ms, 150 ms, and 300 ms; the distributions are based on the diffusion coefficient of water molecules at room temperature. The standard deviations of the Gaussian distributions are marked by dashed lines and indicate the mean diffusion distance, s

ecules do not move linearly in a certain direction but follow a random course that can be visualized in a randomwalk simulation as shown in Fig. 2.8.1a. This figure also demonstrates that, macroscopically, the mean displacement or diffusion distance, s, after a time t is much more interesting than is the linear velocity of the molecule. The mean diffusion distance of a particle is proportional to the square root of the diffusion time t and is described by the diffusion coefficient D: s = 6 Dt . This relation is shown in Fig. 2.8.1b for a water molecule with a diffusion coefficient of D = 2.03 × 10–3 mm2/s at a temperature of 20°C. Since diffusion is a stochastic process, the diffusion distance after the time t is not the same for all molecules but is described by a Gaussian probability distribution as illustrated in Fig. 2.8.2. As shown in this illustration, after a diffusion time t most molecules are still found at or close to their original position; the diffusion distance s, corresponds to the standard deviation of the shown distributions. Typical diffusion distances for free water molecules at room temperature are about 25 µm after a diffusion time of 50 ms and 110 µm after 1 s.

131

132

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Table 2.8.1  Examples of diffusion coefficients of different fluids and tissues ADC –3 2 (10  mm /s)

 Reference

Water, 5°C

1.31

Mills 1973

Water, 20°C

2.02

Tofts et al. 2000

Water, 35°C

2.92

Mills 1973

Ethanol, 20°C

0.98

Tofts et al. 2000

Brain, white matter

0.70

Helenius et al. 2002

Brain, gray matter

0.89

Helenius et al. 2002

Liver

1.83

Boulanger et al. 2003

Kidney

2.19

Cova et al. 2004

In contrast to free diffusion in pure water (Fig. 2.8.3a), the water molecules in tissue cannot move freely, but are hindered by the cellular tissue structure, in particular by cell membranes, cell organelles, and large macromolecules as shown schematically in Fig. 2.8.3b. Due to additional collisions with these obstacles, the mean diffusion distance of water molecules in tissue is reduced compared to that of free water, and a decreased effective diffusion coefficient is found in tissue called the apparent diffusion coefficient (ADC). Obviously, the ADC depends on the number and size of obstacles and therefore on the cell types that compose the tissue. Hence, diffusion properties can be used to distinguish different types of tissue. Examples of diffusion coefficients in different tissues and in fluids at different temperatures are summarized in Table 2.8.1. Not only does the number and size of organelles influence diffusion, but also the geometrical arrangement of the cell membranes. In particular, the diffusion of water molecules can reflect an anisotropic arrangement of cells as indicated in Fig. 2.8.3c. Since cell membranes are barriers for diffusing molecules, water diffuses more freely along the long axis of the cell than perpendicular to it (Beaulieu 2002). Hence, the ADC measured in the direction parallel to the cellular orientation will be greater than that measured in an orthogonal direction. This property, the dependence of a quantity on its orientation in space, is called anisotropy. 2.8.2.2  Diffusion Tensor It has proven useful to illustrate the diffusion properties by spheres for isotropic diffusion and by three-dimensional ellipsoids for anisotropic diffusion as shown

in Fig. 2.8.3d–f. These shapes visualize the probability density function of diffusion distances in space. Isotropic diffusion can be completely described by its (apparent) diffusion coefficient, D, which corresponds to the radius of the sphere (Fig. 2.8.3d,e). More quantities are required for a complete description of anisotropic diffusion, e.g., three angles that define the orientation of the ellipsoid in space, and the length of the three principal axes describing the magnitude of the diffusion coefficients. In physics or mathematics, a quantity that corresponds to such a three-dimensional ellipsoid is called a tensor. The physical object called tensor can also be explained by comparing it to more commonly known objects such as scalars and vectors. A scalar is a quantity that can be measured or described as a single number; typical examples are the temperature, the mass, or the density of an object. In imaging, the image intensity, e.g., in a T2-weighted image, is a scalar: a single number is required for each pixel to describe the intensity. As demonstrated in the last example, scalars can be spatially dependent and be visualized as intensity maps; another example is a temperature map of an object that describes the temperature as scalar quantity for each spatial position of an object. Other physical quantities cannot be described by a single number, such as the velocity or acceleration of a particle in space or the flow of a liquid. These quantities are vectors and require both a direction in space and a magnitude to be fully described. A vector is typically visualized as an arrow. For example, in the case of velocity, the direction of the arrow describes the direction of motion and the length of the arrow represents the magnitude of the vector, e.g., as measured in meters/second. Such an arrow can be mathematically described by three independent numbers: either by its length and two angles defining its orientation or by three coordinates (x-, y-, and z-component of the vector). These coordinates are often presented as a column or row vector, e.g., v = (vx vy vz). Vectors as well as scalars can depend on the spatial position; a flowing liquid can be described by a velocity vector at each position. A full data set consisting of a vector (i.e., an arrow) at each point in space is called a vector field. Some quantities such as the molecular diffusion cannot be fully described as scalars or vectors; they are tensors. As mentioned above, the diffusion properties can be depicted by a three-dimensional ellipsoid and therefore require six independent numbers to define the direction and length of all axes. These six values are visualized in Fig. 2.8.4 as the three lengths of the axes defining the shape of the ellipsoid and the three angles describing its orientation. However, instead of using angles, tensors can equally well be described by six coordinates arranged in a symmetric 3 × 3-matrix in analogy to the three coordinates of a vector. These coordinates are called Dxx, Dyy, Dzz, Dxy, Dxz, and Dyz and form the matrix representation

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging Fig. 2.8.3  Apparent diffusion coefficient and schematic depiction of diffusion properties: a Free diffusion in pure water b Hindered isotropic diffusion in tissue c Hindered anisotropic diffusion in tissue d schematic depiction of free isotropic diffusion as sphere with radius D e schematic depiction of hindered isotropic diffusion as sphere with radius D and f schematic depiction of anisotropic diffusion as ellipsoid with axes of different lengths

of the tensor D:

 Dxx  D =  Dxy D  xz

Dxy D yy D yz

Dxz   D yz  Dzz 

.  This matrix is called symmetric because the elements are mirrored at the diagonal. Of these matrix elements, only the diagonal elements, Dxx, Dyy, Dzz, can be measured directly in MRI and correspond to the diffusion in the x-, y-, and z-directions; the off-diagonal elements must be determined indirectly from further measurements as described in Sect. 2.8.4.

In the case of isotropic diffusion, i.e., if the ellipsoid is a sphere, then this matrix has a very simple form, because a single diffusion coefficient suffices to describe the diffusion. This diffusion coefficient is found on the diagonal of the matrix, and all off-diagonal elements are zero:

Disotropic

D  =0 0 

0 D 0

0  0 D 

The diffusion tensor has some properties that are important to understand in order to measure and interpret diffusion imaging data. The mean diffusivity, i.e., the diffu-

133

134

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

sion coefficient averaged over all spatial orientations, can be derived from the trace of the diffusion tensor, i.e., the sum of its diagonal elements:  Dxx  trace of D = tr D = tr  Dxy D  xz

Dxy D yy D yz

Dxz   D yz  = Dxx + D yy + Dzz Dzz 

The mean diffusivity, D , or ADC is a third of the trace of D: ADC = D = 13 tr D. An MRI measurement of the mean ADC is therefore also called trace imaging. To analyze the non-isotropic properties of the diffusion tensor, a process called diagonalization of the tensor is used. The meaning of tensor diagonalization can be visualized as finding the three axes (i.e., their length and orientation) that define the ellipsoid in Fig. 2.8.4. Mathematically, the tensor matrix is transformed into a form where all off-diagonal elements are zero:

the tensor, in addition to the three diagonal elements, three vectors, v1, v2, v3, are determined which are called eigenvectors. The eigenvectors, which are always orthogonal and have unit length, define the orientation of the ellipsoid and are shown as thick grey arrows in Fig. 2.8.4d. 2.8.2.3  Diffusion Anisotropy

The ratios of the diffusion eigenvalues describe the isotropy or anisotropy of diffusion. In the case of isotropic diffusion, all eigenvalues are the same, D1 = D2 = D3, and diffusion is represented by a sphere; see Fig. 2.8.5a. If the largest eigenvalue is much greater than the two other eigenvalues, D1 >> D2 ≈ D3, then the tensor is represented by a cigar-like shape as in Fig. 2.8.5b. In this case, diffusion in one direction is much less hindered than in the other directions and is sometimes called linear diffusion; this is typically found in white matter fiber tracts, where  D xx D xy D xz  0  the motion of water molecules is restricted by the cell  D1 0     membranes and the glial cells perpendicular to the fiber ation D =  D xy D yy D yz  diagonaliz  →  0 D2 0 , v1 v2 v3 tract orientation. The orientation of the fiber tracts is deD   0 0 D3  scribed by the eigenvector v1 belonging to the large eigen  xz D yz D zz  value, D1. If two large eigenvalues are much greater than  D xx D xy D xz  0   D1 0     diagonaliz ation the third one, D1 ≈ D2 >> D3, then the diffusion tensor is D =  D xy D yy D yz    →  0 D2 0 , v1 v2 v3 represented by a pancake-like shape; see Fig. 2.8.5c. This D    0 D3   0  xz D yz D zz  tensor corresponds to preferred diffusion within a twodimensional plane, which can occur in layered structures The three new diagonal elements, D1, D2, D3, are called and is referred to as planar diffusion. eigenvalues of the tensor; they describe the length of the In order to describe the diffusion anisotropy quantithree axes of the ellipsoid in Fig. 2.8.4. (Some authors de- tatively, several anisotropy indices have been introduced note the eigenvalues as λ1, λ2, λ3 instead of D1, D2, D3.) to reduce the diffusion tensor to a single number, i.e., a Since six parameters are still required to fully describe scalar, measuring the anisotropy. Most frequently used is

Fig. 2.8.4  Tensor visualized as three-dimensional ellipsoid: six independent numbers are required to define a tensor, three lengths (eigenvalues), D1, D2, D3, corresponding to the length of the principal axes of the ellipsoid (shown three-dimensionally

in a and in two-dimensional sections b), and three angles, α, β, γ, describing the spatial orientation of the axes (c,d). The eigenvectors of the tensor are shown as thick gray arrows in d

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging

The relative anisotropy is the magnitude of the anisotropic part of the tensor divided by its isotropic part (Basser and Pierpaoli 1996) and ranges from 0 (isotropy) to 2 ≈ 1.414 (maximum anisotropy). In order to scale the maximum value of the RA to 1 as well, a normalized (or scaled) definition with an additional factor of 1 2 is sometimes used (and often called RA as well):

Less frequently used indices of anisotropy are the volume ratio (VR), Fig. 2.8.5  Typical diffusion tensors in the case of a isotropic diffusion, b linear diffusion, and c planar diffusion

VR =

D1 D2 D3 D3

and the volume fraction (VF),

D1 D2 D3 D3

the fractional anisotropy (FA), defined as

VF = 1 VR = 1

where D = –13 (D1 + D2 + D3) is the mean diffusivity. The fractional anisotropy ranges from 0 (isotropic diffusion) to 1 (maximum anisotropy) and can be interpreted as the fraction of the magnitude of the tensor that can be ascribed to anisotropic diffusion (Basser and Pierpaoli 1996). A similar index is the relative anisotropy (RA), defined as

All these anisotropy indices can be used to describe the diffusion anisotropy (Kingsley and Monahan 2005), but fractional anisotropy may be considered the preferred index that is currently most frequently used. Some typical values of these indices are compared in Table 2.8.2; FA, RA, nRA, and VF start with 0 for isotropic diffusion and increase with increasing anisotropy. The volume ratio is 1 in the case of isotropic diffusion and decreases with increasing anisotropy.

Table 2.8.2  Examples for different anisotropy indices (1, 1, 1)

a

(1, 1, 0.5)

(1, 0.5, 0.5)

(1, 1, 0)

(1, 0.5, 0)

(1, 0, 0)

FA

0

0.33

0.41

0.71

0.77

1

RA

0

0.28

0.35

0.71

0.82

1.41

nRA

0

0.20

0.25

0.5

0.58

1

VF

0

0.14

0.16

1

1

1

VR

1

0.86

0.84

0

0

0

FA fractional anisotropy, RA relative anisotropy, nRA normalized (scaled) anisotropy, VF volume fraction, VR volume ratio  a Eigenvalues (D1, D2, D3)

135

136

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.8.3  MR Measurement of Diffusion-Weighted Images 2.8.3.1  Diffusion Gradients and Diffusion Contrast To introduce diffusion weighting in MRI pulse sequences, today almost exclusively a technique proposed by Stejskal and Tanner in 1965 (Stejskal and Tanner 1965) is used. The basic idea is to insert additional gradients (usually referred to as diffusion gradients) into the pulse sequence in order to measure the stochastic molecular motion as signal attenuation. Originally, these were two identical gradients on both sides of the refocusing 180° RF pulse of a spin-echo sequence: the so-called pulsed-gradient spin echo (PGSE) technique. However, to simplify the explanation, we will replace this scheme with two gradients with opposite signs that do not require a 180° pulse in between, as shown in Fig. 2.8.6. The contrast mechanism is the same for both gradient schemes. As illustrated in Fig. 2.8.6, the diffusion gradients superpose a linear magnetic field gradient over the static field, B0. Since the Larmor frequency of the spins is proportional to the magnetic field strength, spins at different positions now precess with different Larmor frequencies and, thus, become dephased. If the spins are stationary (no diffusion, i.e., diffusion coefficient D = 0) and remain at their position, the second diffusion gradient with opposite sign exactly compensates the effect of the first one and rephases the spins. Hence, without diffusion, the signal after the application of the pair of diffusion gradients is the same as before (neglecting relaxation effects). In the case of diffusing spins, the second diffusion gradient cannot completely compensate the effect of the first one since spins have moved between the first and second gradient. The additional phase the spins gained during the first diffusion gradient is not reverted during the second one. Consequently, rephasing is incomplete after the second diffusion gradient, resulting in diffusion-dependent signal attenuation. As can be deduced from this explanation, the signal attenuation is larger if the diffusivity, i.e., the mobility of the spins, is larger. Quantitatively, the signal attenuation depends exponentially on the diffusion coefficient, Dg, in the direction defined by the diffusion gradient gD: S (Dg , b) = S0 × exp (– b × Dg), where S0 is the original (unattenuated) signal and S(Dg, b) is the attenuated diffusion-weighted signal. The b-value, b, is the diffusion weighting that plays a similar role for diffusion-weighted imaging as the echo time for T2weighted imaging: the diffusion contrast, i.e., the signal difference between two tissues with different ADCs, is low at small b-values and can be maximized by choosing the optimal b-value as discussed below. The b-value is expressed in units of s/mm2 and depends on the timing and

the amplitude, gD, of the diffusion gradients:

As illustrated in Fig. 2.8.6, δ is the duration of each diffusion gradient, and ∆ is the interval between the onsets of the gradients; γ is the gyromagnetic ratio of the diffusing spins. A typical b-value used for diffusion-weighted imaging of the brain is 1,000 s/mm²; for other applications b-values range between 50 s/mm2 (dark blood liver imaging) and about 20,000 s/mm2 (imaging of the diffusion q-space [Assaf et al. 2002; Wedeen et al. 2005]). To obtain b-values of about 1,000 s/mm2, diffusion gradients are required to be much longer (e.g., δ = 25 ms) and have larger amplitudes (e.g., gD = 25 mT/m) than normal imaging gradients applied in MRI; hence, diffusion-weighted imaging can be demanding for the gradient amplifiers and is often acoustically noisy. The formula for the b-value given above is valid only for a pair of Stejskal-Tanner diffusion gradients. The diffusion weighting of arbitrary time-dependent diffusion gradient shapes, gD(t), applied between t = 0 and t = T, can be calculated according to (Stejskal and Tanner 1965)

By applying diffusion gradients, diffusion-weighted images can be acquired in which the signal intensity depends on the ADC, e.g., structures with large ADC such as liquids appear hypointense. To quantify the ADC at least two diffusion-weighted measurements with different diffusion weightings (i.e., different b-values) are required as shown in Fig. 2.8.7. By determining the signal intensity at the lower b-value, S(b1), and the higher b-value, S(b2), the ADC can be calculated as

This can be done either for the mean signal intensities in a region of interest or pixel by pixel in order to calculate an ADC map as in Fig. 2.8.7. The ADC can also be calculated from more than two b-values by fitting an exponential to the measured signal intensities or by linear regression analysis applied to the logarithm of signal intensities. It should be noted that diffusion-weighted images generally exhibit a mixture of different contrasts. Many diffusion-weighted pulse sequences require relatively long echo times between 60  and 120 ms because of the long duration of the diffusion preparation. Thus, diffusion-weighted images are often also T2-weighted, and it can be difficult to differentiate image contrast due to diffusion and T2 effects. This is a typical problem in diffusion-weighted MRI of the brain and known as T2 shinethrough effect (Burdette et al. 1999). A further consequence of the long minimum echo times due to the diffusion preparation is the relatively

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging

Fig. 2.8.6  Diffusion weighting using bipolar diffusion gradients. The diffusion gradients (green) cause a spatially varying magnetic field, ∆B0 (yellow), and thus spatially varying Larmor frequencies. The first diffusion gradient dephases the spins (three spins at different spatial positions are shown as yellow, blue, and

green arrows). If the spins are stationary (no diffusion), then the second diffusion gradient with opposite sign rephases the spins. In the case of diffusing spins, rephasing is incomplete since spins have moved between the first and second gradient; thus, diffusion-dependent signal attenuation is observed (red arrow)

Fig. 2.8.7  Acquisition of two images with different diffusion weightings (b-values b1 and b2) in order to calculate an ADC map. Note the large signal attenuation in CSF at the higher b-value, b2, and the correspondingly high diffusion coefficient in the ADC map

low signal-to-noise ratio of diffusion-weighted images. The combined effects of diffusion weighting, which particularly decreases the signal of fluids, and of T2 weighting, which predominantly reduces the signal of other (non-fluid) tissue, results in globally low signal intensity on diffusion-weighted images. Therefore, signal-increasing techniques such as increasing the voxel volume or (magnitude) averaging are often required for diffusion-

weighted MRI. In addition, ADC calculation can be corrected for the decreasing signal-to-noise ratio at higher b-values (Dietrich et al. 2001a). The range of b-values chosen for a diffusion-weighted MRI experiment should depend on the typical diffusion coefficients that are measured and on the signalto-noise ratio of the diffusion-weighted image data. As a rule of thumb, the signal attenuation should be at least

137

138

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

about 60%, i.e., the product of diffusion coefficient and the b-value range, bmax – bmin, should be approximately 1 (Xing et al. 1997). This corresponds to a b-value difference of about 1,000–1,500 s/mm2 in brain tissue with ADCs between 0.6 × 10–3 and 1.0 × 10–3 mm2/s. However, the choice of the largest b-value is frequently limited by signal-to-noise considerations, and thus, the maximum diffusion weighting is often reduced in order to maintain sufficient signal-to-noise ratio. A second point to consider is the choice of the lowest b-value. Although a b-value of 0 is often chosen, a slightly higher value of, for example, 50 s/mm2 can be advantageous in order to suppress the influence of perfusion effects (LeBihan et al. 1988; van Rijswijk et al. 2002). 2.8.3.2  Pulse Sequences for Diffusion MRI Historically, the first MRI pulse sequences with inserted diffusion gradients were stimulated-echo (Taylor and Bushell 1985; Merboldt et al. 1985) and spin-echo sequences (LeBihan et al. 1986); a schematic spin-echo pulse sequence with diffusion gradients is shown in Fig. 2.8.8. In this diagram, diffusion gradients are added for all three spatial directions (readout, phase, and slice direction); however, they are usually switched on in only one or two of the three directions at a time. Since spins are refocused by a 180° pulse, both diffusion gradients have the same polarity. The main disadvantages of the diffusion-weighted spin-echo sequence are that it requires long acquisition times of many minutes per data set and is extremely sensitive to motion. Examples of images acquired with a diffusion-weighted spin-echo sequence are shown in Fig. 2.8.9a,b. The volunteers were asked to avoid any movements, but no head fixation was applied; severe motion artifacts degrade the images. These artifacts are caused by inconsistent phase information of the complex-valued raw data; stimulated-echo, and spin-echo sequences are particularly sensitive to these effects because

the diffusion preparation must be repeated for each raw data line and different states of motion will occur during these diffusion preparations. More details about the motion sensitivity of diffusion-weighted sequences and about approaches to reduce these artifacts are described in the following section. With techniques such as cardiac gating and navigator-echo correction, image quality of diffusion-weighted spin-echo sequences can be dramatically improved (Fig. 2.8.9c,d). Today, the most commonly used pulse sequence for diffusion-weighted MRI (particularly of the brain) is the single-shot echo planar imaging (EPI) sequence with spin-echo excitation. The diffusion preparation of this sequence is the same as in the conventional spin-echo sequence of Fig. 2.8.8, but instead of acquiring a single echo after each excitation, the full k-space can be read. The advantages of the diffusion-weighted EPI sequence are a very short acquisition time of less than 200 ms per slice and its insensitivity to motion. However, the image resolution is typically limited to 128 × 128 matrices and echo planar imaging is very sensitive to susceptibility variations as demonstrated in Fig. 2.8.10a,b—different susceptibilities of soft tissue, bone, and air, cause severe image distortion and signal cancellation close to interfaces between soft tissue and air or bone. These effects can be reduced with new imaging methods known as parallel imaging or parallel acquisition techniques (see Sect. 2.4). The underlying idea is to use several receiver coil elements with spatially different coil sensitivity profiles to acquire multiple data sets with reduced k-space sampling density in the phase-encode direction. These data sets are used to calculate a single image corresponding to a fully sampled k-space during post-processing. Reducing the number of phase-encode steps shortens the EPI echo train, decreases the minimum echo time as well as the total acquisition time, and increases the effective receiver bandwidth in the phase-encode direction. As a result, susceptibility-induced distortions are reduced as shown in Fig. 2.8.10c,d. Alternatively, the accelerated acFig. 2.8.8  Diffusion-weighted spin-echo pulse sequence. Diffusion gradients (shown in green) are inserted on both sides of the 180° pulse. Conventional imaging gradients are shown in gray

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging

Fig. 2.8.9  Diffusion-weighted spin-echo acquisitions (b = 550 s/mm²) of two healthy and cooperative volunteers. a,b Uncorrected images acquired without cardiac gating. c,d Images after navigator echo correction acquired with cardiac gating

Fig. 2.8.10  Images acquired with a diffusion-weighted EPI sequence. a,b Conventional EPI sequence exhibiting severe distortions (arrows). c,d EPI sequence with parallel imaging (acceleration factor 2) showing reduced susceptibility artifacts. (For better visualization of artifacts, only images without diffusion weighting (b = 0) are shown)

quisition can be used to increase the matrix size and thus enables echo planar imaging with 192 × 192 or 256 × 256 matrices. Diffusion-weighted imaging of the brain is almost exclusively performed with single-shot EPI sequences (with or without parallel imaging). Other organs or body areas, however, are less suited for echo-planar single-shot acquisitions because of much more severe susceptibility effects that often result in images of non-diagnostic quality. Depending on the slice orientation and receiver coil system, these distortions can be reduced either with parallel imaging as described above or with segmented (i.e., multi-shot) EPI sequences that assemble the raw data from multiple shorter echo trains (Holder et al. 2000; Ries et al. 2000; Einarsdottir et al. 2004). A disadvantage of this approach is the increased motion sensitivity since several excitations (and diffusion preparations) are required for a single data set, resulting in potentially inconsistent phase information; segmented EPI sequences are therefore often combined with additional motion correction techniques. Several other pulse sequences have been proposed for diffusion-weighted imaging. Diffusion gradients can be

added to single-shot fast spin-echo sequences with echo trains of multiple spin-echoes (see also Sect. 2.4) such as HASTE or RARE sequences (Norris et al. 1992). However, the additionally inserted diffusion gradients cause an irregular timing of the originally equidistant refocusing RF pulses. In combination with motion-dependent phase shifts, this violates the CPMG condition, which requires a certain phase relation between excitation and refocusing pulses. Thus, in order to avoid artifacts, various modifications to diffusion-weighted fast spin-echo sequences have been suggested such as additional gradients (Norris et al. 1992), a split acquisition of echoes of even and odd parity (Schick 1997), or modified RF pulse trains (Alsop 1997). These modified diffusion-weighted singleshot fast spin-echo sequences are fast and insensitive to motion; disadvantages are a relatively low signal-to-noise ratio, and a certain image blurring that is characteristic for all single-shot fast spin-echo techniques. They have been applied in the brain (Alsop 1997; Lovblad et al. 1998), the spine (Tsuchiya et al. 2003, Clark and Werring 2002), and in several non-neuro applications such as imaging of musculoskeletal (Dietrich et al. 2005) or breast (Kinoshita et al. 2002) tumors. In contrast to echo-planar

139

140

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

imaging, these techniques are insensitive to susceptibility variations and, thus, particularly suited for applications outside the brain. Another, however only infrequently used alternative to echo-planar sequences are fast gradient-echo techniques (FLASH, MP-RAGE) with diffusion preparation (Lee and Price 1994; Thomas et al. 1998). A special sequence type that has successfully been employed for diffusion-weighted imaging is based on steady-state free-precession (SSFP) sequences (see also Sect. 2.4). Pulse sequences known as CE-FAST or PSIF sequences (the acronym PSIF refers to a reverted fast imaging with steady precession, i.e., FISP, sequence) have been adopted to diffusion-weighted imaging by inserting a single diffusion gradient (LeBihan 1988; Merboldt et al. 1989). However, in contrast to all previously described sequences, the diffusion weighting of this technique cannot be easily determined quantitatively. The observed signal attenuation does not only depend on the diffusion coefficient and the diffusion weighting, but also on the relaxation times, T1 and T2, and the flip angle (Buxton 1993). Since these quantities are usually not exactly known, the ADC cannot be determined. Instead, these sequences have been used to acquire diffusion-weighted images that are evaluated based only on visible image contrast. A general advantage of diffusion-weighted PSIF sequences is the relatively short acquisition time due to short repetition times of about 50 ms. Thus, they exhibit only low motion sensitivity. The most important application of this sequence type is the differential diagnosis of osteoporotic and malignant vertebral compression fractures (Baur et al. 1998, 2003). Other applications include diffusion-weighted imaging of the brain (Miller and Pauly 2003) and the cartilage (Miller et al. 2004). 2.8.3.3  Artifacts in Diffusion MRI: Motion and Eddy Currents As mentioned above, an unwanted side effect arising in virtually all diffusion-weighted pulse sequences is extreme motion sensitivity (Trouard et al. 1996; Norris 2001). By introducing diffusion gradients, the pulse sequence is made sensitive to molecular motion in the micro­meter range, but it also becomes susceptible to very small macroscopic motions of the imaged object since the diffusion gradients do not distinguish between stochastic molecular motion and macroscopic bulk motion. Hence, even very small and involuntary movements of the patient, e.g., caused by cardiac motion, cerebrospinal fluid pulsation, breathing, swallowing, or peristalsis, can lead to severe image degradation due to gross motion artifacts. Typical appearances of these artifacts are signal voids and ghosting in phase-encode direction. Several techniques and pulse-sequence modifications have been proposed to reduce the motion sensitivity of diffusion-weighted MRI. On the one hand, any kind of

motion should be minimized. Depending on the body region being imaged, this can be achieved by improved fixation of the patient to the scanner, by imaging during breath hold, or by applying cardiac gating. Effects of motion can also be reduced by decreasing the acquisition time of a pulse sequence, i.e., by using fast acquisition techniques. This is particularly effective if singleshot sequences such as echo planar imaging techniques are applied. Most motion artifacts in diffusion-weighted imaging arise from inconsistent phase information in the complex-valued raw data set. This is caused by different states of motion in the repeated diffusion preparations of the acquisition. In single-shot sequences, only a single diffusion preparation is applied, and thus inconsistent phase information is avoided. It should be noted, however, that even single-shot sequences might be affected by inconsistent phase information if complex data of several measurements is averaged. Instead, only magnitude images should be averaged in diffusion-weighted MRI in order to improve the signal-to-noise ratio. Another approach to reduce motion artifacts is to correct for motion-related phase errors in the acquired raw data. This can be done using navigator echo–correction techniques (Ordidge et al. 1994; Anderson and Gore 1994; Dietrich et al. 2000). The navigator echo is an additional echo without phase encoding acquired after each diffusion preparation. In the absence of motion, all navigator echoes should be identical. Thus, by comparing the acquired navigator echoes, bulk motion can be detected, and degraded image echoes can be discarded or a phase correction can be applied. More advanced navigator-echo techniques acquire several navigator echoes in different spatial directions (Butts et al. 1996) or use spiral navigator readouts (Miller and Pauly 2003). Certain pulse sequences are self-navigated, i.e., a subset of the acquired raw data can be used as navigator echo without the need for an extra navigator acquisition. Examples are pulse sequences with radial or spiral k-space trajectories that acquire the origin of k-space in every readout (Seifert et al. 2001; Dietrich et al. 2001b). An improved self-navigation is possible with the PROPELLER diffusion sequence, which repeatedly acquires a large area around the origin of k-space (Pipe et al. 2002). Some image reconstruction techniques have been proposed that do not use the often-inconsistent phase information of raw data at all. In sequences with radial k-space trajectories, images can be reconstructed by filtered back projection of magnitude projection images (Gmitro and Alexander 1993). Another spin-echo–based approach known as line-scan diffusion imaging assembles the image from one-dimensional lines of magnitude data (Gudbjartsson et al. 1996). In addition to substantially reduced motion sensitivity, repetition times and thus image acquisition time can be considerably reduced since the one-dimensional lines are acquired independently of each other. On the other hand, the signal-to-noise ratio

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging

of line-scan sequences is substantially lower than that of conventional acquisition techniques and the spatial resolution of this approach is limited as well. A second unwanted side effect of diffusion-weighted sequences is eddy current effects caused by the extraordinarily long and strong diffusion gradients. Eddy currents are induced electric currents in coils that occur after switching magnetic fields on or off. These currents then create unwanted additional gradient fields resulting in shifted or distorted images and in incorrect diffusion weightings. Whereas most MRI gradient systems compensate very well for eddy current effects after the switching of short gradients typically used for imaging, the longer diffusion gradients are often not well compensated. Hence, diffusion-weighted images are sometimes distorted depending on the diffusion weighting and the direction of the diffusion gradients, resulting in artifacts on ADC maps such as enhanced edges. To avoid these artifacts, several techniques have been suggested. Diffusion gradients can be shortened by using bipolar diffusion gradients (Alexander et al. 1997) or by adding additional 180° pulses during the diffusion preparation (Reese et al. 2003); eddy currents can be partially compensated for by an additional long gradient before the 90° excitation pulse (Alexander et al. 1997); or diffusion-weighted images can be acquired twice with diffusion gradients of opposite polarity (Bodammer et al. 2004). Other eddycurrent correction schemes are based on the acquisition of diffusion gradient–dependent field maps and data correction in k-space (Horsfield 1999; Papdakis et al. 2005). In general, (automated) image registration as the first step of postprocessing is recommended to reduce influences from both patient motion and eddy-current effects. 2.8.4  MR Measurement of Diffusion Tensor Data 2.8.4.1  Diffusion Trace Imaging Imaging with the Stejskal-Tanner diffusion preparation as described above in Sect. 2.8.3.1, is only sensitive for molecular diffusion parallel to the direction of the diffusion gradient. The diffusion preparation causes a dephasing of spins that move in the direction of the applied field gradient, i.e., between positions with different magnetic field strengths as illustrated in Fig. 2.8.6. Molecular motion perpendicular to this direction does not contribute to the signal attenuation. In general, the diffusion displacement of spins depends on the considered spatial direction; e.g., protons of water molecules in nerve fibers move more freely parallel to the fiber direction than they do in perpendicular directions. This dependence of the diffusion on spatial orientation can be measured by applying diffusion gradients in different spatial directions, e.g., separately in slice, readout, and phase direction as demonstrated in Fig. 2.8.11. The resulting diffusion-

weighted images show substantial signal differences in areas with strong anisotropic diffusion such as the corpus callosum. The signal intensity of the corpus callosum is decreased if diffusion gradients in the left–right direction (readout direction in the example) are applied, but increased for diffusion gradients in the head–foot (slice) direction or the anterior–posterior (phase-encode) direction. This finding is explained by the fact that water mo­ lecules diffuse more freely in the left–right direction (parallel to the nerve fibers) than they do in perpendicular directions, i.e., the effective diffusion coefficient is greater in the left–right direction than it is in other directions, and thus the signal attenuation is increased. This orientation dependence is visible in the ADC maps as well: the ADC in the left–right direction of the corpus callosum is increased compared to the ADCs in perpendicular directions. Other areas such as gray matter or the CSF do not show significant differences depending on the diffusion gradient direction, indicating approximately isotropic diffusion. If the mean (or average) diffusivity of molecules in tissue is to be measured, then diffusion coefficients for all spatial directions must be averaged as shown in Fig. 2.8.11d; the corresponding ADC map is given by the mean value of the three direction-dependent maps. Since the direction-independent or mean ADC of tissue is proportional to the trace of the diffusion tensor, this measurement is also referred to as diffusion trace imaging. The measurement of such a direction-independent diffusion-weighted image can be very important to avoid misinterpretation of hyperintense areas due to high anisotropy as tissue with generally reduced ADC such as areas of focal ischemia. Therefore, diffusion-weighted stroke MRI is generally based on isotropically diffusionweighted images. If only a single direction-independent diffusionweighted image is required for diagnosis, it appears disadvantageous to perform three orthogonal diffusion measurements at the cost of three-times-increased acquisition duration. It should be noted that it is not possible to simply apply gradients in all three directions simultaneously for this purpose; this results in a single magnetic field gradient in diagonal direction, which is again only sensitive for diffusion parallel to this diagonal. However, the Stejskal-Tanner diffusion preparation can be extended by a more sophisticated series of gradient pulses in different directions to achieve an isotropic diffusion weighting within a single diffusion measurement (Wong et al. 1995; Mori and van Zilj 1995; Chun et al. 1998; Cercignani and Horsefield 1999). 2.8.4.2  Basic Diffusion Tensor Imaging Isotropically diffusion-weighted images can thus be acquired by either a single or three orthogonal diffusion

141

142

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.8.11  Diffusion-weighted imaging in different spatial directions. a Diffusion gradients in slice (S), readout (R), and phase (P) direction; the row vectors   (S, R, P) denote the selected gradients.   b Corresponding diffusion-weighted images. c Calculated ADC maps corresponding to the diffusion directions in a and images in b. d Averaged ADC map; all ADCs are in units of 10–3 mm2/s. Note the differing contrast in the diffusion-weighted images and ADC maps depending on the diffusion gradient direction (e.g., in the corpus callosum)

preparations. However, three measurements are not yet sufficient to determine the properties of diffusion an­iso­ tropy in all cases. For example, if a nerve fiber is oriented diagonally to all three coordinate axes, then the diffusion attenuation in this fiber will be the same for the three measurements and cannot be distinguished from isotropic diffusion. The measurement of the full diffusion tensor (cf. Sect. 2.8.2.2) is required to cope with these more general cases. In spite of this limitation, some studies have used the ratio of the largest and the smallest of three perpendicular diffusion coefficients as an estimation of the anisotropy (Holder et al. 2000). However, this approach should be regarded as an inferior method in comparison to diffusion tensor evaluation and is generally not recommended. To determine the diffusion tensor, i.e., to fully measure anisotropic diffusion, more than three diffusionsensitized measurements with diffusion gradients in different spatial directions are required. However, only the diagonal elements of the tensor, i.e., Dxx, Dyy, Dzz, can be meas­ured directly; these elements are exactly the direction-dependent ADCs determined in the example above. The other three (off-diagonal) tensor components Dxy, Dxz, Dyz do not describe diffusion in a spatial direction but

the correlation of diffusion in two different directions; they cannot be measured directly, but must be calculated as linear combinations of several measurements. The minimum number of measurements required to determine the full diffusion tensor can be deduced from the form of the diffusion tensor matrix: The tensor has six independent components Dxx, Dyy, Dzz, Dxy, Dxz, Dyz and, thus, at least six independent diffusion measurements are required. Each of these measurements is based on images of at least two different b-values; in order to reduce the total number of measurements, usually a b-value of 0 is chosen as a direction-independent reference. Thus, this reference image has to be acquired only once instead of separately for each diffusion direction. A possible and frequently used choice of seven diffusion-weighted acquisitions that are sufficient to determine the diffusion tensor (Basser and Pierpaoli 1998) is shown in Fig. 2.8.12. None of the six tensor components Dxx, Dyy, Dzz, Dxy, Dxz, or Dyz is measured directly by this gradient scheme; instead, all components must be calculated as linear combinations of the diffusion coefficients in these six directions. This calculation is based on the so-called b-matrix (Basser and Pierpaoli 1998), a symmetric 3 × 3 matrix describing the diffusion weighting for an arbitrary diffusion gradient

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging Fig. 2.8.12  Diffusion tensor imaging a Choice of diffusion gradients (S slice, R readout, P phase direction; the row vector denotes the selected gradients and their polarity) and b corresponding diffusionweighted images for the determination of the diffusion tensor. Note the different contrast in the diffusion-weighted images depending on the diffusion gradient direction (e.g., in the corpus callosum)

g = ( gx gy gz):

  where gg denotes the dyadic product of these two vectors. This matrix is used to describe the signal attenuation due to the diffusion gradient as

         where bD denotes the matrix product of the b-matrix and the diffusion tensor matrix. The elements of the diffusion tensor Dij can be determined by solving a system of linear equations, since the b-matrix and the signal attenuation are known. The result of this calculation is shown in Fig. 2.8.13. The three calculated diagonal elements correspond to the direct ADC measurements of Fig. 2.8.11. The off-diagonal elements are generally much lower than the diagonal elements (note the differently scaled intensity maps) are and are close to zero in areas with predominantly isotropic diffusion (gray matter and CSF).

2.8.4.3  Optimizing Diffusion Tensor Imaging A simple protocol for diffusion tensor imaging consists of one reference measurement without diffusion weighting (b-value is 0) and six diffusion-weighted measurements with different gradient directions. These gradient directions should be “as different as possible,” i.e., pointing isotropically in all spatial directions. A typical b-value for DTI measurements of the brain is 1,000 s/mm2. Averaging of multiple acquisitions is frequently performed to increase the SNR especially of the images with diffusion weighting. However, all these parameters (b-values, diffusion directions, number of averages) have been evaluated in a number of studies with the aim of optimizing the accuracy of diffusion tensor data. Several studies investigated the optimum choice of the b-values both for conventional diffusion-weighted imaging and for diffusion tensor imaging. Although the results of these studies vary to a certain extent, generally b-values in the range between about 900 and 1400 s/ mm2 have been found to provide the highest accuracy of diffusion measurements in the brain (Jones et al. 1999; Armitage and Bastin 2001; Kingsley and Monahan 2004). The optimum number of averages depends on the b-values, which influence the signal attenuation and, thus, the signal-to-noise ratio of the diffusion-weighted images. In general, a higher number of averages are recommended for the acquisition with the high b-value than for the reference image with low b-value or without

143

144

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.8.4.4  Beyond Diffusion Tensor Imaging

Fig. 2.8.13  Diffusion tensor data calculated from the measurements shown in Fig. 2.8.12; all diffusion coefficients are in units of 10–3 mm2/s. a The three diagonal elements Dxx, Dyy, and Dzz, and b the off-diagonal elements Dxy, Dxz, and Dyz of the tensor matrix. Note the different intensity scales for diagonal and off-diagonal elements. Some remaining eddy-current artifacts can be seen as enhanced edges in the maps of the off-diagonal elements

any diffusion weighting. As shown by Jones et al. (1999) for their choice of b-values, the optimum ratio of the total number of acquisitions with high b-value and low bvalue is about 8.4. The number of diffusion gradients and their directions has also been investigated in several studies. Generally, the accuracy of diffusion tensor data, especially of the diffusion anisotropy and main diffusion direction, is improved when the number of different diffusion directions is increased (Jones et al. 1999; Papadakis et al. 2000; Skare et al. 2000; Jones 2004). If the number of different directions is fixed, then the accuracy of the measurements can be increased by choosing an optimized set of diffusion directions (Skare et al. 2000; Hasan et al. 2001). No final consensus about the optimum number and choice of directions of diffusion gradients has yet been established, but protocols with 20 or more diffusion directions are currently recommended by many research groups.

The diffusion tensor contains complex information about the tissue microstructure that is best visualized as a threedimensional ellipsoid as discussed in Sect. 2.8.2.2. However, diffusion tensor data may be insufficient to describe tissue in certain geometrical situations. A well-known example is the crossing of white-matter fibers within a single voxel as illustrated in Fig. 2.8.14. Water diffusion in such voxels cannot be fully described by a single ellipsoid, i.e., by the diffusion tensor. To overcome this limitation, more complex measurement techniques such as high-angular resolution diffusion imaging (HARDI) (Frank 2001; Tuch et al. 2002) and q-ball imaging (Tuch 2004) have been proposed. All these techniques use a large number of different diffusion directions (e.g., between 43 [Frank 2001] and 253 [Tuch 2004]) distributed isotropically in space. Diffusion data is measured with high-angular resolution in order to determine the spatial distribution of diffusion in more detail as indicated in Fig. 2.8.14d. A further generalization of diffusion tensor measurements loosens the assumption of Gaussian diffusion, which was illustrated in Fig. 2.8.2. If diffusion is severely restricted, e.g., by cell membranes, no or very few mo­lecules will move through this border; the probability distribution of diffusion distances will be limited to distances within the cell volume and will no longer be Gaussian. The exact displacement probabilities in restricted diffusion can be measured with methods called q-space diffusion imaging (Assaf et al. 2002) or diffusion spectrum imaging (Wedeen et al. 2005). Both techniques require the acquisition of images with a large number of different b-values and, in the case of diffusion spectrum imaging, of different diffusion directions; e.g., the total number of diffusion measurements reported by Wedeen et al. (2005) is 515. Obviously, this large number of measurements severely limits the applicability of these new techniques in clinical studies; the studies should therefore be regarded as experimental work. Another approach to overcome the limitations of models based purely on Gaussian diffusion has been proposed by Jensen et al. as diffusional kurtosis imaging (Jensen et al. 2005). Diffusion data is acquired for several b-values over a large range between 0 and 2,500 s/mm2 similarly to the way data is acquired in q-space imaging, but with a different mathematical model of the non-exponential decay. This method is related to several other studies that investigated diffusion properties in tissue at high b-values and found non-mono-exponential diffusion attenuation curves (Inglis et al. 2001; Clark et al. 2002). This observation has frequently been attributed to the simultaneous measurement of water molecules in different environments such as the intracellular and the extracellular space; however, no final agreement on the interpretation of these data has been established (Sehy et al. 2002).

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging Fig. 2.8.14  The diffusion tensor cannot represent diffusion properties in voxels with crossing nerve fibers. Voxels with a single predominant fiber direction (a,b) show diffusion tensor ellipsoids whose longest axes correspond to the fiber orientation. Voxels with crossing fibers (c) result in a diffusion tensor ellipsoid with reduced anisotropy pointing in an averaged fiber direction. Advanced methods such as high angular resolution diffusion imaging (d) can resolve different fiber orientations within a single voxel

2.8.5  Visualization of Diffusion Tensor Data Diffusion tensor imaging results in a large amount of data—a full diffusion tensor, i.e., a symmetric 3 × 3 matrix, is determined for each pixel of the image dataset. Due to this complex data structure, there is no simple way to visualize the complete diffusion tensor as a single intensity or color map. It would be straightforward to display the six independent elements of the tensor as separate maps as shown in Fig. 2.8.13; however, this would not be very helpful for the interpretation or quantitative evaluation of diffusion tensor data. Instead, several techniques are used to reduce the diffusion tensor information to simpler datasets that can as easily be displayed and interpreted as other imaging data. 2.8.5.1  Scalar Diffusion Quantities Most results of imaging examinations are presented as either signal intensity images or scalar parameter maps. These images and maps have the advantage that they can easily be manipulated, e.g., the contrast can be interactively adjusted, and they can be quantitatively evaluated by statistics over regions of interest. In order to obtain similar parameter maps of diffusion tensor data, a single scalar reflecting a certain tensor property must be calculated. The most important examples of such scalars are the mean diffusivity or trace of the diffusion tensor and the anisotropy of the tensor. The mean diffusivity of a diffusion tensor measurement, i.e., the diffusion coefficient averaged over all spatial directions, is displayed as parameter maps in Fig. 2.8.15a,b. The same data can be displayed either as an intensity-coded map (Fig. 2.8.15a) or as a color-coded map (Fig. 2.8.15b). Both maps illustrate, e.g., the high diffusivity of CSF and the typical ADCs of about 0.7 × 10–3 mm2/s in the white matter. Many different scalar measures have been proposed to describe diffusion anisotropy, cf. Sect. 2.8.2.3. The two most important are the fractional anisotropy and the

relative anisotropy shown in Fig. 2.8.15c,d. The maps are very similar; both show the high anisotropy of white matter as hyperintense areas in contrast to low anisotropy in gray matter or CSF. These two scalars derived from the diffusion tensor are by far the most important quantities for the clinical evaluation of diffusion tensor data. The vast majority of clinical studies based on diffusion tensor imaging determine the mean diffusivity and the anisotropy in regions of interest in order, e.g., to statistically compare these data between certain patient groups or between patients and healthy controls. 2.8.5.2  Vector Diffusion Quantities The mean diffusivity and the anisotropy contain certain important information about the diffusion tensor; if the diffusion tensor is visualized as ellipsoid, then the diffusivity reflects the volume of the ellipsoid and the anisotropy its deviation from a spherical shape. However, any information about the main diffusion direction, i.e., the orientation of the longest axis of the diffusion tensor ellipsoid, is missing. This direction corresponds to the micro­ structural orientation of tissue, e.g., the orientation of white-matter tracts, and is determined as the eigenvector of the largest eigenvalue of the tensor (c.f. Sect. 2.8.2.2). There are two common methods to visualize the direction of this eigenvector: color coding and direct vector display. The direction can be color-coded using the red– green–blue (RGB) color model. Each direction in space is defined by a three-component vector v = (vx vy vz) . If this three-component vector is interpreted as an RGB color specification, vectors in x-direction, v = (1 0 0), appear as red pixels, vectors in y-direction as green pixels, and vectors in z-direction as blue pixels. Eigenvectors in other directions are displayed as (additive) mixtures of different colors, e.g., the vector v = (1 0 1) as mixture of red and blue, yielding violet pixels. The resulting color map is finally scaled with the diffusion anisotropy, since the

145

146

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.8.15  Parameter maps displaying scalar quantities calculated from the diffusion tensor. Direction-independent mean diffusivity shown as gray-scaled map (a) and as color-coded map (b); the diffusion coefficients are given in units of 10–3 mm2/s. (c) Fractional anisotropy and (d) relative anisotropy

Fig. 2.8.16  Color-coded visualization of main diffusion orientation in four different slices (a–d). The main diffusion direction (orientation of the longest axes of the diffusion ellipsoid) is shown in red, green, and blue for left–right, anterior–posterior,

and head–foot orientation, respectively, as indicated in e. The green rim at the frontal brain is caused by remaining eddy-current and susceptibility artifacts

main diffusion direction is of interest only in areas with high anisotropy. Some examples of these color-coded vector maps are shown in Fig. 2.8.16. The red color of the corpus callosum demonstrates that the nerve fibers are predominantly oriented in the left–right direction. White-matter areas in green and blue are oriented in the anterior–posterior direction and the head–foot direction, respectively. Alternatively, the main diffusion direction can be directly displayed by a small line in each pixel; some authors refer to this technique as whisker plots. This visualization is on the one hand more intuitive than color coding, but on the other hand difficult to display for large areas because of the large number of pixels (and hence lines) of a complete image. An example is shown in Fig. 2.8.17; the magnified area shows again the corpus callosum, where the diffusion directions follow the anatomical orientation of the nerve fibers.

A general problem and limitation of the visualization of the main diffusion direction is that it is based on the assumption of linear diffusion, i.e., the diffusion ellipsoid is supposed to have a cigar-like shape. This is usually true in white matter tracts, but may lead to deceptive graphical depictions at crossing fibers or if diffusion is described by a planar tensor. Another disadvantage is that vector maps are difficult to compare or to evaluate statistically. 2.8.5.3  Full Tensor Visualization It is also possible to visualize the full diffusion tensor using the diffusion ellipsoid introduced in Sect. 2.8.2.2. As in the vector plots, it is often difficult to visualize the entire amount of data belonging to a single image slice at once. Therefore, this 3D tensor visualization is usually combined with tools to zoom into the illustration and

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging Fig. 2.8.17  Direct vector visualization of main diffusion orientation; the orientation of the longest axis of the diffusion ellipsoid is shown as a yellow line for each pixel

Fig. 2.8.18  Three-dimensional visualization of the full diffusion tensor as colorcoded ellipsoid; the ellipsoids are colored as in Fig. 2.8.16. Visualization was performed with the “DTI Task Card” provided by the MGH/MIT/HMS Athinoula A. Martinos Center for Functional and Structural Biomedical Imaging (Ruopeng Wang)

to rotate the slice in order to view the tensors in specific areas of the brain, as demonstrated in Fig. 2.8.18. The ellipsoids are additionally color-coded to emphasize the direction of their longest axis (the main diffusion direction); their brightness is scaled by the anisotropy. Thus, the ellipsoid visualization combines features of the techniques described in the previous sections and, e.g., CSF is displayed as large but relatively dark spheres (denoting a high diffusion coefficient and low anisotropy), while the tensors in fiber tracts appear as bright elongated ellipsoids corresponding to linear diffusion in a single predominant orientation. The exact depiction of the tensor information is not standardized but may look different depending on the tools used. An alternative visualization may substitute the ellipsoids by cuboids with equivalent dimensions as shown in Fig. 2.8.19. The presented information is the same as before, but the computational cost required to

display cuboids is substantially lower than with smooth ellipsoids. Thus, interactive manipulation of the 3D datasets may be faster using the cuboid visualization. 2.8.5.4  Fiber Tracking Close inspection of the main diffusion directions in Figs. 2.8.17 or 2.8.19 suggests that the shape of white matter tracts can be reconstructed by connecting several diffusion directions in an appropriate way. This process is illustrated in Fig. 2.8.20, based on a magnification of Fig. 2.8.17. By choosing a start point and following the main diffusion direction, trajectories can be constructed that visualize the fiber tracts of white matter. A typical example is shown in Fig. 2.8.21, where a seed region was placed within the corpus callosum, and all fibers through this seed region were reconstructed. The color of the fi-

147

148

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy Fig. 2.8.19  Three-dimensional visualization of the full diffusion tensor as colorcoded cuboids; the cuboids are colored as in Fig. 2.8.16. Visualization was performed with the “DTI Task Card” provided by the MGH/MIT/HMS Athinoula A. Martinos Center for Functional and Structural Biomedical Imaging (Ruopeng Wang)

Fig. 2.8.20  Fiber tracking (white-matter tractography) is performed by connecting pixels along the main diffusion direction. a A start point (or seed region) is selected. b Pixels are connected following the main diffusion orientation. c Several reconstructed fiber tracts starting in the corpus callosum

bers reflects the local anisotropy in this case, but various other color schemes could be used instead. Fiber tracking or diffusion tractography was developed in the late 1990s (Mori et al. 1999; Conturo et al. 1999; Mori and van Zijl 2002; Melhem et al. 2002), and a multitude of different algorithms to reconstruct fibers have been proposed since then. Most techniques include data interpolation to increase the spatial resolution, and all require certain criteria to decide when the tracking of a fiber should be stopped (e.g., at pixels with low anisotropy or at sudden changes of diffusion direction). Fiber tracking is usually based either on a single-region approach, in which all fibers are tracked that go through a user-defined region of interest, or on a two-region approach where connecting fibers between two regions are reconstructed. Fiber tracking depends on good image quality, with sufficient signal-to-noise ratio and without substantial

distortion artifacts. Increased noise can reduce the calculated anisotropy (Jones and Basser 2004) and, thus, the length of the reconstructed fibers. Image distortions cause a mismatch of anatomical fiber orientation and the measured diffusion direction and thus can lead to erroneous tractography results. Therefore, parallel imaging and eddy-current correction techniques can improve the results of white-matter tractography. It is generally assumed that isotropic spatial image resolution is preferable for fiber tracking applications. A typical protocol suggested by Jones et al. acquires data of the whole brain in isotropic 2.5 × 2.5 × 2.5 mm3 resolution (Jones et al. 2002b). Fiber tracking is a valuable tool to visualize white matter structures of the brain. However, it is still very difficult to evaluate tractography results quantitatively, to assess the accuracy of reconstructed fibers, or to compare the results of different examinations. First approaches to these questions include the spatial normalization of ten-

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging Fig. 2.8.21  Reconstruction of white matter tracts starting at a seed region in the corpus callosum. Visualization was performed with the “DTI Task Card” provided by the MGH/MIT/HMS Athinoula A. Martinos Center for Functional and Structural Biomedical Imaging (Ruopeng Wang)

sor data sets (Jones et al. 2002a) and the determination and visualization of uncertainties of diffusion tensor results (Jones 2003; Jones and Pierpaoli 2005).

8.

9.

References 1.

2.

3.

4.

5.

6. 7.

Alexander AL, Tsuruda JS, Parker DL (1997) Elimination of eddy current artifacts in diffusion-weighted echo planar images: the use of bipolar gradients. Magn Reson Med 38:1016–1021 Alsop DC (1997) Phase insensitive preparation of singleshot RARE: application to diffusion imaging in humans. Magn Reson Med 38:527–533 Anderson AW, Gore JC (1994) Analysis and correction of motion artifacts in diffusion weighted imaging. Magn Reson Med 32:379–387 Armitage PA, Bastin ME (2001) Utilizing the diffusion-tonoise ratio to optimize magnetic resonance diffusion tensor acquisition strategies for improving measurements of diffusion anisotropy. Magn Reson Med 45:1056–1065 Assaf Y, Ben-Bashat D, Chapman J, Peled S, Biton IE, Kafri M, Segev Y, Hendler T, Korczyn AD, Graif M, Cohen Y (2002) High b-value q-space analyzed diffusion-weighted MRI: application to multiple sclerosis. Magn Reson Med 47:115–126 Basser PJ, Mattiello J, LeBihan D (1994) MR diffusion tensor spectroscopy and imaging. Biophys J 66:259–267 Basser PJ, Pierpaoli C (1996) Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 111:209–219

10.

11.

12.

13.

14.

15.

Basser PJ, Pierpaoli C (1998) A simplified method to measure the diffusion tensor from seven MR images. Magn Reson Med 39:928–934 Baur A, Stabler A, Bruning R, Bartl R, Krodel A, Reiser M, Deimling M (1998) Diffusion-weighted MR imaging of bone marrow: differentiation of benign versus pathologic compression fractures. Radiology 207:349–356 Baur A, Dietrich O, Reiser M (2003) Diffusion-weighted imaging of bone marrow: current status. Eur Radiol 13:1699–1708 Beaulieu C (2002) The basis of anisotropic water diffusion in the nervous system - a technical review. NMR Biomed 15:435–455 Bodammer N, Kaufmann J, Kanowski M, Tempelmann C (2004) Eddy current correction in diffusion-weighted imaging using pairs of images acquired with opposite diffusion gradient polarity. Magn Reson Med 51:188–193 Boulanger Y, Amara M, Lepanto L, Beaudoin G, Nguyen BN, Allaire G, Poliquin M, Nicolet V (2003) Diffusionweighted MR imaging of the liver of hepatitis C patients. NMR Biomed 16:132–136 Brown R (1866) A brief account of microscopical observations made in the months of June, July and August, 1827, on the particles contained in the pollen of plants; and on the general existence of active molecules in organic and inorganic bodies. In: Bennett JJ (ed) The miscellaneous botanical works of Robert Brown, vol. 1. Hardwicke, London Burdette JH, Elster AD, Ricci PE (1999) Acute cerebral infarction: quantification of spin-density and T2 shinethrough phenomena on diffusion-weighted MR images. Radiology 212:333–339

149

150

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 16. Butts K, de Crespigny A, Pauly JM, Moseley M (1996) Diffusion-weighted interleaved echo planar imaging with a pair of orthogonal navigator echoes. Magn Reson Med 35:763–770 17. Buxton RB (1993) The diffusion sensitivity of fast steady-state free precession imaging. Magn Reson Med 29:235–243 18. Carr HY, Purcell EM (1954) Effects of diffusion on free precession in nuclear magnetic resonance experiments. Phys Rev 94:630–638 19. Cercignani M, Horsfield MA (1999) An optimized pulse sequence for isotropically weighted diffusion imaging. J Magn Reson 140:58–68 20. Cercignani M, Iannucci G, Rocca MA, Comi G, Horsfield MA, Filippi M (2000) Pathologic damage in MS assessed by diffusion-weighted and magnetization transfer MRI. Neurology 54:1139–1144 21. Chien D, Buxton RB, Kwong KK, Brady TJ, Rosen BR (1990) MR diffusion imaging of the human brain. J Comput Assist Tomogr 14:514–520 22. Chien D, Kwong KK, Gress DR, Buonanno FS, Buxton RB, Rosen BR (1992) MR diffusion imaging of cerebral infarction in humans. AJNR Am J Neuroradiol 13:1097–1102 23. Chun T, Ulug AM, van Zijl PC (1998) Single-shot diffusion-weighted trace imaging on a clinical scanner. Magn Reson Med 40:622–628 24. Clark CA, Werring DJ (2002a) Diffusion tensor imaging in spinal cord: methods and applications – a review. NMR Biomed 15:578-586 25. Clark CA, Hedehus M, Moseley ME (2002b) In vivo mapping of the fast and slow diffusion tensors in human brain. Magn Reson Med 47:623–628 26. Conturo TE, Lori NF, Cull TS, Akbudak E, Snyder AZ, Shimony JS, McKinstry RC, Burton H, Raichle ME (1999) Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci USA 96:10422–10427 27. Cova M, Squillaci E, Stacul F, Manenti G, Gava S, Simonetti G, Pozzi-Mucelli R (2004) Diffusion-weighted MRI in the evaluation of renal lesions: preliminary results. Br J Radiol 77:851–857 28. Dietrich O, Heiland S, Benner T, Sartor K (2000) Reducing motion artefacts in diffusion-weighted MRI of the brain: efficacy of navigator echo correction and pulse triggering. Neuroradiology 42:85–91 29. Dietrich O, Heiland S, Sartor K (2001a) Noise correction for the exact determination of apparent diffusion coefficients at low SNR. Magn Reson Med 45:448–453 30. Dietrich O, Herlihy A, Dannels WR, Fiebach J, Heiland S, Hajnal JV, Sartor K (2001b) Diffusion-weighted imaging of the spine using radial k-space trajectories. MAGMA 12:23–31 31. Dietrich O, Raya JG, Sommer J, Deimling M, Reiser MF, Baur-Melnyk A (2005) A comparative evaluation of a RARE-based single-shot pulse sequence for diffusionweighted MRI of musculoskeletal soft-tissue tumors. Eur Radiol 15:772–783

32. Einarsdottir H, Karlsson M, Wejde J, Bauer HC (2004) Diffusion-weighted MRI of soft tissue tumours. Eur Radiol 14:959–963 33. Filippi M, Inglese M (2001) Overview of diffusion-weighted magnetic resonance studies in multiple sclerosis. J Neurol Sci 186:S37–S43 34. Frank LR (2001) Anisotropy in high angular resolution diffusion-weighted MRI. Magn Reson Med 45:935–939 35. Gmitro AF, Alexander AL (1993) Use of a projection reconstruction method to decrease motion sensitivity in diffusion-weighted MRI. Magn Reson Med 29:835–838 36. Gudbjartsson H, Maier SE, Mulkern RV, Morocz IA, Patz S, Jolesz FA (1996) Line scan diffusion imaging. Magn Reson Med 36:509–519 37. Hahn EL (1950) Spin-echoes. Phys Rev 80:580–594 38. Hasan KM, Parker DL, Alexander AL (2001) Comparison of gradient encoding schemes for diffusion-tensor MRI. J Magn Reson Imaging 13:769–780 39. Helenius J, Soinne L, Perkio J, Salonen O, Kangasmaki A, Kaste M, Carano RA, Aronen HJ, Tatlisumak T (2002) Diffusion-weighted MR imaging in normal human brains in various age groups. AJNR Am J Neuroradiol 23:194–199 40. Herneth AM, Philipp MO, Naude J, Funovics M, Beichel RR, Bammer R, Imhof H (2002) Vertebral metastases: assessment with apparent diffusion coefficient. Radiology 225:889–894 41. Holder CA, Muthupillai R, Mukundan S Jr, Eastwood JD, Hudgins PA (2000) Diffusion-weighted MR imaging of the normal human spinal cord in vivo. AJNR Am J Neuroradiol 21:1799–1806 42. Horsfield MA (1999) Mapping eddy current induced fields for the correction of diffusion-weighted echo planar images. Magn Reson Imaging 17:1335–1345 43. Inglis BA, Bossart EL, Buckley DL, Wirth ED III, Mareci TH (2001) Visualization of neural tissue water compartments using biexponential diffusion tensor MRI. Magn Reson Med 45:58–587 44. Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K (2005) Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 53:1432–1440 45. Jones DK (2003) Determining and visualizing uncertainty in estimates of fiber orientation from diffusion tensor MRI. Magn Reson Med 49:7–12 46. Jones DK (2004a) The effect of gradient sampling schemes on measures derived from diffusion tensor MRI: a Monte Carlo study. Magn Reson Med 51:807–815 47. Jones DK, Basser PJ (2004b) “Squashing peanuts and smashing pumpkins”: how noise distorts diffusionweighted MR data. Magn Reson Med 52:979–993 48. Jones DK, Pierpaoli C (2005) Confidence mapping in diffusion tensor magnetic resonance imaging tractography using a bootstrap approach. Magn Reson Med 53:1143–1149 49. Jones DK, Horsfield MA, Simmons A (1999) Optimal strategies for measuring diffusion in anisotropic systems by magnetic resonance imaging. Magn Reson Med 42:515-525

2.8  Diffusion-Weighted Imaging and Diffusion Tensor Imaging 50. Jones DK, Griffin LD, Alexander DC, Catani M, Horsfield MA, Howard R, Williams SC (2002a) Spatial normalization and averaging of diffusion tensor MRI data sets. Neuroimage 17:592–617 51. Jones DK, Williams SC, Gasston D, Horsfield MA, Simmons A, Howard R (2002b) Isotropic resolution diffusion tensor imaging with whole brain acquisition in a clinically acceptable time. Hum Brain Mapp 15:216–230 52. Kingsley PB, Monahan WG (2004) Selection of the optimum b factor for diffusion-weighted magnetic resonance imaging assessment of ischemic stroke. Magn Reson Med 51:996–1001 53. Kingsley PB, Monahan WG (2005) Contrast-to-noise ratios of diffusion anisotropy indices. Magn Reson Med 53:911–918 54. Kinoshita T, Yashiro N, Ihara N, Funatu H, Fukuma E, Narita M (2002) Diffusion-weighted half-Fourier singleshot turbo spin-echo imaging in breast tumors: differentiation of invasive ductal carcinoma from fibroadenoma. J Comput Assist Tomogr 26:1042–1046 55. Le Bihan D (1988) Intravoxel incoherent motion imaging using steady-state free precession. Magn Reson Med 7:346–351 56. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M (1986) MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 161:401–407 57. Le Bihan D, Breton E, Lallemand D, Aubin ML, Vignaud J, Laval-Jeantet M (1988) Separation of diffusion and perfusion in intravoxel incoherent motion MR imaging. Radiology 168:497–505 58. Lee H, Price RR (1994) Diffusion imaging with the MPRAGE sequence. J Magn Reson Imaging 4:837–842 59. Lovblad KO, Jakob PM, Chen Q, Baird AE, Schlaug G, Warach S, Edelman RR (1998) Turbo spin-echo diffusionweighted MR of ischemic stroke. AJNR Am J Neuroradiol 19:201–208 60. Melhem ER, Mori S, Mukundan G, Kraut MA, Pomper MG, van Zijl PC (2002) Diffusion tensor MR imaging of the brain and white matter tractography. Am J Roentgenol 178:3–16 61. Merboldt KD, Hanicke W, Frahm J (1985) Self-diffusion NMR imaging using stimulated echoes. J Magn Reson 64:479–486 62. Merboldt KD, Bruhn H, Frahm J, Gyngell ML, Hanicke W, Deimling M (1989) MRI of “diffusion” in the human brain: new results using a modified CE-FAST sequence. Magn Reson Med 9:423–429 63. Miller KL, Pauly JM (2003) Nonlinear phase correction for navigated diffusion imaging. Magn Reson Med 50:343–353 64. Miller KL, Hargreaves BA, Gold GE, Pauly JM (2004) Steady-state diffusion-weighted imaging of in vivo knee cartilage. Magn Reson Med 51:394–398 65. Mills R (1973) Self-diffusion in normal and heavy water in the range 1–45°. J Phys Chem 77:685–688

66. Mori S, van Zijl PC (1995) Diffusion weighting by the trace of the diffusion tensor within a single scan. Magn Reson Med 33:41–52 67. Mori S, van Zijl PC (2002) Fiber tracking: principles and strategies - a technical review. NMR Biomed 15:468–480 68. Mori S, Crain BJ, Chacko VP, van Zijl PC (1999) Threedimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 45:265–269 69. Moseley M (2002) Diffusion tensor imaging and aging – a review. NMR Biomed 15:553–560 70. Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Kucharczyk J, Wendland MF, Weinstein PR (1990a) Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med 14:330–346 71. Moseley ME, Kucharczyk J, Mintorovitch J, Cohen Y, Kurhanewicz J, Derugin N, Asgari H, Norman D (1990b) Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. AJNR Am J Neuroradiol 11:423–429 72. Moseley M, Bammer R, Illes J (2002) Diffusion-tensor imaging of cognitive performance. Brain Cogn 50:396–413 73. Moteki T, Horikoshi H, Oya N, Aoki J, Endo K (2002) Evaluation of hepatic lesions and hepatic parenchyma using diffusion-weighted reordered turboFLASH magnetic resonance images. J Magn Reson Imaging 15:564–572 74. Neil J, Miller J, Mukherjee P, Huppi PS (2002) Diffusion tensor imaging of normal and injured developing human brain – a technical review. NMR Biomed 15:543–552 75. Norris DG (2001) Implications of bulk motion for diffusion-weighted imaging experiments: effects, mechanisms, and solutions. J Magn Reson Imaging 13:486–495 76. Norris DG, Bornert P, Reese T, Leibfritz D (1992) On the application of ultra-fast RARE experiments. Magn Reson Med 27:142–164 77. Okamoto K, Ito J, Ishikawa K, Sakai K, Tokiguchi S (2000) Diffusion-weighted echo planar MR imaging in differential diagnosis of brain tumors and tumor-like conditions. Eur Radiol 10:1342–1350 78. Ordidge RJ, Helpern JA, Qing ZX, Knight RA, Nagesh V (1994) Correction of motional artifacts in diffusionweighted MR images using navigator echoes. Magn Reson Imaging 12:455–460 79. Papadakis NG, Murrills CD, Hall LD, Huang CL, Adrian Carpenter T (2000) Minimal gradient encoding for robust estimation of diffusion anisotropy. Magn Reson Imaging 18:671–679 80. Papadakis NG, Smponias T, Berwick J, Mayhew JE (2005) k-space correction of eddy current-induced distortions in diffusion-weighted echo planar imaging. Magn Reson Med 53:1103–1111 81. Pierpaoli C, Jezzard P, Basser PJ, Barnett A, Di Chiro G (1996) Diffusion tensor MR imaging of the human brain. Radiology 201:637–648

151

152

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 82. Pipe JG, Farthing VG, Forbes KP (2002) Multishot diffusion-weighted FSE using PROPELLER MRI. Magn Reson Med 47:42–52 83. Reese TG, Heid O, Weisskoff RM, Wedeen VJ (2003) Reduction of eddy current-induced distortion in diffusion MRI using a twice-refocused spin-echo. Magn Reson Med 49:177–182 84. Ries M, Jones RA, Dousset V, Moonen CT (2000) Diffusion tensor MRI of the spinal cord. Magn Reson Med 44:884–892 85. Rijswijk CS van, Kunz P, Hogendoorn PC, Taminiau AH, Doornbos J, Bloem JL (2002) Diffusion-weighted MRI in the characterization of soft-tissue tumors. J Magn Reson Imaging 15:302–307 86. Schick F (1997) SPLICE: sub-second diffusion-sensitive MR imaging using a modified fast spin-echo acquisition mode. Magn Reson Med 38:638–644 87. Sehy JV, Ackerman JJ, Neil JJ (2002) Evidence that both fast and slow water ADC components arise from intracellular space. Magn Reson Med 48:765–770 88. Seifert MH, Jakob PM, Jellus V, Haase A, Hillenbrand C (2000) High-resolution diffusion imaging using a radial turbo-spin-echo sequence: implementation, eddy current compensation, and self-navigation. J Magn Reson 144:243–254 89. Skare S, Hedehus M, Moseley ME, Li TQ (2000) Condition number as a measure of noise performance of diffusion tensor data acquisition schemes with MRI. J Magn Reson 147:340–352 90. Snook L, Paulson LA, Roy D, Phillips L, Beaulieu C (2005) Diffusion tensor imaging of neurodevelopment in children and young adults. Neuroimage 26:1164–1173 91. Stejskal EO, Tanner JE (1965) Spin diffusion measurements: spin-echoes in the presence of a time-dependent field gradient. J Chem Phys 42:288–292 92. Sugahara T, Korogi Y, Kochi M, Ikushima I, Shigematu Y, Hirai T, Okuda T, Liang L, Ge Y, Komohara Y, Ushio Y, Takahashi M (1999) Usefulness of diffusion-weighted MRI with echo planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 9:53–60 93. Sullivan EV, Pfefferbaum A (2003) Diffusion tensor imaging in normal aging and neuropsychiatric disorders. Eur J Radiol 45:244–255 94. Sullivan EV, Adalsteinsson E, Pfefferbaum A (2006) Selective age-related degradation of anterior callosal fiber bundles quantified in vivo with fiber tracking. Cereb Cortex 16:1030–1039 95. Taber KH, Pierpaoli C, Rose SE, Rugg-Gunn FJ, Chalk JB, Jones DK, Hurley RA (2002) The future for diffusion tensor imaging in neuropsychiatry. J Neuropsychiatry Clin Neurosci 14:1–5

96. Taouli B, Vilgrain V, Dumont E, Daire JL, Fan B, Menu Y (2003) Evaluation of liver diffusion isotropy and characterization of focal hepatic lesions with two single-shot echo planar MR imaging sequences: prospective study in 66 patients. Radiology 226:71–78 97. Taylor DG, Bushell MC (1985) The spatial mapping of translational diffusion coefficients by the NMR imaging technique. Phys Med Biol 30:345–349 98. Thomas DL, Pell GS, Lythgoe MF, Gadian DG, Ordidge RJ (1998) A quantitative method for fast diffusion imaging using magnetization-prepared TurboFLASH. Magn Reson Med 39:950–960 99. Tien RD, Felsberg GJ, Friedman H, Brown M, MacFall J (1994) MR imaging of high-grade cerebral gliomas: value of diffusion-weighted echoplanar pulse sequences. Am J Roentgenol 162:671–677 100. Tofts PS, Lloyd D, Clark CA, Barker GJ, Parker GJ, McConville P, Baldock C, Pope JM (2000) Test liquids for quantitative MRI measurements of self-diffusion coefficient in vivo. Magn Reson Med 43:368–374 101. Torrey HC (1956) Bloch equations with diffusion terms. Phys Rev 104:563–565 102. Trouard TP, Sabharwal Y, Altbach MI, Gmitro AF (1996) Analysis and comparison of motion-correction techniques in diffusion-weighted imaging. J Magn Reson Imaging 6:925–935 103. Tsuchiya K, Katase S, Fujikawa A, Hachiya J, Kanazawa H, Yodo K (2003) Diffusion-weighted MRI of the cervical spinal cord using a single-shot fast spin-echo technique: findings in normal subjects and in myelomalacia. Neuroradiology 45:90–94 104. Tuch DS (2004) Q-ball imaging. Magn Reson Med 52:1358–1372 105. Tuch DS, Reese TG, Wiegell MR, Makris N, Belliveau JW, Wedeen VJ (2002) High angular resolution diffusion imaging reveals intravoxel white matter fiber heterogeneity. Magn Reson Med 48:577–582 106. Wedeen VJ, Hagmann P, Tseng WY, Reese TG, Weisskoff RM (2005) Mapping complex tissue architecture with diffusion spectrum magnetic resonance imaging. Magn Reson Med 54:1377–1386 107. Woessner DE (1961) Effects of diffusion in nuclear magnetic resonance spin-echo experiments. J Chem Phys 34:2057–2061 108. Wong EC, Cox RW, Song AW (1995) Optimized isotropic diffusion weighting. Magn Reson Med 34:139–143 109. Xing D, Papadakis NG, Huang CL, Lee VM, Carpenter TA, Hall LD (1997) Optimised diffusion-weighting for measurement of apparent diffusion coefficient (ADC) in human brain. Magn Reson Imaging 15:771–784

2.9  Risks and Safety Issues Related to MR Examinations

2.9  Risks and Safety Issues Related to MR Examinations G. Brix With the rapid development of MR technology and the significant level of growth in the number of patients examined with this versatile imaging modality, the consid­ eration of possi­ble risks and health effects associated with the use of MR procedures is gaining increas­ingly in impor­tance. As described in detail in the previous chapters, three types of fields are employed: • A high static magnetic field generating a macro­scopic nuclear magnetization • Rapidly alternating magnetic gradient fields for spatial encoding of the MR signal • Radiofrequency (RF) electromagnetic fields for excitation and preparation of the spin system In the following, the biophysical interaction mechanisms and biological effects of these fields are summarized as well as exposure limits and precau­tions to be taken to minimize health hazards and risks to patients and volunteers under­going MR proce­dures. In the recent past, a number of excellent reviews and books related to this topic have been published. For details and supplementary information, the reader is referred to these publications quoted in the following and to the bibliographies given therein. Because no ionizing radiation is used in MRI, it is generally deemed safer than diagnos­tic X-ray or nuclear medicine procedures in terms of health protection of patients. In this context, a fun­da­mental difference between ionizing and non-ionizing radiation has to be noted: radia­tion exposure to ionizing radiation—at least at the relatively low doses occurring in medical imaging—results in stochastic effects, whereas biological effects of (electro)magnetic fields are deterministic. A stochastic process is one where the expo­sure determines the probability of the occurrence of an event but not the magnitude of the effect. In contrast, deterministic effects are those for which the magnitude is related to the level of expo­ sure and a threshold may be defined (International Commission on Non-Ionizing Radia­tion Protection [ICNIRP] 2002). As a consequence, the probability of detri­men­tal effects caused by diagnostic X-ray and nuclear medicine examinations performed over many years accumulate, whereas physio­logical stress induced by MR procedures is related to the acute exposure levels of a particular exami­nation and does, to the present knowledge, not accumulate over years. 2.9.1  Safety Regulations and Operating Modes In the recent past, regulations concerning MR safety have been largely harmonized. There are two comprehensive reviews by international commissions that form the basis

for both national safety standards and the implementation of monitor systems by the manufactur­ers of MR devices: 1 The technical product standard IEC 60601-2-33 issued by the International Electrotechnical Commission (IEC) in 2002 (IEC 2002) 2 The safety recommendation issued by the ICNRIP in 2004 (ICNRIP 2004) In order to reflect the existing uncertainty about deleterious effects of (electro)magnetic fields and to offer the necessary flexibility for the development and clinical evaluation of new MR technolo­gies, both safety regulations give exposure limits for three different modes of operation: 1 Normal operating mode. Routine MR examinations that did not cause physiological stress to patients. 2 Controlled operating mode. Specific MR examinations outside the normal operating range where discomfort and/or physiological stress to some patients may occur. Therefore, a clinical decision must be taken to balance such effects against foreseen benefits and exposure must be carried out under medical super­ vision. 3 Experimental operating mode. Experimental MR procedures with exposure levels outside the controlled operating range. In view of the potential risks for patients and volunteers, special ethical approval and adequate medical supervision is required. 2.9.2  Static Magnetic Fields 2.9.2.1  Magnetic Properties of Matter Magnetism in matter originates from the magnetic moments related to the movement and spin states of the electrons in atoms and molecules. The net magnetic moment per unit vol­ume of the material considered is called magnetization Me. This quantity is related to the field strength H of the applied magnetic field by the magnetic susceptibility χ (cf. 2.2.8.1) Me = χ H.

(2.9.1)

The “true” magnetic field in matter is given by the magnetic flux density B (cf. Sect. 2.2.8.1): B = μ0(H + Me) = μ0(1 + χ) H,

(2.9.2)

with μ0 = 1.257 × 10–6 Vs/m the magnetic permeability in vacuum. Due to the covalent binding of atoms, electron shells in most molecules are completely filled and thus all electron spins are paired. Nevertheless, these diamagnetic materials can be weakly magnetized in an external magnetic field. As described in Sect. 2.2.8.1, this universal effect

153

154

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

is caused by changes in the orbital motion of electrons in an external magnetic field. The induced magnetization is very small and in a direction opposite to that of the applied field (χ < 0). Paramagnetic materials, on the other hand, contain molecules with single, unpaired electrons. The intrinsic magnetic moments related with these electrons tend—comparable to the much weaker nuclear magnetic moment (cf. Sect. 2.2.3)—to align in an external magnetic field. This effect increases the magnetic field in paramagnetic materi­als (0 < χ < 0,01). In ferromagnetic materials—such as iron, cobalt, or nickel—unpaired electron spins align spontaneously with each other in the absence of a magnetic field in a region called a domain. These materials are characterized by a large positive magnetic susceptibility (χ > 0,01). Biomolecules are in general diamagnetic and contain at most some paramagnetic centers. In almost all human tissues, the concentration of paramagnetic components is so low that they are characterized by susceptibilities differing by no more than 20% from that of water (χ = –9,05 · 10–6) (Schenck 2000, 2005). As a consequence, there is virtually no effect of the human body on an applied magnetic field (B ≅ μ0 H ). 2.9.2.2  Biophysical Interaction Mechanisms There are several established physical mechanisms through which static magnetic fields can in­teract with biological tissues and organisms. The most relevant mechanisms are discussed in the following. Magneto-mechanical interactions. Even in a uniform magnetic field, molecules or structurally ordered molecule assemblies with either a field-induced (diamagnetic) or permanent (paramagnetic) magnetic moment mmol experience a mechanical torque that tends to align their magnetic moment parallel (or antiparallel) to the external magnetic field and thus to minimize the poten­tial energy (Fig. 2.9.1a). Orientation effects, however, can only occur when molecules or molecule clusters have a nonspherical structure and/or when the magnetic properties are anisotropically distributed. Moreover, the alignment must result in an appreciable reduction of the potential energy (Emag ∝ – mmol B) of the molecules in the external field with respect to their thermal energy (Etherm ∝ kT). At higher temperatures, as for exam­ple in the human body,

the alignment of molecules with small magnetic moments is prevented by their thermal movement (Brownian movement). In a non-uniform magnetic field, as for example in the periph­ery of an MR system, paramagnetic and ferromagnetic materials, moreover, are attracted and thus can quickly become dangerous projectiles (Fig. 2.9.1b). Magneto-hydromechanical interactions. Static magnetic fields also exert forces (called Lorentz forces) on moving elec­tro­lytes (ionic charge carriers) giving rise to induced electric fields and currents. For an electrolyte with charge q, the Lorentz force, which acts perpendicular to the direc­tion of the magnetic field, B, and the velocity, v, of the electrolyte is given by F = q · (v × B).

(2.9.3)

Since electrolytes with a positive or negative charge moving, for example, through a cylindrical blood vessel orientated perpendicular to a magnetic field are accelerated into opposite directions, this mechanism gives rise to an electrical voltage across the vessel, which is commonly referred to as blood flow potential (Fig. 2.9.2). Moreover, the induced transversal velocity component also interacts with the magnetic field according to Eq. 2.9.3, which results in a Lorentz force that is directed antiparallel to the longi­tudinal velocity component. At very high magnetic field strengths, this secondary effect can reduce the flow velocity and the flow profile of blood in large vessels (Tenforde 2005). Theoretically modeling of magneto-hydromechanical interaction processes was performed by Tenforde (2005) based on the Navier-Stokes equation describing the flow of an electrically conductive fluid in the presence of a magnetic field using the finite element technique. Induced current densities in the region of the sinoatrial node are predicted to by greater than 100 mA/m2 at field levels of more than 5 T in an adult human. Moreover, magneto-hydro­mechanical interactions were predicted to reduce the volume flow rate of blood in the human aorta by a maximum of 1.3, 4.9, and 10.4% at field levels of 5, 10, and 15 T, respectively. Magnetic effects on chemical reactions. As shown by in vitro studies, several classes of organic chemical reactions can be influ­enced by static magnetic fields under Fig. 2.9.1  Magneto-mechanical effects. a Orientation of a molecule with a magnetic moment m in a uniform magnetic field. b Attraction of a paramagnetic or ferromag­netic object in a non-uniform magnetic field. The direction of the acting forces F is indicated by arrows

2.9  Risks and Safety Issues Related to MR Examinations

appropriate, non-physiological conditions (Grissom 1995; World Health Organization [WHO] 2006). An established effect consists in the modifi­cation of the kinetics of chemical reactions with radicals as intermediate products, brought about by splitting and modification of electron spin states in the magnetic field. An example is the conversion of ethanolamine to acetaldehyde by the bacterial enzyme ethanolamine ammonia lyase. Radical pair magnetic field effects are thus used as a tool for invitro studies of enzyme reactions (WHO 2006). 2.9.2.3  Biological Effects For individual macromolecules, the extent of orientation in strong magnetic fields is very small. For example, measurements on DNA in solution have been shown that a magnetic flux density of 12 T is required to produce orientation of about 1% of the molecules (Maret et al. 1975). In contrast, there are several examples of molecular aggregates that can be oriented to a large extend by static magnetic fields, such as outer segments of retinal rod cells, muscle fibers, and filamentous virus particles (ICNIRP 2003; WHO 2006). An example of an intact cell that can be oriented magnetically is the erythrocyte. It has been shown that both resting and flowing sickled erythrocytes align in fields of more than 0.35 T with their long axis perpen­dicu­lar to the magnetic flux lines (Brody et al. 1985; Murayama 1965). Highashi et al. (1993) reported that normal erythro­cytes could be oriented with their disk planes parallel to the magnetic field direction. This effect was detectable even at 1 T, and almost 100% of the cells were oriented when exposed to 4 T. On the other hand, calculations performed by Schenck (2005) yielded that all of these orientation effects observed in vitro are probably too small to affect the ori­entation of the equivalent structures in vivo. However, although biophysical models make it possible to roughly estimate the magnitude of static magnetic field effects, the reality is so complex that calculations can in principle not rule out physiological effects (Hore 2005). Based on the evidence at present, there is no strong likelihood of major physiologi­cal consequences arising from radical-pair magnetic field effects on enzymatic reactions. Reasons against are the efficacy of homeo-

static buffering and the fact that the contrived conditions needed to observe a magnetic field response in the labora­ tory are unlikely to occur under physiological conditions (Hore 2005). There have been only a few studies on the effects of static magnetic fields at the cellular level. They reveal that exposure to static magnetic fields alone has no or extremely small effects on cell growth, cell cycle distribution, and the frequency of genetic damage, regardless of the magnetic flux density. However, in combination with other external factors such as ionizing radiation or some chemicals, there is evidence to suggest that a static magnetic field modifies their effects (Miyakoshi 2005). With regard to possible effects on reproduction and development, no adverse effects of static magnetic fields have been consistently demonstrated; few good studies however have been carried out, especially to fields in excess of 1 T (ICNIRP 2003; Saunders 2005; WHO 2006). Several studies indicate that implantation as well as prenatal and postnatal development of the embryo and fetus is not affected by exposure for varying periods during gestation to magnetic fields of flux densities between 1 and 9.4 T (Konermann and Mönig 1986; Murakami et al. 1992; Okazaki et al. 2001; Sikov et al. 1979). On the other hand, Mevissen et al. (1994) reported that continuous exposure of rats to a 30-mT field slightly decreased the numbers of viable fetuses per litter. Electric flow potentials generated across the aorta and other major arteries by the flow of blood in a static magnetic field can routinely seen in the ECG of animals and humans, exposed to fields in excess of 100 mT. In humans, the largest potentials occur across the aorta after ventricular contraction and appear superimposed on the T-wave amplitude of the ECG. Different animal studies demonstrated effects of static magnetic fields on blood flow, arterial pressure, and other parameters of the cardiovascular system, often at fields with flux densities much less than 1 T (Saunders 2005). The results of these studies, however, have to be interpreted with caution because it is difficult to reach any firm conclusion from cardio­vascular responses observed in anaesthetized animals (Saunders 2005; WHO 2006). On the other hand, two recent studies on humans exposed to a maximum flux density of 8 T (Chakeres et al. 2003; Kangarlu et al. 1999) did not yield clinically relevant changes in the heart rate, respiratory Fig. 2.9.2  Magneto-hydrodynamic effect. Positively and negatively charged electrolytes moving with a velocity v through a blood vessel oriented perpendicular to a magnetic field are accelerated into opposite directions and thus induce an electric voltage UH across the vessel (blood flow potential). Cross-hatches indicate the direction of the magnetic field into the paper plane

155

156

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

rate, diastolic blood pressures, finger pulse oxygenation levels, and core body temperature The only physiol­ogic parameter that was found to be altered significantly by high-field exposure was a change in measured systolic blood pressure. This is consistent with a hemodynamic compensatory mechanism to counteract the drag on blood flow exerted by magneto-hydrodynamic forces as described in Sect. 2.9.2.2 (Chakeres and de Vocht 2005). Various behavioral studies yielded that the movement of laboratory rodents in static magnetic fields above 4 T may be unpleasant, inducing aversive responses and condi­ tioned avoidance (WHO 2006). Such effects are thought to be consistent with magneto-hydro­dynamic effects on the endolymph of the vestibular apparatus (WHO 2006). This is in line with reports that some volunteers and patients exposed in static magnetic fields with flux densities above 1.5 T experienced sensations of vertigo, nausea, and a metallic taste in the mouth (Chakeres et al. 2003; Kangarlu et al. 1999; Schenck 2000, 2005). More­over, some of them reported on magnetophosphenes occurring during rapid eye move­ment in a field of at least 2 T, which may be attribut­able to weak electric fields induced by movements of the eye, resulting in an excitation of structures in the retina (Reilly 1998; Schenck 2000). Two recent studies evaluated neurobehavioral effects among subjects exposed to static magnetic fields of 1.5 and 8 T, respectively, using a neurobehavioral test battery. Performance in an eye–hand coordination test and a near-visual contrast sensi­tivity task slightly declined at 1.5 T (de Vocht et al. 2003), whereas a small negative effect on short-term memory was noted at 8 T (Chakeres et al. 2003). Taking also into account the results of other neurobehavioral studies, it can be concluded that there is at present no evidence of any clinically relevant modifi­ cation in human cognitive function related to static magnetic field exposure (Chakeres and de Vocht 2005). There are only a few epidemiological studies available that were specifically designed to study health effects of static magnetic fields. The majority of these have been focused on cancer risks. In 2002, the International Agency for Research on Cancer (IARC) (2002) reviewed epidemiological studies focused on cancer risks. Generally, these studies have not pointed to higher risks, although the number of studies was small, the num­bers of cancer cases were limited, and the information on individual exposure levels was poor. Therefore, the available evidence from epidemiological studies is at present not suffi­cient to draw any conclusions about potential health effects of static magnetic field exposure (Feychting 2005; WHO 2006). Some epidemiological studies have investigated reproductive outcome for workers involved in alumi­num industry or in MRI. Kanal et al. (1999), for example, evaluated 1,421 pregnancies of women working at clinical MR facilities. Comparing these preg­nancies with those occurring in employees at other jobs, they did not find significant increased risks for spontaneous abortions, delivery before 39 weeks, reduced birth weight, and male gender

Table 2.9.1  Limits for the magnetic flux density B0 for volunteers and patients undergoing MR procedures (ICNIRP 2004; IEC 2002) Operating mode Normal

Controlled

Experimental

B0 ≤ 2 T

2 T < B0 ≤ 4 T

B0 > 4 T

of the offspring. However, no studies of high quality have been carried out of workers occupationally exposed to fields greater than 1 T. 2.9.2.4  Exposure Limits Although there are initial experiences concerning the examination of volunteers and patients in ultra-high MR systems with magnetic flux densities of up to 8 T, most clinical MR procedures have been performed so far at static magnetic fields below 3 T. As summa­rized in Sect. 2.9.2.3, the literature does not indicate any serious adverse health effects from the expo­sure of healthy human subjects up to a flux density of 8 T (ICNIRP 2004). However, because movements in static magnetic fields above 2 T can produce nausea and vertigo, both the IEC standard and the ICNIRP recommendation (Table 2.9.1) regulate that MR examinations above this static magnetic flux density should be performed in the con­trolled operat­ ing mode under medical supervision. The recom­mended upper limit for the operating mode is 4 T, due to the limited information concerning possible effects above this magnetic flux density. For MR examina­tions performed in the experimental oper­ating mode, there is no upper limit for the magnetic flux density. In a safety document issued in 2003, the US Food and Drug Administration (FDA) (2003) deemed MR devices significant risk only when a static magnetic field of more than 8 T is used. 2.9.3  Time-Varying Magnetic Gradient Fields 2.9.3.1  Electric Fields and Currents Induced by Time-Varying Magnetic Fields According to Faraday’s law, a time-varying magne­tic field B(t) induces an electric field E(t), which has two important character­istics: the field strength is proportional to the time rate of change of the magnetic flux density, dB(t)/dt, and the field lines form closed loops around the direction of the mag­netic field. Time-varying magnetic fields are used in MRI— among others—to spatially encode MR signals arising from the different volume elements within the human body. To this end, three independent gradient coils are

2.9  Risks and Safety Issues Related to MR Examinations Fig. 2.9.3  Schematic representation of the electric field induced by time-varying magnetic fields B(t) that are directed par­ allel (left) and perpendicular (right) to the long axis of the human body. The electric field lines form closed loops around the direction of the mag­netic field

used to produce magnetic fields directed parallel to the static magnetic field B0 = (0, 0, B0) with a field strength varying in a linear manner along the x-, y- and z-direction as shown in Fig. 2.3.1. For the special case of a spatially uniform magnetic field directed in the z-direction, Bz(t), the electric field strength along a circular (conductive) loop of radius r in the x–y-plane is given by

(2.9.4)

This equation reveals that the electric field strength in the considered circular loop increases linearly with its radius as well as with the rate of change, dB(t)/dt. This model gives, for example, the electric field induced by the mag­ netic gradi­ent field B = (0, 0, G z · z) of the z-gradient coil. In contrast, the distribution of the electric fields induced by the time-varying magnetic gradient fields B = (0, 0, G x · x) and B = (0, 0, G y · y) is much more complex, since the magnetic flux density of these fields is not uniform over the x–y-plane. Moreover, the generation of these gradient fields is inevitable connected – due to fundamental prin­ciples of electrodynam­ics – with the occurrence of magnetic fields directed in the x- and y-direction, i.e., B = (Bx, 0, 0) and B = (0, By, 0), respec­tively. Although these “Maxwell terms” are of no rele­vance for the acquisition of MR images, they have to be considered carefully with respect to biologi­cal effects. The distribution of electric fields induced by timevarying magnetic fields directed par­allel and perpendicular to the long axis of the human body is schematically shown in Fig. 2.9.3. The precise spatial and temporal

distri­bution of the electric fields in the human body, of course, strongly depends on both the techni­cal characteristics of the gra­dient coils implemented at a specific MR system and the morphology of the body region exposed, and thus cannot be described by a simple mathematical expression. For worst-case estimations, however, it can be assumed that the electric field induced by a non-uniform mag­netic field is equal or smaller than the electric field produced by a uniform magnetic field with field strength equal to the maximum magnetic flux density of the nonuni­form field (Schmitt et al. 1998). For a uniform magnetic field, the electric field strength reaches, in general, a maximum when the magnetic field is orientated perpendicular to the coro­nal plane of the body (see Fig. 2.9.3, right) since the extension of conductive loops is largest in this direction (Reilly 1998). In conductive media, such as biological tissues, the internally induced electric field E(t), results in circulating eddy currents, j(t). Both quantities are related by the elec­ tric conductivity of the medium, σ, j(t) = σ E(t).

(2.9.5)

Calculation of the current distribution in the human body is complicated due to widely differing conductivities of various tissue components. For rough estimations, however, the body can be treated as a homogeneous medium with an average conductivity of σ = 0.2 S/m (Reilly 1998). According to Eqs. 2.9.4 and 2.9.5, for example, a current density of 20 mA/m2 is induced at a radius of 20 cm by a rate of change in the magnetic flux density of dBz/dt = 1 T/s.

157

158

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

2.9.3.2  Biophysical Interaction Mechanisms The magnetic flux density of gradient fields used in MRI is about two orders of magni­tude lower than that of the static magnetic field B0. Therefore, time-varying magnetic fields produced by gradient coils in MRI can be neglected compared to the strong static mag­netic field as far as interactions of magnetic fields with biological tissues and organisms are concerned (cf. Sect. 2.9.2.2). In contrast, however, biophysical effects related to the electric fields and currents induced by the magnetic fields have to be considered carefully. In general, rise times of magnetic gradients in MRI are longer than 100 µs, resulting in time-varying electric fields and currents with frequencies below 100 kHz. In this frequency range, the conductivity of cell membranes is several orders of magnitude lower than that of the extra- and intracellular fluid (Foster and Schwan 1995). As illustrated in Fig. 2.9.4, this has two important consequences. First, the cell membrane tends to shield the interior of cells very effectively from current flow, which is thus restricted to the extracellular fluid. Second, voltages are induced across the membrane of cells. When the electric voltages are above a tissue-specific threshold level, they can stimulate nerve and muscle cells (Foster and Schwan 1995). Theoretical models describing cardiac and peripheral nerve stimu­lation have been presented by various scientists (i.e., by Irnich, Mansfield, and Reilly). A detailed discus­sion of the underlying assumptions of the different models and the differences between them can be found, among others, in (Schaefer et al. 2000; Schmitt et al. 1998). The best approximation to experimental data is given by a hyper­bolic strength-duration expression

(2.9.6)

which relates the stimulation threshold, expressed as rate of change dB/dt of the magnetic flux density, with the

stimulus duration t, i.e., the ramp time of the magnetic gradient field (Schaefer et al. 2000; Schmitt et al. 1998). A hyperbolic model comparable to Eq. 2.9.6 was first established by G. Weiss in 1901 for an electric current pulse and the corresponding electric charge. This “fundamental law of electrostimulation” has been confirmed meanwhile in numerous studies for neu­ral and cardiac excitation as well as for defibrillation (Schaefer et al. 2000). As shown in Fig. 2.9.5, the threshold for the strength of a stimulus decreases with its duration. The asymptotic stimulus strength, B•∞, for an infinite duration is denoted as “rheobase”; the characteristic response time constant, τc , as “chronaxie”. It should be mentioned that ac­cording to a model presented by Irnich et al. stimulation depends on mean (rather than peak) dB/dt changes and is thus independent on the special shape of the gradient pulse (Schaefer et al. 2000). In current safety regulations, however, exposure limits are unanimously expressed as maximum dB/dt values. 2.9.3.3  Biological Effects In accordance with the biophysical mechanisms described in the previous section, there is now a strong body of evidence suggesting that the transduction processes through which induced electric fields and currents can influence cellular properties involve interactions at the level of the cell membrane (ICNIRP 2003). In addition to the stimulation of electrically excitable tissues, changes in membrane pro­perties—such as ion-binding to membrane macromolecules, ion transport across the membrane, or ligand–receptor inter­actions—may trigger transmembrane signaling events. Cardiac and peripheral nerve stimulation. Experimental studies with magnetic stimulation of the heart have been carried out since about 1991, with the introduction of improved gradient hardware for EPI. Experiments

Fig. 2.9.4  In the frequency range below 100 kHz, the conductivity of cell membranes (σm) is several orders of magnitude lower than that of the extra- and intracellular fluid (σext and σint, respectively) so that the induced electric fields (and also the resulting electric currents) are mainly restricted to the extracellular fluid (Eext > Eint). As a result, electric voltages are generated across the membrane of cells that can stimulate nerve and muscle cells

2.9  Risks and Safety Issues Related to MR Examinations Fig. 2.9.5  Hyper­bolic strength-duration expression that relates the stimulation threshold, expressed as rate of change dB/dt of the magnetic flux density, with the stimulus dura­tion t, i.e., the ramp time of the magnetic gradient field. The asymptotic • stimulus strength, B∞ , for an infinite duration is denoted as rheobase; the characteristic response time constant, τc , as chronaxie

were not performed of course with humans, but rather with dogs. The data, which are listed and reviewed by Reilly (1998), reveal that magnetic stimulation is most effective, when it is deliv­ered during the T wave of the cardiac cycle. Moreover, excitation thresholds for the heart are substantially greater than that for nerve as long as the pulse duration is suffi­ciently less than the chronaxie time of the heart of about 3 ms. There­fore, the avoidance of peripheral sensations in a patient provides a conservative safety margin with respect to cardiac stimulation. Bourland et al. (1991) determined a mean value of 14.1 ± 6.7 for the ratio of cardiac (the induction of ectopic beats) to muscle stimula­tion in dogs for a pulse duration of 530 µs, which is quite close to the theo­retical heart/nerve ratio of 14.0 estimated by Reilly (1998). Various stud­ies yielded that the cardiac threshold variability of healthy persons is surprisingly low, which is confirmed by experimental and clinical experience that pacing thresholds are rather uniform (Schmitt et al. 1998). Drugs and changes in electrolyte concentrations can lower thresholds, but not below about 80% of the normal value (Schmitt et al. 1998). Peripheral nerve stimulation has been investigated in various volunteer studies. A systematic evaluation of the available data was presented by Schaefer et al. in 2000. They recalculated published threshold levels—often reported for different gradient coils and shapes in differ­ent terms—to the maximum dB/dt occur­ring during the maximum switch rate of the gradient coil at a radius of 20 cm from the central axis of the MR system, i.e., at the border of the volume normally acces­sible to patients. In Fig. 2.9.6, the recalculated threshold levels are plotted for the y- (ante­rior/posterior) and z-gradient coils (superior/inferior) as compared to model estimates by Reilly. As expected, y-gradient coils have lower stimulation threshold for a given ramp time than x-gradient

coils since the x–z cross-sections of the body are usually larger than are x–y cross-sections. By fitting the hyperbolic strength-duration relationship defined in Eq. 2.9.6 to mean peripheral nerve stimulation thresholds measured by Bourland et al. (1999) in 84 human subjects, Schaefer et al. estimated the following values for the rheobase/chro­naxie: 22.1 T/s/0.365 ms for the y-gradient and 31.7 T/s /0.378 ms for the z-gradi­ent. As shown in Fig. 2.9.6, the dB/dt intensity to induce a sensation that the subject described as uncomfortable or even painful was significantly above the sensation threshold. Bourland et al. (1999) also analyzed their stimulation data in the form of cumulative frequency distributions, that gives for a dB/dt level the number of subjects that had already reported on perceptible, uncomfortable, or even intolerable sensations. They found that the dB/dt level needed for the lowest percentile for uncomfortable stimulation is approximately equal to the median threshold for perception. The lowest percentile for intolerable stimulation occurs at a dB/dt level approximately 20% above the median perception threshold. Time-varying magnetic fields can also result in the perception of magnetophosphenes due to the induction of electrical currents, presumably in the retina (cf. Sect. 2.9.2.3). A unique feature of phosphenes, which are not considered to be hazardous to humans, is their low excitation threshold and sharply defined excitation frequency of about 20 Hz as compared to other forms of neural stimulation (Reilly 1998). In general, a combination of magnetic gradient fields from all three gradient coils is used in MRI. In this case, the biological relevant time-varying magnetic field is approximately given by the vector sum of the magnetic field components. A detailed discussion of the effect of stimulus shape, number of stimuli, and other experimental set-

159

160

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

Fig. 2.9.6  Mean human nerve stimulation thresholds determined in different volunteer studies for y-gradient coils (anterior–posterior; left) and z-gradient coils (supe­rior–inferior;

right) as compared to model estimates by Reilly. For details see text. (From Schaefer et al. 2000; reprinted by permission of John Wiley & Sons, Ltd. All Rights Reserved)

tings on stimulation thresholds can be found in (Reilly 1998; Schmitt et al. 1998).

of amyotrophic lateral sclerosis (ALS) with occupational EMF exposure although confounding is a potential explanation. Whether there are asso­cia­tions with breast cancer and cardio­vascular disease remains unsolved.

Other biological endpoints. A comprehensive review of the current scientific evidence on biological effects of low-frequency electromagnetic fields in the frequency range up to 100 kHz has been pub­lished by ICNIRP in 2003. The majority of the reviewed studies focus on extremely low–frequency (ELF) magnetic fields associated with the use of electricity at power frequencies of 50 or 60 Hz. According to the ICNRIP review, cellular studies do not provide convincing evidence that low-frequency magnetic fields alter cell division, calcium homeostasis, and signaling pathways. Furthermore, no consistent effects were found in animals and humans with respect to genotoxicity, reproduction, development, immune system function, as well as endocrine and hematological parameters. On the other hand, a number of laboratory and field studies on humans demonstrated an effect of low-frequency magnetic fields at higher exposure levels on the power spectrum of different EEG frequency bands and on sleep structure. In the light of cognitive and performance studies yielding a number of biolo­gical effects, further studies are neces­sary to clarify the significance of the observed effects for human health. Over the last two decades, a large number of high quality epidemiological investigations of long-term disease endpoints such as cancer, cardiovascular and neurodegenerative disorders have been performed in relation to time-varying—mainly ELF—magnetic fields. Following the mentioned ICNIRP review (2003), the results can be summarized as follows. Among all the outcomes evalu­ated, childhood leukemia in relation to postnatal exposures to 50 or 60 Hz magnetic fields at flux densities above 0.4 µT is the one for which there is most evidence of an associa­tion. However, the results are difficult to interpret in the absence of evidence from cellular and animal studies. There is also evi­dence for an association

2.9.3.4  Exposure Limits From a safety standpoint, the primary concern with regard to rapid switching of magnetic gradients is cardiac fibrillation, because it is a life-threatening condition. In contrast, peri­pheral nerve stimulation is of practical concern because uncomfortable or intolerable stimulations would interfere with the examination (e.g., patient movements) or would even result in a termination of the examina­tion. In the current safety recommendations issued by IEC (2002) and and ICNRIP (2004), maximum dB/dt values for time-varying mag­netic fields created by gradient coils is limited for patient and volunteer examinations performed in the normal and the controlled operating mode by the dB/dt level of 80% and 100% of the mean perception threshold for peripheral nerve stimulation, respec­ tively. To this end, mean perception threshold levels have to be determined by the manufacturers for any given type of gradient system by means of experimental studies on human volunteers. As an alternative, the following empirical hyperbolic strength-duration expression for the mean threshold for peripheral nerve stimulation (expressed as maxi­mum change of the magnetic flux density in T/s) can be used:

(2.9.7)

In this equation, teff is the effective stimulation duration (in milliseconds), i.e., the duration of the period of monotonic increasing or decreasing gradient. A mathematical definition for arbitrary gradient shapes can be found in the IEC standard (2002).

2.9  Risks and Safety Issues Related to MR Examinations

2.9.4  Radiofrequency Electromagnetic Fields 2.9.4.1  Biophysical Interaction Mechanisms Time-varying magnetic fields used for the excitation and preparation of the spin system in MRI (B1 fields, cf. Sect. 2.2.4) have typically frequencies above 10 MHz. In this RF range, the conductivity of cell membranes is comparable to that of the extra- and intracellular fluid, which means that no substantial voltages are induced across the membranes (Foster and Schwan 1995). Due to this reason, stimulation of nerve and muscle cells is no longer a matter of concern. Instead, thermal effects due to tissue heating are of impor­tance. Energy dissipation of RF fields in tissues is described by the frequency-dependent con­ductivity σ(ω), which characterizes energy losses due to the induction and orienta­tion of electrical dipoles as well as the drift of free charge carriers in the induced time-varying electric field (Foster and Schwan 1995). The energy absorbed per unit of tissue mass and time, the so-called specific absorption rate (SAR, in W/kg), is approximately given by

SAR =

j⋅E σ ⋅ E2 = ρ ρ

(2.9.8)

where E is the induced electric field, j the corresponding current density, and ρ the tissue density (cf., 2.9.3.1). Absorption of energy in the human body strongly depends on the size and orientation of the body with respect to the RF field as well as on the frequency and polarization of the field. Theoretical and experimental considerations reveal that RF absorption in the body approaches a maximum when the wavelength of the field is in the order of the body size. Unfortunately, the wavelength of the RF fields used in MRI falls into this “reso­ nance range.” In order to discuss the effect of various measurement parameters on RF absorption, let us consider a simple MR sequence with only one RF pulse—such as a 2D or 3D FLASH sequence. In this case, the time-averaged SAR can approximately be described by the expres­sion

SAR ∝ B 02 ⋅ a 2 ⋅

tP ⋅ NS TR

(2.9.9)

According to this equation, the time-averaged SAR is proporti­onal • To the square of the static magnetic field, B0, which means that energy absorption is markedly higher at high-field as compared to low-field MR systems • To the square of the pulse angle, α, so that a sequence with a 90° or even a 180° pulse will result in a much higher SAR value than a sequence with a low-angle excitation pulse • To the duty cycle, tP  / TR, of the sequence, e.g., the ratio of the pulse duration tp and the repetition time TR of the pulse or sequence

• To the number of slices, NS, subsequently excited within the repetition time of a 2D sequence (multi-slice technique, cf. Sect. 2.3.5; NS = 1 for 3D sequences) In case of a more complex MRI sequence with various RF pulses, e.g., a spin-echo or a turbo spin-echo sequence, the contribution of the different RF pulses to patient exposure has to be summed up. The most relevant quantity for the characterization of physiological effects related to RF exposure is the temperature rise in the various body tissues, which is not only dependent on the localized SAR and the duration of exposure, but also on the thermal conductivity and microvascular blood flow (perfusion). In case of a partial-body RF exposure, the latter two factors lead to fast temperature equalization within the body (Adair 1996). Based on the bioheat equation, it can be shown (Brix et al. 2002) that for this particular exposure scenario the temperature response in the center of a homogenous tissue region, which is larger in each direction than the so-called thermal equilibration length, λ, is given by a convolu­tion of the exposure-time course, SAR(t), with a tissue-specific system func­tion, exp (–t/τ), (2.9.10) where τ is the thermal equilibration time, Ta the constant temperature of arterial blood, and c the specific heat capacity of the tissue. For representative tissues, equilibration lengths and times are between 1.5 and 12 mm and 0.2 and 25 min, respectively (Brix et al. 2002). Both parameters are inversely related to tissue perfusion and thus vary considerably. In case of a continuous RF exposure, the temperature rise even in poorly perfused tissues is less than 0.5°C for each W/kg of power dissipated. Using a simple model of power deposition in the head, Athey (1989) showed that continu­ous RF expo­sure over 1 h is unlikely to raise the temperature of the eye by more than 1.6°C when the average SAR to the head is less than 3.2 W/kg. More complex compu­tations were performed by Gandhi and Chen (1999) for a high-resolution model of the human body using the finite-difference time domain in order to assess SAR distribu­tions in the body for different RF coils. Their calcu­lations indicate that the maximum SAR averaged over 100 g of tissue can be ten times greater than the whole-body average SAR (“hot spots”). 2.9.4.2  Biological Effects Established biological effects of RF fields used for MR examinations are primarily caused by tissue heating. Therefore, it is important to critically evaluate the numerous number of studies focused on temperature effects, from the cellular and tissue level to the whole-body level, in-

161

162

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

cluding potential effects on vulnerable persons. In contrast, non-thermal (or athermal) effects are not well understood but seem—as far as this can be assessed at moment—to have no relevance with respect to the assessment of adverse effects associated with MR examinations. Non-thermal effects are those which can only be explained in terms of mechanisms other than increased random molecular motion (i.e., heating) or which occur at SAR levels so low that a thermal mechanism seems unlikely (ICNIRP 1997). As summarized in a review by Lepock (2003), relative short exposures of mammalian cells to temperatures in excess of 40–41°C result in a variety of effects, such as inhibition of cell growth, cytotoxic changes, alteration of signal transduction pathways, and an increased sensitivity to other stresses such as ionizing radiation and chemical agents. This suggests that damage is not localized to a single target, but that multiple heat-labile targets are damaged. Extensive protein denaturation has been observed at temperatures of 40–45°C for moderate periods. The most sensitive animal responses to heat loads are thermoregulatory adjustments, such as reduced metabolic heat production, vasodilatation, and increased heart rate. The corresponding SAR thresholds are between about 0.5 and 5 W/kg (ICNIRP 1998). The observed cardiovascular changes reflect normal thermoregulatory responses that facilitate the conduction of heat to the body surface in order to maintain normal body temperatures. Direct quantitative extrapolation of the animal (including primate) data to humans, however, is difficult given the marked species differences in the basal metabolism and thermoregu­latory ability (WHO 1993). At levels of RF exposure that cause body temperature rises of 1°C or more, a large number of additional, in most cases reversible, physiological effects have been observed in animals, such as alterations in neural and neuromuscular functions, increased blood-brain barrier permeability, stress-associated changes in the immune system, and hema­tological changes (ICNIRP 1998; Michaelson and Swicord 1996; WHO 1993). Thermal sensitivities and thresholds for irreversible tissue damage from hyperthermia have been summarized by Dewhirst et al. (2003). The most sensitive organs to acute damage are the testes and brain as well as portions of the eye (lens opacities and corneal abnormalities). The SAR thresh­old for irreversible effects even in the most sensitive tissues caused by RF exposure, however, is greater than 4 W/kg under normal environmental conditions (ICNIRP 1998). Effects of heat on embryo and fetus have been thoroughly reviewed by Edwards et al. (2003). Processes critical to embryonic development, such as cell proliferation, migration, differentiation, and apoptosis are adversely affected by elevated maternal temperatures. Therefore, hyperthermia of animals during pregnancy can cause embryonic death, abor­tion, growth retardation, and developmental defects. Especially the development of the

central nervous system is susceptible to heat. However, most animal data indicate that implantation and the development of the embryo and fetus are unlikely to be affected by RF exposures that increase maternal body temperature by less than 1°C (WHO 1993). In humans, epidemiological studies suggest that an elevation of maternal body tempera­ture by 2°C for at least 24 h during fever can cause a range of developmental defects, but there is little information on temperature thresholds for shorter exposures (Edwards et al. 2003). Humans possess comparatively effective heat loss mechanisms. In addition to a well-developed ability to sweat, the dynamic range of blood flow rates in the skin is much higher than it is in other species. Studies focused on RF-induced heating of patients during MR procedures have been summarized and evaluated in a review by Shellock (2000). They indicate that exposure of resting humans for 20–30 min to RF fields producing a whole-body SAR of up to 4 W/kg results in a body temperature increase between 0.1 and 0.6°C (WHO 1993). Of special interest is an extensive MR study reported by Shellock et al. (1994). In this study, thermal and physiologic responses of healthy volunteers undergoing an MR examination over 16 min at a whole-body averaged SAR of 6.0 W/kg were investigated in a cool (22°C) and a warm (33°C) environment. In both cases, signifi­cant variations of various physiologic parameters were observed, such as an increase in the heart rate, systolic blood pressure, or skin temperature. However, all variations were in a range that can be physiologically tolerated by an individual with normal thermo­regulatory function (Shellock et al. 1994). Generally, these studies are supported by mathematical modeling of human thermoregulatory responses to MR exposure (Adair 1996; Adair and Berglund 1986, 1989). It should be noted, however, that heat tolerance or thermoregu­lation may be compromised in some individuals undergoing an MR examination, such as the elderly, the very young and people with certain medical conditions (e.g., obesity, hyper­tension, impaired cardiovas­cular functions, diabetes, fever, etc.) and/or taking certain medications (e.g., beta-blockers, calcium channel blockers, sedatives, etc.) (Donaldson et al. 2003; Goldstein et al. 2003; ICNIRP 2004; Shellock 2000). Some regions of the human body, in particular the brain, are particularly vulnerable to raised temperatures. Mild-to-moderate hyperthermia (body temperature less than 40°C) induced thermal stress. For example, it affects cognitive performance (Sharma and Hoopes 2003) and can produce specific alterations in the CNS that may have long-term physiological and neuropathological consequences (Hancock and Vasmatzidis 2003). There have been a large number of epidemiological studies over several decades, par­ticularly on cancer, cardiovascular disease, and cata­ract, in relation to occupational, residential, and mobile-phone RF exposure. As

2.9  Risks and Safety Issues Related to MR Examinations

pairment, it is desirable to limit body core temperature increases to 0.5°C. As indicated in Table 2.9.2, these values have been laid down in the current safety recommendations (IEC, ICNIRP) to limit the body core tempera­ ture for the normal and controlled operating mode. Additionally, local tempera­tures under exposure to the head, trunk, and extremi­ties are limited for each of the two oper­ating modes to the values given in Table 2.9.2. However, temperature changes in the different parts of the body are difficult to control during an MR procedure in clinical routine. Therefore, SAR limits have been derived on the basis of experimental and theoretical studies, which should not be exceeded in order to limit the temperature rise to the values given in Table 2.9.2. As only parts of the body—at least in the case of adult patients—are exposed simultaneously during an MR procedure, not only the whole-body SAR but also partial-body

summa­rized in a review published by the ICNIRP Standing Committee on Epidemiology (Ahlbom et al. 2004), results of these studies give no consistent or convincing evidence of a causal relation between RF exposure and adverse health effect. It has to be noted, however, that the studies considered not only have too many deficiencies to rule out an association but also focus on chronic exposures at rela­tively low levels—an exposure scenario that is not comparable to MR examinations of patients. 2.9.4.3  Exposure Limits As reviewed in the previous section, no adverse health effects are expected if the RF-induced increase in body core temperature does not exceed 1°C. In case of infants, pregnant women, or persons with cardiocirculatory im-

Table 2.9.2  Basic restrictions for body temperature rise and partial-body temperatures for volunteers and patients undergoing MR procedures (ICNIRP 2004; IEC 2002) Operating mode

Rise of body core tempera­ture (°C)

 Spatially localized temperature limits Head (°C)

Trunk (°C)

Extremities (°C)

Normal

  0.5

38

39

40

Controlled

  1

38

39

40

Experimental

 > 1

> 38

> 39

> 40

Table 2.9.3  SAR limits for volunteers and patients undergoing MR procedures (ICNIRP 2004; IEC 2002), which holds at environmental temperatures below 24°C Body region →

Averaging time: 6 min

Operating mode

 Whole-body SAR (W/kg)

 Partial-body SAR (W/kg)

Whole-body

Any region, except head

↓

Local SAR  (averaged over 10 g tissue) (W/kg)

Head

a

Head

Trunk

Extremities

Normal

2

2–10b

3.2

10c

10

20

Controlled

4

b

4–10

3.2

c

10

10

20

Experimental

>4

> (4–10)b

> 3.2

10c

> 10

> 20

Short-term SAR

The SAR limit over any 10-s period shall not exceed 3 times the corresponding average SAR limit

Partial-volume SARs given by IEC; ICNIRP limits SAR exposure to the head to 3 W/kg Partial-body SARs scale dynamically with the ratio r between the patient mass exposed and the total patient mass: normal operating mode, SAR = (10–8 . r) W/kg; controlled operating mode, SAR = (10–6 . r) W/kg c  In cases where the eye is in the field of a small local coil used for RF transmission, care should be taken to ensure that the temperature rise is limited to 1°C

a 

b 

163

164

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy

SARs for the head, the trunk, and the extremities have to be estimated by means of suitable patient models (e.g., Brix et al. 2001) and limited to the values given in Table 2.9.3 for the normal and controlled operating mode. With respect to the application of the SAR levels defined in Table 2.9.3, the following points should be taken into account: • When a volume coil is used to excite a greater field-of view homogeneously, the partial-body and the wholebody SARs have to be controlled: in the case of a local RF transmit coil (e.g., a surface coils), the local and the whole-body SAR (IEC 2002). • Partial-body SARs scale dynamically with the ratio r between the patient mass exposed and the total patient mass. For r → 1 they converge against the correspond­ ing whole-body values, for r → 0 against the localized SAR level of 10 W/kg estab­lished by ICNIRP for occupational exposure of head and trunk (ICNIRP 1998). • The recommended SAR limits do not relate to an individual MR sequence, but rather to running SAR averages computed over each 6-min-period, which is assumed a typi­cal thermal equilibration time of smaller masses of tissue. But even if MR examinations are performed within the established SAR limits, severe burns can occur under unfavorable conditions at small focal skin-to-skin contact zones. The potential danger is illustrated in Fig 2.9.7 by the case of a patient who developed third-degree burns at the calves after conventional MR imaging. In this case, the contact between the calves resulted in the formation of a closed conducting loop and high current densities near the small contact zone. Therefore, patients should always be positioned in such a way that focal skin-to-skin

Fig. 2.9.7  Current-induced third-degree burns due to a small focal skin-to-skin contact between the calves during the MR examination. (From Knopp et al. 1998, with permission by Springer-Verlag)

contacts are avoided (e.g., by foam pads) (Knopp et al. 1998). 2.9.5  Special Safety Issues, Contraindications To protect volunteers, patients, accompanying persons, and uninformed healthcare workers from possible hazards and accidents associated with the MR environment, it is indispensable to perform a proper control of access to the MR environment. The greatest potential hazard comes from metallic, in parti­cular ferromagnetic materials (such as coins, pins, hair clips, pocketknives, scissors, nail clippers, etc.), that are acceler­ated in the inhomogeneous magnetic field (cf. Sect. 2.9.2.2) in the periph­ery of an MR system and quickly become dangerous projectiles (missile effect). This risk can only be minimized by a strict and careful screening of all individuals entering the MR environ­ment for metallic objects. Every patient or volunteer should complete a detailed questionnaire prior to the MR examination to ensure that every item posing a potential safety issue is considered. An example of such a form can be found, for example, in Shellock and Crues (2004), or can be downloaded from http://www.MRIsafety.com. Next, an oral interview should be conducted to verify the infor­mation of the form and to allow discussion of any question or concern that the patient may have before undergoing the MR procedure. An in-depth discussion of the various aspects of screening patients for MR procedures and individuals for the MR environ­ment can be found in various publications by Shellock (e.g., Shellock 2005; Shellock and Crues 2004) and the webpage mentioned above. Here only a condensed summary of the most important risks and con­ traindications can be given. All patients (and volunteers) undergoing MR procedures should—at the very least—be visually (e.g., by using a camera system) and/or acoustically (using an intercom system) monitored. Moreover, physiologic monitoring is indicated whenever a patient requires observation of vital functions due to a health problem or whenever the patient is unable to communicate with the MR technologist regarding pain, respiratory problems, cardiac stress, or other difficulty that might arise during the examination (Shellock 2001). This holds espe­cially in the case of sedated or anesthetized patients. For patient monitoring, special MR-compatible devices are available (Shellock 2001). Pregnant patients undergoing MR examinations are exposed to the static magnetic field, time-varying gradient fields and RF fields. The few studies concerning the combined effects of these fields on pregnancy outcome in humans follow­ing MR examinations have not revealed any adverse effects, but are very limited due to the small numbers of patients involved and difficulties in the interpretation of the results (Colletti 2001; ICNIRP 2004). It

2.9  Risks and Safety Issues Related to MR Examinations

is thus advised that MR procedures may be per­formed in pregnant patients, in particular in the first trimester, only after critical risk/benefit analysis and with verbal and written informed consent of the mother or parents (Colletti 2001). The standard of care is that MR imaging may be used in pregnant woman, if other non-ionizing forms of diagnostic imaging (e.g., sonography) are inade­quate or if the examination provides important information that would otherwise require exposure to ionizing radiation (e.g., fluoroscopy or CT) (Colletti 2001; Shellock and Crues 2004). In any case, however, exposure levels of the normal operating mode should not be exceeded and the duration of exposure should be reduced as far as possi­ble (ICNIRP 2003). MR examinations of patients with implants or metallic objects (such as bullets, pellets) are always associated with a serious risk, even if all procedures are performed within the established exposure limits summarized in the previous sections. This risk can only be minimized by a careful interview of the patient, evaluation of the patient’s file and contacting the implanting clinician and/or the manufacturer for advice on MR safety and compatibility of the implant (Medical Devices Agency 2002). In any case, MR procedures should be per­formed only after critical risk/benefit analysis. It should be noted that having undergone a previous MR procedure without incident does not guarantee a safe subsequent MR examination, since various factors (type of MR system, orientation of the patients, etc.) can substan­tially change the scenario (Shellock and Crues 2004). In the case of passive implants—e.g., vascular clips and clamps, intravascular stents and filters, vascular access ports and catheters, heart valve prostheses, orthopedic prostheses, sheets and screws, intrauterine contraceptive devices, etc.—it has to be clarified if they are made of or contain ferromagnetic materials. As already mentioned, strong forces act on ferromagnetic objects in a static magnetic field. These forces (ASTM 2005a) may result in a movement and dislodgment of ferromagnetic objects that could injure vessels, nerves or other critical tissue structures. Comprehensive information on the MR compatibility (ASTM 2005b) of more than 1,200 implants and other metallic objects is available in a reference manual published by Shellock (2005) and online at http://www. MRIsafety.com. MR examinations are deemed relatively safe for patients with implants or objects that have been shown to be non-ferromagnetic or weakly ferromagnetic (Shellock and Sawyer-Glover 2001). Furthermore, patients with certain implants that have relatively strong ferromagnetic qualities may safely undergo MR procedures when the object is held in place by sufficient retentive forces, is not located near vital structures, and will not heat excessively (Shellock and Sawyer-Glover 2001). However, such examinations should be restricted to essential cases and should be performed at MR systems with a low magnetic field strength.

Examinations of patients with active implants or lifesupport systems are strictly contraindicated at conventional MR systems, if the patient implant card does not explicitly state their safety in the MR environ­ment. In addition to the risks already mentioned above, there is the possibility that the function of the active implant is changed or perturbed, which may result in a health hazard for the patient. Clinically important examples are cardiac pacemakers, implantable cardioverter defibrillators, infusion pumps, programmable hydrocephalus shunts, neurostimulators, and cochlear implants, etc. (Medical Devices Agency 2002; Shellock and Sawyer-Glover 2001). The induction of electric currents by RF fields during imaging in implants made from conducting materials can result in excessive heating and thus may pose risks to patients. Excessive heating is typically associated with implants that have elongated configura­tions and/or are electronically activated, as for example the leads of cardiac pacemakers or neurostimulation systems (Shellock and Crues 2004). The same holds for electrically conductive objects (e.g., ECG leads, cables, wires, etc.), in particular when they form conductive loops in the bore of the MR system. To avoid severe burns, the instructions for proper operation of the equipment provided by the manufacturer of the implant or device have strictly to be followed. Practical recommendations concerning this issue can be found in (Shellock and Sawyer-Glover 2001). In various reports, transient skin irritations, cutaneous swellings or heating sensations were described in relation to the presence of both permanent (cosmetic) and decorative tattoos. These findings seem to be associated with the use of iron oxide or other metal-based pigments that are prone to magnetic field–related interactions and/or RF-induced heating, in particular when the pigments are organized as loops or rings. According to a survey performed by Tope and Shellock (2002), however, this side effect has an extremely low rate of occurrence in a population of subjects with tattoos and should not prevent patients—after informed consent—from under­going a clinically indicated MR procedures (Shellock and Crues 2004). As a precautionary measure, a cold compress may be applied to the tattoo site during the MR examination (Tope and Shellock 2002). References 1.

2.

Adair ER (1996) Thermoregulation in the presence of microwave fields. In: Polk C, Postow E (eds) Handbook of biological effects of electromagnetic fields. CRC, Boca Raton, pp 403–433 Adair ER, Berglund LG (1986) On the thermoregulatory consequences of NMR imaging. Magn Reson Imaging 4:321–333

165

166

2  Basics of Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy 3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Adair ER, Berglund LG (1989) Thermoregulatory consequences of cardiovascular impairment during NMR imaging in warm/humid environments. Magn Res Imaging 7:25–37 Ahlbom A, Green A, Kheifets L, Savitz D, Swerdlow A, ICNIRP Standing Committee on Epidemiology (2004) Epidemiology of health effects of radiofrequency exposure. Environ Health Perspect 112:1741–1754 ASTM International Standard F2052-02 (2005a) Standard test method for measurement of magnetically induced displacement force on medical devices in the magnetic resonance environment ASTM International Standard F2503-05 (2005b) Standard practice for marking medical devices and other items for safety in the magnetic resonance environment Athey TW (1989) A model of the temperature rise in the head due to magnetic resonance imaging procedures. Magn Reson Med 9:177–184 Bourland JD, Nyenhuis JA, Mouchawar GA, Geddes LA, Schaefer DJ, Riehl ME (1991) Z-gradient coil and eddycurrent stimulation of skeletal and cardiac muscle in the dog. Society for Magnetic Resonance in Medicine, Proc. 10th Annual Meeting, San Francisco Bourland JD, Nyenhuis JA, Schaefer DJ (1999) Physiologic effects of intense MR imaging gradient fields. Neuroimaging Clin N Am 9:363–377 Brix G, Reinl M, Brinker G (2001) Sampling and evaluation of specific absorption rates during patient examinations performed on 1.5-Tesla MR systems. Magn Reson Imaging 19:769–779 Brix G, Seebass M, Hellwig G, Griebel J (2002) Estimation of heat transfer and temperature rise in partial-body regions during MR procedures: an analytical approach with respect to safety considerations. Magn Reson Imaging 20:65–76 Brody AS, Sorette MP, Gooding CA et al (1985) Induced alignment of flowing sickle erythrocytes in a magnetic field: a preliminary report. Invest Radiol 20:560–566 Chakeres DW, de Vocht F (2005) Static magnetic field effects on human subjects related to magnetic resonance imaging systems. Prog Biophys Molec Biol 87:255–265 Chakeres DW, Kangarlu A, Boudoulas H, Young DC (2003) Effect of static magnetic field exposure of up to 8 Tesla on sequential human vital sign measurements. J Magn Reson Imaging 18:346–352 Colletti PM (2001) Magnetic resonance procedures and pregnancy. In: Shellock FG (ed) Magnetic resonance procedures: health effects and safety. Boca Raton: CRC, Boca Raton, pp 149–182 Dewhirst MW, Viglianti BL, Lora-Michiels M, Hanson M, Hoopes PJ (2003) Basic principles of thermal dosimetry and thermal thresholds for tissue damage from hyperthermia. Int J Hyperthermia 19:267–294 Donaldson GC, Keatinge WR, Saunders RD (2003) Cardiovascular responses to heat stress and their adverse consequences in healthy and vulnerable human polulations. Int J Hyperthermia 19:225–235

18. Edwards MJ, Saunders RD, Shiota K (2003) Effects of heat on embryos and foetuses. Int J Hyperthermia 19:295–324 19. Feychting M (2005) Health effects of static magnetic fields – a review of the epidemiological evidence. Prog Biophys Molec Biol 87:241–246 20. Foster KR, Schwan HP (1995).Dielectrical properties of tissues. In: Polk C, Postow E (eds) Handbook of biological effects of electromagnetic fields. CRC, Boca Raton, pp 25–102 21. Gandhi OP, Chen XB (1999) Specific absorption rates and induced current densities for an anatomy-based model of the human for exposure to time-varying magnetic fields of MRI. Magn Reson Med 41:816–823 22. Goldstein LS, Dewhirst MW, Repacholi M, Kheifets L (2003) Summary, conclusions and recommendations: adverse temperature levels in the human body. Int J Hyperthermia 19:373–384 23. Grissom CB (1995) Magnetic field effects in biology – a survey of possible mechanisms with emphasis on radicalpair recombination. Chemical Reviews 95:3–24 24. Hancock PA, Vasmatzidis I (2003) Effects of heat stress on cognitive performance: the current state of knowledge. Int J Hyperthermia 19:355–372 25. Higashi T, Yamagishi A, Takeuchi T et al (1993) Orientation of erythrocytes in a strong static magnetic field. Blood 82:1328–1334 26. Hore PJ (2005) Rapporteur’s report: sources and interactions mechanisms. Prog Biophys Molec Biol 87:205–212 27. International Agency for Research on Cancer (IARC) (2002) Static and extremely low frequency electric and magnetic fields. IARC Monographs on the evaluation of carcinogenic risks to humans, vol. 80 28. International Commission on Non-Ionizing Radiation Protection (ICNIRP) (1997) Non-thermal effects of rf electromagnetic fields. ICNIRP Report 3/97 29. ICNIRP (1998). Guidelines for limiting exposure to timevarying electrical, magnetic, and electromagnetic fields (up to 300 GHz). Health Physics 74:494–522 30. ICNIRP (2002) General approach to protection against non-ionizing radiation. Health Physics 74:494–522 31. ICNIRP (2003). Matthes R, Vecchia P, McKinlay AF, Veyret B, Bernhardt JH (eds) Review of the scientific evidence on dosimetry, biological effects, epidemiological observations, and health consequences concerning exposure to static and low frequency electromagnetic fields (0–100 kHz). Märkl, Munich 32. ICNIRP (2004) Medical magnetic resonance (MR) procedures: protection of patients. Health Physics 87:197–216 33. International Electrotechnical Commission (IEC) (2002) Particular requirements for the safety of magnetic resonance equipment for medical diagnosis. IEC 60601-2-33 (2nd edn.) 34. Kanal E, Evans JA, Savitz DA, Shellock FG (1993) Survey of reproductive health among female MR workers. Radiology 187:395–399

2.9  Risks and Safety Issues Related to MR Examinations 35. Kangarlu A, Burgess RE, Zhu H et al (1999) Cognitive, cardiac, and physiological safety studies in ultra high field magnetic resonance imaging. Magn Reson Imaging 17:1407–1416 36. Knopp MV, Metzner R, Brix G, van Kaick G (1998) Sicherheitsaspekte zur Vermeidung strominduzierter Hautverbrennungen in der MRT. Radiologe 38:759–763 37. Konermann G, Mönig H (1986) Untersuchungen über den Einfluß statischer Magnetfelder auf die pränatale Entwicklung der Maus. Radiologe 26:490–497 38. Lepock JR (2003) Cellular effects of hyperthermia: relevance to the minimum dose for thermal damage. Int J Hyperthermia 19:252–266 39. Maret G, von Schlickfus M, Mayer A, Dransfeld K (1975) Orientation of nucleic acids in high magnetic fields. Phys Rev Lett 35:397–400 40. Medical Devices Agency (2002) Guidelines for magnetic resonance equipment in clinical use. http://www.mhra.gov. uk 41. Mevissen M, Buntenkötter S, Löscher W (1994) Effects of static and time-varying (50 Hz) magnetic fields on reproduction and fetal development in rats. Teratology 50:229–237 42. Michaelson SM, Swicord ML (1996) Interaction of nonmodulated and pulse modulated radio frequency fields with living matter: experimental results. In: Polk C, Postow E (eds) Handbook of biological effects of electromagnetic fields. CRC, Boca Raton, pp 435–533 43. Miyakoshi J (2005) Effects of static magnetic fields at the cellular level. Prog Biophys Molec Biol 87:213–223 44. Murakami J, Torii Y and Masuda K (1992) Fetal development of mice following intrauterine exposure to a static magnetic field of 6.3 T. Magn Reson Imaging 10:433–437 45. Murayama M (1965) Orientation of sickled erythrocytes in a magnetic field. Nature 206:420–422 46. Okazaki R, Ootsuyama A, Uchida S, Norimura T (2001) Effects of a 4.7 static magnetic field on fetal development in ICR mice. J Radiat Res 42:273–283 47. Reilly JP (1998) Applied bioelectricity. From electrical stimulation to electropathology. Springer, Berlin, Heidelberg, New York 48. Saunders R (2005) Static magnetic fields: animal studies. Prog Biophys Molec Biol 87:225–239 49. Schaefer DJ, Bourland JD, Nyenhuis JA (2000) Review of patient safety in time-varying gradient fields. J Magn Reson Imaging 12:20–29 50. Schenck JF (2000) Safety of strong, static magnetic fields. J Magn Reson Imaging 12:2–19 51. Schenck JF (2005) Physical interactions of static magnetic fields with living tissues. Prog Biophys Molec Biol 87:185–204 52. Schmitt F, Irnich W, Fischer H (1998) Physiological side effects of fast gradient switching. In: Schmitt F, Stehling ML, Turner R (eds) Echo planar imaging. Springer, Berlin Heidelberg New York

53. Sharma HS, Hoopes PJ (2003) Hyperthermia-induced pathophysiology of the central nervous system. Int J Hyperthermia 19:325–354 54. Shellock FG (2000) Radiofrequency energy-induced heating during MR procedures: a review. J Magn Reson Med 12:30–36 55. Shellock FG (2001) Patient monitoring in the magntic resonance environment. In: Shellock FG (ed) Magnetic resonance procedures: health effects and safety. CRC, Boca Raton pp 217–240 56. Shellock FG (2005) Reference manual for magnetic resonance safety, implants, and devices: 2005 edn. Biomedical Researc, Los Angeles 57. Shellock FG, Crues JV (2004) MR procedures: biologic effects, safety, and patient care. Radiology 232:635–652 58. Shellock FG, Sawyer-Glover AM (2001) The magnetic resonance environment and implants, devices, and materials. In: Shellock FG (ed) Magnetic resonance procedures: health effects and safety. CRC, Boca Raton, pp 271–326 59. Shellock FG, Schaefer DJ, Kanal E (1994) Physiological responses to MR imaging at an SAR Level of 6.0 W/kg. Radiology 192:865–868 60. Sikov MR, Mahlum DD, Montgomery LD, Decker JR (1979) Development of mice after intrauterine exposure to directcurrent magnetic fields. In: Phillips RD, Gillis MF, Kaune WT, Mahlum DD (eds) Biological effects of extremely low frequency electromagnetic fields. 18th Hanford Life Sciences Symposium, Richland, Washington, October 1978. Springfield, Virginia, US Department of Energy, National Technical Information Service, pp 462–473, 1979 61. Tenforde TS (2005) Magnetically induced electric fields and currents in the circulatory system. Prog Biophys Molec Biol 87:279–288 62. Tope WD, Shellock FG (2002) Magnetic resonance imaging and permanent cosmetics (tattoos): survey of complications and adverse events. J Magn Reson Imging 15:180–184 63. US Food and Drug Administration (2003) Center for Devices and Radiological Health. Criteria for significant risk investigations of magnetic resonance diagnostic devices. http://www.fda.gov/cdrh/ode/guidance/793.pfd 64. Vocht F de, van-Wendel-de-Joode B, Engels H, Kromhout H (2003).Neurobehavioral effects among subjects exposed to high static and gradient magnetic fields from a 1.5 Tesla magnetic resonance imaging system: case-crossover pilot study. Magn Reson Med 50:670–674 65. World Health Organization (WHO) (1993) United Nations Environment Programme/Word Health Organisation/International Radiation Protection Association: environmental health criteria 137, electromagnetic fields (300Hz to 300 GHz). WHO Press, Brussells 66. WHO (2006) Environmental health criteria 232, static fields. WHO Press, Brussells

167

Chapter 3

Brain, Head, and Neck

3

Brain: Modern Techniques   and Anatomy  .. . . . . . . . . . . . . . . . . . . . . . . .   172 M. Wintermark, M.D. Wirt, P. Mukherjee, G. Zaharchuk, E. Barbier, and W.P. Dillon

3.2.4.7  Chiari Malformations  . . . . . . . . . . . . . . . . .   211

3.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   172

3.2.4.10  Rhombencephalosynapsis  . . . . . . . . . . . . .   214

3.1.2 

Diffusion-Weighted Imaging  .. . . . . . . . . .   173

3.2.4.11  Joubert Syndrome  .. . . . . . . . . . . . . . . . . . . .   215

3.1.3 

Diffusion Tensor Imaging  .. . . . . . . . . . . . .   175

3.2.4.12  Lhermitte-Duclos Syndrome  .. . . . . . . . . .   215

3.1.4 

Dynamic Susceptibility   Contrast Imaging  . . . . . . . . . . . . . . . . . . . . .   178

3.2.5 

3.1.5 

Arterial Spin Labeling  . . . . . . . . . . . . . . . . .   179

3.1.6 

Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . . .   183

3.1 

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   189 3.2 

Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics  .. . . . . . . . . . . . . . . . . . . . . . . . .   193 B.B. Ertl-Wagner and C. Rummeny

3.2.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   193

3.2.2 

Examination Technique  . . . . . . . . . . . . . . .   193

3.2.2.1  Patient Preparation  .. . . . . . . . . . . . . . . . . . .   193 3.2.2.2  Imaging Protocols  .. . . . . . . . . . . . . . . . . . . .   193 3.2.3 

Normal Development of the Brain  .. . . . .   194

3.2.3.1  Normal Intrauterine Development of the Brain  .. . . . . . . . . . . . . . . . . . . . . . . . . .   194 3.2.3.2  Normal Postnatal Myelination  . . . . . . . . .   195 3.2.4 

Congenital Disorders of the Brain  . . . . . .   197

3.2.4.8  Dandy-Walker Complex  .. . . . . . . . . . . . . .   213 3.2.4.9  Cerebellar Hypoplasia  .. . . . . . . . . . . . . . . .   214

Phakomatoses  . . . . . . . . . . . . . . . . . . . . . . . .   216

3.2.5.1  Neurofibromatosis Type 1  .. . . . . . . . . . . . .   216 3.2.5.2  Neurofibromatosis Type 2  .. . . . . . . . . . . . .   219 3.2.5.3  Tuberous Sclerosis  . . . . . . . . . . . . . . . . . . . .   219 3.2.5.4  von Hippel-Lindau Disease  . . . . . . . . . . . .   222 3.2.5.5  Sturge-Weber Syndrome  .. . . . . . . . . . . . . .   223 3.2.5.6  Other Phakomatoses  . . . . . . . . . . . . . . . . . .   225 3.2.6 

Hypoxic–Ischemic Injuries to the Pediatric Brain  .. . . . . . . . . . . . . . . . .   226

3.2.6.1  Hypoxic–Ischemic Injury in the Premature Infant  .. . . . . . . . . . . . . . .   228 3.2.6.2  Hypoxic–Ischemic Injury in the Term Infant  .. . . . . . . . . . . . . . . . . . . .   231 3.2.6.3  Kernicterus  .. . . . . . . . . . . . . . . . . . . . . . . . . .   231 3.2.7 

Metabolic Diseases   of the Pediatric Brain  .. . . . . . . . . . . . . . . . .   231

3.2.7.1  Metabolic Disorders Primarily   Involving the Deep White Matter  .. . . . . .   233

3.2.4.1  Disorders of Cortical Development  .. . . .   197

3.2.7.2  Metabolic Disorders Primarily   Involving the Subcortical White Matter    235

3.2.4.2  Agenesis or Hypogenesis of the Corpus Callosum  . . . . . . . . . . . . . . .   204

3.2.7.3  Metabolic Disorders Primarily   Involving the Gray Matter  . . . . . . . . . . . . .   236

3.2.4.3  Callosal Agenesis with Interhemispheric Cyst  . . . . . . . . . . . .   206

3.2.7.4  Metabolic Disorders Involving   Gray and White Matter  .. . . . . . . . . . . . . . .   238

3.2.4.4  Intracranial Lipomas  . . . . . . . . . . . . . . . . . .   207

3.2.7.5  Mitochondrial Disorders  . . . . . . . . . . . . . .   240

3.2.4.5  Holoprosencephaly  .. . . . . . . . . . . . . . . . . . .   208

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   241

3.2.4.6  Cephaloceles  . . . . . . . . . . . . . . . . . . . . . . . . .   210

170

3  Brain, Head, and Neck 3.3 

Intracranial Tumors   .. . . . . . . . . . . . . . . . .   243 M. Essig

3.4 

Cerebrovascular Disease   .. . . . . . . . . . . . .   310 D.C. Bergen, J.M. Fagnou, and R.J. Sevick

3.3.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   243

3.4.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   310

3.3.2 

The WHO Classification   of Brain Tumors  . . . . . . . . . . . . . . . . . . . . . .   243

3.4.2 

MR Technique  .. . . . . . . . . . . . . . . . . . . . . . .   310

3.3.3 

Practical Aspects of MR Imaging in Brain Tumors  . . . . . . . . . . . . . . . . . . . . . .   244

3.4.3 

Acute Ischemic Stroke  .. . . . . . . . . . . . . . . .   311

3.3.4 

Blood–Brain Barrier and Tumor Enhancement: Mechanisms and Applications  .. . . . . . . . .   245

3.3.4.1  Mechanisms of Contrast Enhancement in Brain Tumors  . . . . . . . . . . . . . . . . . . . . . .   245 3.3.4.2  Contrast Media Dosage  .. . . . . . . . . . . . . . .   246 3.3.4.3  Contrast Agents Used at Different Field Strength  . . . . . . . . . . . . .   247 3.3.5 

Intra-Axial Cerebral Tumors   . . . . . . . . . .   248

3.3.5.1  Neuroepithelial Tumors  . . . . . . . . . . . . . . .   248 3.3.5.2  Lymphoma  . . . . . . . . . . . . . . . . . . . . . . . . . . .   268 3.3.5.3  Cerebral Metastases  . . . . . . . . . . . . . . . . . . .   269

3.4.3.1  MR Findings in Acute Ischemic Stroke     311 3.4.3.2  Stroke Etiology  . . . . . . . . . . . . . . . . . . . . . . .   322 3.4.4 

Intracerebral Hemorrhage  . . . . . . . . . . . . .   327

3.4.4.1  Parenchymal Hematoma: Temporal Evolution of MR Findings  .. . .   327 3.4.4.2  Anatomical Location of Hemorrhage and Its Appearance on MR  .. . . . . . . . . . . .   332 3.4.5 

Intracerebral Hemorrhage Etiology  .. . . .   334

3.4.5.1  Vascular Malformations  . . . . . . . . . . . . . . .   334 3.4.5.2  Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   341 3.4.5.3  Saccular Aneurysms  .. . . . . . . . . . . . . . . . . .   341 3.4.5.4  Spontaneous Hemorrhages  .. . . . . . . . . . . .   342

Extra-Axial Cerebral Tumors  .. . . . . . . . . .   273

3.4.5.5  Hemorrhagic Transformation of Ischemic Stroke  .. . . . . . . . . . . . . . . . . . . .   343

3.3.6.1  Meningeal Tumors  . . . . . . . . . . . . . . . . . . . .   273

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   344

3.3.6 

3.3.6.2  Choroid Plexus Tumors  .. . . . . . . . . . . . . . .   280 3.3.6.3  Tumors of the Sellar Region  .. . . . . . . . . . .   281 3.3.6.4  Tumors of Uncertain Histogenesis  .. . . . .   285 3.3.6.5  Tumors of Peripheral Nerves That Affect the CNS  .. . . . . . . . . . . . . . . . . .   286 3.3.7 

Non-Tumorous Changes  .. . . . . . . . . . . . . .   288

3.3.7.1  Arachnoid Cysts   .. . . . . . . . . . . . . . . . . . . . .   288 3.3.7.2  Epidermoid Cysts  . . . . . . . . . . . . . . . . . . . . .   289 3.3.7.3  Dermoid Cysts  .. . . . . . . . . . . . . . . . . . . . . . .   289 3.3.7.4  Cerebral Lipomas  . . . . . . . . . . . . . . . . . . . . .   290 3.3.8 

Functional Imaging   in Intracranial Tumors  .. . . . . . . . . . . . . . . .   290

3.5 

Intracranial Infections  . . . . . . . . . . . . . . . .   348 E. Turgut Tali and Serap Gültekin

3.5.1 

Meningitis  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   348

3.5.1.1  Bacterial Meningitis  .. . . . . . . . . . . . . . . . . .   348 3.5.1.2  Viral Meningitis  . . . . . . . . . . . . . . . . . . . . . .   349 3.5.1.3  Meningitis Caused by Specific   Organisms (Tuberculous Meningitis)  . . .   351 3.5.1.4  Complications of Meningitis   .. . . . . . . . . .   353 3.5.2 

Empyema  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   355

3.5.2.1  Epidural Empyema  .. . . . . . . . . . . . . . . . . . .   355 3.5.2.2  Subdural Empyema  . . . . . . . . . . . . . . . . . . .   355 3.5.3 

Cerebritis and Abscess  .. . . . . . . . . . . . . . . .   357

3.3.8.1  MR Spectroscopy  . . . . . . . . . . . . . . . . . . . . .   293

3.5.3.1  Pyogenic Cerebritis and Abscess  .. . . . . . .   357

3.3.8.2  Contrast-Enhanced Perfusion MR Imaging in Brain Tumors   .. . . . . . . . .   293

3.5.3.2  Tuberculosis  .. . . . . . . . . . . . . . . . . . . . . . . . .   361

3.3.8.3  Dynamic Contrast-Enhanced   MR Imaging  .. . . . . . . . . . . . . . . . . . . . . . . . .   297 3.3.8.4  Diffusion-Weighted and Diffusion Tensor MRI  . . . . . . . . . . . . .   298 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   302

3.5.3.3  Fungal Infections  . . . . . . . . . . . . . . . . . . . . .   363 3.5.3.4  Parasitic Infections  .. . . . . . . . . . . . . . . . . . .   365 3.5.4 

Encephalitis  . . . . . . . . . . . . . . . . . . . . . . . . . .   373

3.5.4.1  Herpes Encephalitis  .. . . . . . . . . . . . . . . . . .   373 3.5.4.2  HIV Encephalitis  .. . . . . . . . . . . . . . . . . . . . .   376 3.5.4.3  Cytomegalovirus Encephalitis  .. . . . . . . . .   378 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   378

3  Brain, Head, and Neck 3.6 

Neurodegenerative Disorders  . . . . . . . . .   381 S. Karimi and A.I. Holodny

3.6.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   381

3.6.1.1  Normal Aging Brain  .. . . . . . . . . . . . . . . . . .   381

3.7.4.3  Pituitary Adenoma  .. . . . . . . . . . . . . . . . . . .   405 3.7.4.4  Microadenomas  .. . . . . . . . . . . . . . . . . . . . . .   405 3.7.4.5  Macroadenomas  . . . . . . . . . . . . . . . . . . . . . .   405

Dementia  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   381

3.7.4.6  Cavernous Sinus Invasion by Pituitary Adenoma  . . . . . . . . . . . . . . . . .   408

3.6.2.1  Alzheimer’s Disease  . . . . . . . . . . . . . . . . . . .   381

3.7.4.7  Postoperative Pituitary Adenoma  .. . . . . .   408

3.6.2.2  Pick’s Disease  .. . . . . . . . . . . . . . . . . . . . . . . .   383

3.7.4.8  Pituitary Adenocarcinoma  .. . . . . . . . . . . .   408

3.6.2.3  Multiple-Infarct Dementia, Leukoencephalopathy of Aging, and Binswanger’s Disease  .. . . . . . . . . . . . .   384

3.7.4.9  Pituitary Apoplexy  . . . . . . . . . . . . . . . . . . . .   408

3.6.2 

3.6.2.4  CADASIL  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   385 3.6.2.5  Creutzfeldt-Jakob Disease  .. . . . . . . . . . . . .   386 3.6.3 

Disorders with Prominent Motor Disability  . . . . . . . . . . . . . . . . . . . . . .   387

3.7.4.10  Craniopharyngioma  .. . . . . . . . . . . . . . . . . .   408 3.7.4.11  Pilomyxoid Astrocytoma  . . . . . . . . . . . . . .   411 3.7.4.12  Meningioma  .. . . . . . . . . . . . . . . . . . . . . . . . .   412 3.7.4.13  Rathke’s Cleft Cyst  . . . . . . . . . . . . . . . . . . . .   413 3.7.4.14  Arachnoid Cyst  .. . . . . . . . . . . . . . . . . . . . . .   414

3.6.3.1  Huntington’s Disease  . . . . . . . . . . . . . . . . . .   387

3.7.4.15  Metastasis  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   414

3.6.3.2  Parkinson’s Disease  .. . . . . . . . . . . . . . . . . . .   388

3.7.4.16  Germ Cell Tumors  . . . . . . . . . . . . . . . . . . . .   416

3.6.3.3  Multiple-System Atrophy  . . . . . . . . . . . . . .   388

3.7.4.17  Primary Sellar Lymphoma  . . . . . . . . . . . . .   417

3.6.3.4  Amyotrophic Lateral Sclerosis  .. . . . . . . . .   389

3.7.4.18  Granular Cell Tumor   of the Sellar and Suprasellar Regions  .. . .   417

3.6.4 

Hydrocephalus  .. . . . . . . . . . . . . . . . . . . . . . .   389

3.6.4.1  Normal Pressure Hydrocephalus  . . . . . . .   394 3.6.5 

Mesial Temporal Sclerosis  . . . . . . . . . . . . .   395 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   396

3.7.4.19  Hypothalamic Hamartoma  .. . . . . . . . . . . .   417 3.7.4.20  Langerhans Cell Histiocytosis  .. . . . . . . . .   418 3.7.4.21  Sarcoidosis  . . . . . . . . . . . . . . . . . . . . . . . . . . .   419 3.7.4.22  Lymphocytic Hypophysitis  .. . . . . . . . . . . .   420

3.7 

Pituitary Gland and Parasellar Region    399 M. Kanagaki, N. Sato, and Y. Miki

3.7.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   399

3.7.4.24  Hemochromatosis  .. . . . . . . . . . . . . . . . . . . .   422

3.7.2 

Examination Techniques  .. . . . . . . . . . . . . .   399

3.7.4.25  Pituitary Dwarfism  .. . . . . . . . . . . . . . . . . . .   422

3.7.2.1  Patient Positioning  .. . . . . . . . . . . . . . . . . . .   399

3.7.4.26  Cavernous Sinus Diseases  .. . . . . . . . . . . . .   424

3.7.2.2  Selection of Coils  . . . . . . . . . . . . . . . . . . . . .   399

3.7.4.27  Chordomas   . . . . . . . . . . . . . . . . . . . . . . . . . .   428

3.7.2.3  Examination Sequences  .. . . . . . . . . . . . . . .   399

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   429

3.7.2.4  Imaging Planes  . . . . . . . . . . . . . . . . . . . . . . .   399 3.7.2.5  Thickness of Slices  . . . . . . . . . . . . . . . . . . . .   399 3.7.2.6  Preferred Coverage  .. . . . . . . . . . . . . . . . . . .   399 3.7.3 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   400

3.7.3.1  Adenohypophysis and Neurohypophysis    400 3.7.3.2  Hypothalamic–Pituitary Axis  . . . . . . . . . .   401 3.7.3.3  Cavernous Sinus  . . . . . . . . . . . . . . . . . . . . . .   402 3.7.3.4  Liliequist’s Membrane  . . . . . . . . . . . . . . . . .   403 3.7.4 

Pathological Conditions  . . . . . . . . . . . . . . .   404

3.7.4.1  Empty Sella  .. . . . . . . . . . . . . . . . . . . . . . . . . .   404 3.7.4.2  Central Diabetes Insipidus  .. . . . . . . . . . . .   404

3.7.4.23  Abscess of the Pituitary Gland  .. . . . . . . . .   422

3.8 

The Orbits  . . . . . . . . . . . . . . . . . . . . . . . . . . .   433 N. Hosten, C. Zwicker, and M. Langer

3.8.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   433

3.8.2 

Examination Techniques  .. . . . . . . . . . . . . .   433

3.8.2.1  Patient Preparation   and Selection of Coils   .. . . . . . . . . . . . . . . .   433 3.8.2.2  Imaging Planes  . . . . . . . . . . . . . . . . . . . . . . .   433 3.8.2.3  Pulse Sequences  .. . . . . . . . . . . . . . . . . . . . . .   434 3.8.2.4  Use of Contrast Media  .. . . . . . . . . . . . . . . .   434 3.8.3 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   434

3.8.4 

Pathological Lesions  .. . . . . . . . . . . . . . . . . .   434

3.8.4.1  Intrabulbar Space-Occupying Lesions  .. .   434

171

172

3  Brain, Head, and Neck 3.8.4.2  Retrobulbar Space-Occupying Lesions     438

3.9.4.2  Traumatic Lesions  .. . . . . . . . . . . . . . . . . . . .   475

3.8.4.3  Non-Tumorous Retrobulbar Lesions   . . .   441

3.9.4.3  Inflammatory Lesions  . . . . . . . . . . . . . . . . .   475

3.8.5 

Differential Diagnosis  . . . . . . . . . . . . . . . . .   442

3.9.4.4  Benign Neoplasms  . . . . . . . . . . . . . . . . . . . .   476

3.8.6 

Diagnostic Procedures  .. . . . . . . . . . . . . . . .   442

3.9.4.5  Malignant Neoplasms  . . . . . . . . . . . . . . . . .   480

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   444

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   481

3.9 

3.9.1 

Magnetic Resonance of the Skull Base and Petrous Bone  .. . . . . . . . . . . . . . . . . . . .   445 R. Maroldi, D. Farina, A. Borghesi, E. Botturi, and C. Ambrosi Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   445

3.9.1.1  Examination Techniques  .. . . . . . . . . . . . . .   445 3.9.2 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   450

3.9.2.1  Key Anatomy of the Anterior Skull Base    450 3.9.2.2  Key Anatomy of the Central Skull Base     450 3.9.2.3  Key Anatomy of the Posterior Skull Base    452 3.9.2.4  Key Anatomy of the Temporal Bone  . . . .   452 3.9.3 

Lesions of the Skull Base  .. . . . . . . . . . . . . .   454

3.9.3.1  Congenital Lesions  .. . . . . . . . . . . . . . . . . . .   454 3.9.3.2  Inflammatory Lesions  . . . . . . . . . . . . . . . . .   456 3.9.3.3  Benign Neoplasms   .. . . . . . . . . . . . . . . . . . .   460 3.9.3.4  Malignant Neoplasms  . . . . . . . . . . . . . . . . .   467 3.9.4 

Lesions of the Temporal Bone   . . . . . . . . .   474

3.9.4.1  Congenital Lesions  .. . . . . . . . . . . . . . . . . . .   474

3.1  Brain: Modern Techniques and Anatomy M. Wintermark, M.D. Wirt, P. Mukherjee, G. Zaharchuk, E. Barbier, and W.P. Dillon 3.1.1  Introduction MRI in neuroradiology has evolved in the last 30 years, becoming faster, more precise, and more specific. The latest additions, including magnetic resonance spectroscopy (MRS), diffusion imaging, diffusion tensor imaging, functional MRI, and dynamic susceptibility contrast perfusion imaging, have expanded the applications for MR imaging. Currently, fluid attenuation inversion recovery (FLAIR) imaging, thin-section 3D volumetric imaging with spoiled gradient techniques, and the others mentioned above permit not only the precise localization of brain lesions, but also the evaluation of their metabolic profile, their location relative to eloquent regions of the

3.10 

Head and Neck  . . . . . . . . . . . . . . . . . . . . . . .   483 H.E. Stambuk and N.J. Fischbein

3.10.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   483

3.10.1.1  Examination Technique   .. . . . . . . . . . . . . .   483 3.10.2 

Mucosal Diseases of the Head and Neck    484

3.10.2.1  Paranasal Sinuses and Nasal Cavity  .. . . .   485 3.10.2.2  Nasopharynx  . . . . . . . . . . . . . . . . . . . . . . . . .   489 3.10.2.3  Oropharynx  . . . . . . . . . . . . . . . . . . . . . . . . . .   493 3.10.2.4  Oral Cavity  .. . . . . . . . . . . . . . . . . . . . . . . . . .   495 3.10.2.5  Hypopharynx  .. . . . . . . . . . . . . . . . . . . . . . . .   500 3.10.2.6  Larynx  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   502 3.10.2.7  Lymph Nodes  .. . . . . . . . . . . . . . . . . . . . . . . .   508 3.10.3 

Non-Mucosal Diseases   of the Head and Neck  . . . . . . . . . . . . . . . . .   509

3.10.3.1  Spatial Approach  .. . . . . . . . . . . . . . . . . . . . .   512 3.10.3.2  Thyroid and Parathyroid Glands  .. . . . . . .   526 3.10.3.3  The Pediatric Neck  . . . . . . . . . . . . . . . . . . . .   529 Suggested Reading  . . . . . . . . . . . . . . . . . . . .   533

cortex and subcortical white matter, and the relative blood volume and permeability of the vasculature that supplies the lesion. Thus, cellular, vascular, functional and anatomic information are obtained in one examination session and are available to treating physicians in their office, operating room, or radiation therapy suite. The use of MR spectroscopy is in many ways in its early development. The hope for MR imaging at 3 T and greater is that newer, more specific metabolites will increase the sensitivity and specificity for detection of disease states. MRS has proved a useful tool in the diagnosis of several metabolic brain diseases and enzymatic deficiency syndromes such as mitochondrial diseases, even in areas of brain that appear normal on conventional MR images. Diffusion is not only the most sensitive technique for detection of infarcted brain, but it has been helpful in the detection of normal, and abnormal, white matter tracts in a host of disease entities. In this section, we summarize the new techniques and the information they reveal about the brain.

3.1  Brain: Modern Techniques and Anatomy

3.1.2  Diffusion-Weighted Imaging Technical Description Diffusion-weighted imaging (DWI) is an MR imaging technique in which contrast between normal and abnormal structures is determined by differences in the degree of random Brownian motion of water molecules (Carr and Purcell 1954; Woessner 1961; Stejskal 1965; Stejskal and Tanner 1965). DWI relies on the Fick principle and the mass conservation principle, which combine to lead to the basic equation underlying DWI: T −  S =  S� × e T 

E

2

 −b × D  ×e ,  

where S is the observed MR signal, S0 the MR signal generated by the maximal value of the longitudinal component of the magnetic moment and TE the time of echo. The term in brackets, added to long repetition times used in echo planar imaging, is responsible for what some investigators have called the “T2 shine-through.” The latter is observed especially within cerebral ischemic lesions, because of their increased T2 values (Stejskal 1965; Stejskal and Tanner 1965). D is called diffusion coefficient, and characterizes the diffusion properties of a substance (Woessner 1961; Stejskal 1965; Stejskal and Tanner 1965). Typical D values are 4.5 × 10–3 (mm2/s) for 37 °C water and 2.9 × 10–3 (mm2/s) for cerebrospinal fluid (CSF). D appears to be relatively the same in gray matter and in white matter. However, unlike diffusion within gray matter, which is characterized by an invariant D of 0.8 × 10–3 (mm2/s), diffusion within white matter has a strong directional preponderance, referred to as diffusion anisotropy. White matter D ranges from 0.6 × 10–3 up to 1.1 × 10–3 (mm2/s) depending on whether it is measured parallel or perpendicular to fiber direction (Moseley et al. 1990a,b). b is the other determinant of the equations underlying DWI. It describes the intensity of diffusion weighting in obtained DWI images. Admitting a single water compartment due to fast exchange across cell membranes, and thus a monoexponential decay in voxels, two b values, in the 0 –1,500 (s/mm2) range, are enough to accurately calculate D values according the abovementioned equation (Woessner 1961; Stejskal 1965; Stejskal and Tanner 1965; Warach et al. 1992; Baird and Warach 1998; Beaulieu et al. 1999). Technical Requirements Most 1.5-T MR scanners are now equipped with the fastimaging capabilities required for DWI. Post-processing software available on the operator console for the processing of the DWI images and the calculation of ADC maps is recommended.

Imaging Protocol DWI Sequence type

SE-EPI

TR (ms)

≥ 3,000

TE (ms)

≤ 130

TI (ms)

N/A

Flip angle

90°

NA

1

Slice direction

Axial

FOV (cm2)

24 × 24

No. of slices

20

Total slab thickness (cm)

≥ 13

Resolution (readout × phase resolution)

Minimum: 64 × 64 Optimum: 128 × 128

Rect. FOV applicable

Yes

Slice thickness (mm)

6

Gap (mm)

0.4–0.6

Dosage (mmol/kg BW)

N/A

Injection rate (ml/s)

N/A

b Value (s/mm2)

0 and 1,000 (b = 1,000 in 3 orthogonal directions)

Presaturation slice

No

Total acquisition time

≤ 2 min

Interpretation Clinical practice uses different representations of the results of DWI data processing: diffusion-weighted images, DWI trace, and ADC maps, which are all equivalent. Diffusion coefficients are typically measured in one direction at a time, thanks to adequate bipolar gradient pulses. Because of anisotropy, data obtained from using a one-direction diffusion gradient only can result in misleading information and diagnostic pitfalls in the detection and delineation of cerebral ischemia. As a result, bipolar gradient pulses are usually applied along at least three orthogonal sampling directions, and the diffusion coefficients thus sampled allow for calculation of a DWI trace, which represents the average of diffusionweighted images in all three directions. On the DWI trace, the anisotropy and orientation of myelin fibers are completely removed. Such images show small contrast between white and gray matter and, as discussed later, the

173

174

3  Brain, Head, and Neck

diffusion coefficient of white matter is decreased in acute stroke to approximately the same extent as that of gray matter (Gonzalez et al. 1999; van Gelderen et al. 1994; Ulug et al. 1997). DWI trace signal intensity is not solely the result of diffusion characteristics but also of the T2 decay (Neil et al. 1998). So called T2 shine-through effect can be eliminated through computing isotropic or directionally encoded ADC. As described above, the ADC map can be calculated from at least two b values, typically 500 and 1,000 (s/mm2). The ADC map shows relative signal intensities due only to differences in tissue diffusion (van Gelderen et al. 1994; Neil et al. 1998). On ADC maps, tissues have an appearance opposite to that seen on the DWI trace maps, reflecting the negative exponent in the equation relating signal intensity to the ADC. Thus, on diffusion-weighted images, lesions with restricted water diffusion are hyperintense relative to normal tissue, while on ADC maps such abnormalities appear hypointense (van Gelderen et al. 1994). Information carried by DWI trace and ADC maps is complementary. The DWI trace affords sharp delineation of pathologic processes, through both the removal of anisotropy of myelin fibers and the absence of contrast between gray and white matter, while ADC maps remove the “T2 shine-through effect.” Feasibility A DWI sequence is typically performed in less than 1–2 minutes and is easily obtained in all patients stable enough to undergo an MRI examination. As it is the most sensitive technique for detection of acute ischemic infarction, DWI should typically be the first sequence performed in patients with suspected stroke.

Clinical Applications of Diffusion DWI provides information that is not available on standard T1- and T2-weighted MR images. It has the ability to show areas of brain ischemia within minutes after onset (Fig. 3.1.1), whereas CT or T2-weighted MR images become positive only several h, usually 5 or 6, after stroke onset (Provenzale et al. 1999; Mohr et al. 1995; Lutsep et al. 1997). In an animal model, sensitivity of diffusionweighted imaging in the detection of acute infarction has been reported as 60% within 50 min and 100% within 2 h after symptom onset Lansberg et al. 2000; Kumon et al. 1999). During the first 48 h after symptom onset, the addition of DWI to conventional MRI in acute stroke patients improves the accuracy of identifying acute ischemic brain lesions from 71 to 94% (Oppenhaim et al. 2000). There are, however, several reports documenting an absence of DWI abnormalities early in the course of an infarction (NINDS t-Pa Stroke Study Group 1997; Hacke et al. 1995). DWI is abnormal not only in acute stroke, but also in any condition in which neurons are acutely damaged or in which high levels of proteinaceous debris reduces the diffusability of water. Examples include vasculitis, herpes encephalitis (Kuker et al. 2004), Creutzfeldt-Jakob disease (Fig. 3.1.2), abscesses (Fig. 3.1.3), epidermoid cysts, and some brain tumors, such as lymphoma. In CreutzfeldtJakob disease, the caudate and putamen, as well as areas of the archicortex (insula, hippocampus, and cingulum) and parietal cortex, show reduced diffusion (Young et al. 2005). DWI also can help distinguish between brain abscess (reduced diffusion) and necrotic metastasis (normal or increased diffusion) (Dorenbeck et al. 2003). Similarly, the reduced diffusion typical of epidermoid cysts easily differentiates these lesions from arachnoid cysts, which Fig. 3.1.1  64-year-old male patient admitted with aphasia and right-body motor deficit. Admission MR angiography shows a proximal stenosis of the left middle cerebral artery (MCA) (arrow). Diffusion-weighted images (DWI) and apparent diffusion coefficient (ADC) maps feature a focus of restricted diffusion in the deep territory of MCA (arrowheads) consistent with acute stroke. Time-to-peak (TTP) and mean transit time (MTT) maps demonstrate an extensive alteration of brain hemodynamics. The DWI-PWI mismatch is classically considered as a hallmark of tissue at risk or penumbra. (Figure courtesy of Dr. Salvador Pedrazza, Girona, Spain

3.1  Brain: Modern Techniques and Anatomy Fig. 3.1.2  58-year-old male patient with Creutzfeldt-Jakob disease, showing a typical pattern of reduced diffusion in the caudates, anterior putamens, and in the bifrontal and right insular cortex

Fig. 3.1.3  47-year-old male patient with a rim-enhancing lesion in the left thalamus. The reduced diffusion within this rimenhancing lesion makes it typical for an abscess, and allows it to be distinguished from a necrotic tumor

show normal CSF diffusion characteristics (Hakyemez et al. 2005). 3.1.3  Diffusion Tensor Imaging Technical Description Diffusion tensor imaging (DTI) uses the anisotropy of water motion within structures such as axons to create image contrast, which is used then to depict the orientation and volume of white matter fiber tracts. This technique has been propelled by the fast imaging techniques of echo planar imaging, better gradients, and software post-processing techniques. Clinical applications include not only anatomic localization of subcortical tracts in the

brainstem and supratentorial areas, but also measurements of the developing brain and the effect of disease entities on white matter anatomy. Single-shot spin-echo, echo planar imaging (EPI) is the most widely used technique for acquiring DTI. The rapid imaging capability of EPI freezes bulk macroscopic motion, thereby permitting imaging of water diffusion at microscopic spatial scales. DTI requires a high signal-tonoise ratio (SNR) (preferably exceeding 20) for accurate assessment of diffusion anisotropy. DTI fiber tractography also requires high spatial resolution for detailed visualization of small white matter tracts, preferably 2.5 mm cubic voxels or smaller. The use of cubic voxels, which have the same length in all three orthogonal dimensions, is recommended for tractography in order to avoid bi-

175

176

3  Brain, Head, and Neck

asing the 3D tracking algorithm toward the direction of poorer spatial resolution. EPI can provide sufficient spatial resolution with adequate SNR for DTI tractography at 1.5 T in a clinically feasible acquisition time (Virta et al. 1999). Higher field magnets (3 T and above) enable DTI at higher spatial resolution and/or shorter acquisition times; however, geometric warping artifacts common to EPI may limit anatomic fidelity, especially in areas of high magnetic field susceptibility, due to brain–air–bone interfaces such as the skull base and the posterior fossa. These susceptibility artifacts can be problematic even at 1.5 T, and increase markedly at higher field strengths. Pulsation artifacts from cerebrospinal fluid also create artifacts, especially in the posterior fossa and in regions of the supratentorial brain bordering the lateral ventricles (Virta et al. 1999). Technical Requirements Perhaps the most important hardware consideration for performing DTI is the performance of the diffusion and the EPI readout gradients. Stronger and faster gradients enable better diffusion weighting in a shorter period, as well as reducing the time required to form an EPI image. This permits shorter echo times (TE), which improves SNR and reduces geometric warping artifacts. Hence, the latest generation of MR imagers, which contain 4 G/cm gradients, allows DTI with high spatial resolution and anatomic fidelity. New multi-channel head radiofrequency (RF) coils with better SNR characteristics than the standard birdcage head RF coils have also enhanced DTI. The multi-channel RF coils also enable parallel imaging, a technical advance that can improve the image quality of DTI (Bammer et al. 2002). Parallel imaging is instrumental for ameliorating the greater EPI susceptibility artifacts that occur at 3 T and above, thereby permitting high-field DTI with improved image quality. Other variables that may affect the quality of DTI include the b value (diffusion-weighting factor), and the number of directions in 3D space in which diffusion gradients are applied. A b value of 1,000 s/mm2 has become the standard for clinical diffusion-weighted imaging (DWI), and has also been employed for DTI in many studies. Six diffusion-sensitizing directions are the minimum needed to solve for the diffusion tensor, although to provide sufficient SNR at high enough spatial resolution on a 1.5-T scanner for DTI tractography, each six-direction, whole-brain acquisition must be averaged several times. At present, there is no consensus on the optimum number and geometry of diffusion-encoding directions for DTI. Some studies indicate that there is no advantage to acquiring more than the minimum six directions (Hasan et al. 2001), while others indicate that SNR per unit time improves with greater numbers of directions isotropically distributed in 3D space (Papadakis et al. 2000; Skare et al. 2000).

Imaging Protocol DTI Sequence type

SE-EPI

TR (ms)

> 3,000

TE (ms)

< 130

TI (ms)

N/A

Flip angle

90

NA

1

Slice direction

Axial

FOV (cm2)

28 × 28

No. of slices

60

Total slab thickness (cm)

13

Resolution (readout × phase resolution)

128 × 128

Rect. FOV applicable

Yes

Slice thickness (mm)

2.2

Gap (mm)

0

Dosage (mmol/kg BW)

N/A

Injection rate (ml/s)

N/A

b Value (s/mm2)

0 and 1,000 (b = 1,000 in at least 6 directions)

Presaturation slice

No

Total acquisition time

>10 min

Interpretation Post-processing and visualization of DTI data requires the generation of parametric maps, the most popular of which are diffusion anisotropy, direction-encoded color anisotropy, and the eigenvalues of the diffusion tensor. At present, calculation of DTI parametric maps and 3D tractography requires postprocessing on a dedicated image workstation, although vendors are increasingly incorporating online DTI visualization tools into their latest MR scanner software releases. The three eigenvalues of the diffusion tensor represent the magnitude of diffusion along the three principal directions in 3D space, which are mutually orthogonal. The eigenvalue with the maximum value (the “major eigenvalue”) is the magnitude of diffusion along the orientation in which water diffuses most freely, while the two other eigenvalues (the “minor eigenvalues”) represent the magnitude of diffusion along the directions orthogonal

3.1  Brain: Modern Techniques and Anatomy Fig. 3.1.4  Directionally-encoded color DTI axial image in a normal volunteer shows the fiber orientation of white matter tracts as red if they are predominantly oriented left–right, green if anteroposterior, and blue if superoinferior. The corpus callosum, cingulum bundle, centrum semiovale, and superior longitudinal fasciculus are labeled

Fig. 3.1.5  3D DTI fiber tractography overlaid on a midline sagittal directionally encoded color DTI image shows the short and long association fibers of the cingulum bundle better than does Fig. 3.1.4

to this preferred orientation. The mean of the three eigenvalues is equivalent to the ADC, and the variance of the three eigenvalues is related to the diffusion anisotropy. The projection of the major eigenvector on each of three orthogonal axes (left–right, anteroposterior, and craniocaudal) can be encoded by different colors. In the most widely accepted directional encoding scheme, the left–right direction is assigned red, the anteroposterior dimension is assigned green, and the craniocaudal direction is assigned blue (Fig. 3.1.4) (Pajevic and Peirpaoli 1999). This works well for differentiating large association tracts, which are usually green since they connect anterior and posterior cortical regions within a single cerebral hemisphere, from projection pathways, which are often blue since they connect superior cortical areas to inferior subcortical regions, and also from commissural

fibers, which appear red because of their left–right orientation across the two hemispheres. Since white matter pathways in the brain exist in three dimensions, even sophisticated 2D representations such as direction-encoded color anisotropy maps are intrinsically limited. Moreover, these color anisotropy maps cannot differentiate adjacent white matter tracts that usually have the same fiber orientation. These obstacles can be overcome with 3D fiber tractography. There are many different techniques for performing fiber tractography that have been described in the literature, but most of them are variations on the same underlying idea of tracking bidirectionally along the orientation of the primary eigenvector of the diffusion tensor from voxel to voxel in three dimensions (Fig. 3.1.5) (Basser et al. 2002).

177

178

3  Brain, Head, and Neck

Feasibility DTI has no age limitation and can be performed even in neonates, in whom the development of white matter tracts may be studied. The use of DTI in emergency settings such as acute stroke or head trauma is limited by the logistic difficulties of performing extended MR imaging protocols in these settings. Clinical Applications DTI has been used in the evaluation of traumatic brain injury (TBI). In a small series of patients with mild traumatic brain injury examined within 24 h of the event, significant reductions in anisotropy in brain regions adjacent to the injury when compared with the contralateral side, as well as when compared to normal controls, were observed (Arfanakis et al. 2002). DTI may be more sensitive for acute TBI than conventional MR imaging or even DWI with ADC mapping, although further investigation is needed to confirm this initial observation. DTI has been employed extensively in research on demyelinating and dysmyelinating white matter diseases. For instance, DTI has been frequently applied to the evaluation of patients with multiple sclerosis (MS), where elevated ADC and reduced anisotropy have been documented in normal appearing white matter (NAWM) on conventional MR imaging (Filippi et al. 2001). DTI with 3D tractography may aid in the presurgical evaluation of brain tumor patients by delineating functionally important white matter pathways such as the pyramidal tract so that they may be spared during the tumor resection (Holodny and Ollenschlager 2002). The use of DTI has also been reported for epilepsy (Rugg-Gunn et al. 2002), in neurodevelopmental (Mukherjee & McKinstry 2005), neurodegenerative, and neuropsychiatric diseases (O’Sullivan et al. 2001), and in stroke (Mukherjee 2005). 3.1.4  Dynamic Susceptibility Contrast Imaging Technical Description Dynamic susceptibility contrast imaging (DSC) relies on the measurement of the T2 or T2* decrease during the first pass of an exogenous endovascular tracer through the capillary bed (Rosen et al. 1990). The technique requires ultrafast imaging such as echo planar imaging (EPI), principles of echo-shifting with a train of observations (PRESTO) or spiral imaging. Gradient-echo (GRE) or spin-echo (SE) sequences can be used, but the signal change measured with GRE (∆T2*) is greater than that measured with SE (∆T2), allowing one to use a shorter echo time (TE) and a lesser amount of contrast agent. The sequence duration is about 1 minute and the sampling rate (repetition time [TR] of the sequence) should be kept below 2 s. This can be achieved with GRE sequences for whole-brain coverage (up to 24 slices) (Speck et al. 2000).

The tracer used is a conventional chelate of gadolinium, injected through an 18-ga catheter into a peripheral vein at a regular dose of 0.1 mmol/kg for GRE or 0.2 mmol/ kg for SE, and at an injection rate of 5–10 ml/s, immediately followed by a 20- to 30-ml saline flush (Rosen et al. 1990). DSC does not expose patients to ionizing radiation. Intolerance to gadolinium chelates is very rare. The contraindications are those of MRI in general: pacemakers and some other implanted metallic or electronic devices, obesity (more than 150 kg). Fixed ferromagnetic dental devices and intracranial clips generate prominent artifacts. Claustrophobia or agitation may require sedation. Technical Requirements Most 1.5-T MR scanners are now equipped with the fastimaging capabilities required for DSC. The use of a power injector is recommended. Imaging Protocol PWI Sequence type

GRE-EPI

TR (ms)

1,300–2,000 minimum for 20 slices

TE (ms)

50–70

TI (ms)

N/A

Flip angle

90°

NA

1

Slice direction

Axial

FOV (cm2)

24 × 24

No. of slices

20

Total slab thickness (cm)

≥13 Full-brain coverage

Resolution (readout × phase resolution)

Minimum: 64 × 64 Optimum: 128 × 128

Rect. FOV applicable

Yes

Slice thickness (mm)

6

Gap (mm)

0.4–0.6

Dosage (mmol/kg BW)

0.15 at 1.0 T 0.1 at ≥1.5 T

Injection rate (ml/s)

3–5

b Value (s/mm2)

N/A

Presaturation slice

No

Total acquisition time

90 s

3.1  Brain: Modern Techniques and Anatomy

Interpretation DSC relies on the application of the indicator dilution theory (Meier and Zierler 1954). In case of blood–brain barrier rupture, the indicator dilution theory must be modified and a measure of the permeability must be considered (Harrer et al. 2004). For qualitative hemodynamic measurements, a preload dose of contrast before the bolus is an easier way to minimize the effect of contrast leakage on calculations. Using commercially available software, various parameters can be calculated in a few minutes from the time–intensity curves measured in each pixel, allowing one to reconstruct parametric maps. The most commonly calculated parameters are: time-to-peak (TTP), apparent mean transit time (apMTT), corresponding to the first moment of the curves, cerebral blood volume (CBV) calculated from the area under the curve, and cerebral blood flow index (CBFi), equal to CBV/apMTT. These maps do not afford quantitative assessment of brain hemodynamics, but provide indicators of hemodynamic disturbances that are very useful in a clinical setting. They can be interpreted visually or semi-quantitatively by calculating the ratio or difference between the values in a ROI placed in the abnormal area and a mirror ROI placed in the contralateral area considered as a normal reference. Note that, for the moment, there is no standardization in the interpretation of the DSC parametric maps (Cha 2003). The quantification of CBF by DSC requires the deconvolution of the measured tissue curves by an arterial input function (AIF) (Østergaard et al. 1996). If the AIF is adequately calibrated, absolute quantitative CBV and CBF maps can be obtained (Østergaard et al. 1998). This step is more complex with DSC than with perfusion CT (PCT) because the relationship between the signal intensity and the gadolinium concentration is not always linear (Kiselev 2001). Several studies have demonstrated a good correlation between the absolute CBV and CBF values obtained according to this approach as compared to PET or XeCT (Hagen et al. 1999). Note that, in order to interpret the DSC perfusion maps quantitatively or semiquantitatively, it is necessary to mask out the large vessels (very prominent with GRE sequences). The reproducibility of the method ranges around 10–15%. Feasibility DSC has no age limitation and can be performed in children. MRI is not a bedside technique. A spatial resolution around 1.5 × 1.5 × 4 mm is routinely available, but the actual in-plane resolution is usually closer to 2 mm, considering the degradation of the point spread function during the bolus passage (Grandin 2003). A delay of 25 min between successive contrast injections has been shown to be sufficient for repeated CBF measurements, allowing one to assess the cerebral vascular reserve using acetazolamide. A longer delay may be

required for successive CBV measurements (Levin et al. 1995). Clinical Applications The main clinical applications of DSC MRI are in patients with acute stroke, chronic cerebrovascular disease, or 1q tumors. MRI can be performed in the emergency setting of hyperacute stroke. DSC is used in association with DWI and magnetic resonance angiography (MRA) for the early evaluation of stroke patients. Prolonged TTP and MTT values are the most sensitive parameters for detecting a hemodynamic disturbance. CBV and CBF maps are more difficult to visually interpret (especially in white matter) but better reflect brain perfusion (Thomalla et al. 2003). A threshold can be applied to DSC maps in order to identify the area at risk for infarction and predict outcome, but no consensus has been achieved regarding the specific thresholds that might distinguish reversible and irreversible ischemia (Shih et al. 2003). The presence of a mismatch (DSC abnormality larger than the diffusion abnormality), vessel occlusion on MRA and absence of hemorrhage should prompt consideration of interventions directed at reperfusion (Fig. 3.1.1) (Schellinger et al. 2003). DSC has also been successfully used to assess the cerebrovascular reserve (Gükel et al. 1996) and vasospasm (Rordorf et al. 1999). In the management of patients with brain tumors, DSC can be combined with anatomic imaging, DWI, spectroscopy and DTI to provide comprehensive and cost beneficial information in one examination (Law et al. 2003). DSC can differentiate high-grade glial tumors with neovascular proliferation and high CBV values from low-grade glial tumors, which typically demonstrate lower levels of CBV (Fig. 3.1.6). DSC can also evaluate the response to treatment (CBV decrease), differentiate tumor recurrence (high CBV) from radiation necrosis (low CBV) (Fig. 3.1.7), and distinguish tumor (high CBV) from infection (low CBV, depending on the etiology) or tumefactive multiple sclerosis lesions (low CBV) (Holmes et al. 2004). 3.1.5  Arterial Spin Labeling Technical Description Arterial spin labeling (ASL), also called arterial spin tagging, relies on the detection of magnetically labeled water perfusing the brain. Once the magnetization of the inflowing water has been modified (generally inverted) upstream, it induces a small MR signal change (a few percent of the tissue magnetization) downstream. Meanwhile, the magnetization of the perfusion tracer—i.e., the labeled water—is relaxing back toward its equilibrium values with the time constant T1 (Detre et al. 1992; Kwong et al. 1995).

179

180

3  Brain, Head, and Neck

Fig. 3.1.6  28-year-old male patient with prior history of seizures. Conventional FLAIR and post-gadolinium T1-weighted images demonstrated a right mesiofrontal, heterogenous, but well-delineated, mass. Perfusion-weighted dynamic susceptibility imaging demonstrated increased area over the curve in a (green) region of interest placed within the mass compared to a region of interest (purple) placed within the normal contralateral

frontal white matter. This increased area over the curve, corresponding to increased cerebral blood volume, combined with the low-grade appearance of the lesion on conventional imaging sequences, raised the suspicion of an oligodendroglioma, which is low-grade tumor with rich blood supply. This diagnosis was confirmed histologically

3.1  Brain: Modern Techniques and Anatomy Fig. 3.1.7  82-year-old male patient radiated for external ear cancer underwent an MRI study that showed an abnormally enhancing lesion in the right temporal lobe, without significant mass effect. Perfusionweighted dynamic susceptibility imaging demonstrated no increased cerebral blood volume (no red) within this lesion, confirming the suspicion of radiation necrosis based on the clinical history

Numerous ways to apply the spin label have been described (Calamante et al. 1999). Existing techniques fall into two categories depending on how the spin labeling is performed: pulsed techniques and continuous techniques. In both cases, blood is labeled before it irrigates the tissue of interest. With a “pulsed” labeling technique, blood magnetization is inverted within a thick slab of tissue located proximal to the slices of interest using a short (a few milliseconds) shaped RF pulse. To better define the tail of this inverted blood bolus, a saturation pulse can advantageously be used (Wong et al. 1998). With a “continuous” labeling technique, blood is inverted continuously as it crosses a plane defined by the RF frequency offset and a constant gradient along the arterial flow direction during the TR interval. Labeling may be applied either with the imaging coil or with a separate RF coil (Zaharchuk et al. 1999). More recently a “pseudo-continuous” ASL method has been proposed, which breaks the continuous pulse into shorter discrete RF and gradient pulses; it is hoped that such a method will be useful for reducing SAR for high-field ASL without loss in the efficacy of labeling (Garcia et al., Proc. ISMRM 2005; 37). Another recent approach has been proposed where the blood magnetization is selectively inverted based on the blood velocity rather than spatial loca-

tion. This new technique, dubbed velocity selective ASL (VS-ASL) differs from the classical pulse inversion and opens a new range of possibilities, including insensitivity to especially long blood arrival times, such as might be seen in ischemic stroke (Duhamel et al. 2003; Wong et al. 2006). The different ASL techniques share one characteristic: no contrast media is needed to determine cerebral blood flow. ASL uses endogenous water as a tracer. A typical acquisition lasts between 5 and 10 min, depending on the scanner quality (magnetic field strength, RF coil sensitivity, etc.). Multiple pairs of label and control images are averaged together in order to obtain the required signalto-noise ratio. Technical Requirements Performed in MRI scanners, ASL requires that the subject can be placed in a magnetic field and hence patients are subject to MRI contraindications. Since ASL is a subtraction technique, it is sensitive to subject movement. Recently, background suppression techniques have been proposed, in which the static tissue signal is reduced as far as possible. In these approaches, the sensitivity of the technique to motion is greatly reduced (Talagala et al. 2004; Ye et al. 2000).

181

182

3  Brain, Head, and Neck

Imaging Protocol Sequence type

Continuous ASL

TR (ms)

4s

TE (ms)

Minimum

TI (ms)

N/A, except as background suppression

Flip angle

90°

NA

20–80, typically

Slice direction

All planes

FOV (cm2)

20–24 cm

No. of slices

1–20

Total slab thickness (cm)

1–10 cm

Resolution (readout × phase resolution)

64 × 64, optional 128 × 128

Rect. FOV applicable

As needed

Slice thickness (mm)

3–10 mm

Gap (mm)

0–5 mm

Dosage (mmol/kg BW)

N/A

Injection rate (ml/s)

N/A

b Value (s/mm2)

N/A

Presaturation slice

N/A, except as background suppression

Total acquisition time

5–10 min (1.5 T) 2–10 min (3 T)

Interpretation For both pulsed and continuous ASL, a control acquisition is necessary. The control acquisition must yield the same tissue signal without blood magnetization inversion. A simple subtraction between the averaged control and label images yields a flow-weighted map. Various models have been proposed to convert this flow-weighted image into a quantitative perfusion map. The data processing can be performed within a few minutes (Buxton et al. 1998). The quantitative accuracy of the ASL technique has been addressed extensively in the literature. Computer simulations using an extensive model (St. Lawrence et al. 2000) and direct comparisons with other, non-MRI methods have been performed (Ewing et al. 1999). It appears that blood flow is correctly estimated in the gray matter. In white matter, numerical simulations predict an overestimation while direct measurements show an underestimation. Quantitative ASL perfusion maps show

less than 10% change when rescanning the same subject (Parkes et al. 2004). The difference in signal between the label and control acquisitions is around 1% of the control images. Perfusion monitoring using ASL therefore requires a very high signal-to-noise ratio (SNR). As the signal difference is low, ASL cannot accurately map blood flow below ~10ml/100g/min. On the other hand, as flow increases (>150 ml/100g/min), ASL will underestimate blood flow. This is due in part to the fact that the labeled protons in the blood are not completely extracted during the first pass; this reduced extraction fraction effect has been observed in animal models (Barbier et al. 2001), though such an effect would only likely be seen in humans in brain neoplasms. In ASL techniques, the MRI sequences include spoiler gradients that suppress the signal arising from large vessels. Hence, ASL is not sensitive to large arteries or veins. The signal obtained with ASL comes mainly from water located in small vessels and in the surrounding tissue, due to the water exchange between blood and tissue (Ye et al. 1997). In the case of cerebrovascular disease, where the time for blood to travel from the labeling plane to the imaged plane is long and possibly spatially heterogeneous, special methods involving long post-labeling delay times or acquisition of images at multiple inversion times have been suggested. Even so, for very long arrival times (e.g., when flow is provided by collateral networks), the magnetic label of the blood will be essentially completely relaxed, such that no information about perfusion can be reliably ascertained (Schepers et al. 2004). Feasibility ASL examinations can be performed in any patient that can tolerate MRI. It is of particular interest for imaging cerebral blood flow in infants and children, given the lack of ionizing radiation or need for intravenous access. Perfusion maps obtained with ASL can cover the entire brain. In humans, the typical voxel size is 2 × 2 × 4 mm or 4 × 4 × 8 mm. Since fast imaging techniques are generally used, it can be challenging to obtain a good image quality in regions with strong magnetic susceptibility gradients (i.e., magnetic field distortions), like the base of the brain near the frontal sinuses, although it is likely that parallel imaging techniques will mitigate this problem. Successive ASL CBF maps can be acquired every other 5–8 s. ASL techniques can thus be applied to functional MRI applications, where control and label images are alternatively acquired. It has been suggested that the changes in blood flow may be more closely spatially registered to neuronal activation than standard BOLD fMRI (Detre et al. 1999). Clinical Applications While ASL techniques have not entered widespread clinical usage, their utility has been demonstrated for a va-

3.1  Brain: Modern Techniques and Anatomy

Fig. 3.1.8  Pre- and post-acetazolamide arterial spin label CBF images in a 9-year-old girl with vasculitis, high-grade right proximal internal carotid artery stenosis, and multiple subacute strokes (arrows). Significant CBF decrease is noted in the entire

right hemisphere following acetazolamide, indicative of cerebrovascular steal. The effect is particularly marked in the regions of the subacute strokes

riety of acute and chronic cerebrovascular diseases. ASL is feasible for acute stroke patients and other emergency patients in hospitals with MRI access, assuming that they are stable enough to undergo an MRI examination. Initial studies were performed in the setting of ischemic cerebrovascular disease (stroke and transient ischemia attack [TIA]), which demonstrated the feasibility of acquiring CBF maps using ASL in both the acute and chronic setting (Fig. 3.1.8) (Alsop et al. 2003). ASL has also been used to study temporal lobe epilepsy (Wolf et al. 2003) and brain tumor perfusion (Warmuth et al. 2003). A study of malignant gliomas demonstrated ASL perfusion imaging to be equally effective in determining tumor grade compared with DSC (Detre et al. 1999). One significant advantage of ASL is the ability to perform multiple repeated measurements, as might be ne­ cessary before and after a cerebrovascular dilator (such as acetazolamide [Ances et al. 2004] or CO2 inhalation), or before and after a neurointerventional procedure (such as carotid endarterectomy [Ances et al. 2004] or stenting). ASL methods also afford evaluation of functional brain activation (Detre et al. 2002; Silva and Kim 2003), useful for instance for the planning of neurosurgical procedures. The ability of ASL to image the perfusion territories of individual cerebral arteries yields completely new information, and as such, offers much promise for imaging the functional significance of aberrant cerebral circulation and the patency of bypass grafts (Hendrikse et al. 2004).

standard magnetic resonance imaging (MRI) scanner with a vender-supplied software package. The technique is based on the physical principles of proton nuclear magnetic resonance (1H NMR) spectroscopy, where absorption of electromagnetic radiation in the radiofrequency spectrum generates a plot of peak intensities versus absorption frequency, modulated by the molecular composition of the sample. Clinical MRS of the brain provides a diagnostic modality for the biochemical characterization of developmental and pathologic neurological conditions. Theoretically, MRS could be performed on any nuclei where the sum of the number of protons and neutrons results in an odd number—e.g., (1H) hydrogen, (15N) nitrogen, (13C) carbon, (19F) fluorine, (23Na) sodium and (31P) phosphorus—however, nuclei must also be in an adequate number to generate sufficient signal-to-noise ratio. Hydrogen is present in adequate concentrations in the brain for routine clinical spectroscopy (Burtscher and Holtas 2001; Bujar et al. 2005). Higher field strength magnets may permit enough signal to allow clinically relevant MR spectroscopy of other elements such as sodium. MR spectra are depicted in a plot of peak intensities on the y-axis, which corresponds to the relative number of chemically equivalent protons, and resonance frequency on the x-axis, represented in parts per million (ppm). Protons in different local chemical environments will have different resonance frequencies allowing for the spectroscopic differentiation of physiologically relevant chemical compositions. Spectral resonances are “normalized” for the magnetic field strength at which they are acquired (e.g., 1.5 or 3 T). The resonances are then plotted on the x-axis in ppm, which is a dimensionless unit (Silverstein et al. 1981).

3.1.6  Spectroscopy Technical Description Magnetic resonance spectroscopy (MRS) is a non-invasive diagnostic technique that can be performed with a

183

184

3  Brain, Head, and Neck

Either single voxel (SV) or multi-voxel techniques can be utilized for spectroscopic data acquisition. Of the single voxel techniques, two commonly used acquisition sequences are stimulated echo acquisition mode (STEAM) and point-resolved spectroscopy (PRESS). The STEAM technique is faster, utilizing shorter TE values, but has a twofold smaller signal-to-noise ratio than the PRESS technique (Ballestero 2001; Danielsen and Ross 1999). Single voxel techniques typically utilize a cubic box (2 cm × 2 cm × 2 cm) that provides a volume of 8 cm3 and good signal-to-noise ratios. Either short or long echo times can be employed to evaluate the brain parenchyma. In short echo spectroscopy, which utilizes TE of 20–35 ms, metabolites with both short and long T2 relaxation times are observed. Short TE spectra are useful in evaluating complex metabolic abnormalities. Glutamine, glutamate, myoinositol, and most amino acids are better evaluated at short TE (Cecil and Jones 2001). With a long TE, such as 288 ms, only metabolites with a long T2 are observed. Long TE studies produce spectra where N-acetyl aspartate (NAA), creatine (Cr), and choline (Cho) are the dominant peaks. Evaluation for the presence of a lactate, a doublet peak at 1.3 ppm, can also be performed at long echo. At an intermediate TE of 144 ms, the characteristic lactate doublet peak is inverted, which can aid in its identification. Multi-voxel techniques allow for the acquisition of data over a larger volume of brain than single-voxel techniques. This facilitates comparison of normal and abnormal tissue without requiring completion of a separate single voxel acquisition in a new location. Multi-voxel techniques include two- and three-dimensional chemical shift imaging (CSI). The time of acquisition is longer than with single voxel techniques. Multiple voxels can be acquired in single or multiple slices using either the PRESS or STEAM technique. Since the concentration of hydrogen atoms from physi­ ologic water in the brain is approximately 10,000 times that of metabolites of clinical interest, suppression of the water peak (resonance) at approximately 4.7 ppm is required to facilitate spectral interpretation (Burtscher and Holtas 2001). Water suppression is commonly accomplished using a chemically selective saturation (CHESS) radiofrequency pulse applied at the water resonance prior to implementation of the selected localization technique (Danielsen and Ross 1999). To produce a more homogeneous magnetic field within the voxels, shim coils are used. The effect of this process reduces spectral peak broadening and improves signal-to-noise ratios. Technical Requirements Most 1.5- and 3-T MRI scanners are equipped with vender-supplied software packages capable of performing MRS.

Imaging protocol: Single Voxel MRS Protocol (for a GE 3T Signa MRI Scanner, Using the Excite Software Package, Version 11) Single Voxel

MRS

Sequence type

PRESS (Probe-P)

TR (ms)

1,500

TE (ms)

26, 144, 288

Flip angle

90–180–180°

NEX

8

Slice direction

Axial

FOV (cm2)

24 × 24

No. of slices

1

Total slab thickness (cm)

2

Resolution (readout × phase resolution)

1×1

Rect. FOV applicable

No

Slice thickness (mm)

20

Gap (mm)

N/A

Phase

1

Frequency

1

Presaturation slice

6 saturation bands

Total acquisition time

3:48

Imaging Protocol: Multiple Voxel MRS Example of 2D MRS Protocol (for a GE 3T Signa MRI Scanner, Using the Excite Software Package, Version 11) Multiple Voxel

MRS

Sequence type

PRESS (Probe-P)

TR (ms)

1,000

TE (ms)

144

Flip angle

90–180–180°

NEX

1

Slice direction

Axial

FOV (cm2)

24 × 24

No. of slices

1

Total slab thickness (cm)

1

3.1  Brain: Modern Techniques and Anatomy

Imaging Protocol: Multiple Voxel MRS Example of 2D MRS Protocol (for a GE 3T Signa MRI Scanner, Using the Excite Software Package, Version 11) (continued) Multiple Voxel

MRS

Sequence type

PRESS (Probe-P)

Resolution (readout × phase resolution)

1×1

Rect. FOV applicable

No

Slice thickness (mm)

10

Gap (mm)

N/A

Phase

16

Frequency

16

Presaturation slice

6 saturation bands

Total acquisition time

4:20

Interpretation The main metabolites in brain MRS include NAA, Cr, Cho, and the lactate and lipid groups (LL). These metabolites are best evaluated at long TE (Fig. 3.1.9a). Additional metabolites, best seen at short TE, include glutamine (Glu) and glutamate (Gln) (often referred to as

Fig. 3.1.9  a Normal long echo (TE = 288 ms) single-voxel white matter spectrum demonstrating the N-acetyl aspartate (NAA), creatine (Cr), and choline (Cho) resonances at 2.02, 3.02 and 3.22 ppm, respectively. There is no evidence of lactate or lipid contributions at 0.9–1.3 ppm. b Normal short echo (TE = 26 ms) sin-

Glx in combination), myoinositol (mI), and amino acids (Fig. 3.1.9b). These key metabolites are discussed below. N-Acetyl Aspartate The NAA resonance is seen as a single peak at 2.02 ppm and is believed to be a marker of neuronal health or function (Birken and Oldendorf 1989). Since NAA is located primarily in the neurons and their axons, reduction or loss of the NAA peak is seen in most neurodegenerative and neurodestructive processes. It is probably better to describe NAA as a marker proportional to the health of the neurons, as it is theoretically possible that energy depletion without permanent neuronal damage could result in a transient reduction in the NAA peak (Barkovich 2000). The presence of a large NAA peak indicates more normal neuronal presence and function, while diminished peaks occur in situations where neural damage or replacement (such as tumor) has occurred. A gradual physiologic increase in NAA is seen in the normally developing infant brain for approximately the first two years of life (Horska et al. 2002). The only pathological process where abnormally elevated NAA is seen is with Canavan’s disease (Cecil and Jones 2001). Creatine Cr and phosphocreatine resonance peaks are present at 3.03 and 3.94 ppm, respectively. These compounds are a marker of metabolic brain energy. Cr serves as an inter-

gle-voxel white matter spectrum demonstrating the NAA, Cr, and Cho resonances at 2.02, 3.02 and 3.22 ppm, respectively. Contributions from glutamine and glutamate (Glx) are seen as a complex set of peaks at 2.2–2.5 ppm. Primary resonant peaks of myoinositol (mI) are present at 3.56 and 4.06 ppm

185

186

3  Brain, Head, and Neck

nal standard. The concentration of creatine is highest in the cerebellum, then gray matter, and finally, white matter (Michaelis et al. 1993). Comparison of the Cr:NAA and Cr:Cho ratios can be helpful in many disease processes to indicate a relative loss or gain of NAA or Cho (Ross and Bluml 2001). Care must be taken when evaluating these ratios however, since Cr levels are subject to variability in disease process such as creatine deficiency syndromes and in the setting of stroke, trauma, and necrotic tumors (Danielsen and Ross 1999). Choline The dominant Cho resonance is a combination of several choline containing molecules, and is located at 3.22 ppm as a single peak. Phosphocholine is present in cell walls of normal brain tissue. Elevation of the Cho peak can be identified when degenerative or inflammatory disease processes release membrane-bound Cho. Elevation of the Cho peak is also seen in cellular processes with high cell turnover such as high-grade tumors. Demyelinating processes or acute infarctions will also release Cho, or cause lysis of cell walls, and increase the concentration of choline. This can be a transient effect however, while tumors will demonstrate persistent Cho elevation (Danielsen and Ross 1999). Lactate and lipids Lipid and lactate (lactic acid) are seen as multiple peaks in the 0.9- to 1.3-ppm range. Lactate (Lac) is a doublet (two peaks close to one another) at 1.33 ppm and is a byproduct of anaerobic metabolism. Lipids resonate between 0.9 and 1.2 ppm. Both are released with cell destruction or synthesized in necrosis. Increased lipid–lactate (LL) can be seen in necrotic tumors and in stroke, due to destruction of cells. Elevated Lac is a sign of hypoxic ischemic injury and can also be seen as a marker for mitochondrial disorders (Lin et al. 2003). Abscess can also demonstrate elevations in LL. To deconvolute the contribution of Lac in the presence of a prominent lipid peak, an intermediate TE spectrum (TE = 144 ms) can be completed to invert the lactate doublet, while the lipid resonance will remain positive at this TE. Lac is present in cerebral spinal fluid (CSF) as a normal finding (Danielsen and Ross 1999). Care must be taken in placement and interpretation of MRS voxels to avoid or account for contributions of CSF from adjacent tissue. Artifactually elevated lipid contributions can occur if voxels improperly include portions of the skull or scalp soft tissues. This is especially problematic when evaluating lesions near the skull base or at the convexity. A small Lac concentration can also be seen in the premature neonatal brain as a normal finding, which varies with location and gestational age (Barkovich 2000). It is important to note that solvents in common sedation agents, such as propylene glycol, will be detected in MR spectroscopy as

a doublet at 1.1 ppm and should not be confused with the doublet of Lac at 1.3 ppm (Cady et al. 1994). Glutamine and Glutamate Glutamine and glutamate (Glx) contribute to resonant peaks between approximately 2.2 and 2.5 ppm and also at 3.75 ppm (Govindaraju et al. 1998). The most abundant amino acid in the human brain is the neurotransmitter Glu. Gln, the primary derivative of Glu, is found primarily in astrocytes. Destructive neuronal processes, such as hypoxia–ischemic injury (Pu et al. 2000) and hyperammonemia (Wang and Zimmerman 1998) can cause an elevation of the Glx resonances. Myoinositol mI has primary resonant peaks at 3.56 and 4.06 ppm as is a glial cell marker. It has been proposed that elevations in mI can be seen in processes such as gliosis, tuberous sclerosis, and cortical dysplasias (Cecil and Jones 2001). Feasibility Most 1.5- and 3.0-T MRI scanners are equipped with vender-supplied software packages capable of performing MRS. There are no age limitations, though infants and children often require conscious sedation to reduce patient motion. Clinical Applications Clinical applications vary with the age of the patient. In adult patients, the primary clinical utility is the evaluation of suspected intracranial neoplasms; monitoring the response to treatment in patients with brain and certain demyelinating processes; and in the differentiation of abscess versus tumor. The bulk of the clinical utility of MRS in adult patients lies in the primary evaluation of brain tumors and assessment of surgical and chemotherapeutic treatment effects. The evaluation of brain tumors is best-accomplished utilizing long echo (TE = 288 ms) two- or three-dimensional chemical shift imaging (CSI) technique that includes both voxels overlying abnormal brain and normal brain for comparison. In high-grade neoplasms, such as glioblastoma multiformae (GBM), a complex series of spectral changes can be seen, including elevation of Cho, with reduction of NAA as normal brain is replaced with tumor (Fig. 3.1.10). The presence of lactate and lipids is not unexpected in high-grade neoplasms, in which central necrosis may be present. In neonates and children, MRS is employed to evaluate for the presence of Lac in suspected hypoxic/ischemic injury or for metabolic abnormalities. For the evaluation of hypoxic ischemic injury, single voxels (SV) placed over the frontal white matter and basal ganglia, at long or intermediate echo (TE = 288 or 144 ms) are recommended (Fig. 3.1.11).

3.1  Brain: Modern Techniques and Anatomy

Fig. 3.1.10  Long echo (TE = 288 ms) point resolved three-dimensional point-resolved MR spectroscopy in a patient with glioblastoma multiformae. The inset green voxel corresponds to an area of abnormal enhancement on the diagnostic image at the margin of a cavitary lesion. An abnormal elevation of Cho (the dominant peak in the spectrum) with a notable absence of NAA is seen. Additionally, the second largest peak in the spectrum (at the far right of the voxel) represents lipid contributions from tumor necrosis. The three-dimensional acquisition facilitates comparison of normal and abnormal tissues in a single acquisition. z Values represent abnormal primary peak ratios (NAA, Cr, Cho) which are more than two standard deviations greater than the expected ratios. Voxels with abnormal z values are shaded gray for identification

187

188

3  Brain, Head, and Neck Fig. 3.1.11  Long echo (TE = 288) singlevoxel white matter spectrum in a newborn infant with hypoxic–ischemic injury. The presence of a lactate doublet (Lac) at 1.33 ppm is consistent with hypoxic/ischemic injury. The NAA, Cr, and Cho ratios are normal for an infant of this age. As brain development progresses, the NAA:Cho ratio increases to an adult spectrum seen   in Fig. 3.1.9a.

Fig. 3.1.12  Short echo (TE = 26 ms) singlevoxel spectrum placed over the left basal ganglia in a newborn child with maple syrup urine disease. A doublet at 0.9 ppm (arrow) demonstrates the characteristic branched-chain amino acid resonance. This peak should not be confused with the typical Lac doublet seen at 1.33 ppm (arrowhead)

3.1  Brain: Modern Techniques and Anatomy

The use of MRS for the evaluation of metabolic abnormalities is extensively represented in the literature (Danielsen and Ross 1999; Cecil and Jones 2001; Barkovich 2000). Metabolic abnormalities are generally best-evaluated utilizing short echo (TE = 26) single-voxel spectroscopy. As an example, short echo SV MRS can be an important diagnostic imaging technique for the diagnosis of maple syrup urine disease, an abnormality of oxidative phosphorylation of branched-chain amino acids (Gujar et al. 2005; Cecil and Jones 2001; Barkovich 2000). The early clinical symptoms include poor feeding, vomiting, dystonia, and seizure. Short echo spectroscopy demonstrates a characteristic doublet at 0.9 ppm (Fig. 3.1.12) (which should not be confused with the doublet of lactate at 1.3 ppm). MRS has been reported to be positive, even in the absence of conventional imaging findings, and improves following treatment (Zimmerman and Wang 1997). References 1.

Alsop DC, McGarvey ML, Maldjian JA, Wang J, Detre JA (2003) Susceptibility contrast and arterial spin label perfusion MRI in cerebrovascular disease. J Neuroimaging 13:17–27 2. Ances BM, McGarvey ML, Abrahams JM et al (2004) Continuous arterial spin labeled perfusion magnetic resonance imaging in patients before and after carotid endarterectomy. J Neuroimaging 14:133–138 3. Arfanakis K, Haughton VM, Carew JD et al (2002) Diffusion tensor MR imaging in diffuse axonal injury. AJNR Am J Neuroradiol 23:794–802 4. Baird A, Warach S (1998) Magnetic resonance imaging of acute stroke. J Cereb Blood Flow Metab 18:583–609 5. Ballestero J (2001) Essentials of proton magnetic resonance spectroscopy and applications in space-occupying lesions of the brain. Applied Radiology 30:55–63 6. Bammer R, Auer M, Keeling SL et al (2002) Diffusion tensor imaging using single-shot SENSE-EPI. Magn Reson Med 48:128–136 7. Barbier EL, Silva AC, Kim SG, Koretsky AP (2001) Perfusion imaging using dynamic arterial spin labeling (DASL). Magn Reson Med 45:1021–1029 8. Barkovich AJ (2000) Pediatric neuroimaging, 3rd edn. Lipincott Williams and Wilkins, Philadelphia, pp 55–61 9. Basser PJ, Mattiello J, Le Bihan D (1994) Estimation of the effective self-diffusion tensor from the NMR spin-echo. J Magn Reson 103:247–254 10. Basser PJ, Pajevic S, Pierpaoli C et al (2002) In vivo fiber tractography using DT-MRI data. Magn Reson Med 44:625–632 11. Beaulieu C, de Crespigny A, Tong DC, Moseley DC, Albers GW, Marks MP (1999) Longitudinal magnetic resonance imaging study of perfusion and diffusion in stroke: evolution of lesion volume and correlation with clinical outcome. Ann Neurol 46:568–578

12. Birken DL, Oldendorf WH (1989) N-acetyl-l-aspartic acid: a literature review of a compound prominent in 1h-nmr spectroscopic studies of the brain. Neurosci Biobehav Rev 13:23–31 13. Burtscher IM, Holtas S (2001) Proton MR spectroscopy in clinical routine. J Magn Reson Imaging 13:560–567 14. Buxton RB, Frank LR, Wong EC, Siewert B, Warach S, Edelman RR (1998) A general kinetic model for quantitative perfusion imaging with arterial spin labeling. Magn Reson Med 40:383–396 15. Cady EB, Lorek A, Penrice J et al (1994) Detection of propan-1,2-diol in the neonatal brain by in vivo proton magnetic resonance spectroscopy. Magn Reson Med 32:764–767 16. Calamante F, Thomas DL, Pell GS, Wiersma J, Turner R (1999) Measuring cerebral blood flow using magnetic resonance imaging techniques. J Cereb Blood Flow Metab 19:701–735 17. Carr HY, Purcell EM (1954) Effects of diffusion on free precession in nuclear magnetic resonance experiments. Phys Rev 94:630–638 18. Cecil KM, Jones BV (2001) Magnetic resonance spectroscopy of the pediatric brain. Magn Reson Imaging 12:435–452 19. Cha S (2003) Perfusion MR imaging: basic principles and clinical applications. Magn Reson Imaging Clin N Am 11:403–413 20. Danielsen ER, Ross B (1999) Magnetic resonance spectroscopy diagnosis of neurological disease. Dekker, New York 21. Detre JA, Wang J (2002) Technical aspects and utility of fMRI using BOLD and ASL. Clin Neurophysiol 113:621–634 22. Detre JA, Leigh JS, Williams DS, Koretsky AP (1992) Perfusion imaging. Magn Reson Med 23:37–45 23. Detre JA, Samuels OB, Alsop DC, Gonzalez-At JB, Kasner SE, Raps EC (1999) Noninvasive magnetic resonance imaging evaluation of cerebral blood flow with acetazolamide challenge in patients with cerebrovascular stenosis. J Magn Reson Imaging 10:870–875 24. Dorenbeck U, Butz B, Schlaier J, Bretschneider T, Schuierer G, Feuerbach S (2003) Diffusion-weighted echo-planar MRI of the brain with calculated ADCs: a useful tool in the differential diagnosis of tumor necrosis from abscess? J Neuroimaging 13:330–338 25. Duhamel G, de Bazelaire C, Alsop DC (2003) Evaluation of systematic quantification errors in velocity-selective arterial spin labeling of the brain. Magn Reson Med 50:145–153 26. Ewing JR., Wei L, Knight R, Nagaraja TN, Fenstermacher JD (1999) A direct comparison between MRI arterial spintagging and quantitative autoradiography for measured cerebral blood flow in rats with experimental cerebral ischemia. Copenhagen: Brain ‘99, 19th annual meeting 1999, p 595 27. Filippi M, Cercignani M, Inglese M et al (2001) Diffusion tensor magnetic resonance imaging in multiple sclerosis. Neurology 56:304–311

189

190

3  Brain, Head, and Neck 28. Gelderen P van, de Vleeschouwer MHM, DesPres D, Pekar J, van Zijl PCM, Moonen CTW (1994) Water diffusion and acute stroke. Magn Reson Med 31:154–163 29. Gonzalez RG, Schaefer PW, Buonanno F et al (1999) Diffusion-weighted MR imaging: diagnostic accuracy in patients imaged within 6 h of stroke symptom onset. Radiology 210:155–162 30. Govindaraju V, Basus VJ, Matson GB, Maudsley AA (1998) Measurement of chemical shifts and coupling constants for glutamate and glutamine. Magn Reson Med 39:1011–1013 31. Grandin CB (2003) Assessment of brain perfusion with MRI: methodology and application to acute stroke. Neuroradiology 45:755–766 32. Gujar SK, Maheshwari S, Bjorkman-Burtscher I, Sundgren PC (2005) Magnetic resonance spectroscopy. J NeuroOphthalmol 25:217–226 33. Gükel FJ, Brix G, Schmiedek P et al (1996) Cerebrovascular reserve capacity in patients with occlusive cerebrovascular disease: assessment with dynamic susceptibility contrastenhanced MR imaging and the acetazolamide stimulation test. Radiology 201:405–412 34. Hacke W, Kaste M, Fieschi C et al (1995) Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA 274:1017–1025 35. Hagen T, Bartylla K, Piepgras U (1999) Correlation of regional cerebral blood flow measured by stable xenon CT and perfusion MRI. J Comput Assist Tomogr 23:257–264 36. Hakyemez B, Aksoy U, Yildiz H, Ergin N (2005) Intracranial epidermoid cysts: diffusion-weighted, FLAIR and conventional MR findings. Eur J Radiol 54:214–220 37. Harrer JU, Parker GJ, Haroon HA et al (2004) Comparative study of methods for determining vascular permeability and blood volume in human gliomas. J Magn Reson Imaging 20:748–757 38. Hasan KM, Parker DL, Alexander AL (2001) Comparison of gradient encoding schemes for diffusion-tensor MRI. J Magn Reson Imaging 13:769–780 39. Hendrikse J, van der Grond J, Lu H, van Zijl PC, Golay X (2004) Flow territory mapping of the cerebral arteries with regional perfusion MRI. Stroke 35:882–888 40. Holmes TM, Petrella JR, Provenzale JM. Distinction between cerebral abscesses and high-grade neoplasms by dynamic susceptibility contrast perfusion MRI. AJR Am J Roentgenol 183:1247–1252 41. Holodny AI, Ollenschlager M (2002) Diffusion imaging in brain tumors. Neuroimaging Clin N Am 12:107–124 42. Horska A, Kaufmann WE, Brant LJ, Naidu S, Harris JC, Barker PB (2002) In vivo quantitative proton MRSI study of brain development from childhood to adolescence. J. Magn Reson Imaging 15:137–143 43. Jones SC, Perez-Trepichio AD, Xue M, Furlan AJ, Awad IA (1994) Magnetic resonance diffusion-weighted imaging: sensitivity and apparent diffusion constant in stroke. Acta Neurochir 60:207–210

44. Kiselev VG (2001) On the theoretical basis of perfusion measurements by dynamic susceptibility contrast MRI. Magn Reson Med 46:1113–1122 45. Kuker W, Nagele T, Schmidt F, Heckl S, Herrlinger U (2004) Diffusion-weighted MRI in herpes simplex encephalitis: a report of three cases. Neuroradiology 46:122–125 46. Kumon Y, Zenke K, Kusunoki K et al (1999) Diagnostic use of isotropic diffusion-weighted MRI in patients with ischaemic stroke: detection of the lesion responsible for the clinical deficit. Neuroradiology 41:777–784 47. Kwong, KK, Chesler, DA, Weisskoff, RM et al (1995) MR perfusion studies with T1 weighted echo planar imaging, Magn Reson Med 34:878–887 48. Lansberg MG, Norbash AM, Marks MP et al (2000) Advantages of adding diffusion-weighted magnetic resonance imaging to conventional magnetic resonance imaging for evaluation of acute stroke. Arch Neurol 57:1311–1316 49. Law M, Yang S, Wang H et al (2003) Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR 24:1989–1998 50. Levin JM, Kaufman MJ, Ross MH, Mendelson JH, Maas LC, Cohen BM, Renshaw PH (1995) Sequential dynamic susceptibility contrast MR experiments in human brain: residual contrast agent effect, steady state, and hemodynamic perturbation. Magn Reson Med 34:655–663 51. Lin DM, Crawford TO, Barker PB (2003) Proton MR Spectroscopy in the Diagnostic Evaluation of Suspected Mitochondrial Disease, AJNR Am J Neuroradiol:24:33–41 52. Lutsep HL, Albers GW, DeCrespigny A, Kamat GN, Marks MP, Moseley ME (1997) Clinical utility of diffusionweighted magnetic resonance imaging in the assessment of ischemic stroke. Ann Neurol 41:574–580 53. Meier P, Zierler KL (1954) On the theory of the indicatordilution method for measurement of blood flow and volume. J Appl Physiol 6:731–744 54. Michaelis T, Merboldt KD, Bruhn H, Hanicke W, Frahm J (1993) Absolute concentrations of metabolites in the human brain in-vivo: quantification of localized proton MR spectra. Radiology 197:219–227 55. Mohr JP, Biller J, Hilal SK et al (1995) Magnetic resonance versus computed tomographic imaging in acute stroke. Stroke 26:807–812 56. Moseley M, Cohen Y, Mintorovitch J et al (1990a) Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighred MRI and spectroscopy. Magn Reson Med;14:330–346 57. Moseley M, Kucharczyk J, Mintorovirch J et al (1990b) Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility enhanced MR imaging in cats. AJNR 11:423–429 58. Mukherjee P, McKinstry RC (2006) Diffusion tensor imaging and tractography of human brain development. Neuroimaging Clin N Am 16:19–43

3.1  Brain: Modern Techniques and Anatomy 59. Mukherjee P (2005) Diffusion tensor imaging and fiber tractography in acute stroke. Neuroimaging Clin N Am 15:655–665 60. Neil JJ, Shiran Sl. McKinstry RC et al (1998) Normal brain in human newborns: apparent diffusion coefficient and diffusion anisotropy measured by using diffusion tensor MR imaging. Radiology 209:57–68 61. NINDS t-PA Stroke Study Group The (1997) Intracerebral hemorrhage after intravenous t-PA therapy for ischemic stroke. Stroke 28:2109–2118 62. Oppenheim C, Stanescu R, Dormont D et al (2000) Falsenegative diffusion-weighted MR findings in acute ischemic stroke. AJNR 21:1434–1440 63. Østergaard L, Weisskoff RM, Chesler DA, Gyldensted C, Rosen RR (1996) High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part I: mathematical approach and statistical analysis. Magn Reson Med 36:715–725 64. Østergaard L, Smith DF, Vestergaard-Poulsen P, Hansen SB, Gee AD, Gjedde A, Gyldensted C (1998) Absolute cerebral blood flow and volume measured by magnetic resonance imaging bolus tracking: comparison with positron emission tomography values. J Cereb Blood Flow Metab 18:425–432 65. O’Sullivan M, Jones DK, Summers PE et al (2001) Evidence for cortical “disconnection” as a mechanism of age-related cognitive decline. Neurology 57:632–638 66. Pajevic S, Pierpaoli C (1999) Color schemes to represent the orientation of anisotropic tissues from diffusion tensor data: application to white matter fiber tract mapping in the human brain. Magn Reson Med 42:526–540 67. Papadakis NG, Murrills CD, Hall LD et al (2000) Minimal gradient encoding for robust estimation of diffusion anisotropy. Magn Reson Imaging 18:671–679 68. Parkes LM, Rashid W, Chard DT, Tofts PS (2004) Normal cerebral perfusion measurements using arterial spin labeling: reproducibility, stability, and age and gender effects. Magn Reson Med 51:736–743 69. Provenzale JR, Engelter ST, Petrella JR, Smith JS, MacFall JR (1999) Use of MR exponential diffusion-weighted images to eradicate T2 shine-through effect. AJR Am J Roentgenol 172:537–539 70. Pu Y, Li QF, Zeng CM et al (2000) Increased detectability of alpha brain glutamate/glutamine in neonatal hypoxic-ischemic encephalopathy. AJNR Am J Neuroradiol 21:203–212 71. Rordorf G, Koroshetz WJ, Copen WA et al (1999) Diffusion- and perfusion-weighted imaging in vasospasm after subarachnoid hemorrhage. Stroke 30:599–605 72. Rosen BR, Belliveau JW, Vevea JM, Brady TJ (1990) Perfusion imaging with NMR contrast agents. Magn Reson Med 14:249–265 73. Ross B, Bluml S (2001) Magnetic resonance spectroscopy of the human brain. Anat Rec 265:54–84

74. Rugg-Gunn FJ, Eriksson SH, Symms MR et al (2002) Diffusion tensor imaging in refractory epilepsy. Lancet 359:1748–1751 75. Schellinger PD ,Fiebach JB, Hacke W (2003) Imagingbased decision making in thrombolytic therapy for ischemic stroke: present status. Stroke 34:575–575 76. Schepers J, Veldhuis WB, Pauw RJ et al (2004) Comparison of FAIR perfusion kinetics with DSC-MRI and functional histology in a model of transient ischemia. Magn Reson Med 51:312–320 77. Shih LC, Saver JL, Alger JR, Starkman S et al (2003) Perfusion-weighted magnetic resonance imaging thresholds identifying core, irreversibly infarcted tissue. Stroke 34:1425–1430 78. Silva AC, Kim SG (2003) Perfusion-based functional magnetic resonance imaging. Concepts in Magn Reson Part A 16A:16–27 79. Silverstein RM, Bassler GC, Morrill TC (1981) Spectrometric identification of organic compounds, 4th edn. Wiley, New York, pp 181–247 80. Skare S, Hedehus M, Moseley ME, Li TQ (2000) Condition number as a measure of noise performance of diffusion tensor data acquisition schemes with MRI. J Magn Reson 147:340–352 81. Speck O, Chang L, DeSilva NM, Ernst T (2000) Perfusion MRI of the human brain with dynamic susceptibility contrast: gradient-echo versus spin-echo techniques. J Magn Reson Imaging 12:381–387 82. St. Lawrence KS, Frank JA, McLaughlin AC (2000) Effect of restricted water exchange on cerebral blood flow values calculated with arterial spin tagging: a theoretical investigation. Magn Reson Med 44:440–449 83. Stejskal EO (1965) Use of spin-echoes in a pulsed magnetic-field gradient to study anisotropic restricted diffusion and flow. J Chem Phys 43:3597–3603 84. Stejskal EO, Tanner JE (1965) Spin diffusion measurements: spin-echoes in the presence of a time-dependent field gradient. J Chem Phys 42:288–292 85. Talagala SL, Ye FQ, Ledden PJ, Chesnick S (2004) Wholebrain 3D perfusion MRI at 3.0 T using CASL with a separate labeling coil. Magn Reson Med 52:131–140 86. Thomalla GJ, Kucinski T, Schoder V et al (2003) Prediction of malignant middle cerebral artery infarction by early perfusion- and diffusion-weighted magnetic resonance imaging. Stroke 34:1892–1899 87. Ulug AM, Beauchamp N, Bryan RN, van Zijl PCM (1997) Absolute quantitation of diffusion constants in human stroke. Stroke 28:483–490 88. Virta A, Barnett A, Pierpaoli C (1999) Visualizing and characterizing white matter fiber structure and architecture in the human pyramidal tract using diffusion tensor MRI. Magn Reson Imaging 17:1121–1133 89. Wang ZJ, Zimmerman RA (1998) Proton MR spectroscopy of pediatric metabolic disorders. Neuroimag Clin N Am 8:781–807

191

192

3  Brain, Head, and Neck 90. Warach S, Chien D, Li W, Ronthal M, Edelman RR (1992) Fast magnetic resonance diffusion-weighted imaging of acute human stroke. Neurology 42:1717–1723 91. Warmuth C, Gunther M, Zimmer C (2003) Quantification of blood flow in brain tumors: comparison of arterial spin labeling and dynamic susceptibility-weighted contrast-enhanced MR imaging. Radiology 228:523–532 92. Woessner DE (1961) Effects of diffusion in nuclear magnetic resonance spin-echo experiments. J Chem Phys 34:2057–2061 93. Wolf RL, Alsop DC, McGarvey ML, Maldjian JA, Wang J, Detre JA (2003) Susceptibility contrast and arterial spin label perfusion MRI in cerebrovascular disease. J Neuroimaging 13:17–27 94. Wong EC, Buxton RB, Frank LR (1998) Quantitative imaging of perfusion using a single subtraction (QUIPSS and QUIPSS II). Magn Reson Med 39:702–708 95. Wong EC, Cronin M, Wu W-C, Inglis B, Frank LR, Liu TT (2006) Velocity-selective arterial spin labeling. Magn Reson Med 55:1334–1341

96. Ye FQ, Mattay VS, Jezzard P, Frank JA, Weinberger DR, McLaughlin AC (1997) Correction for vascular artifacts in cerebral blood flow values measured by using arterial spin tagging techniques. Magn Reson Med 37:226–237 97. Ye FQ, Frank JA, Weinberger DR, McLaughlin AC (2000) Noise reduction in 3D perfusion imaging by attenuating the static signal in arterial spin tagging (ASSIST). Magn Reson Med 44:92–100 98. Young GS, Geschwind MD, Fischbein NJ et al (2005) Diffusion-weighted and fluid-attenuated inversion recovery imaging in Creutzfeldt-Jakob disease: high sensitivity and specificity for diagnosis. AJNR 26:1551–1562 99. Zaharchuk G, Ledden PJ, Kwong KK, Reese TG, Rosen BR, Wald LL (1999) Multislice perfusion and perfusion territory imaging in humans with separate label and image coils. Magn Reson Med 41:1093–1098 100. Zimmerman RA, Wang ZJ (1997) The value of proton MR spectroscopy in pediatric metabolic brain disorders. AJNR Am J Neuroradiol 18:1872–1879

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics B.B. Ertl-Wagner and C. Rummeny 3.2.1  Introduction MR imaging is of paramount importance in the diagnostic evaluation of congenital and acquired diseases of the brain in children. Its superior delineation of gray and white matter, and its lack of ionizing radiation make it the method of choice in diagnostic imaging of the brain in children. When assessing the central nervous system (CNS) in a pediatric population, the radiologist is faced with specific challenges. Both the normal development and age-specific disorders of the brain must be taken into account. In addition, many pediatric patients need a special preparation prior to undergoing MR imaging, such as sedation or deep anesthesia. Informed consent must be obtained from the parents and/or legal guardians who also play a crucial role in the success of the procedure, especially in children undergoing MR imaging without sedation. However, it is not only the pediatric (neuro-)radiologist who is faced with congenital disorders of the brain or with CNS diseases acquired during childhood. Many affected children reach adulthood and subsequently present with their own subset of imaging features quite different from those commonly encountered in the adult population. Therefore, it is also important for the diagnostic radiologist primarily working with adults to be familiar with diseases of the pediatric brain. 3.2.2  Examination Technique 3.2.2.1  Patient Preparation Both the age and the potential degree of mental impairment of the child need to be taken into account when planning an MR examination of the brain. The overall goal is to acquire all diagnostically necessary images without interference from motion artifacts. To reach this end, most children under the age of 6- to 8-yearsold need to either be sedated or undergo anesthesia with intubation. However, there are several exceptions to this rule. Neonates can oftentimes be scanned without sedation when MR imaging is performed directly after feeding. Moreover, quite a few 5- to 8-year-olds are able to undergo a full MR examination when the procedure is properly explained to them and a parent is present in the scanner room. Sometimes it can be helpful—albeit a bit uncomfortable for the parent—to allow the parent to lie

in the scanner with the child. In addition, if only major changes need to be excluded and a high spatial resolution is unnecessary, such as in the evaluation of CSF shunt malfunction, it is often possible to image the non-sedated child with rapid sequences, such as HASTE, RARE, or SSFSE sequences. It is feasible to construct a protocol that takes little over one minute in such instances. Sedation can be administered via an oral, rectal, intramuscular, or intravenous route. Commonly used sedatives for children are chloral hydrate and pentobarbital. Another agent used in children is propofol. The American Academy of Pediatrics recommends that before sedation infants be kept NPO for at least 4 h and older children for at least 6 h (American Academy of Pediatrics Committee on Drugs 1992, 2002). Children should be monitored for arterial oxygen saturation, heart and respiratory rates, and blood pressure when undergoing sedation. From about school age on, many children are able to undergo MR imaging without sedation. It is important to adequately explain the procedure to the child and also to involve the accompanying parent in all stages of the examination. Moreover, the individual development of the child and potential developmental delays need to be taken into account when making a decision for or against sedation. 3.2.2.2  Imaging Protocols Generally, sagittal sequences should be performed in all children undergoing MR imaging of the brain in order to adequately evaluate the midline structures. We generally prefer to use T1-weighted volumetric sequences with isotropic or near isotropic voxels (e.g., MP-RAGE) covering the whole brain, as these can be reformatted in any desired imaging plane when abnormalities are found. T2-weighted sagittal sequences can also be helpful, e.g., when an early leukodystrophic involvement of the corpus callosum is sought. In addition, axial T2- and T1-weighted sequences should always be performed. We prefer to use inversion recovery sequences with a strong T1 weighting, as these allow for a good differentiation of gray and white matter even in young infants. Additional FLAIR sequences are also helpful especially when seeking areas of gliosis or edema in children with complete or near complete myelination. T2*-weighted sequences can aid in delineating foci of old hemorrhage and calcification, e.g., in children after traumatic injury to the brain or in vascular malformations. In children with temporal lobe epilepsy, it is mandatory to perform sequences with a thin-slice thickness of the temporal lobe in axial and coronal orientations, which should follow the axis of the temporal lobe and be perpendicular to it, respectively. FLAIR sequences and T1-weighted inversion recovery sequences are especially

193

194

3  Brain, Head, and Neck

helpful when looking for areas of hippocampal sclerosis and cortical dysplasia in these instances. In children with a suspected neoplasm of the brain, a gadolinium-based contrast medium should be applied intravenously at a standard dose of 0.1 mmol/kg. T2weighted and/or FLAIR sequences as well as unenhanced sequences should be acquired prior to the administration of contrast medium. After the administration of contrast medium, T1-weighted sequences should be acquired in at least two, preferably three planes. In our experience, it can be helpful to also acquire a rapid diffusion weighted sequence in order to aid in differentiating a cerebral abscess from a primary neoplasm of the brain. Pediatric brain tumors are covered in another chapter of this book. In children with signs and symptoms of acute stroke, diffusion- and perfusion-weighted sequences can be helpful in delineating the area of focal cytotoxic edema and in defining regions of perfusion–diffusion mismatch. MR angiography can delineate the site of vascular occlusion and aid in the diagnosis of pathologies of the dural veins and sinuses. Diffusion tensor imaging (DTI), a relatively new technique, provides parameters of anisotropy and of fiber connectivity. In children, DTI can aid in assessing the degree of maturation of the brain. An increased anisotropy may also be a sensitive parameter in children with developmental delays of undetermined etiology (Filippi et al. 2003). Magnetization transfer techniques can also be helpful when evaluating normal myelination or demye­ linating and dysmyelinating processes. Phase contrast studies of CSF flow can aid in evaluating children with Chiari malformations or aqueductal stenosis. MR spectroscopy is especially helpful when assessing children with metabolic disorders of the brain. Moreover, it can aid in the characterization and follow-up of pediatric brain tumors. Functional MRI (fMRI), which is based on the blood oxygen level dependent (BOLD) effect, can be difficult to perform in young children and in children with developmental delays, as this technique generally requires a high degree of cooperation. In older children, fMRI can often be successfully performed, when the child is adequately prepared. However, it still mostly belongs in the realm of research and is rarely performed in the routine clinical setting. 3.2.3  Normal Development of the Brain 3.2.3.1  Normal Intrauterine Development of the Brain The knowledge of the embryonic and fetal development of the brain is of paramount importance for the understanding of congenital disorders. Early in the embryonic development, the neural plac-

ode arises, which subsequently develops into the neural tube. From the rostral part of the neural tube, three primary brain vesicles form: the prosencephalon, the mesencephalon, and the rhombencephalon. The prosencephalon subsequently develops two outpouchings, forming the telencephalon. These outpouchings later develop into the two cerebral hemispheres and the lateral ventricles. A disorder of this cleavage of the prosencephalon is called holoprosencephaly (see below). In addition, the diencephalon arises from the prosencephalon, giving rise to the thalamus, hypothalamus, and the third ventricle. The mesencephalon develops into the structures of the midbrain including the aqueduct. The rhombencephalon—also termed hindbrain—develops into the metencephalon and the myelonencephalon. The former gives rise to the pons and cerebellum, the latter to the medulla oblongata. The fourth ventricle also develops from the rhombencephalon. The lamina terminalis lies between the outpouchings of the telencephalon. Around the seventh week of development, the dorsal part thickens giving rise to the corpus callosum and the cerebral commissures. The ventral part of the lamina terminalis forms into the meninx primitiva, which later develops into the cerebral meninges. During the development of the corpus callosum, the dorsal part of the genu and the corpus form first. Subsequently, the anterior part of the genu and then the splenium arise. Last, the rostrum develops. As a rule of thumb, the corpus callosum is formed from anterior to posterior, with the exception of the anterior part of the genu and the rostrum. Thus, when assessing a child with a partially missing corpus callosum, this sequence of development needs to be taken into account. The cortical development consists of three processes. The neuronal stem cells are formed in the germinal matrix zone, a zone of high proliferative activity, which lies adjacent to the ependyma of the lateral ventricles. The neurons then radially migrate to the cortical surface, where they are organized into their respective layers. Thus, the process of cortical development consists of neuronal proliferation, neuronal migration, and cortical organization. It is important to remember that the formation of gyri and sulci does not take place until relatively late in fetal cerebral development. Until about the fifth month of intrauterine development, the cerebral hemispheres appear smooth. It is not until then that the Sylvian fissure appears. Subsequently, the other primary, secondary, and tertiary sulci develop. It is not until term, however, that the gyration of the brain appears approximately mature, though even then the sulci are shallower than in older children and adults. When interpreting fetal MR images or images of premature infants, it is important to take these processes of gyration into account in order to be able to differentiate disorders of gyration from normal cerebral development.

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

3.2.3.2  Normal Postnatal Myelination For the radiologist primarily involved in MR imaging of the adult brain, the MR appearance of the brain of the neonate and small child may be unfamiliar due to its inherent lack of myelination. Large parts of the process of myelination take place postnatally and continue throughout infancy. When assessing cranial MR images in young children, it is important to keep the child’s age in mind in order to correctly interpret the individual state of cerebral maturation. If the child formerly was a premature infant, the corrected age in relation to the term date needs to be considered. Myelination begins during the fifth month of intrauterine development and generally continues throughout life. However, the most active periods of myelination occur during the first 2 years of life. T2- and T1-weighted images reflect the changes in myelination at a different rate (Barkovich et al. 1988). As a general rule, myelination appears slightly more advanced on T1-weighted images as compared with T2-weighted sequences. In the

first 6 months of life, T1-weighted sequences tend to be more helpful in assessing the myelination, while in the second year of life T2-weighted images tend to be more reliable. Myelination generally adheres to a specific pattern with a set chronological order. It is important to keep this pattern in mind in order to be able to differentiate delayed myelination from myelinoclastic disorders. From the 25th week of gestation to term birth, the medial and lateral lemnisci, the median longitudinal fasciculus, and the superior cerebellar peduncle usually mye­ linate, as demonstrated both in T2- and in T1-weighted sequences (Fig. 3.2.1). Moreover, the posterior portion of the posterior limb of the internal capsule is usually mye­ linated at term birth, while the anterior portion is demonstrated to be myelinated in the first month of postnatal life in T1-weighted sequences and in the first 6 months in T2-weighted sequences. The centrum semiovale and the anterior limb of the internal capsule usually demonstrate signs of myelination by 2–4 months in T1-weighted sequences and by 7–11 months in T2-weighted sequences (Fig. 3.2.2). The central parts of the frontal and the

Fig. 3.2.1  T2-weighted (a,b) and T1-weighted IR (c,d) axial sequences in a premature infant (imaging performed 3 weeks before term) demonstrate a yet largely absent myelination. Beginning myelination is discerned in the dorsal brainstem and cerebellar hemispheres

Fig. 3.2.2  T2-weighted (a,b) and T1-weighted IR (c,d) axial sequences in a 6-month-old girl demonstrate myelination of the internal capsule and the pericentral regions

195

196

3  Brain, Head, and Neck

Fig. 3.2.3  T2-weighted (a,b) and T1-weighted IR (c,d) axial sequences in a 12-month-old boy demonstrate myelination of large parts of the cerebral white matter, with the exception of the subcortical U fibers

Fig. 3.2.4  T2-weighted sagittal (a) and axial (b) sequences in an 18-month-old girl demonstrate a largely myelinated cerebral white matter with the exception of the directly subcortical U fibers in the temporal lobe

occipital white matter usually demonstrate signs of myelination by 3–6 months in T1-weighted sequences and by 9–16 months in T2-weighted sequences, with myelination occurring slightly later in the frontal white matter (Fig. 3.2.3). The peripheral, subcortical white matter with its U fibers generally matures last, with the exception of the perirolandic and calcarine areas. While T1-weighted sequences usually demonstrate signs of subcortical mye­ lination by 4–7 months in the peripheral occipital white matter and by 7–11 months in the frontal white matter, T2-weighted sequences mostly demonstrate signs of subcortical myelination by about 11–15 months in the occipital white matter and by about 14–18 months in the frontal white matter. The white matter of the temporal lobe generally matures last. T1-weighted sequences usu-

ally demonstrate signs of myelination in the subcortical temporal white matter until about 7–11 months, while T2-weighted sequences display myelination of the subcortical fibers at a later stage, from about 12 months to about 22–24 months of age (Fig. 3.2.4). By a child’s second birthday, the process of myelination as visualized by conventional MR imaging sequences appears nearly complete with the exception of the so-called terminal zones of myelination. These terminal zones consist of areas of persistently increased signal intensity in T2-weighted sequences in the peritrigonal zones dorsal and superior to the ventricular trigones and to a lesser degree along the bodies of lateral ventricles. These fibers with a physiologically delayed myelination belong to the so-called association areas in parts of the parietal and temporal lobes. Histologically, these fibers are known to

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

develop myelin as late as the fourth decade of life (Barkovich 2005). It can sometimes be challenging to differentiate these zones of physiologically delayed myelination from pathological processes such as periventricular leukomalacia. However, a layer of normal myelination is usually present between the terminal zone and the ventricular ependyma and other signs of brain injury such as a focal thinning of the corpus callosum or an irregularity of the ventricular wall are absent (Baker et al. 1988). 3.2.4  Congenital Disorders of the Brain 3.2.4.1  Disorders of Cortical Development As outlined above, cortical development consists of three basic processes: neuronal stem cell proliferation in the periventricular germinal matrix zone, neuronal migration along radial pathways and organization of the cortex into the respective cellular layers. Disorders of cortical development have therefore classically been divided into three categories: disorders of cellular proliferation, disorders of neuronal migration, and disorders of cortical organization (Barkovich et al. 1996). These classifications have been somewhat modified and refined, as additional knowledge has become available, especially on a molecular level (Barkovich et al. 2001a). Disorders of cellular proliferation are now further divided into disorders with decreased proliferation and increased apoptosis (i.e., programmed cell death), disorders with increased proliferation and decreased apoptosis, and disorders with abnormal proliferation. The first category encompasses the microlissencephalies, microcephalies with polymicrogyria or cortical dysplasia, and microcephalies with normal or thin cortex. The megalen-

cephalies belong to the second category, while the hemimegalencephalies as well as cortical dysplasias with balloon cells and cortical hamartomas in tuberous sclerosis are part of the third category. Moreover, certain tumors, such as gangliogliomas, gangliocytomas, and dysembryoplastic neuroepithelial tumors are considered disorders of abnormal neuronal proliferation. Heterotopia are classical disorders of neuronal migration. They can be subependymal or subcortical. Band heterotopia are now considered part of the lissencephaly spectrum, which is also classified as a disorder of neuronal migration. In addition, the cobblestone lissencephalies are considered disorders of neuronal migration. Polymicrogyria—including the bilateral polymicrogyria syndromes—and schizencephalies are considered malformations of cortical organization. In addition, cortical dysplasias without balloon cells and microdysgenesis belong to the category of disorders of cortical organization. A fourth category of not otherwise classified malformations encompasses malformations secondary to inborn errors of metabolism, such as mitochondrial or peroxisomal disorders (Barkovich et al. 2001a). 3.2.4.1.1  Microlissencephaly A child is considered microcephalic when the head circumference is at least two standard deviations below the norm, and severely microcephalic when the head circumference is at least three standard deviations below the norm (Barkovich et al. 1998). Microlissencephaly is caused by an increased apoptosis and a decreased proliferation of the neuronal stem cells. The gyral pattern is simplified with a paucity of gyri and shallow sulci (Fig. 3.2.5).

Fig. 3.2.5  Sagittal T2-weighted (a) and axial T1-weighted (b) sequences in a 10-month-old boy with microlissencephaly de­ monstrate an extreme microcephaly with a strikingly reduced gyral pattern

197

198

3  Brain, Head, and Neck

Microlissencephalies were further subdivided into initially five, later six groups (Barkovich 2005; Barkovich et al. 1998)). Strictly speaking, only the latter two groups are called microlissencephaly, while the first four groups are termed microcephaly with simplified gyral pattern. Children with a group 1 microlissencephaly are usually born after an uneventful pregnancy and initially have normal neonatal courses. MR imaging demonstrates a volume reduction of the cerebral hemispheres and diminished gyri and sulci, while the cortex itself appears normal. Infants with a group 2 microlissencephaly usually demonstrate signs and symptoms of spasticity and often of epilepsy soon after birth. The MR appearance resembles that of children with a group 1 lissencephaly; myelination is commonly delayed, though. Children with a group 3 microlissencephaly usually become symptomatic early with diminished reflexes, hypotonia, and seizures. Gyri and sulci are usually fewer in number compared to groups 1 and 2. Heterotopia and arachnoid cysts are common. Patients with a group 4 microlissencephaly usually have various additional malformations such as arthrogryposis multiplex congenital or jejunal atresia. MR imaging resembles that of group 1 or 2 microlissencephalies. In group 5 microlissencephaly, there is a severe reduction both of head circumference and of gyration. Neonates are hypotonic and develop seizures early in development. No more than five sulci are found in each cerebral hemisphere. Group 6 microlissencephaly is characterized by a complete or near complete lack of gyration. Moreover, the corpus callosum is frequently absent and the cerebellum can be hypogenetic. The development of affected children is usually severely compromised. 3.2.4.1.2  Hemimegalencephaly Hemimegalencephaly is a complex disorder of stem cell proliferation, neuronal migration, and cortical organization resulting in a hamartomatous overgrowth and increased volume of an entire cerebral hemisphere or parts thereof (Barkovich and Chuang 1990). The disorder may be associated with a hemihypertrophy of the body. Some, but not all affected children are symptomatic from infancy with intractable seizures and developmental delays. MR imaging demonstrates a variable degree of volume augmentation of affected parts—or an entire—cerebral hemisphere. The lateral ventricle is often enlarged; however, it may also appear compressed. Heterotopia, abnormalities of gyration, and glioses are often found as well. The white matter–gray matter junction is often blurred. It is always important to thoroughly evaluate both the affected and the apparently unaffected hemispheres, especially when considering a surgical resection of the

affected region, as hemimegalencephaly can in selected cases have bilateral manifestations. Moreover, one should remember that the appearance of the affected cerebral regions can change during the process of brain maturation (Barkovich 2005). 3.2.4.1.3  Transmantle Cortical Dysplasia (Cortical Dysplasia with Balloon Cells) Focal transmantle cortical dysplasia is also referred to as cortical dysplasia with balloon cells or as Taylor type dysplasia. The term balloon cells is a histopathological description of large cells with an extensive cytoplasm, which are considered neuronal stem cells (Mackay et al. 2003). As cortical hamartomas in tuberous sclerosis present with the exact same histopathological appearance, affected children should be screened for potential other signs of tuberous sclerosis (Mackay et al. 2003). Clinically, affected patients usually present with a seizure disorder, which is oftentimes refractory to medical therapy. MR imaging characteristically demonstrates a funnelshaped signal abnormality extending from the cortex to the ventricular surface (compare Fig. 3.2.11c). This signal abnormality is hyperintense to myelinated white matter on T2-weighted and FLAIR images. The adjacent cortex appears dysplastic. 3.2.4.1.4  Heterotopia Heterotopia are abnormal collections of gray matter within the white matter. They result from an arrest in neuronal migration. Heterotopia are differentiated into subependymal heterotopia, focal subcortical heterotopia, and band heterotopia. According to the newer classification, band heterotopia are now considered part of the lissencephaly spectrum (Barkovich et al. 2001a). Most patients with heterotopia present with seizure disorders. Affected children may display signs of a developmental delay as well. Heterotopia are not uncommonly associated with other disorders of cortical development such as polymicrogyria or cortical dysplasias and also with visceral abnormalities (Barkovich and Kuziecky 2000). However, if heterotopia represent an isolated finding without other associated abnormalities, the symptoms of the patient may be mild. Subependymal heterotopia are collections of neurons that remained within the germinal matrix zone directly adjacent to the ventricular ependyma. In some cases, subependymal heterotopia line the entire surface of the lateral ventricles. Subependymal heterotopia are often sporadic, but may be X-linked or autosomal recessive as well, especially if they are extensive (Barkovich and Kuziecky 2000). On MR imaging, subependymal heterotopia present as round to ovoid nodules that often slightly protrude into

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.6  Sagittal T2-weighted sequences in an 11-year-old boy with subependymal heterotopia demonstrate multiple foci of gray matter directly adjacent to the lateral ventricle

the lateral ventricles (Fig. 3.2.6). They are isointense to cortical gray matter in all sequences. The most important differential diagnosis is subependymal nodules in tuberous sclerosis. However, subependymal nodules in tuberous sclerosis are not isointense to gray matter, frequently display signs of calcification and often enhance after administration of contrast medium. In addition, their long axis is usually perpendicular to the ventricular surface, while it is mostly parallel to the ventricular surface in subependymal heterotopia (Barkovich 2005). Focal subcortical heterotopia are abnormal collections of neurons along the radial path of migration from the subependymal germinal matrix zone to the cortical surface. They are therefore nests of gray matter surrounded by white matter. Associated malformations of the brain, such as agenesis or hypogenesis of the corpus callosum, are common. Affected children commonly present with epilepsy and developmental delay (Barkovich and Kuziecky 2000). However, if the focal subcortical heterotopia are small, development may be normal. On MR imaging, focal subcortical heterotopia are round to ovoid signal abnormalities within the white matter that are isointense to the cortical gray matter in all sequences. Their long axis is usually oriented along the radial path of migration (Fig. 3.2.7). They never demonstrate perifocal edema and never enhance after application of contrast medium, which aids in differentiating heterotopia from tumors (Barkovich and Kuziecky 2000). The overlying cortex is usually thinned and may show signs of dysplasia.

Fig. 3.2.7  Axial T2-weighted (a) and T1-weighted IR (b) sequences in a 7-year-old boy with focal subcortical heterotopia demonstrate nests of gray matter along the radial tracts of migration

199

200

3  Brain, Head, and Neck

Band heterotopia are now considered part of the lissencephaly spectrum, but shall be discussed in this section for didactic purposes. They are also called double cortex, as a band of cortex-isointense signal intensity is found within the white matter. Symptoms can be quite variable, with seizure disorders and developmental delays representing the most common complaints. Girls are affected more often than are boys. An X-chromosomal genetic locus is known, but there are sporadic cases as well (Leventer et al. 2000). As in other heterotopia, band heterotopia are isointense to cortical gray matter in all MR sequences. The band lies between the ependymal surface and the cerebral cortex and is surrounded by white matter on both sides. Band heterotopia can be complete or partial. The overlying cortex is usually of normal thickness, but sulci tend to be shallower. 3.2.4.1.5  The Lissencephaly Spectrum The lissencephaly spectrum is characterized by a reduced or absent gyration, i.e., the gyri and sulci are too few and too shallow or completely absent. Classical lissencephaly is also called the agyria–pachygyria complex, with agyria being a complete lissencephaly, i.e., a complete absence of gyri, and pachygyria an incomplete lissencephaly, i.e., a reduced gyral pattern. At the same time, the cortical band is thickened. The underlying process in classical lissencephaly is an arrest in neuronal migration. Affected children usually

present with a marked developmental delay and with a seizure disorder. However, symptoms may vary depending on the underlying genetic cause. Numerous genetic aberrations have been identified in classical lissencephaly. The most common aberrations are mutations in the so-called LIS gene (locus at 17p13.3.) or in the DCS/XLIS gene (locus at Xq22.3–q23) (Barkovich 2005). Patients with a LIS gene mutation may have characteristic dysmorphic features, which are then called Miller-Dieker syndrome. Mothers of boys with a DCS/ XLIS mutation, i.e., an X-chromosomal mutation, commonly have band heterotopia (Barkovich 2005). A mutation in the gene encoding for reelin (RELN) also results in a lissencephaly with usually severe developmental delay and a generalized seizure disorder. The respective cortical layers are disorganized. Another subset of patients suffers from mutations in the ARX gene (locus at Xp22.13), which causes lissencephaly with callosal dysgenesis and ambiguous genitalia (Barkovich 2005). On MR imaging, classical lissencephaly is always characterized by a simplified or absent gyral pattern with a concomitant thickening of the cortical band. In complete lissencephaly, the surface of the brain appears smooth. Due to shallow Sylvian fissures, the brain usually has the appearance of the figure-of-eight. In incomplete lissencephaly, shallow sulci and gyri are present; there may be areas of normal and/or absent gyration as well (Fig. 3.2.8). Incomplete lissencephaly is much more common than complete lissencephaly. When differentiating incomplete lissencephaly from polymicrogyria, high-resolution MR

Fig. 3.2.8  Coronal T2-weighted (a, b) sequences in a 12-month-old boy with incomplete lissencephaly demonstrate a reduced pattern of gyration with shallow sulci

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

sequences are helpful. In lissencephaly (pachygyria), the junction between the gray and the white matter appears smooth, while in polymicrogyria it is always irregular. Depending on the underlying genetic cause, associated malformations may be present such as agenesis or hypogenesis of the corpus callosum and cerebellar hypoplasia or dysplasia. 3.2.4.1.6  Cobblestone Lissencephaly In contrast to classical lissencephaly, cobblestone lissencephaly is characterized by an overmigration of neurons. Cobblestone malformations are characteristically associated with congenital muscular dystrophies (van der Knaap et al. 1997). Among these are Walker-Warburg syndrome, muscle–eye–brain disease, and Fukuyama muscular dystrophy. Children with Walker-Warburg syndrome are usually severely affected from birth on (van der Knaap et al. 1997). The cortical band is thickened with only few, shallow sulci. In contrast to classical lissencephaly, the junction between gray and white matter is characteristically irregular, hence the name cobblestone lissencephaly (Fig. 3.2.9). In addition, affected children usually have hydrocephalus, severe hypomyelination and a hypogenesis of the corpus callosum. Moreover, ocular malformations such as microphthalmos or a persistent hyperplastic primary vitreous are common. Pontine and cerebellar hypogenesis may also be present.

Patients with muscle–eye–brain disease usually present early with hypotonia, developmental delay, seizures, and visual disturbances (van der Knaap et al. 1997). Most cases are described in children of Finnish heritage. MR imaging again demonstrates a cobblestone appearance of the cortex. There is a characteristic inverse pattern of myelination starting in the subcortical U fibers and continuing in a centripetal fashion. Moreover, patchy signal alterations within the white matter remain after completion of myelination. Pons and cerebellum are usually hypoplastic with dysplastic cysts within the cerebellar cortex. Hypogenesis of the corpus callosum and agenesis of the septum pellucidum are common. Hydrocephalus may be present. Fukuyama muscular dystrophy is mostly seen in children of Japanese heritage. Again, children present with severe developmental delay and a seizure disorder (van der Knaap et al. 1997). The severity, however, is somewhat milder than with the other two diseases. On MR imaging, a cobblestone appearance of the cortex is mostly seen in the temporal and occipital regions, while the frontal regions are typically affected by polymicrogyria. On the inner surface of the cobblestone cortex, nodules of gray matter can frequently be identified. The cerebellum tends to be dysplastic with subcortical cysts. Again, myelination is usually delayed and there is an inverse pattern of myelination starting in the subcortical U fibers.

Fig. 3.2.9  Axial T2-weighted (a) and coronal T1-weighted IR (b) sequences in a 9-month-old boy with Walker-Warburg syndrome demonstrate cobblestone lissencephaly and pronounced hypomyelination

201

202

3  Brain, Head, and Neck

3.2.4.1.7  Polymicrogyria Polymicrogyria results from a disturbance in late neuronal migration and cortical organization. Its hallmark feature are too many and too small gyri and sulci. Affected children present with variable degrees of developmental delay and seizure disorders with the severity of symptoms depending on the region and on the extent of the cortex involved.

Several syndromes have been described in association with polymicrogyria, especially when it is present in a bilaterally symmetrical fashion. The congenital bilateral perisylvian syndrome is characterized by a bilateral opercular polymicrogyria (Kuzniecky et al. 1993). It can be inherited or sporadic. Pseudobulbar palsy, dysarthria, and epilepsy are usually present. Another syndrome with an identified genetic locus (16q12.2–21) has been described as bilateral symmetrical frontoparietal polymicrogyria (Guerrini et al. 2000). The children usually suffer from esotropia, seizures bilateral pyramidal tract signs, and a developmental delay. Bilateral polymicrogyria syndromes in other regions of the cortex have been described as well. In addition, polymicrogyria can be sporadic or arise secondarily, e.g., after a congenital infection, such as a CMV infection, or after ischemic events in utero. MR imaging in polymicrogyria demonstrates the typical, irregular interface between white and gray matter (Fig. 3.2.10). As this irregularity can be slight and difficult to discern on routine MR images, it is important to perform high-resolution sequences in three planes in order to ascertain the diagnosis and to differentiate it from pachygyria. The insular and peri-insular regions are involved most commonly; however, polymicrogyria can affect any region of the cortex. The cortical surface is characteristically irregular in polymicrogyria as well; however, it can paradoxically appear smooth, as the outer cortical layer may fuse over the polymicrogyria. It is therefore always mandatory to scrutinize the junction between cortex and white matter. The affected cortical region is usually isointense to physiologic cortex in all sequences. However, there may be a patchy increase in signal intensity in the adjacent subcortical white matter in T2-weighted sequences. In addition, calcification may be present, especially in children with secondary causes, such as a congenital CMV infection. Developmental venous anomalies may be present in proximity to the area of polymicrogyria. 3.2.4.1.8  Schizencephaly

Fig. 3.2.10  Axial (a) and coronal (b) T2-weighted sequences in a 2-month-old boy with bilateral perisylvian polymicrogyria demonstrate too many and too small gyri in the perisylvian region with an irregular gray matter–white matter interface

The hallmark feature of schizencephaly is a—unilateral or bilateral—cleft extending from the ependymal lining of the ventricles to the cortical surface. As polymicrogyria is also usually found at the outer rim of the schizencephalic cleft, these two disorders are now usually summarized as the polymicrogyria–schizencephaly complex (Barkovich et al. 2001a). However, classifications vary and some authors place it under the category of disorders of stem cell proliferation as opposed to disorders of cortical organization. Most likely, processes of all three stages of cortical development are involved in schizencephaly. Schizencephaly can be familial (e.g., with a mutation of the EMX2 gene) or acquired, e.g., by an in utero focal transmantle injury of the developing brain (Granata et

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

al. 2005). The clinical symptoms of the affected children vary, depending on the extent of the clefts and on the region of the brain involved. Most patients present with a seizure disorder, though. In addition, developmental delays and signs of hemiparesis or hemianopsia may be present as well. Schizencephalies can be unilateral in about 60% and bilateral in about 40%. They are further subdivided into

open clefts that are filled with CSF (open-lip schizencephaly) and clefts with fused lips (closed-lip schizencephaly) (Barkovich 2005). Open-lip schizencephalies can vary with regard to the size of the cleft. On MR imaging, a cleft can be seen that extends from the ventricular surface to the cortical surface (Fig. 3.2.11). This cleft is always lined with dysplastic gray matter, which aids in differentiating schizencephaly from a

Fig. 3.2.11  Axial T2-weighted (a) and T1- weighted (b) and coronal T1-weighted IR (c, d) sequences in a 15-year-old boy with bilateral schizencephaly demonstrate a bilateral cleft lined with dysplastic gray matter. In addition, heterotopia, polymicrogyria, and cortical dysplasias including transhemispheric cortical dysplasias are found

203

204

3  Brain, Head, and Neck

Fig. 3.2.12  Axial T2-weighted (a) and T1-weighted IR (b) sequences in a 7-year-old boy with cortical dysplasia demonstrate two areas of focal cortical dysplasia in the left frontal lobe. The gray matter–white matter junction is blurred and adjacent heterotopia are found

simple cerebral defect. This dysplastic gray matter usually has an irregular surface and a somewhat uneven interface to the adjacent white matter. The adjacent cortex often displays signs of polymicrogyria. The septum pellucidum is commonly absent, especially in cases of open lip schizencephaly or bilateral schizencephaly. An open-lip schizencephaly is usually easily discerned as it is filled with CSF along its entire extent. A large open-lip schizencephaly often leads to a focal bulging of the calvarium due to direct CSF pulsations. A closed-lip schizencephaly is often more difficult to discern. However, a focal dimple is usually seen at the ventricular surface, which points to the site of schizencephaly. Moreover, the dysplastic cortex lining the cleft can be followed through the entire hemisphere. As schizencephalies are bilateral in 40% of cases, it is always important to scrutinize the contralateral hemisphere as well. In addition to a contralateral cleft, cortical dysplasia may be present as well, often in a mirror image. 3.2.4.1.9  Focal Cortical Dysplasia without Balloon Cells In contrast to transmantle cortical dysplasia, which is histologically characterized by balloon cells and classified as a disorder of stem cell proliferation, focal cortical dysplasia is categorized as a disorder of cortical organization

(Barkovich et al. 2001a). The cortical dysplasia lies focally on the cortical surface and usually involves the directly underlying white matter; however, it does not extend through the hemisphere. Affected patients usually present with partial epilepsy. Focal cortical dysplasia without balloon cells can be challenging to diagnose. High-resolution sequences with thin sections are usually required. Typically, a focal blurring of the interface between cortex and subcortical white matter can be discerned (Fig. 3.2.12). Moreover, signal alterations are often present, with a focal hyperintensity being most easily detected on thin- section FLAIR sequences. However, in some cases with focal cortical dysplasia even high-resolution MR imaging can be completely normal (Kim et al. 2000). Additional PET scanning may be helpful in these instances. 3.2.4.2  Agenesis or Hypogenesis of the Corpus Callosum As mentioned earlier, the corpus callosum is formed during a critical period of cerebral development. There is a specific sequence of the formation of the corpus callosum with the dorsal genu being formed first, followed by the anterior genu and the corpus. Subsequently, the splenium and finally the rostrum are formed. Thus, in a hypogenesis of the corpus callosum, the callosal parts that are formed later in the course of development will be absent,

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.13  Sagittal T1-weighted (a) and axial T2-weighted (b) sequences in a 2.5-year-old boy with callosal agenesis demonstrate a complete absence of the corpus callosum. The sulci of the medial hemispheres radiate directly into the third ventricle. The corpora of the lateral ventricles appear straight and parallel

while the parts formed earlier will be present. If embryologically “older” portions are absent, while more recently formed portions are present, the diagnosis of a secondary lesion of the corpus callosum, and not of a primary malformation, must be made. An example would be the absence or reduction of the peri-isthmic region with more dorsal regions of the splenium being present, which often occurs in the setting of periventricular leukomalacia (PVL) with an injury of the respective fibers. An exception to this rule is holoprosencephaly in its semilobar form, in which the anterior portion of the corpus callosum is absent. This malformation is discussed in further detail in the next section. Clinically, the symptoms of affected children can be quite variable, ranging from a normal or near normal neurological exam to a severe developmental delay. The severity of the symptomatology also depends on associated or concomitant malformations. Malformations of cortical development are quite common in children with callosal agenesis or hypogenesis and should be specifically sought. In addition, a multitude of syndromes is associated with agenesis or hypogenesis of the corpus callosum. The most frequent of these are Chiari II malformation and Dandy-Walker malformation, which are both discussed in further detail farther below. Other syndromes include Morning Glory syndrome, Apert syndrome, Goltz syndrome, Neu-Laxova syndrome, Cogan syndrome, and Shapiro syndrome (Barkovich 2005). Especially note-

worthy is Aicardi syndrome, which is a spontaneous, balanced translocation of the X-chromosome and therefore primarily occurs in girls (Rosser 2003). Aicardi syndrome is characterized by callosal agenesis or hypogenesis, an interhemispheric cyst, polymicrogyria, heterotopia, extraaxial cysts, especially in the posterior fossa, cerebellar hypoplasia, microphthalmos, and papillomas of the choroid plexus (Rosser 2003). An absence of the corpus callosum leads to a characteristic appearance on MR imaging. To facilitate diagnostic decision-making, axial, coronal, and sagittal sequences should be acquired in all children with suspected callosal agenesis or hypogenesis. The sagittal orientation is the most straightforward for identifying callosal agenesis, as the corpus callosum—or the lack thereof—can be imaged in its longitudinal axis (Fig. 3.2.13). Occasionally, an enlarged hippocampal commissure can mimic the splenium in callosal agenesis; coronal sequences, however, reliably identify the course of the respective fiber tracts (Barkovich 2005). In newborns or infants with largely unmyelinated white matter, the corpus callosum can sometimes be difficult to identify due to its small volume. A helpful sign in this setting is that the sulci of the medial cerebral hemispheres can be followed into the third ventricle in a radial fashion, which is due to a persistent eversion and volume reduction of the cingulate gyrus (Fig. 3.2.13). On axial sequences, the corpora of the lateral ventricles usually have a straight and parallel appearance (Fig.

205

206

3  Brain, Head, and Neck

Fig. 3.2.15  Coronal T1-weighted sequences in a 17-year-old boy with callosal agenesis demonstrate a characteristic indentation of the lateral ventricles leading to a crescent-shaped appearance

On coronal sequences, there usually is a characteristic indentation of the lateral ventricles leading to a crescent shape (Fig. 3.2.15). This indentation is due to the presence of Probst’s bundle on the medial walls of the lateral ventricles (Lee et al. 2004). Probst’s bundle consists of fibers that would normally cross the midline through the corpus callosum and in its absence turn and course parallel to the interhemispheric fissure (Lee et al. 2004). However, if the underlying pathology of a callosal agenesis is an improper guidance of the axons to the midline, Probst’s bundle is absent (Barkovich 2005).

Fig. 3.2.14  T2-weighted sagittal (a) and axial (b) sequences in a 5-year-old boy with callosal hypogenesis demonstrate a partial absence of the corpus callosum. The occipital horns of the lateral ventricles are dilated

3.2.13). The frontal horns of the lateral ventricles appear convex in their lateral parts. The posterior horns of the lateral ventricles, in contrast, are usually dilated, as parts of their morphological stability are lost due to the absence of the tightly packed corpus callosum (Fig. 3.2.14). This ventricular configuration is also called colpocephaly. In addition, the temporal horns of the lateral ventricles are commonly dilated due to an eversion and volume loss of the cingulate gyrus.

3.2.4.3  Callosal Agenesis with Interhemispheric Cyst An agenesis or hypogenesis of the corpus callosum can be associated with an interhemispheric cyst. This condition is generally viewed as a separate entity from simple callosal agenesis and may have a different pathogenesis. According to a more recent classification, agenesis of the corpus callosum with interhemispheric cyst is now usually subdivided into two groups depending on the communication with the ventricle (Barkovich et al. 2001b). While type 1 interhemispheric cysts generally communicate with the ventricles (Fig. 3.2.16), type 2 interhemispheric cysts do not. Children with type 1a and type 2d cysts usually have the best prognosis. Generally, developmental delay and focal neurologic deficits appear to be less frequent in type

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.16  Axial T2-weighted (a) and T1-weighted IR (b) sequences in a 8-year-old boy with an interhemispheric cyst demonstrate an asymmetric cyst that directly communicates with the ventricles

1 cysts than in type 2 cysts. Fenestration or shunting will frequently need to be performed, especially in type 2 cysts. In adequately decompressed cysts, the prognosis is often good (Barkovich et al. 2001b). Type 1 cysts are further subdivided into type 1a, 1b, and 1c cysts. All type 1 cysts are unilocular and isointense to CSF. Type 1a cysts are characterized by neonatal macrocephaly and hydrocephalus. Type 1b cysts communicate with the third ventricle. There is an associated fusion of the thalami. Type 1c cysts communicate with the lateral and third ventricles. The ipsilateral hemisphere is usually small. Affected infants are usually microcephalic and suffer from early seizures. Type 2 cysts are further subdivided into type 2a, 2b, 2c, and 2d cysts, with type 2a, 2b, and 2c cysts being multiloculated and type 2d cysts being unilocular. Type 2a cysts are characterized by neonatal macrocephaly and hydrocephalus. In contrast to other interhemispheric cysts, type 2b cysts are not isointense to CSF, but hyperintense on T1-weighted MR sequences. Associated abnormalities include polymicrogyria, an absence of the falx cerebri and subependymal heterotopia. Affected infants usually suffer from seizures and a developmental delay. Type 2c cysts again are isointense to CSF. Associated abnormalities include subcortical heterotopia. Affected children often suffer from developmental delays and seizures. In contrast to other type 2 cysts, type 2d cysts are unilocular. They represent arachnoid cysts without associated anomalies (Barkovich et al. 2001b).

3.2.4.4  Intracranial Lipomas Intracranial lipomas do not represent true neoplasms, but rather are viewed as malformations stemming from the meninx primitiva that surrounds the brain during early cerebral development. These undifferentiated mesenchymal cells undergo a faulty differentiation into fat cells (Truwit and Barkovich 1990). Intracranial lipomas do not have a neoplastic potential with cell multiplication and rarely exert a mass effect. Moreover, blood vessels usually course through the lesion. A therapeutic approach is therefore only very rarely indicated; in most instances, intracranial lipomas are better left alone (Truwit and Barkovich 1990). Intracranial lipomas are most commonly located in the interhemispheric fissure and are usually associated with an agenesis or hypogenesis of the corpus callosum. Interhemispheric lipomas belong to the category of midline malformations, and are commonly associated with other developmental disorders of the midline, such as midline facial clefts. Other locations of intracranial lipomas are the supracerebellar cisterns and quadrigeminal plate, the suprasellar and interpeduncular cisterns, the cerebellopontine angel and the perisylvian region. On MR imaging, intracranial lipomas present as fat-isointense, subarachnoid lesions. They are typically distinctly hyperintense in T1-weighted sequences (Fig. 3.2.17). Chemical shift artifacts are present in larger lipomas. In addition, sequences with a fat-saturation pulse

207

208

3  Brain, Head, and Neck

Fig. 3.2.17  Sagittal T1-weighted IR (a) and coronal T2-weighted (b,c) sequences in a 3-year-old boy with an interhemispheric lipoma demonstrate callosal agenesis and a fat-isointense lesion of the midline. The vessels course directly through the lipoma. In addition, areas of polymicrogyria and pachygyria are discerned (c)

may be helpful, leading to a pronounced reduction in signal within the lesion. Interhemispheric lipomas usually lie posterior to a hypogenetic corpus callosum. They have a tendency to calcify in the periphery, often in a curvilinear or punctate fashion. Vessels coursing through the lipoma can often be identified as flow voids on MR imaging (Fig. 3.2.17). 3.2.4.5  Holoprosencephaly Holoprosencephaly is a disorder of cleavage of the prosencephalon during early intrauterine cerebral development. As outlined under Sect. 3.3, the prosencephalon usually develops two outpouchings that differentiate into the telencephalon and, ultimately, into two cerebral hemispheres. In holoprosencephaly, the medial aspects of the hemispheres fail to form and therefore appear fused with the contralateral side.

The group of holoprosencephaly is further divided into alobar, semilobar, and lobar holoprosencephaly with alobar holoprosencephaly being the most severe form. In addition, there is a middle interhemispheric variant of holoprosencephaly that is also referred to as syntelencephaly (Barkovich and Quint 1993). The clinical symptoms of affected children depend on the form and severity of the holoprosencephaly. Midline facial dysmorphism, including clefts, may be present. In addition, seizures, developmental delays, and movement disorders are common. The causes for holoprosencephaly can be genetic or teratogenic. Known causes include maternal diabetes, trisomy 13, trisomy 18, and a variety of mutations in different genetic loci (Roessler and Muenke 1998). Alobar holoprosencephaly is the most severe form of holoprosencephaly. Affected children usually only have a short life span or are stillborn. Diagnosis is usually already made in prenatal ultrasonography and MR imag-

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.18  Axial T2-weighted (a), coronal FLAIR (b) and sagittal T1-weighted sequences in a 2-year-old boy with semilobar holoprosencephaly demonstrate a fusion of the cerebral hemispheres in the anterior portion. The anterior part of the corpus callosum is absent, while the posterior part is present (c). In addition, areas of polymicrogyria and heterotopia are found

ing is rarely performed. In alobar holoprosencephaly, the cerebral hemispheres are not differentiated. Falx cerebri, interhemispheric fissure, and corpus callosum are absent. Dorsal to the undifferentiated cerebrum, a crescentshaped holoventricle is found, which usually communicates with a large cyst. In semilobar holoprosencephaly, a partial separation of the cerebral hemispheres can be seen. Children are less severely affected than with alobar holoprosencephaly. The degree of the cerebral anomaly can be variable. Posterior parts of the interhemispheric fissure and of the corpus callosum are formed, while the anterior portions are not (Fig. 3.2.18). Thus, the splenium is usually present, while the genu and often at least parts of the corpus are absent. This is the only exception to the abovementioned rule that in partial malformations of the corpus callosum the parts formed later rather than the parts formed earlier in cerebral development are absent. In semilobar holoprosencephaly, the frontal horns of

the lateral horns are usually lacking, while the posterior horns are at least partially formed. The temporal horns are commonly rudimentary with underdeveloped hippocampi. There may be cysts in the posterior interhemispheric fissure. As in all forms of holoprosencephaly, the septum pellucidum is absent. Lobar holoprosencephaly is the least severe form of holoprosencephaly. Affected children usually present with varying degrees of developmental delay, visual disturbances, and often endocrinologic disorders of the hypothalamic–pituitary axis. On MR imaging, the frontal horns of the lateral ventricles are at least partially formed. The third ventricle, hippocampi, and at least the posterior part of the corpus callosum are present, even though the third ventricle and the hippocampi can be partially hypoplastic. The septum pellucidum, however, is absent. The spectrum of mild forms of lobar holoprosencephaly appears closely related to septo-optic dysplasia.

209

210

3  Brain, Head, and Neck

A variant of holoprosencephaly is syntelencephaly, which is also called the middle interhemispheric variant of holoprosencephaly (Barkovich and Quint 1993). In this subgroup of patients the interhemispheric fissure is formed both in the anterior and posterior portions of the brain, but absent in the middle. The brain therefore appears fused in mid-part of the cerebrum, between the posterior frontal and the parietal lobes. Clinically, the symptoms of affected children are less severe than in alobar or semilobar holoprosencephaly, with often only relatively mild developmental delays and visual symptoms. On MR imaging, an absent callosal body is demonstrated, while the genu and splenium are usually present. The cerebral hemispheres appear fused in the region of the posterior frontal and anterior parietal lobes with an absent falx cerebri in this region. Septo-optic dysplasia is usually considered a separate entity from holoprosencephaly. However, continuity between lobar holoprosencephaly and a subgroup of septo-optic dysplasia has been proposed (Barkovich et al. 1989a). Septo-optic dysplasia is characterized by an agenesis of the septum pellucidum and a hypoplasia or aplasia of the optic nerves. Affected children usually have disturbances of the visual system; however, visual acuity may be normal as well, depending on the degree of optic hypoplasia. In addition, a dysfunction of the hypothalamic-pituitary axis is commonly present leading to an impaired growth. On MR imaging, a complete or partial absence of the septum pellucidum is demonstrated. The fornices usually

lie comparatively low. Optic hypoplasia is usually more difficult to demonstrate, especially when mild. Highresolution MR sequences of the optic nerves and chiasm with thin sections can be helpful in identifying the conditions. However, MR imaging should always be performed in conjunction with an ophthalmological exam, when septo-optic dysplasia is suspected, as up to 50% of optic hypoplasias are known to be missed on MR imaging alone (Barkovich et al. 1989a). Two subgroups of septooptic dysplasia have been proposed. One group is characterized by a partial hypogenesis of the septum pellucidum and associated anomalies of cortical development such as heterotopia or schizencephaly, while the other is characterized by a complete absence of the septum pellucidum and hypoplastic white matter with relative ventriculomegaly. The latter group is considered to fall into the spectrum of lobar holoprosencephaly. 3.2.4.6  Cephaloceles Cephaloceles are characterized by a defect in the skull or skull base, through which intracranial structures extend extracranially. Depending on the herniating intracranial structure, meningoencephaloceles, meningoceles, atretic celes, and glioceles are differentiated (Fig. 3.2.19). Atretic cephaloceles contain fibrous tissue, dura, and degenerated cerebral tissue, while glioceles are characterized by a CSF-filled cyst with glial lining. Regarding their location, cephaloceles are further classified into occipital, fronto-

Fig. 3.2.19  Sagittal (a) and axial (b) T2-weighted sequences in a newborn girl with a frontal meningocele demonstrate a focal, frontal, CSF-filled outpouching of meninges. In addition, polymicrogyria is found in the left frontal lobe (b)

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

ethmoidal, occipitocervical, frontal, parietal, temporal, lateral, sphenomaxillary, spheno-orbital, and nasopharyngeal cephaloceles, with occipital and frontoethmoidal being by far the most common locations (Barkovich 2005). While occipital cephaloceles are considered most common among children of Caucasian descent, frontoethmoidal cephaloceles are more common in children from Southeast Asian ancestry (Naidich et al. 1992). Occipital encephaloceles are usually clinically obvious malformations, which are often already diagnosed in prenatal ultrasonography. MR imaging is typically performed in a prone or lateral decubitus position. A special focus should be placed on the identification and topography of the dural sinuses in relation to the cele, as this is an important factor in treatment planning. Therefore, a venous MR angiography should always be performed in these children. The cephalocele can contain supratentorial or infratentorial tissue with the tentorium often being dysplastic. Associated malformations, such as heterotopia, cerebellar dysplasia, interhemispheric cysts, agenesis, or hypogenesis of the corpus callosum as well as Chiariand Dandy-Walker malformations should be specifically sought. Frontoethmoidal cephaloceles are not as obvious as occipital cephaloceles. They are most commonly found in a nasoethmoidal location, but can also occur nasofrontally or naso-orbitally. Related conditions include dermal sinus tracts opening into a small dimple on the nasal surface and nasal gliomas, which consist of dysplastic ce-

Fig. 3.2.20  Sagittal T1-weighted (a) and axial T2-weighted (b) sequences in a 12-month-old girl with a frontoethmoidal cephalocele demonstrate an extension of cerebral tissue through a nasoethmoidal bone defect

rebral tissue within the nose without an associated bone defect (Naidich et al. 1992). Dermal sinus tracts commonly contain dermoid or epidermoid cysts. Symptoms of affected children are often comparatively minor if associated cerebral malformations are absent. Occasionally, only a nasal congestion prompts the diagnostic evaluation. MR imaging should be performed with thin sections in coronal and sagittal orientations, when a frontoethmoidal cephalocele, a dermal sinus tract, or a nasal glioma is suspected. In children with a nasal dimple, the extent of the dermal sinus tract should be demonstrated and dermoid or epidermoid cysts should be specifically sought. In children with frontoethmoidal cephaloceles, MR imaging demonstrates the direct extension of intracranial tissues into the cephalocele through a bone defect (Fig. 3.2.20), whereas this bone defect is absent in case of a nasal glioma (Naidich et al. 1992). As in children with occipital cephaloceles, associated malformations, such as agenesis or hypogenesis of the corpus callosum, anomalies of cortical development, or intracranial lipomas should be specifically sought. 3.2.4.7  Chiari Malformations Chiari malformations have classically been divided into three subtypes, with Chiari I and Chiari II malformations being by far the most common. Chiari III malformations, characterized by a spina bifida aperta of the first

211

212

3  Brain, Head, and Neck

and second cervical vertebral bodies with posterior fossa contents within a cele, are exceedingly rare and will not be elaborated on in further detail. The hallmark of Chiari I malformations is a caudal displacement of the cerebellar tonsils into the foramen magnum. The size of the posterior fossa is often small and associated craniovertebral anomalies, such as remnants of a proatlas, median basilar invaginations, and C1 assimilations are quite common. Clinical symptoms can be variable. Occipital headaches are among the most common presenting complaints. In addition, vertigo or tinnitus and a downbeat nystagmus may be present. In small children, oropharyngeal dysfunction has been described. Moreover, as associated syringohydromyelia is quite commonly present as well; the spine should therefore also always be examined in patients with Chiari I malformations. On MR imaging, the primary diagnostic criterion to diagnose a Chiari I malformation is tonsillar ectopia within the foramen magnum (Fig. 3.2.21). To reach the diagnosis, thin-section sequences should be performed in a mid-sagittal location. Several lines have been described to quantify the degree of tonsillar ectopia with the most commonly used being a line connecting basion and opisthion. Physiologically, the tip of the cerebellar tonsils should not lie over 5 mm below this line in small children and adults, and over 6 mm below this line in children between the ages of 5 and 15 years (Mikulis et al. 1992). In addition, CSF flow studies with gated phase-contrast measurements can be helpful demonstrating a reduced CSF flow and an increased motion of the brain stem and cerebellar tonsils (Alperin et al. 2005). The most important differential diagnosis to Chiari I malformations is that of intracranial hypotension resulting from a chronic CSF leak. While patients may present with comparatively similar clinical symptoms, the therapeutic approaches to these two disease entities are completely different. In both disorders, a tonsillar ectopia within the foramen magnum is found. In case of intracranial hypotension, this tonsillar ectopia is, however, due to a “sagging” of the brain secondary to the CSF leak. To differentiate these two disease entities, it is helpful to intravenously administer contrast medium. Patients with intracranial hypotension have a thickened dura with a markedly increased enhancement. In addition, the brainstem and the third ventricle also demonstrate a degree of downward displacement. Chiari II malformations—also called Arnold-Chiari malformations—are characterized by an ectopia and downward displacement of the cerebellar tonsils, as well. However, in contrast to Chiari I malformations, the anomalies are more complex and involve the spinal column. Virtually all children with Chiari II malformations also suffer from a spina bifida aperta with concomitant meningomyelocele, which is usually operated on right after birth. A high percentage of children with Chiari II

Fig. 3.2.21  Sagittal T1-weighted sequences in a 39-year-old man with a Chiari I malformation demonstrate a tonsillar ectopia within the foramen magnum

malformations subsequently develop hydrocephalus. In addition, brainstem dysfunction such as dysphagia or breathing difficulty may occur (Barkovich 2005). MR imaging in children with a Chiari II malformation should always include thin-section sagittal sequences through the midline to assess the situation of the hindbrain and the corpus callosum, and axial sections through the entire neurocranium to evaluate the brain for signs of hydrocephalus or additional associated malformations. In addition, if a child with a Chiari II malformation and a status post repair of a meningomyelocele presents with a worsening symptomatology, both cranial and spinal MRI should be performed to look for potential causes. On MR imaging, the posterior fossa appears unphysiologically small in patients with a Chiari II malformation. The cerebellar tonsils and often the cerebellar vermis are downwardly displaced into the foramen magnum and upper cervical spine (Fig. 3.2.22). The cerebellum may also wrap around the brainstem, which is best appreciated on axial sequences. The medulla oblongata is usually displaced downwardly, while the cervical myelon is relatively fixed in its position by ligaments, resulting in a characteristic medullary kinking. In addition, the quadrigeminal plate is commonly upwardly displaced resulting in a characteristic tectal “beaking” (Fig. 3.2.22). Both the pons and the fourth ventricle usually appear compressed. The pons loses its convex shape and is commonly squeezed against the clivus. The fourth ventricle is small and elongated and may assume a slit-like configuration. However, as a result of an altered CSF circulation,

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.22  Sagittal T2-weighted (a) and T1-weighted (b) sequences in a 3-year-old girl with a Chiari II malformation demonstrate a downward displacement of the cerebellar tonsils, as well as a medullary kinking and tectal beaking. The fourth ventricle appears compressed

the fourth ventricle may become trapped resulting in an encysted fourth ventricle. As the fourth ventricle is characteristically small in patients with Chiari II malformations, even a normal sized fourth ventricle is suspicious of an altered CSF passage (Barkovich 2005). Associated malformations are common in patients with a Chiari II malformation. Almost 80% of affected patients develop hydrocephalus. CSF shunting is commonly already performed in early infancy. Up to two thirds of patients have a concomitant hypogenesis of the corpus callosum. The falx cerebri commonly appears fenestrated with an interdigitation of gyri across the midline. The occipital horns of the lateral ventricles are often enlarged even after shunting. An enlarged supratentorial midline CSF space may also be present. Moreover, the medial parts of the occipital lobes often demonstrate multiple small gyri, a process called stenogyria. Occasionally, malformations of cortical development are seen as well. 3.2.4.8  Dandy-Walker Complex Dandy-Walker complex is characterized by cystic malformations of the posterior fossa. It is classically divided into true Dandy-Walker malformation, Dandy-Walker variant, and megacisterna magna (Barkovich et al. 1989b). However, the definition of Dandy-Walker variant has been subject to debate and the category itself has been questioned. The clinical presentation of affected children can be

quite variable. An association with a variety of chromosomal aberrations and syndromes has been described. Developmental delay, impaired hearing and/or vision and seizures are comparatively common presenting complaints. The severity of symptoms appears to be related to the presence of associated malformation, the presence and control of hydrocephalus and the degree of lobulation of the vermis (Barkovich 2005). MR imaging in children with a suspected malformation of Dandy-Walker complex should always include sagittal and axial sequences in order to evaluate the situation of the hindbrain and to search for signs of hydrocephalus and associated malformations. In patients with a Dandy-Walker malformation, there is an increased volume of the posterior fossa. The tentorium has a high insertion point. The fourth ventricle is dilated by a large cystic malformation. The volume of the cerebellar vermis is markedly reduced, with either a hypogenesis or an agenesis being present. The cerebellar hemispheres are hypoplastic to a variable degree. The pons is often compressed against the clivus with a reduction of its physiologic convexity. The term Dandy-Walker variant is usually defined as a cystic dilatation of the fourth ventricle with vermian hypogenesis in the absence of an enlargement of the posterior fossa (Fig. 3.2.23). However, other authors relate the term to the patency of the fourth ventricular outflow tracts. In patients with a megacisterna magna, there again is an increased volume of the posterior fossa, as the cisterna

213

214

3  Brain, Head, and Neck

Fig. 3.2.23  Sagittal (a) and axial (b) T2-weighted sequences in a 15-month-old girl with a Dandy-Walker variant demonstrate a cystic malformation in the posterior fossa that directly communicates with the fourth ventricle

magna is markedly enlarged. The fourth ventricle and the vermis, however, are normal in size and configuration. Associated malformations are common in children with a disorder of Dandy-Walker complex. Hydrocephalus is present in the vast majority of affected children at the time of diagnosis. In addition, agenesis or hypogenesis of the corpus callosum is present in about a third of affected patients. Other relatively frequent associated malformations include disorders of cortical development such as heterotopia or polymicrogyria, occipital cephaloceles, and the presence of syringohydromyelia. 3.2.4.9  Cerebellar Hypoplasia Cerebellar hypoplasia can sometimes be difficult to distinguish from cerebellar atrophy. However, in cerebellar hypoplasia the size of the cerebellar fissures is normal, whereas they appear enlarged in cerebellar atrophy. In addition, if progressive volume loss of the cerebellum is documented over time, the assumption of cerebellar atrophy as opposed to cerebellar hypoplasia must be made. There exist a variety of genetic, metabolic, inflammatory, and paraneoplastic causes for cerebellar hypoplasia (Barkovich 2005). Cerebellar hypoplasia is relatively often associated with hypoplasia of the pons, an association that is also called pontocerebellar hypoplasia. Causes and clinical presentation can be quite variable. However, two distinct syndromes have been identified, which are called pontocerebellar hypoplasia type 1 and type 2. In both types, the volumes of the cerebellum and of the pons are markedly reduced.

Type 1 pontocerebellar hypoplasia is commonly associated with hypotonia, developmental delay, progressive microcephaly, and abnormal eye movements. Concomitant callosal agenesis or hypogenesis and disorders of cortical developmental commonly occur. Type 2 pontocerebellar hypoplasia has an autosomal recessive mode of inheritance. Affected patients usually present with severe, progressive neurological deficits and epilepsy. Another cause of cerebellar hypoplasia is MarinescoSjögren syndrome, which has been mapped to chromosome 5q31 (Barkovich 2005). This syndrome is characterized by ataxia, with markedly reduced volumes both of the cerebellum and of the posterior fossa, mental retardation, congenital cataracts, and hypogonadism. There is also an X-linked form of cerebellar hypoplasia, which is non-progressive in nature. Affected boys suffer from cerebellar ataxia, dysarthria, and external ophthalmoplegia without signs of mental retardation. Hǿyeraal-Hreidarsson syndrome is also X-linked. Both cerebellum and pons are small. In addition, myelination is delayed and there is a volume loss of the supratentorial white matter as well. Patients are severely affected with a pronounced developmental delay, microcephaly, ataxia, spasticity, and pancytopenia. 3.2.4.10  Rhombencephalosynapsis In rhombencephalosynapsis, the cerebellar hemispheres are not completely separated, with a concomitant aplasia or hypoplasia of the vermis. In most instances, both the cerebellar hemispheres, the dentate nuclei, and the supe-

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

rior cerebellar peduncles are fused. Associated supratentorial malformations are common (Barkovich 2005). The clinical symptomatology of affected children can be quite variable, ranging from only mild abnormalities to severe defects. On MR imaging, the abnormalities are often most easily appreciated on coronal sequences. The characteristic hallmark feature of rhombencephalosynapsis is a fusion of the cerebellar hemispheres with cerebellar folia continuously crossing the midline, most commonly in the dorsal portion (Fig. 3.2.24). The vermis is absent or at least hypoplastic and the superior cerebellar peduncles are usually fused as well. Ventriculomegaly is common and an agenesis of the septum pellucidum is noted in about half of the cases. In addition to the cerebellar fusion, the thalami may appear merged as well. Structural alterations of the limbic system and olivary hypoplasia are frequently noted. Moreover, malformations of cerebral cortical development and premature synostoses of cranial sutures may occur. 3.2.4.11  Joubert Syndrome Joubert syndrome is characterized by a dysgenesis of the vermis with concomitant structural alterations of the cerebellar nuclei and of the pyramidal tracts. Affected children usually suffer from periods of hyperpnea, cerebellar ataxia, and disorders of the oculomotor functions (Kumandas et al. 2004). The disorder may actually be more heterogeneous than initially thought, with a subgroup of affected patients suffering from a chromosomal mutation (genetic locus 9q34.3), others having no signs of hyperpnea during infancy and yet other patients having various associated intracranial malformations, such as hypothalamic hamartomas, polymicrogyria, and other disorders of cortical development, and extracranial abnormalities, such as renal or hepatic malformations, limb abnormalities or ocular anomalies (Barkovich 2005; Kumandas et al. 2004). A variety of syndromal names is known for this latter group of patients depending on the associated anomaly. However, the MR imaging appearance of the hindbrain malformation is the same in all instances. The cranial MR imaging presentation of children with Joubert syndrome is very characteristic. It is helpful to acquire thin-section axial sections through the hindbrain, accompanied by coronal and sagittal sequences. The Joubert syndrome has also been called the molar tooth midbrain–hindbrain malformation. “Molar tooth” refers to the characteristic shape of the midbrain on axial sections resulting from enlarged, parallel superior cerebellar peduncles and a reduced volume of the midbrain itself. The diameter of the midbrain is small. The superior part of the fourth ventricle assumes a “bat-wing” appearance on axial sections. More caudally, the shape of the fourth ventricle is triangular resulting from the reduced size of the vermis. Sagittal sequences demonstrate a hypoplasia

Fig. 3.2.24  Axial T2-weighted sequences in a 3-year-old boy with rhombencephalosynapsis demonstrate a fusion of the cerebellar hemispheres

of the vermis with dysplastic folia and an unusually high position. The pontomesencephalic junction often appears small. In patients with an MR imaging appearance of Joubert syndrome associated malformations, such as disorders of cortical development, hypothalamic hamartomas, and ocular malformations should be specifically sought for. In addition, affected children should be screened for extracranial abnormalities as well. 3.2.4.12  Lhermitte-Duclos Syndrome Lhermitte-Duclos syndrome is also known as dysplastic cerebellar gangliocytoma. There has recently been evidence that many cases of Lhermitte-Duclos syndrome are associated with Cowden syndrome, a disease with multiple hamartomas in various organ systems and an increased frequency of malignancy, including breast cancer and thyroid cancer (Perez-Nunez et al. 2004). Symptoms from a dysplastic cerebellar gangliocytoma can arise at any age. However, it is not rare for the Lhermitte-Duclos syndrome to be discovered as an incidental finding on MR imaging. The MR imaging appearance of Lhermitte-Duclos syndrome is highly characteristic. The imaging protocol should include axial T2-weighted sequences and

215

216

3  Brain, Head, and Neck

Fig. 3.2.25  Sagittal (a) and coronal (b) T1-weighted sequences after the administration of paramagnetic contrast medium in a 48year-old woman with Lhermitte-Duclos syndrome demonstrate a cerebellar mass with curvilinear streaks of gray matter

T1-weighted sequences before and after the intravenous administration of paramagnetic contrast medium. On T2-weighted sequences, there is a characteristic cerebellar mass with a “zebra-like” pattern. Curvilinear streaks of gray matter are found to course through the mass, alternating with hyperintense areas (Fig. 3.2.25). The mass is usually sharply delineated. Contrast enhancement is rare; however, there have been reports of cases with pial or intraaxial enhancement. 3.2.5  Phakomatoses Phakomatoses are neurocutaneous disorders. They primarily affect structures of embryologically ectodermal origin, i.e., the nervous systems, the eyes, and the skin. Other organ systems may be affected as well, though. Five common, “classic” phakomatoses are known: neurofibromatosis type 1 (von Recklinghausen disease), neurofibromatosis type 2, tuberous sclerosis (BournevillePringle disease), Sturge-Weber syndrome (encephalotrigeminal angiomatosis), and von Hippel-Lindau disease (retinocerebellar angiomatosis). In addition, a multitude of comparatively rare phakomatoses has been described. 3.2.5.1  Neurofibromatosis Type 1 Neurofibromatosis type 1 is also known as von Recklinghausen disease. It is the most common phakomatosis, with an incidence of about 1:5,000. The mode of inheritance is autosomal dominant with a genetic locus on chromosome 17.

There are several important clinical criteria to diagnose neurofibromatosis type 1. According to the National Institutes of Health (NIH) Consensus development conference on neurofibromatosis, at least two of the following criteria need to be present in order to establish the diagnosis (NIH 1988): • Six or more café-au-lait spots (diameter >5 mm in children or >15 mm in adults) • Two or more neurofibromas or one plexiform neurofibroma • Axillary or inguinal freckling • Two or more Lisch nodules • Optic glioma • A distinctive osseous lesion, e.g., cortical thinning of long bones or sphenoid dysplasia • A first-degree relative with neurofibromatosis type 1 While café-au-lait spots are already common during infancy and early childhood, other manifestations such as cutaneous neurofibromas and Lisch nodules—being pigmented iris hamartomas—tend to appear in older children and adults. Cognitive deficits are present in about half of the affected patients. MR imaging of patients with a neurofibromatosis type 1 should always include T1- and T2-weighted sequences of the entire brain and dedicated, thin-section sequences through the anterior optic pathway. When imaging children with neurofibromatosis type 1, white matter abnormalities are commonly seen. These probably represent areas of dysplastic myelin—they are commonly called myelin vacuoles. These areas of myelin vacuolization are typically discerned on T2-weighted and FLAIR sequences and tend to “disappear” on T1-weighted

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.26  Axial T2-weighted (a) and FLAIR (b) sequences in a 5-year-old boy with neurofibromatosis type I demonstrate socalled myelin vacuoles

sequences. Typical locations include the splenium of the corpus callosum, the pallidum, the brain stem, and the cerebellar white matter (Fig. 3.2.26). These white matter abnormalities are usually not seen in newborns and infants. They commonly start to appear around the third year of life and diminish and finally disappear in the teenage years. It is important to distinguish these areas of myelin vacuolization from neoplastic lesions such as gliomas. This can be impossible to accomplish in a single MR study, as there are no completely accurate criteria to differentiate these two entities. Therefore, in cases with white matter abnormalities, a follow-up study including the administration of contrast medium should be performed six to twelve months after the initial study in order to rule out an increasing mass effect or a pathological contrast enhancement (Barkovich 2005). Areas of myelin vacuolization typically do not enhance and show no mass effect or perifocal edema. In addition, they usually cannot be discerned on T1-weighted imaging as opposed to T2-weighted and FLAIR sequences. Moreover, locations in the pons or mesencephalon, in the cerebellar white matter, in the globus pallidus or in the splenium of the corpus callosum are quite typical of myelin vacuoles. The incidence of cerebral astrocytomas is increased in patients with neurofibromatosis type 1 and these tu-

mors should always specifically be sought for on MR images. Most astrocytomas are pilocytic astrocytomas or low-grade gliomas, but high-grade gliomas may also occur. The most common location is the brainstem, especially the medulla oblongata and the midbrain. They may, however, also arise in the cerebral or cerebellar hemispheres. Generally, brainstem gliomas have a better prognosis in patients with neurofibromatosis type 1 than in the general population, even though the imaging presentations are identical. As mentioned above, cerebral gliomas need to be differentiated from areas of myelin vacuolization. Optic pathway gliomas are the most common gliomas in neurofibromatosis type 1. They occur in about 15% of patients with neurofibromatosis type 1, but become symptomatic in only about half of the cases. If symptomatic, they commonly cause visual disturbances. However, large optic pathway gliomas may also lead to a compression of the hypothalamus with subsequent endocrine dysregulations such as precocious puberty. Optic pathway gliomas can involve the optic nerves, the chiasm, and/or the optic tracts. Most of these are of the juvenile pilocytic subtype of astrocytomas. The prognosis of these tumors is good and they frequently even spontaneously involute over time. However, high-grade optic pathway gliomas occur.

217

218

3  Brain, Head, and Neck

As mentioned above, every MR examination in a patient with neurofibromatosis type 1 should include dedicated, thin-section sequences of the optic pathway. Optic gliomas can diffusely or focally involve the optic pathway. The affected area is enlarged and shows a profuse contrast enhancement. The optic nerves can be diffusely infiltrated by tumor or the tumor can show a subarachnoid spread around the nerve (Barkovich 2005). Care should be taken to distinguish non-enhancing dural ectasias of the optic nerve from spread of enhancing tumor tissue. Neurofibromas are nerve sheath tumors with a higher proportion of connective tissue than schwannomas. They are commonly found in spinal MR examinations of patients with neurofibromatosis type 1 and can affect all segments of the spine. Moreover, they are often found in the neck region. In children with neurofibromatosis type 1, they are much less commonly encountered than in adults. In addition, they are only rarely symptomatic in children. On MR imaging, spinal neurofibromas are most commonly found in an extradural location (Fig. 3.2.27). They can appear as small nodules of enhancing tissue, or, when larger, may widen the neural foramina. Neurofibromas typically demonstrate high signal intensity in the periph-

ery of the lesion in T2-weighted sequences with variable signal intensity in the center, giving the appearance of a target sign. On T1-weighted images, they usually appear slightly hyperintense to muscle. When evaluating the spine of a patient with neurofibromatosis type 1, it is always important to look for possible signs of cord compression, which may especially occur in the situation of bilateral neurofibromas on the same spinal level. In 3–18% of patients with neurofibromatosis type 1, neurofibrosarcomas (i.e., malignant neurofibromas) are found (Barkovich 2005). A neurofibrosarcoma is usually suspected on clinical grounds with increasing pain or neurological deficits. On MR imaging, however, there are no specific signs to differentiate a neurofibrosarcoma from a benign neurofibroma, even though neurofibrosarcomas tend to be more heterogeneous in signal intensity and contrast enhancement. Plexiform neurofibromas usually grow diffusely along a nerve. They are locally aggressive tumors with an unorganized intracellular matrix. On cranial MR imaging, they are commonly found in the periorbital region, in the temporal region or in the masticator space. Plexiform neurofibromas are usually more heterogenous and grow more profusely than solitary neurofibromas. On

Fig. 3.2.27  Coronal (a) and axial (b) T1-weighted sequences after administration of paramagnetic contrast medium in a 39year-old man with neurofibromatosis type 1 demonstrate numerous neurofibromas

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

T2-weighted sequences, their signal intensity is usually high. Enhancement after the administration of contrast medium can be variable. Patients with neurofibromatosis type 1 additionally have a higher incidence of dysplasias of the cerebral and extracranial vessels. Proliferations of the vascular intima may lead to arterial stenoses, especially in the circle of Willis, but also in the carotid arteries. A Moya Moya pattern of the lenticulostriate arteries may ensue (Barkovich 2005). In addition, aneurysms and arteriovenous malformations are found with a higher incidence. MR angiography can be helpful in detecting these vascular abnormalities. Another typical imaging feature of neurofibromatosis type 1 is the presence of osseous lesions of the orbits and/ or of the calvarium. Most commonly, sphenoid dysplasias are found. In addition, osseous defects may be found. In addition to the above-described neurofibromas, other malformations of the spine may be encountered as well. Dysplasias of the vertebral bodies and of the pedicles and transverse and spinous processes are quite common. These dysplasias commonly lead to the situation of a dextroscoliosis. If a child, however, presents with a levoscoliosis, other possible causes should be sought for as well (Barkovich 2005). Spinal lipomas, intramedullary spinal tumors, such as astrocytomas, and syringohydromyelia are all found with an increased frequency in patients with neurofibromatosis type 1. Scalloping of the posterior aspects of the vertebral bodies is another commonly encountered finding in patients with neurofibromatosis type 1. In addition, the incidence of dural ectasias and lateral meningoceles is increased. In contrast to neurofibromas, these dural ectasias are isointense to CSF and do not enhance after the administration of contrast medium. 3.2.5.2  Neurofibromatosis Type 2 Neurofibromatosis type 2 is a much less frequently occurring disease than is neurofibromatosis type 1. Its mode of inheritance is autosomal dominant with a genetic locus on chromosome 22. In many instances, neurofibromatosis type 2 is not diagnosed until adulthood. It has also been called neurofibromatosis with bilateral acoustic schwannomas, which already points to one of the hallmark features of the disease. Hearing loss is among the most common presenting complaints in adults with neurofibromatosis type 2. Affected children tend to rather present with facial palsy or epilepsy. Skin manifestations, such as café-au-lait spots, are less common than in neurofibromatosis type 2 than in neurofibromatosis type 1. According to the NIH Consensus development conference on neurofibromatosis, at least one of the following criteria need to be present in order to establish a definite diagnosis of neurofibromatosis type 2 (NIH 1988):

• Bilateral acoustic schwannomas demonstrated by contrast-enhanced MRI • A first-degree relative with neurofibromatosis type 2 plus either – A unilateral acoustic schwannoma at less than 30 years of age or – The presence of two of the following: – Neurofibroma – Glioma – Schwannoma – Meningioma – Juvenile cataract   (posterior subcapsular or cortical) When neurofibromatosis type 2 is suspected, MR imaging should always include T1-weighted sequences before and after the administration of contrast medium covering the entire brain. In addition, thin sections through the cerebellopontine angle should be acquired. As the presence of bilateral acoustic schwannomas is a disease-defining criterion in and by itself, the radiologist may be the first to make the diagnosis. In addition, if a schwannoma or a meningioma is encountered in a child or young adult, neurofibromatosis type 2 should always be suspected and other manifestations of the disease should be specifically sought. These include schwannomas of other cranial nerves, such as the fifth or the seventh cranial nerve, meningiomas, or gliomas (Fig. 3.2.28). In addition, spinal manifestations are common. As in the neurocranium, schwannomas and to a lesser extent neurofibromas are commonly encountered, especially in adulthood. These nerve sheath tumors can occur in an intraspinal, intraforaminal or paraspinal location. They are usually slightly hyperintense to the spinal cord and markedly enhance after the administration of contrast medium. The incidence of intraspinal meningiomas is also distinctly increased, especially in the thoracic region. Moreover, intramedullary tumors, such as gliomas, ependymomas, and intramedullary schwannomas are potential manifestations of the disease. 3.2.5.3  Tuberous Sclerosis Tuberous sclerosis—also called Bourneville-Pringle disease—is an autosomal dominant disorder with two distinct genetic loci, namely on chromosome 9 (9q34) and on chromosome 16 (16p13.3). Mutations in either gene appear to produce an almost identical phenotype with only relatively minor differences (Barkovich 2005). The classically described clinical triad of tuberous sclerosis is that of mental retardation, adenoma sebaceum, and epilepsy. However, mental retardation is only noted in about half and epilepsy in about three fourth of affected patients. Adenoma sebaceum is usually only encountered in older children or adults.

219

220

3  Brain, Head, and Neck

Fig. 3.2.28  Axial T1-weighted sequences after administration of paramagnetic contrast medium in a 24-year-old man with neurofibromatosis type 2 demonstrate bilateral acoustic schwannomas and a trigeminal schwannoma

The Tuberous Sclerosis Complex Consensus Conference has issued diagnostic criteria for the disease (Roach et al. 1998). To establish a definite diagnosis of tuberous sclerosis, either two major features or one major and two minor features must be present. A probable diagnosis of tuberous sclerosis is made, when one major and one minor feature are present, a possible diagnosis, when one major feature or two minor features are found. The following criteria have been defined as major features of the disease (Roach et al. 1998): • Cortical tubers • Subependymal nodules • Subependymal giant cell astrocytoma • Facial angiofibromas • Ungual or periungual fibromas • Shagreen patch • More than three hypomelanotic macules (white spots) • Retinal nodular hamartomas • Cardiac rhabdomyoma • Lymphangioleiomyomatosis • Renal angiomyolipoma If the latter two criteria are present, then other criteria are required to make a definite diagnosis. Minor features of the disease are defined as follows (Roach et al. 1998): • Affected first-degree relative

• • • • • • • • •

Radial migration lines in the cerebral white matter Gingival fibromas Dental enamel pits in a random distribution Retinal achromic patch “Confetti” skin lesions Hamartomatous rectal polyps Non-renal hamartomas Multiple renal cysts Bone cysts

When looking at these disease-defining criteria, it becomes clear that MR imaging plays a major role in diagnosing tuberous sclerosis. The presence of cortical tuber and of subependymal nodules are both major features of the disease—the combination of these two features therefore suffices to establish the diagnosis. Cranial MR imaging in patients with suspected tuberous sclerosis should always include FLAIR sequences and pre- and post-contrast T1-weighted sequences, preferably in a coronal orientation. FLAIR sequences are especially sensitive in the detection of cortical tubers, which represent cerebral hamartomas (Fig. 3.2.29b). These tubers may histologically contain balloon cells and therefore resemble the above described focal cortical dysplasias with balloon cells (transmantle cortical dysplasias) (Mackay et al. 2003). Cortical tubers are typically focal, cortex-based lesions with a hyperintense signal on FLAIR and T2-weighted sequences

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.29  Axial T1-weighted IR (a) and FLAIR (b) sequences in an 11-year-old girl with tuberous sclerosis demonstrate multiple subependymal nodules that are not isointense to gray matter as well as multiple cortical tubers

and a hypointense signal on T1-weighted sequences. In older children and adults, cortical tubers may calcify and sometimes become relatively hyperintense on T1-, and hypointense on T2-weighted sequences. In infants and neonates with largely unmyelinated white matter, however, the tubers appear hypointense on T2-weighted images and hyperintense on T1-weighted images as compared to the surrounding unmyelinated white matter. The number of cortical tubers may widely vary from patient to patient. There may be only very few or up to 20 and more cortical tubers. It has been subject of discussion, whether the number of cortical tubers correlates with the prognosis of the child with regard to the presence of mental retardation and seizures with some authors favoring a correlation (Shepherd et al. 1995), while others dispute it (Takanashi et al. 1995). Despite this apparent uncertainty, it appears sensible to report the approximate number of cortical tubers when evaluating a cranial MR examination of a child with tuberous sclerosis and to perform sequences that are sensitive for their detection such as FLAIR or magnetization transfer sequences. Another hallmark feature of tuberous sclerosis is the presence of subependymal nodules, which are small, nodular hamartomas in a subependymal location (Fig. 3.2.29a). These hamartomas are usually found along the lateral walls of the lateral ventricles. They may be nodular or slightly irregular in shape and tend to protrude into the ventricular lumen. Subependymal hamartomas are

usually mostly isointense with myelinated white matter and are most easily discerned on T1-weighted sequences. Calcification commonly occurs with advancing age leading to areas of signal loss on T2- and T2*-weighted sequences. In neonates and young infants with a largely unmyelinated white matter, subependymal hamartomas appear relatively hypointense on T2-, and hyperintense on T1-weighted sequences. The most important differential diagnosis to subependymal nodules in tuberous sclerosis are subependymal heterotopia. However, heterotopia are always isointense to gray matter on all sequences, while subependymal nodules have variable signal intensity with mostly isointensity to white matter (Barkovich 2005; Barkovich and Kuziecky 2000). After the administration of contrast medium, subependymal nodules may show a variable degree of enhancement. This enhancement, however, does not aid in distinguishing subependymal nodules from giant cell astrocytomas. Giant cell tumors are mostly situated in or near the foramen Monroi, but may occur in other subependymal regions as well. They are found in about 10% of patients with tuberous sclerosis. In contrast to subependymal nodules, giant cell tumors have a tendency to enlarge over time and to lead to a disturbance of CSF passage in the foramen Monroi with subsequent hydrocephalus of the respective lateral ventricle. On MR imaging, it can

221

222

3  Brain, Head, and Neck

Fig. 3.2.30  Axial FLAIR sequences (a) and sagittal T1-weighted (b) sequences after administration of paramagnetic contrast medium in a 38-year-old male patient with tuberous sclerosis demonstrate a giant cell tumor in the left foramen Monroi

be difficult to distinguish simple subependymal nodules from giant cell tumors. Neither the signal intensity nor the presence or degree of contrast enhancement provides a reliable criterion. Subependymal nodules with a diameter of 12 mm or more are generally suggestive of a giant cell tumor (Fig. 3.2.30). However, the most reliable criterion seems to be the demonstration of growth over time on serial MR examinations (Barkovich 2005). Rarely, giant cell tumors may degenerate into higher-grade tumors that tend to infiltrate the surrounding brain parenchyma. In addition, giant cell tumors may also occur in patients without other signs of tuberous sclerosis. White matter lesions are also commonly encountered in patients with tuberous sclerosis. These represent clusters of heterotopic dysplastic neuronal cells within the white matter. Their imaging appearance is identical to that of cortical tubers and they may also calcify over time. These white matter lesions are commonly observed to transverse the entire hemisphere in a radial fashion. As in cortical tubers, these lesions are typically best discerned within the myelinated white matter on FLAIR sequences. In addition to the abovementioned classical features, cerebral parenchymal cysts, cerebellar tubers and vascular lesions, such as aneurysms, have been described in patients with tuberous sclerosis. 3.2.5.4  von Hippel-Lindau Disease The von Hippel-Lindau disease—also called CNS angiomatosis—is an autosomal dominant disorder with in-

complete penetrance and a genetic locus on chromosome 3 (3p25–p26) (Latif et al. 1993). The disease is usually not diagnosed until the teenage years or early adulthood, when cerebellar or spinal symptoms or visual disturbances are common complaints. von Hippel-Lindau disease is characterized by the presence of hemangioblastomas, which most commonly occur in a cerebellar or spinal location, and of retinal angiomas. In addition, papillary cystadenomas of the endolymphatic sac commonly occur. Extracranial manifestations of the disease include an increased incidence of renal cysts, renal cell carcinoma, pheochromocytoma, and papillary cystadenomas and cysts of the testis. MR imaging in patients with von Hippel-Lindau disease should always include pre- and post-contrast sequences of the entire brain and spine. In addition, thin sections through the petrous bone should be performed, if a history of hearing loss is elicited. Hemangioblastomas typically present as fluid-filled cystic structures with an enhancing, vascular mural nodule in the majority of cases (Fig. 3.2.31). However, in up to one third of cases only a solid tumor is found, which may subsequently grow and develop a cystic component. The content of the cyst is usually isointense to CSF. Occasionally, hemorrhage may develop, though, leading to signal alterations within the cyst. The solid portion of the tumor is well vascularized and demonstrates an enhancement on post-contrast studies. As mentioned above, hemangioblastomas most commonly occur in the cerebellum and in a spinal location (Fig. 3.2.32). They may, however, also be found in the brainstem or in the cerebral hemispheres.

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.31  Axial T2-weighted sequences (a) as well as axial (b) and coronal (c) T1-weighted sequences after administration of paramagnetic contrast medium in a 40-year-old female patient with von Hippel-Lindau disease demonstrate a hemangioblastoma of the posterior fossa

Retinal angiomas are usually diagnosed during an ophthalmologic exam. Due to their small size, they are not reliably detected on MR imaging. Secondary effects, such as retinal detachment, may, however, be noted. Papillary cystadenomas of the endolymphatic sac occur in about 7% of patients with von Hippel-Lindau disease (Mukherji et al. 1997). Affected patients most commonly present with unilateral hearing loss. On MR

imaging, the tumor typically has highly heterogeneous signal intensity. Contrast enhancement is variable as well and flow voids may be present. 3.2.5.5  Sturge-Weber Syndrome Sturge-Weber syndrome represents an encephalotrigeminal angiomatosis, which preferentially involves the sen-

223

224

3  Brain, Head, and Neck

Fig. 3.2.32  Sagittal T2-weighted sequences (a) and sagittal T1-weighted sequences (b) after administration of paramagnetic contrast medium in a 25-year-old male patient with von HippelLindau disease demonstrate a spinal hemangioblastoma

sory region of the trigeminal nerve, the meninges, and the choroid of the eye. No clear hereditary pattern has been described, even though some familial cases exist. Affected patients most commonly present with a facial nevus flammeus (port wine stain), which is composed of capillary-like vessels. However, cases without evidence of a facial nevus have been reported. In addition, an intracranial pial angioma is present, which again is composed of thin-walled vessels and venous channels. Angiomas of the choroid and sclera are noted in about one third of patients (Barkovich 2005). Clinically, most patients with Sturge-Weber syndrome present with seizures that are often difficult to control with medication. In addition, hemiparesis and/or hemianopsia as well as developmental delays may ensue. MR imaging in patients with Sturge-Weber syndrome should always include T2- and T2*-weighted sequences as well as pre- and post-contrast T1-weighted sequences. In neonates and small infants with a largely un­ myelinated white matter, T2-weighted sequences commonly demonstrate a relative hypointensity in the white matter adjacent to the angioma, the exact cause of which has not been determined. This signal alteration may represent an accelerated myelination or an increased presence of deoxyhemoglobin (Barkovich 2005). With increasing

age, calcifications of the subcortical white matter and of the cortical region directly adjacent to the angioma commonly ensue, presumably resulting from chronic venous ischemia (Fig. 3.2.33). These calcifications are best demonstrated in T2*-weighted sequences, in which they typically present as linear hypointensities. Over time, atrophy of the affected regions commonly ensues. Post-contrast T1-weighted sequences usually demon­ strate a pronounced enhancement of the subarachnoid space in the region affected by the angioma (Fig. 3.2.33). However, cases of “burned-out” angiomas have been described, in which only calcifications and atrophy, but no enhancement were present (Fischbein et al. 1998). In addition to the leptomeningeal enhancement, an enlargement of the ipsilateral choroid plexus is commonly noted, probably resulting from an increased venous drainage. About one third of patients with Sturge-Weber syndrome also present with angiomas of the choroid or sclera of the eye. This can lead to glaucoma, buphtalmos, or retinal detachment. On MRI, the posterior wall of the globe typically appears thickened with a pronounced enhancement in cases of choroid angioma. Fat-suppressed, contrast-enhanced, T1-weighted sequences are especially helpful in this setting.

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.33  Axial T2-weighted sequences (a) as well as axial (b) and coronal (c) T1-weighted sequences after administration of paramagnetic contrast medium in a 9-year-old girl with SturgeWeber syndrome demonstrate a leptomeningeal angioma with concomitant calcifications and atrophy

3.2.5.6  Other Phakomatoses There is a multitude of other, far less frequent neurocutaneous disorders (Barkovich 2005). To discuss these in detail would by far exceed the scope of this book chapter. The following text therefore provides only a very brief outline, with the most characteristic imaging features. Hypomelanosis Ito—also called incontinentia pigmenti achromians—is one of the more common neurocutaneous syndromes. MR imaging of the brain can be completely normal. However, disorders of cortical development, such as polymicrogyria, heterotopia or hemimegalencephaly, as well as cerebral atrophy may occur. Ataxia telangiectasia is an autosomal recessive disorder (genetic locus chromosome 11q22–23). Characteris-

tic MR imaging features include an atrophy of the cerebellum, including both vermis and cerebellar hemispheres. In addition, hemorrhage may occur from teleangiectatic lesions. Moreover, pulmonary vascular malformation can lead to cerebral emboli resulting in cerebral infarcts. Nijmegen breakage syndrome again is an autosomal recessive disorder (genetic locus chromosome 8q21). MR imaging demonstrates microcephaly and hypogenesis of the corpus callosum in about half of the cases. Affected children have DNA repair defects predisposing to lymphoma, leukemia, and infection. Exposure to ionizing radiation should be avoided. Gorlin syndrome—also referred to as basal cell nevus syndrome—is characterized by basal cell carcinomas and other neoplasms that arise at a very young age. On neuro-

225

226

3  Brain, Head, and Neck

imaging, typical features include odontogenic cysts and early calcifications of the dura. The incidence of medulloblastoma is increased in this patient population. Epidermal nevus syndrome is characterized by malformations of cortical development such as polymicrogyria or hemimegalencephaly. Hemiatrophy of the brain, intracranial lipomas, or vascular malformations may also be observed. Patients with an epidermal nevus syndrome also have an increased incidence of secondary lesions, such as porencephaly or cerebral infarcts. Ocular malformations, such as microphthalmos, colobomas, or retinal dysplasia may be seen as well. Incontinentia pigmenti is also known as Bloch-Sulzberger syndrome. MR imaging features are variable and include hemodynamic cerebral infarcts, cerebral atrophy, and ocular malformations. However, MR imaging of the brain may also be completely normal. Neurocutaneous melanosis is characterized by multiple congenital nevi and often-giant melanocytic nevi. On MR imaging of the brain and spine, regional areas of hyper­intensity are noted on T1-weighted sequences, which correspond to melanin deposits. These may degenerate into malignant melanoma. Parry-Romberg syndrome is also called progressive facial hemiatrophy. There is hemiatrophy of both the skin and the subcutaneous tissue in the face. Ipsilateral dilatation of the lateral ventricle and cortical calcifications may be seen on cranial imaging studies. In addition, cysts may occur in the ipsilateral hemisphere and in the corpus callosum and the sulcal pattern may be abnormal.

The term PHACES syndrome is an acronym that stands for posterior fossa malformations, facial hemangiomas, arterial anomalies, cardiac anomalies and aortic coarctation, eye anomalies and sternal clefting/supraumbilical raphe. On MR imaging of the brain, hypoplasia of a cerebellar hemisphere or a disorder of Dandy-Walker complex may be seen. In addition, vascular anomalies and/or aberrant arteries, such as a persistent trigeminal artery or an absence of the carotid or vertebral arteries are common. These may involve any part of the cervicocranial vasculature. 3.2.6  Hypoxic–Ischemic Injuries to the Pediatric Brain There are several quite distinctive patterns of injury in the immature brain following hypoxic–ischemic events that fundamentally differ from those patterns observed in the adult brain. This difference is due both to the pattern of development of the human brain, with a shift in watershed territories and metabolic activity and to the brain´s increasing ability to mount an astrocytic response and therefore form gliotic scar tissue. In the following section, only the patterns of injury characteristic to the pediatric brain should be outlined, with a special focus on injuries in preterm and term infants. Other patterns of injury, such as territorial infarcts or generalized hypoxia in an older child will not be outlined here, as the MR imaging characteristics in these conditions resemble those of adults.

Fig. 3.2.34  Axial T2-weighted (a) and coronal FLAIR (b) sequences in a 12-month-old girl with porencephaly demonstrate a CSFisointense defect directly adjacent to the left lateral ventricle

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Several terms are used to describe different patterns of damage to the developing brain. The term porencephaly is generally used for an injury that results in the formation of a cavity with no or minimal perifocal gliosis. Concomitant disorders of cortical development may be present, if the injury occurred early during in utero development, i.e., usually prior to the 26th week of pregnancy. Some authors use the term encephaloclastic porencephaly for the above-described cavity in order to differentiate it from agenetic porencephaly corresponding to schizencephaly (Barkovich 2005). On MR imaging, encephaloclastic porencephaly usually presents as a CSFisointense defect in the brain parenchyma. No or only minimal perifocal gliosis is present. The porencephalic

cavity commonly lies adjacent to the lateral ventricles and may communicate with them (Fig. 3.2.34). Hydranencephaly may be considered an extreme form of porencephaly. The term is used when the majority of the supratentorial brain has been afflicted by the injury and has subsequently been resorbed. The etiology of hydranencephaly has been subject to debate. In most instances, a generalized, profound hypoxic-ischemic injury to the brain is the most likely etiology. However, genetic and infectious causes have been proposed as well (Barkovich 2005). On MR imaging, hydranencephaly presents as an extensive defect to the supratentorial brain (Fig. 3.2.35). Large parts of the cerebrum are destroyed and replaced

Fig. 3.2.35  Axial T2-weighted (a) and T1-weighted (b) sequences as well as sagittal (c) T2-weighted sequences in a 12-month-old boy with hydranencephaly demonstrate a nearly complete destruction of the supratentorial brain with extensive CSF-isointense defects

227

228

3  Brain, Head, and Neck

Fig. 3.2.36  Axial T2-weighted (a) and T1-weighted (b) sequences in a 5-month-old boy with a status post infarct of the right middle cerebral artery (MCA) demonstrate encephalomalacia in the territorial region of the infarct

by CSF. Most commonly, parts of the inferior and medial frontal lobes as well as the inferior parts of the temporal lobes, the thalami, and the cerebellum are preserved. The volume of the brain stem is usually reduced, most likely due to Wallerian degeneration. The term encephalomalacia is applied to an injury that is characterized by astroglial proliferation and often subseptations. This pattern of destruction arises, when an injury occurs later in gestational development or around term. When multiple cystic cavities of variable size are found, the process is commonly referred to as multicystic encephalomalacia. Encephalomalacia may present in a localized or in a diffuse fashion. When a diffuse or non-specific pattern is found, the etiology may be infectious (Barkovich 2005). Moderate hypotension tends to lead to encephalomalacia in the hemodynamic watershed areas of the brain, while profound hypotension may lead to an injury of the basal ganglia or of large parts of the cerebral white and gray matter, depending on the gestational age and the severity of the injury. Thrombembolic infarctions instead generally lead to encephalomalacia in the territorial region of the occluded vessel (Fig. 3.2.36). On MR imaging, encephalomalacia typically presents as often multiple cavities with septations. In contrast to porencephalic defects, signs of reactive gliosis are present, which are usually best appreciated on FLAIR sequences.

3.2.6.1  Hypoxic–Ischemic Injury in the Premature Infant For a variety of reasons, premature infants generally are at a higher risk for cerebral hypoxic-ischemic injury than infants born at term are. These include the relative immaturity of the lung, the absent or reduced cerebral autoregulation, and the sensitivity of the developing, immature brain. Moreover, the brain of the premature infant reacts with a different pattern of injury than the brain at term. A common pattern of injury in the premature infant selectively involves the white matter adjacent to the lateral ventricles. It is commonly referred to as periventricular leukomalacia. However, other areas of the white matter may be affected as well and causes are not purely hypoxic-ischemic. Therefore, the term white matter injury of prematurity may be more recommendable (Barkovich 2005). Clinically, premature infants affected by a hypoxicischemic cerebral injury most commonly present with so-called spastic diplegia. This is a condition, in which spasticity develops that is typically more pronounced in the lower than in the upper extremities. Concomitant visual disturbances are common as well. The cause for this characteristic clinical picture lies in the pronounced sensitivity of the directly periventricular white matter to hypoxic events in the premature infant. Several reasons have been proposed to explain the heightened sensitivity of precisely this region in prema-

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

ture infants. Most commonly, this area is defined as an early watershed area of the brain that later changes in location when maturity progresses. The sensitivity of this region to hypoxia may be exacerbated by a limited capacity for vasodilation. In addition, the periventricular white matter is a site of oligodendrocytic proliferation and therefore of a highly metabolically active region requiring large amounts of oxygen. Another, less common pattern of hypoxic-ischemic injury in premature infants is the so-called pontosubicular necrosis. It is usually seen in children that are also affected by periventricular leukomalacia. In pontosubicular necrosis, the ventral part of the pons as well as the subiculum and fascia dentata of the hippocampus are characteristically damaged (Barkovich 2005). As outlined under Sect. 3.3.1, the germinal matrix zone is a zone of neuronal stem cell proliferation that lies directly adjacent to the ventricular ependyma. It is particularly active from the 8th to the 28th week of gestation and slowly involutes until the 34th week of gestation. When this zone is hypoperfused, a periventricular and intraventricular hemorrhage commonly ensues, which is subdivided into four grades: • In grade I hemorrhage, only a hemorrhage of the germinal matrix zone is seen without significant amounts of intraventricular blood. • Infants with a grade II hemorrhage demonstrate a germinal matrix hemorrhage with intraventricular hemorrhage; however, the ventricles are not dilated. • Grade III hemorrhage is referred to as a condition in which there is both periventricular and intraventricular hemorrhage with subsequent ventricular dilatation. • Grade IV hemorrhage is considered to stem from a different pathophysiological principle. It is most likely the result from venous infarctions with subsequent hemorrhage. Hemorrhage is found in the deep white matter. The outcome of the affected children worsens with increasing grade. If a premature infant suffers a circulatory arrest or very severe hypotension, the pattern of injury may be different with the thalami, the anterior vermis, and the dorsal brainstem being most commonly affected. Concomitant germinal matrix hemorrhage or leukomalacia may occur as well, though. The diagnostic radiologist may be confronted with hypoxic–ischemic injuries in preterm infants at different stages. In the acute setting, the primary diagnostic method of choice is ultrasonography. However, MR imaging is increasingly being performed in preterm infants even at an early stage, as both imaging techniques and methods of monitoring are increasingly becoming more refined (American Academy of Pediatrics Committee on Drugs 2002). In addition, the radiologist may encounter

this group of patients at a later stage or even as adults, when the changes stemming from an early injury need to be correctly interpreted. MR imaging in the acute setting should always include T2- and T1-weighted images as well as T2*-weighted sequences. In older children and adults, FLAIR sequences should be included as well. In the acute setting, diffusionweighted sequences with ADC maps can be helpful in addition. In the subacute setting, germinal matrix zone hemorrhage can commonly be seen as hyperintense material on T1-weighted sequences (corresponding to extracellular met-hemoglobin) and areas of markedly low signal intensity on T2- and T2*-weighted sequences. In the hyperacute setting within the first 3 days, the hemorrhage usually appears iso- to hypointense on T1-weighted images, before slowly becoming hyperintense. This T1 hyperintensity subsequently diminishes over the next 3 weeks. As outlined above, special note should be made of the presence of intraventricular blood and the presence and extent of ventricular dilation. Ventricular enlargement can be the result of hydrocephalus or of an ex vacuo dilation due to a destruction of white matter. In addition to assessing peri- and intraventricular hemorrhage, a potential pontosubicular necrosis should be specifically sought. Moreover, hemorrhage of the choroid plexus is commonly found. Periventricular hemorrhagic infarctions present as areas of hemorrhage in the deep white matter following the signal characteristics outlined above. When the blood is resorbed, a porencephalic defect adjacent to the lateral ventricle remains. White matter injury of prematurity (periventricular leukomalacia) is commonly primarily diagnosed on sonography, where it typically presents as periventricular flares with variable degrees of echogenicity depending on the sonographic grade. On MR imaging, punctate areas of slightly hyperintense signal on T1-weighted images and larger areas of relative hyperintensity on T2-weighted images are commonly seen in the acute setting. These signal alterations are considered to represent areas of reactive astroglial proliferation (Felderhoff-Mueser et al. 1999). Cavitations may or may not develop. Over the following weeks to months, a gradual volume loss of the periventricular white matter is observed resulting in the characteristic findings of end-stage leukomalacia. In a routine clinical setting, the radiologist is much more commonly confronted with this end-stage picture of periventricular leukomalacia than with the above-described acute findings. MR imaging typically demonstrates an irregular contour of the lateral ventricles, especially of the corpora and trigones with a concomitant volume loss of the periventricular white matter (Fig. 3.2.37). In severe cases, sulci may closely approximate the lateral ventricles due to a pronounced loss of white matter. On FLAIR and T2-weighted sequences, areas of hyperintensity are found

229

230

3  Brain, Head, and Neck

Fig. 3.2.37  Axial T2-weighted (a), FLAIR (b), and T1-weighted IR (c) sequences in a 3-year-old boy with a status post–white matter injury of prematurity demonstrate an irregular contour of the lateral ventricles with a dilatation and adjacent gliosis

adjacent to the lateral ventricles, especially in the peritrigonal region, if the injury occurred at a time when the developing brain was already able to mount a significant astroglial response (Fig. 3.2.37). In contrast to zones of terminal myelination, these periventricular signal alterations lie directly subependymal, adjacent to the lateral ventricles. In very early white matter injuries, this signal alteration may be absent, while the characteristic configuration of the ventricles will be present. While the above-described periventricular signal alterations are quite characteristic for white matter injury of prematurity, they are by no means specific, as a variety of other disorders may cause an analogous picture. As parts

of the white matter are destroyed as a result of the periventricular leukomalacia, the corpus callosum is typically thinned. Most commonly, the isthmic and peri-isthmic regions demonstrate the most pronounced volume loss. In addition to the above-described patterns of supratentorial injury, cerebellar injury is probably more common in the preterm infant than previously thought. Up to 8% of premature infants are found to have focal or diffuse cerebellar injury (Mercuri et al. 1997). However, many of these children are asymptomatic and only about half of these cases have associated supratentorial hemorrhage. Larger areas of hemorrhage may evolve into focal or diffuse cerebellar atrophy.

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

3.2.6.2  Hypoxic–Ischemic Injury in the Term Infant At term, the brain of an infant reacts differently to hypoxic–ischemic injuries than in the premature setting. This altered pattern may in part be due to a shift in the watershed areas of the brain. In addition, the pattern of metabolic activity significantly changes during fetal cerebral development. As with premature infants, the pattern of injury depends on the severity and duration of the hypoxia/hypotension. In term neonates suffering from mild to moderate hypotension/hypoxia at the time of birth, the most commonly injured areas are the cerebral watershed areas, i.e., the regions between the territories of the anterior and middle cerebral arteries and between the middle and posterior arteries. Therefore, a parasagittal pattern of astrogliosis is quite typical. Affected patients commonly present with variable degrees of spasticity of the proximal limbs and cognitive impairment. MR imaging in the acute setting should include MR spectroscopy and diffusion-weighted sequences with ADC maps. MR spectroscopy typically demonstrates an elevated lactate in the affected regions, while NAA may be reduced (Barkovich 2005). Diffusion-weighted imaging shows a restricted diffusion with reduced ADC values in the affected regions. Within the first hours after the injury, an underestimation of the extent of injury or a falsenegative study is possible, however. Therefore, a repeat MR study may become necessary. In the further course, edema in the affected regions will become noticeable on T2- and T1-weighted images with a hyperintense signal on T2-, and a hypointense signal on T1-weighting. In the end-stage, gliosis and volume loss develop. As a secondary effect, the overlying gyri tend to appear mushroomshaped, a pattern that has been referred to as ulegyria. In term infants suffering from circulatory arrest or severe hypotension/hypoxia, a different pattern of injury typically evolves. In these children, the most metabolically active regions of the brain are typically damaged. Therefore, the injuries are most commonly found in the lateral thalami, the posterior parts of the putamina, the corticospinal tracts and in the hippocampi. In severe cases, additional injury may be found in the brainstem nuclei and in the optic radiations. If affected children survive the profound hypoxia, neurologic sequelae typically include choreoathetosis, spasticity of the limbs, cognitive impairment, and epilepsy. Symptoms may not evolve or not be noticed until later in development, sometimes as late as school age (Barkovich 2005). MR imaging in the acute setting, i.e., in the first 24 h after birth, should include MR spectroscopy, as this technique may show elevated lactate levels in the affected regions as early as several hours after the injury (Barkovich et al. 2001c). However, MR spectroscopy tends to tran-

siently normalize after about 24 h. Diffusion-weighted imaging with ADC maps typically demonstrates a restricted diffusion with low ADC values; however, falsenegative results are possible within the first hours, and the diffusivity usually normalizes within about a week (Barkovich et al. 2001c). Conventional sequences usually become positive after about one to two days. However, care must be taken to specifically evaluate the most commonly affected structures as subtle changes are commonly overlooked, even though they serve as an important prognostic indicator. An abnormally high signal is usually found in the affected structures, especially in the lateral thalami, in the posterior putamina and in the perirolandic areas (Fig. 3.2.38). Signal in T1-weighted images may be heterogeneous with foci of hyperintensity. In the end-stage, atrophy of the affected region evolves and areas of gliosis develop (Fig. 3.2.38). When evaluating children who underwent profound hypotension/hypoxia at the time of term birth, it is important to remember that changes on MR imaging may be quite subtle, but highly prognostically relevant. Even very small lesions of the lateral thalami and dorsal putamina may cause quite severe symptoms and need to be reported. 3.2.6.3  Kernicterus Kernicterus is also called bilirubin encephalopathy. It represents a brain injury that arises in the setting of severe neonatal hyperbilirubinemia. There are a variety of causes of hyperbilirubinemia, the most common being hemolysis, e.g., secondary to blood-group incompatibility or hemoglobin defects. The consecutive damage most commonly occurs in the globi pallidi, in the subthalamic nuclei and in the hippocampi. Clinically, affected infants present with initial hypotonia and consecutive hypertonia with backward arching. Developmental delays and extrapyramidal symptoms ensue (Barkovich 2005). MR imaging in the acute, neonatal setting typically demonstrates an increased signal both in T1- and in T2-weighted sequences in the globi pallidi and subthalamic nuclei (Govaert et al. 2003). In older children who underwent kernicterus as neonates, atrophy and gliosis with a hyperintense signal on T2-weighted images are commonly found in the globi pallidi, subthalamic nuclei, and hippocampi (Fig. 3.2.39). The signs may, however, be quite subtle and should specifically be sought when a history of bilirubin encephalopathy exists. 3.2.7  Metabolic Diseases of the Pediatric Brain Metabolic disorders of the brain constitute a complex group of disorders with a multitude of differential di-

231

232

3  Brain, Head, and Neck

Fig. 3.2.38  Axial (a, b) and sagittal (c) T2-weighted sequences in an 18-month-old boy with a status post–profound hypotension at the time of a term birth demonstrate bilateral signal hyperintensities in the dorsal putamina, in the lateral thalami and in the perirolandic areas

agnoses (van der Knaap and Valk 1995). In the following, only a concise outline of the most common diseases among this heterogeneous group is provided, as a full review of metabolic disorders affecting the CNS would by far exceed the scope of this section. To facilitate an approach to metabolic disorders of the brain, it is usually helpful to apply a topographic classification regarding the primarily involved structures, as follows (Barkovich 2005; van der Knaap et al. 1999): • Disorders primarily involving white matter – Disorders primarily involving the deep white matter – Disorders primarily involving the subcortical white matter

• Disorders primarily involving gray matter • Disorders involving gray and white matter • Disorders primarily involving the cerebellum Another potential approach to metabolic disorders of the brain is to classify them according to the site of the metabolic defect on a cellular level, i.e., to divide them into • Mitochondrial disorders • Peroxisomal disorders • Lysosomal disorders • Organic and aminoacidopathies, and • Other disorders

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.39  Axial T2-weighted (a) and FLAIR (b) sequences in a 12-month-old boy with a status post–bilirubin encephalopathy (kernicterus) demonstrate bilateral signal hyperintensities in the globi pallidi

We believe that the former classification facilitates the approach for the radiologist, as he or she will be primarily confronted with the topographical pattern of disease on MR imaging, and not with the underlying metabolic defect. We therefore decided to adhere to the topographical classification in this section, with the exception of mitochondrial disorders, which are difficult to classify on a purely topographical basis. It is, however, important to keep in mind, that a topographical classification of metabolic disorders, e.g., into disorders primarily involving the deep versus those affecting the more peripheral white matter, is oftentimes only possible in the early or mid-course of the disease. The end-stage of a variety of disorders may be indistinguishable on MR imaging. 3.2.7.1  Metabolic Disorders Primarily Involving the Deep White Matter Metabolic disorders that primarily affect the white matter are also called leukodystrophies. Among the most common leukodystrophies is the metachromatic leukodystrophy (van der Knaap and Valk 1995). The underlying pathogenetic mechanism in metachromatic leukodystrophy is a reduced activity of arylsulfatase A or its cofactor (genetic locus on chromosome 22) (Barkovich 2005). It is a lysosomal storage disorder.

MR imaging typically demonstrates symmetric areas of hyperintensity on T2-weighted and FLAIR sequences in the deep white matter with an early involvement of the peritrigonal areas (Fig. 3.2.40). Initially, the perivascular white matter is commonly spared, resulting in a tigroid pattern. In addition, the subcortical U fibers are initially unaffected (Kim et al. 1997). X-linked adrenoleukodystrophy is another comparatively common leukodystrophy that primarily affects the deep white matter (Barkovich 2005; van der Knaap et al. 1999). It is a peroxisomal disorder with an impaired transport of very-long-chain fatty acids mapping to chromosome X. The most common form is the childhood cerebral form that usually presents in school age boys. Another subtype is adrenomyeloneuropathy that typically presents in young adults. On MR imaging, X-linked adrenoleukodystrophy primarily affects the posterior deep white matter including the peritrigonal area as well as the splenium of the corpus callosum in most instances. The subcortical U fibers are usually initially spared. The affected areas demonstrate a marked hyperintensity on T2-weighted and a concomitant hypointensity on T1-weighted images. After administration of a paramagnetic contrast agent, enhancement is commonly seen along the borders of this signal alteration. In a subgroup of patients, a primary involvement of the frontal as opposed to the posterior white matter may be seen. On MR spectroscopy, NAA is typically de-

233

234

3  Brain, Head, and Neck

Fig. 3.2.40  Axial T2-weighted (a) and FLAIR (b) sequences in a 2-year-old girl with metachromatic leukodystrophy demonstrate bilateral signal hyperintensities in the peritrigonal white matter with relative sparing of the subcortical U fibers

creased even in areas that appear unaffected in conventional sequences, with a concomitant increase in choline, glutamate, and glutamine (Eichler et al. 2002). Globoid cell leukodystrophy is also called Krabbe disease (Barkovich 2005; van der Knaap et al. 1999). It is an autosomal recessive disorder involving the lysosomal enzyme galactosyceramide β-galactosidase (genetic locus on chromosome 14). Affected infants usually develop acute symptoms within the first 6 months of life and die within several years. The disease may, however, also manifest itself at a later age depending on the site of the mutation. Early in the course of the disease, CT may show a bilateral hyperdensity of the basal ganglia and thalami, while MR imaging may still be normal. In the further course of the disease, a diffuse T2-hyperintensity in the white matter and a global atrophy usually ensue affecting both the cerebral and the cerebellar hemispheres, as well as the cerebellar nuclei, while the subcortical white matter remains spared. In the later stages of the disease, an involvement of the thalami is commonly noted on MR imaging as well. Both cranial nerves and cauda equina may show an enhancement after paramagnetic contrast administration (Bernal and Lenn 2000). Childhood ataxia with diffuse CNS hypomyelination is also called vanishing white matter disease. Some authors include it to the group of van der Knaap leukencephalopathies. A genetic locus has been identified on chromosome 3 (Barkovich 2005). After an initial period of normal development, affected children develop progressive ataxia and spasticity and eventually die from

the disease. Disease progression is often episodic and may be triggered by minor infection or trauma. On MR imaging, a markedly increased signal intensity of the white matter is found on T2-weighted sequences with signal intensity close to CSF. White matter involvement is diffuse and symmetric. It usually starts in the deep white matter, but also involves the subcortical U fibers early in the course of the disease. The degree of cerebellar atrophy may vary with a predominant involvement of the vermis. The Lowe syndrome is also called oculocerebrorenal syndrome. It is an X-linked recessive disorder with an enzymal defect in the Golgi apparatus (Barkovich 2005). MR imaging characteristically demonstrates multiple cyst-like, CSF-isointense lesions in the deep and subcortical white matter. Additionally, confluent areas of T2 hyper­intensity and concomitant T1 hypointensity are found primarily involving the deep white matter. Cockayne syndrome may be classified as a disorder primarily affecting the deep white matter or as one primarily affecting the subcortical white matter. It has an autosomal recessive mode of inheritance and affected children typically present within the first year of life. T2-hyperintensity and atrophy are usually initially found in the deep white matter, in the basal ganglia and in the cerebellar nuclei. However, the subcortical white matter may also be involved comparatively early in the course of the disease. Calcifications of the cerebral and cerebellar nuclei are commonly present that are best discerned on T2*-weighted sequences. Among the other leukodystrophies that primarily af-

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Fig. 3.2.41   Axial FLAIR (a) and T1-weighted (b) sequences in a 9-year-old boy with van der Knaap leukodystrophy demonstrate pronounced signal alterations with preferential involvement of the subcortical white matter. The gyri appear swollen

fect the deep white matter are phenylketonuria, maple syrup urine disease, and homocysteinuria (Barkovich 2005). Most of these disorders are now commonly detected early due to metabolic neonatal screening programs; however, screening programs are not universally performed. Maple syrup urine disease is a disorder with a characteristic MR imaging presentation. It typically manifests itself acutely during the first week of life. However, milder forms with a later onset exist. MR imaging characteristically demonstrates a marked edema during the acute stage. In the early stage, this edema is diffuse, while in the subacute phase the deep cerebellar white matter, the dorsal brainstem, the cerebral peduncles, the posterior limb of the internal capsule and the perirolandic regions are preferentially involved. 3.2.7.2  Metabolic Disorders Primarily Involving the Subcortical White Matter Among the leukodystrophies that primarily affect the subcortical white matter is the megalencephalic leukoencephalopathy with cysts. It has also been called leukoencephalopathy with macrocephaly and mild clinical course or van der Knaap disease (Barkovich 2005). Some authors subsume four fairly recently described leukencephalopathies under the term van der Knaap

leukencephalopathies, namely megalencephalic leukoencephalopathy with cysts, vanishing white matter disease, white matter disease with lactate as well as hypomyelination with atrophy of the basal ganglia and cerebellum (Schiffmann and van der Knaap 2004). Megalencephalic leukoencephalopathy with cysts has been identified as an autosomal recessive disorder with a genetic locus on chromosome 22. There is a discrepancy between a comparatively slow clinical course and highly abnormal imaging features. Affected children usually present with developmental delay, macrocephaly, and a slowly deteriorating neurological status. On MR imaging, the subcortical white matter appears swollen with a marked hyperintensity on T2-weighted and hypointensity on T1-weighted images (Fig. 3.2.41). Characteristically, subcortical cysts are found, especially in the frontal and temporal lobes. The deep white matter and the cerebellar white matter are relatively spared, even though some degree of cerebellar involvement may be present. Diffusivity is usually increased in affected areas. Pelizaeus-Merzbacher disease is a pseudo-leukodystrophy that also primarily involves the subcortical white matter. In its classic form, the mode of inheritance is X-linked recessive. However, a so-called connatal form exists with either X-linked or autosomal inheritance, and there are other subtypes as well (Barkovich 2005). Children affected by the classic form usually present early

235

236

3  Brain, Head, and Neck

with motor developmental delays, increasing spasticity, and involuntary movements. The disease progression is relatively slow, but relentless and patients usually die during early adulthood. Children affected by the connatal form usually have a more rapid course of the disease and present with extrapyramidal symptoms, nystagmus, and spasticity (Barkovich 2005). MR imaging characteristically shows the appearance of a neonate with a striking lack of myelination. Only parts of the internal capsule and optic radiations initially appear myelinated; in the further course of the disease, total absence of myelination may be noted. In addition, progressive cerebral and cerebellar atrophy may be seen. 18q- syndrome is caused by a deletion of the long arm or chromosome 18, on which one of the genes encoding for myelin basic protein is situated. Affected children commonly present with developmental delays, dysmorphic features, and genital hypoplasia. MR imaging characteristically demonstrates a hypomyelination on T2-weighted images, while T1-weighted images appear normal (Fig. 3.2.42). A reduced volume of the anterior pituitary gland may be seen as well. A multitude of other disorders present with leukodystrophic changes preferentially affecting the subcortical white matter or with unspecific patterns. These include galactosemia, Salla disease, leukodystrophy with trichothiodystrophy, pseudo-TORCH syndromes, nonketotic hypergylcinemia, dihydropyrimidine dehydrogenase deficiency, and 3-HMG-CoA lyase deficiency (Barkovich

2005; van der Knaap et al. 1999). To discuss all of these disorders in detail would far exceed the scope of this section. 3.2.7.3  Metabolic Disorders Primarily Involving the Gray Matter Neuronal ceroid lipofuscinoses are lysosomal storage disorders. They are comparatively common metabolic encephalopathies. A number of subtypes have been identified to date, among which are the Santavuori-Haltia disease (infantile onset), the Jansky-Bielschowsky disease (late infantile onset), the Batten or Spielmeyer-Vogt disease (juvenile onset), and a variety of others (Goebel and Wisniewski 2004). Several genetic loci have been identified for the different subtypes. Among the affected chromosomes are chromosomes 1, 8, 11, 13, 15, and 16 (Goebel and Wisnewski 2004). Affected children usually develop progressive visual, motor, and cognitive symptoms. On MR imaging the most common feature is both a cerebral and cerebellar atrophy of variable degree. In addition, a hypointense signal is usually discerned in the globi pallidi and thalami on T2-weighted sequences. MR spectroscopy typically demonstrates an absence of the NAA peak in the infantile form and a reduced NAA peak in the late infantile form, while spectra may be normal in other variants (Barkovich 2005).

Fig. 3.2.42  Axial T2-weighted (a,b) sequences in a 3-year-old girl with an 18q- syndrome demonstrate a pronounced hypomyelination of the white matter

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

Hallervorden-Spatz disease is also called pantothenate kinase–associated neurodegeneration. Some of the affected patients have been demonstrated to have a mutation on chromosome 20 involving the pantothenate kinase gene (PANK2) (Hayflick et al. 2003). While the age of presentation may vary, most children become symptomatic during school age. A classic form with an onset before the age of 10 and an atypical form with a slightly older age of onset have been described. Affected children usually present with progressive extrapyramidal symptoms. A concomitant retinopathy is common, and pyramidal signs may ensue (Hayflick et al. 2003). On MR imaging, a characteristic hypointensity in the globus pallidus is usually discerned on T2- and T2*-weighted images corresponding to iron depositions. Within these areas of hypointensity focal areas of hyperintensity may be present resulting in the characteristic “eye-of-the-tiger” sign. This sign has been demonstrated to correlate well with the presence of a PANK2 mutation (Hayflick et al. 2003). Niemann-Pick disease is an autosomal recessive disorder leading to a sphingomyelinase deficiency. Genetic loci have been mapped to chromosomes 11 and 18 (Barkovich 2005). MR imaging demonstrates cortical as well as white matter atrophy. There is usually a comparatively slight, diffuse white matter hyperintensity on T2-weighted images. Rett syndrome is a comparatively common X-chromosomal disorder with a mutation in the methyl-CpG-

binding protein 2 (MECP2) gene (Barkovich 2005). It almost exclusively affects girls, as the disease is usually fatal in boys. Rare cases of mosaicism are an exception. Affected patients usually become symptomatic between 6 and 12 months of age. Progressive motor and cognitive deterioration ensue, with cognitive function usually stabilizing on a level of severe disability around age 4. Epilepsy is also common. MR imaging in affected girls most commonly demonstrates largely normal findings. However, quantitative evaluations may show a reduction in parenchymal volume mostly affecting the frontal and temporal lobes and the caudate nucleus (Fig. 3.2.43). Huntington’s disease only very rarely presents in the pediatric population, as patients usually do not become symptomatic until adulthood. The mode of inheritance is autosomal dominant with a genetic locus on the short arm of chromosome 4. Affected children commonly develop hypokinesia, rigidity and a cognitive decline, while choreiform movements usually only develop later (Barkovich 2005). MR imaging characteristically demonstrates an atrophy of the caput nuclei caudati leading to enlarged, convex frontal horns of the lateral ventricles. In addition, atrophy of the putamen and frontal cortical atrophy are common findings. Other metabolic disorders that primarily involve the gray matter include isovaleric acidemia, creatine deficiency syndromes, the Aicardi-Goutières syndrome, succinic semialdehyde dehydrogenase deficiency, aspar-

Fig. 3.2.43  Axial (a) and sagittal (b) T2-weighted sequences in a 3-year-old girl with Rett syndrome demonstrate a reduction in the parenchymal volume of the frontal lobes

237

238

3  Brain, Head, and Neck

tylglucosaminuria and infantile neuroaxonal dystrophy (Barkovich 2005; van der Knaap and Valk 1995). 3.2.7.4  Metabolic Disorders Involving Gray and White Matter Alexander disease is also called fibrinoid leukodystrophy. In most instances, a mutation of the glial fibrillary acidic protein (GFAP) with a genetic locus on chromosome 17 is found, leading to an accumulation of so-called Rosenthal fibers (Barkovich 2005; van der Knaap and Valk 1995). Affected children usually present early with macrocephaly and developmental delays. Most children with Alexander disease die in infancy or early childhood. Neonatal, juvenile, and adult forms of the disease appear to be much less common. On MR imaging, there usually is a pronounced signal hyperintensity on T2-weighted images predominantly affecting the frontal lobes with an early involvement of subcortical U fibers. Around the lateral ventricles, a subependymal rim of T2 hypointensity and T1 hyperintensity is commonly discerned. The putamina and the heads of the caudate nuclei as well as the periaqueductal region and the dorsal medulla oblongata are commonly hyper­ intense on T2-weighted sequences and may show an uptake of paramagnetic contrast. In addition, a contrast enhancement of the trigonal ependyma may be seen. Enhancement is usually quite pronounced in the early stages of the disease, while it tends to diminish later on. In the

further course of the disease intraparenchymal cysts may develop. Canavan disease, also called spongiform leukodystrophy, is an autosomal recessive deficiency of aspartoacylase, which is particularly common in Jews of Ashkenazi descent. Affected children usually become symptomatic in early infancy with hypotonia, seizures, and macrocephaly, with subsequent spasticity, cognitive decline, and optic atrophy. Most children die in the second year of life (Barkovich 2005; van der Knaap and Valk 1995). MR imaging demonstrates diffuse signal hyperintensities on T2-weighted sequences and concomitant hypointense areas on T1-weighted images in the white matter, with an early involvement of subcortical fibers and a centripetal spread. The gyri may appear edematous. The globi pallidi and thalami commonly demonstrate a hyperintense signal on T2-weighted images as well. In the course of the disease, diffuse cerebral atrophy ensues involving both the gray and the white matter. On MR spectroscopy, a pronounced elevation of the NAA peak is noted. Glutaric acidurias are divided into type I and a type II glutaric acidurias. As the deficient enzymes are located in the mitochondria, they are strictly speaking mitochondrial disorders, but shall be discussed in this section. Glutaric aciduria type I is the result of a glutarylCoA dehydrogenase deficiency with a genetic locus on chromosome 19. Affected children usually present with an acute crisis during the first year of life, followed by a gradual neurologic deterioration and development of dystonia (Barkovich 2005; van der Knaap and Valk 1995).

Fig. 3.2.44  Axial T1-weighted (a) and coronal FLAIR (b) sequences in a 3-year-old boy with glutaric aciduria demonstrate a large right-sided subdural hematoma

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

MR imaging characteristically demonstrates marked frontotemporal atrophy with a pronounced dilation of the Sylvian fissures. The striking widening of the outer CSF spaces may resemble cysts. Myelination is commonly delayed and cerebral atrophy is progressive over the course of the disease. About one third of affected children develop chronic subdural hematomas, which are often bilateral and need to be differentiated from child abuse (Figs. 3.2.44, 3.2.45). These subdural hematomas are the result of easily torn bridging veins due to enlarged outer CSF spaces. They may result from only minor trauma. Glutaric aciduria type II is caused by a multiple acylCoA dehydrogenase deficiency; several subtypes have been published (Barkovich 2005; van der Knaap and Valk 1995). MR imaging demonstrates hyperintense signal on T2-weighted images in the putamen and caudate head as well as in the supratentorial white matter. Mucopolysaccharidoses are lysosomal disorders characterized by an inability to degrade glycosaminoglycan and an accumulation of mucopolysaccharides. A variety of subtypes is known, in which different lysosomal enzymes are deficient. The mode of inheritance is usually autosomal recessive, but may be X-linked recessive as well (Barkovich 2005; van der Knaap and Valk 1995). MR imaging characteristically demonstrates widened, CSF-isointense perivascular Virchow-Robin spaces with an accumulation of mucopolysaccharides in several of the mucopolysaccharidoses. A cribriform pattern may ensue. The incidence of arachnoid cysts is also relatively high. In addition, diffuse leukodystrophic changes with T2-hyperintensities as well as cerebral atrophy are commonly found. In several subtypes, compression of the spinal cord on the atlantoaxial level may be seen resulting either from a shortened odontoid, relatively loose ligaments and pronounced reactive soft-tissue changes or from intradural mucopolysaccharide deposits. Every cranial MR imaging study in children with mucopolysaccharidoses should therefore include a visualization of the upper cervical spine, preferably in a sagittal orientation. Zellweger syndrome is a peroxisomal disorder, in which peroxisomal activity is almost completely absent. Several genetic loci have been described. Affected children are usually already symptomatic at the time of birth. Progressive neurologic deterioration along with other features such as hepatomegaly and facial dysmorphism are commonly seen. Most children die within the first year of life (Barkovich 2005; van der Knaap and Valk 1995). On MR imaging, one of the most characteristic features are subependymal cysts that are most commonly found in a caudothalamic location. These are called subependymal germinolytic cysts. The white matter is usually markedly hypomyelinated. In addition, microgyri are commonly seen in the frontal and temporal lobes and polymicrogyria is frequently present in the perirolandic areas. The cortex may also appear partially thickened and smooth, resembling pachygyria. MR spectroscopy usu-

Fig. 3.2.45  A CT scan in a 4-year-old boy with glutaric aciduria demonstrates bilateral subdural hematomas

ally demonstrates a reduced lactate peak and elevated lipid and lactate peaks. The gangliosidoses are inherited lysosomal glycosphingolipid disorders that are divided into GM1 and GM2 gangliosidoses (Barkovich 2005; van der Knaap and Valk 1995). In GM1 gangliosidosis, the lysosomal beta-galactosidase is deficient. A genetic locus has been mapped to chromosome 3. Onset may be infantile, late infantile/juvenile or chronic. Affected children usually present with developmental delays, seizures, and—in the infantile form—hepatomegaly and facial dysmorphism. In GM2 gangliosidoses, the lysosomal hexosaminidase is deficient. A genetic locus has been mapped to chromosome 15. Depending on the affected isoenzyme or enzymal subunit, the disorder is also called Tay-Sachs disease or Sandhoff disease. Affected children present with neurologic deterioration with spasticity, dystonia, ataxia, and blindness (Barkovich 2005; van der Knaap and Valk 1995). MR imaging only marginally differs between the different types of gangliosidoses. There is usually a signal hypointensity in the thalami on T2-weighted imaging, with a concomitant signal hyperintensity on T1-weighted sequences. In Tay-Sachs disease, the signal may be focally hypointense in the ventral thalamus and hyperintense in the dorsal thalamus on T2-weighted sequences. The dorsal putamina may be affected as well. In addition, a mild, diffuse T2-hyperintensity is found in the cerebral white matter. In the course of the disease, progressive cerebral and cerebellar atrophy is found. Other metabolic disorders that involve both the gray

239

240

3  Brain, Head, and Neck

and the white matter include fucosidosis, pyruvate dehydrogenase deficiency, Molybdenum cofactor deficiency, disorders of the urea cycle, methylmalonic and proprionic acidemias, hypomyelination with atrophy of the basal ganglia and cerebellum, l-2-hydroxyglutaric aciduria,a and 3-methylglutaconic aciduria (Barkovich 2005; van der Knaap and Valk 1995). 3.2.7.5  Mitochondrial Disorders Mitochondrial disorders are a complex and relatively common group of diseases with defects in the respiratory chain. An overlap with other metabolic syndromes exists and the distinction between mitochondrial disorders and other metabolic disorders is not always completely clear. Mitochondrial disorders most commonly affect the CNS and the skeletal muscle; other organ systems may, however, be involved as well. They tend to affect both the gray and the white matter of the neurocranium. The acronym MELAS stands for mitochondrial encephalopathy with lactic acidosis and stroke-like episodes. Affected patients may become symptomatic at any age. They most commonly present with stroke-like events; serum lactate levels are elevated. MR imaging typically demonstrates focal areas of T2-hyperintensity and T1-hypointensity in the affected areas, most commonly in the occipital and parietal lobes (Fig. 3.2.46). These areas of abnormal signal intensity do not correspond to classic vascular territories. Studies on diffusion-weighted imaging in patients with MELAS have reported conflicting results. MR spectroscopy demonstrates elevated lactate both in areas affected on conventional sequences and in seemingly unaffected cerebral regions (Barkovich 2005). Kearns-Sayre syndrome is a mitochondrial encephalopathy characterized by external ophthalmoplegia and retinitis pigmentosa. Patients usually demonstrate beginning signs of neurologic or neuromuscular dysfunction before reaching adulthood. Cerebellar ataxia, hearing loss, cognitive decline, endocrine dysfunction, and cardiac symptoms are other potential complaints (Barkovich 2005). MR imaging demonstrates leukodystrophic changes with an early involvement of the subcortical white matter, while the deep, periventricular white matter is initially spared. In the further course of the disease, the globi pallidi, the thalami, and the dorsal midbrain typically demonstrate a hyperintense signal on T2-weighted and T1-weighted sequences. However, diffuse symmetric calcifications may be found in the basal ganglia, the caudate nuclei, and the cerebellar nuclei as well. Again, MR spectroscopy demonstrates an elevated lactate, while the NAA peak is reduced. Ophthalmoplegia plus is another mitochondrial syndrome characterized by external ophthalmoplegia.

Fig. 3.2.46  Axial T2-weighted sequences in an 11-year-old boy with MELAS demonstrate bilateral signal hyperintensities in the occipital lobes

In addition, affected patients usually demonstrate pyramidal and extrapyramidal as well as cerebellar symptoms, cognitive decline, myopathic changes that may be exercise-induced or permanent, and signs of neuropathy (Barko­vich 2005; van der Knaap and Valk 1995). The MR imaging appearance does not significantly differ from that of the Kearns-Sayre syndrome. Leigh syndrome is sometimes also referred to as subacute necrotizing encephalopathy. It actually represents a complex of symptoms with a high genetic heterogeneity. Underlying defects include a defective pyruvate dehydrogenase complex, a deficiency of the cytochrome oxidase, mutations of the ATPase 6 gene or a complex I deficiency (Barkovich 2005; van der Knaap and Valk 1995). Inheritance patterns include autosomal recessive, X-linked and maternal mitochondrial inheritance. Patients typically present in infancy with hypotonia and signs of psychomotor delay or regression. Progressive neurologic degeneration ensues and most affected patients die in early childhood. The MR imaging appearance may vary depending on the underlying pathophysiological process. A characteristic, but also variable feature is a bilateral hyperintensity of the putamina and caudate nuclei on T2-weighted and FLAIR sequences, especially when the pyruvate dehydro-

3.2  Normal Development, Congenital, Hereditary, and Acquired Diseases of the Central Nervous System in Pediatrics

genase complex is defective. The thalami may be involved as well. Myelination is commonly delayed. In cytochrome oxidase deficiency, the subthalamic nuclei, dorsal pons, and periaqueductal gray matter are more commonly involved, while complex I deficiency may lead to a pronounced cavitation of the white matter. The acronym MERRF refers to myoclonic epilepsy with ragged red fibers. Affected children present with signs of a mitochondrial encephalopathy, seizures, and myopathic changes. The so-called ragged red fibers are a characteristic feature on muscle biopsy. MR imaging most commonly demonstrates signal alteration in the basal ganglia and in the caudate nuclei. In addition, infarcts in watershed areas of the brain are commonly seen. Alpers disease is a mitochondrial disorder with an autosomal-recessive mode of inheritance (Barkovich 2005; van der Knaap and Valk 1995). Affected children usually present early with refractive epilepsy, developmental regression with an episodic pattern, cortical blindness, and hepatic cirrhosis. MR imaging demonstrates progressive cortical atrophy and delayed myelination. Other mitochondrial disorders include the glutaric acidurias, discussed above, trichpoliodystrophy, as well as familial mitochondrial encephalopathy with macrocephaly, cardiomyopathy and complex I deficiency (Barkovich 2005; van der Knaap and Valk 1995).

9.

References

17.

1.

2.

3.

4.

5. 6.

7. 8.

Alperin N, Sivaramakrishnan A, Lichtor T (2005) Magnetic resonance imaging-based measurements of cerebrospinal fluid and blood flow as indicators of intracranial compliance in patients with Chiari malformation. J Neurosurg 103:46–52 American Academy of Pediatrics Committee on Drugs (1992) Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures. Pediatrics 89:1110–1115 American Academy of Pediatrics Committee on Drugs (2002) Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures: addendum. Pediatrics 110:836–838 Baker LL, Stevenson DK, Enzmann DR (1988) End-stage periventricular leukomalacia: MR evaluation. Radiology 168:809–815 Barkovich AJ (2005) Pediatric neuroimaging. Lippincott Williams & Wilkins, Philadelphia Barkovich AJ, Chuang SH (1990) Unilateral megalencephaly: correlation of MR imaging and pathologic characteristics. AJNR Am J Neuroradiol 11:523–531 Barkovich AJ, Kuziecky RI (2000) Gray matter heterotopia. Neurology 55:1603–1608 Barkovich AJ, Quint DJ (1993) Middle interhemispheric fusion: an unusual variant of holoprosencephaly. AJNR Am J Neuroradiol 14:431–440

10. 11.

12.

13.

14.

15.

16.

18.

19.

20.

21.

22.

23.

24.

25.

Barkovich AJ, Kjos BO, Jackson DE Jr, Norman D (1988) Normal maturation of the neonatal and infant brain: MR imaging at 1.5 T. Radiology 166:173–180 Barkovich AJ, Fram EK, Norman D (1989a) Septo-optic dysplasia: MR imaging. Radiology 171:189–192 Barkovich AJ, Kjos BO, Norman D, Edwards MS (1989b) Revised classification of posterior fossa cysts and cystlike malformations based on the results of multiplanar MR imaging. AJR Am J Roentgenol 153:1289–1300 Barkovich AJ, Kuzniecky RI, Dobyns WB, Jackson GD, Becker LE, Evrard P (1996) A classification scheme for malformations of cortical development. Neuropediatrics 27:59–63 Barkovich AJ, Ferriero DM, Barr RM et al (1998) Microlissencephaly: a heterogeneous malformation of cortical development. Neuropediatrics 29:113–119 Barkovich AJ, Kuzniecky RI, Jackson GD, Guerrini R, Dobyns WB (2001a) Classification system for malformations of cortical development: update 2001. Neurology 57:2168–2178 Barkovich AJ, Simon EM, Walsh CA (2001b) Callosal agenesis with cyst: a better understanding and new classification. Neurology 56:220–227 Barkovich AJ, Westmark KD, Bedi HS, Partridge JC, Ferriero DM, Vigneron DB (2001c) Proton spectroscopy and diffusion imaging on the first day of life after perinatal asphyxia: preliminary report. AJNR Am J Neuroradiol 22:1786–1794 Bernal OG, Lenn N (2000) Multiple cranial nerve enhancement in early infantile Krabbe’s disease. Neurology 54:2348–2349 Eichler FS, Barker PB, Cox C et al (2002) Proton MR spectroscopic imaging predicts lesion progression on MRI in X-linked adrenoleukodystrophy. Neurology 58:901–907 Felderhoff-Mueser U, Rutherford MA, Squier WV et al (1999) Relationship between MR imaging and histopathologic findings of the brain in extremely sick preterm infants. AJNR Am J Neuroradiol 20:1349–1357 Filippi CG, Lin DD, Tsiouris AJ et al (2003) Diffusion-tensor MR imaging in children with developmental delay: preliminary findings. Radiology 229:44–50 Fischbein NJ, Barkovich AJ, Wu Y, Berg BO (1998) SturgeWeber syndrome with no leptomeningeal enhancement on MRI. Neuroradiology 40:177–180 Goebel HH, Wisniewski KE (2004) Current state of clinical and morphological features in human NCL. Brain Pathol 14:61–69 Govaert P, Lequin M, Swarte R et al (2003) Changes in globus pallidus with (pre)term kernicterus. Pediatrics 112:1256–1263 Granata T, Freri E, Caccia C, Setola V, Taroni F, Battaglia G (2005) Schizencephaly: clinical spectrum, epilepsy, and pathogenesis. J Child Neurol 20:313–318 Guerrini R, Barkovich AJ, Sztriha L, Dobyns WB (2000) Bilateral frontal polymicrogyria: a newly recognized brain malformation syndrome. Neurology 54:909–913

241

242

3  Brain, Head, and Neck 26. Hayflick SJ, Westaway SK, Levinson B et al (2003) Genetic, clinical, and radiographic delineation of HallervordenSpatz syndrome. N Engl J Med 348:33–40 27. Kim TS, Kim IO, Kim WS et al (1997) MR of childhood metachromatic leukodystrophy. AJNR Am J Neuroradiol 18:733–738 28. Kim SK, Na DG, Byun HS et al (2000) Focal cortical dysplasia: comparison of MRI and FDG-PET. J Comput Assist Tomogr 24:296–302 29. Knaap MS van der, Valk J (1995) Magnetic resonance of myelin, myelination, and myelin disorders. Springer, Berlin Heidelberg New York 30. Knaap MS van der, Smit LM, Barth PG et al (1997) Magnetic resonance imaging in classification of congenital muscular dystrophies with brain abnormalities. Ann Neurol 42:50–59 31. Knaap MS van der, Breiter SN, Naidu S, Hart AA, Valk J (1999) Defining and categorizing leukoencephalopathies of unknown origin: MR imaging approach. Radiology 213:121–133 32. Kumandas S, Akcakus M, Coskun A, Gumus H (2004) Joubert syndrome: review and report of seven new cases. Eur J Neurol 11:505–510 33. Kuzniecky R, Andermann F, Guerrini R (1993) Congenital bilateral perisylvian syndrome: study of 31 patients. The CBPS Multicenter Collaborative Study. Lancet 341:608–612 34. Latif F, Tory K, Gnarra J et al (1993) Identification of the von Hippel-Lindau disease tumor suppressor gene. Science 260:1317–1320 35. Lee SK, Mori S, Kim DJ, Kim SY, Kim DI (2004) Diffusion tensor MR imaging visualizes the altered hemispheric fiber connection in callosal dysgenesis. AJNR Am J Neuroradiol 2004; 25:25–28 36. Leventer RJ, Pilz DT, Matsumoto N, Ledbetter DH, Dobyns WB (2000) Lissencephaly and subcortical band heterotopia: molecular basis and diagnosis. Mol Med Today 6:277–284 37. Mackay MT, Becker LE, Chuang SH et al (2003) Malformations of cortical development with balloon cells: clinical and radiologic correlates. Neurology 60:580–587

38. Mercuri E, He J, Curati WL, Dubowitz LM, Cowan FM, Bydder GM (1997) Cerebellar infarction and atrophy in infants and children with a history of premature birth. Pediatr Radiol 27:139–143 39. Mikulis DJ, Diaz O, Egglin TK, Sanchez R (1992) Variance of the position of the cerebellar tonsils with age: preliminary report. Radiology 183:725–728 40. Mukherji SK, Albernaz VS, Lo WW et al (1997) Papillary endolymphatic sac tumors: CT, MR imaging, and angiographic findings in 20 patients. Radiology 202:801–808 41. Naidich TP, Altman NR, Braffman BH, McLone DG, Zimmerman RA (1992) Cephaloceles and related malformations. AJNR Am J Neuroradiol 13:655–690 42. Perez-Nunez A, Lagares A, Benitez J et al (2004) LhermitteDuclos disease and Cowden disease: clinical and genetic study in five patients with Lhermitte-Duclos disease and literature review. Acta Neurochir (Wien) 146:679–690 43. Roach ES, Gomez MR, Northrup H (1998) Tuberous sclerosis complex consensus conference: revised clinical diagnostic criteria. J Child Neurol 13:624–628 44. Roessler E, Muenke M (1998) Holoprosencephaly: a paradigm for the complex genetics of brain development. J Inherit Metab Dis 21:481–497 45. Rosser T (2003) Aicardi syndrome. Arch Neurol 60:1471–1473 46. Schiffmann R, van der Knaap MS (2004) The latest on leukodystrophies. Curr Opin Neurol 17:187–192 47. Shepherd CW, Houser OW, Gomez MR (1995) MR findings in tuberous sclerosis complex and correlation with seizure development and mental impairment. AJNR Am J Neuroradiol 16:149–155 48. Takanashi J, Sugita K, Fujii K, Niimi H (1995) MR evaluation of tuberous sclerosis: increased sensitivity with fluidattenuated inversion recovery and relation to severity of seizures and mental retardation. AJNR Am J Neuroradiol 16:1923–1928 49. Truwit CL, Barkovich AJ (1990) Pathogenesis of intracranial lipoma: an MR study in 42 patients. AJR Am J Roentgenol 155:855–864; discussion, p 865 50. US Department of Health and Human Services, National Institutes of Health (1988) Consensus Development Conference. Neurofibromatosis. Conference statement. Arch Neurol 45:575–578

3.3  Intracranial Tumors

3.3  Intracranial Tumors M. Essig 3.3.1  Introduction The goals and requirements for brain tumor imaging are multiple and complex. They involve providing a diagnosis and a differential diagnosis, and, if possible, a specific diagnosis, as well as accurate grading of the tumor. Neuroimaging is an essential part of the decision-making process for therapy and later for precise planning of surgical or radiological interventions. Prior to neurosurgery, neuroimaging can precisely define the location and accurately delineate the lesion prior to intervention. It can also aid in radiotherapy planning by precisely defining the lesion margins. In addition, neuroimaging is mandatory after therapeutic intervention to monitor disease and possible side effects. Due to its high tissue contrast and its noninvasiveness, magnetic resonance imaging (MRI) is accepted as the most sensitive method for diagnosing brain tumors (Brandt-Zawadzki et al. 1984; Muroff and Runge 1995; Levin et al. 2001). The ruling out of brain tumors is one of the most common indications for neuroimaging using MRI. In the past years, it has become generally recognized that MRI would be the imaging study of choice in the evaluation of intracranial tumors if availability and cost were not an issue. The method therefore plays the most important role in the early diagnosis of brain tumors. Worldwide, approximately 176,000 new cases of brain and other central nervous system (CNS) tumors were diagnosed in the year 2000. Available registry data from the Surveillance, Epidemiology, and End Results (SEER) database for 1996– 2000 indicate that the combined incidence of primary invasive CNS tumors in the United States is 6.6 per 100,000 persons per year, with an estimated mortality of 4.7 per 100,000 persons per year (Ries et al. 2003). Not included are the benign brain tumors and the high number of patients diagnosed with brain metastases from malignancy elsewhere in the body (Johnson and Young 1996). In general, the incidence of primary brain tumors is higher in whites than in blacks, and mortality is higher in males than in females (Levin et al. 2001). For 2007, the estimated numbers of new cases for the United States are 20,000, with a total of 12,500 deaths (American Cancer Society 2007). The classification of brain tumors is still controversial, including classification by location or histology. The most common classification is that of the World Health Organization (WHO) (Kleihues et al. 1993; Lopes et al. 1993) (Table 3.3.1); in the following sections, however, to better reflect the imaging findings, the brain tumors are divided into intra-axial cerebral lesions and extra-axial cerebral lesions. This compartmental localization of cerebral tumors is of great clinical importance and is fundamental

for diagnosis because it determines the pathway for correct differential diagnosis and impacts treatment selection, treatment planning and outcome. Differentiation based on MRI is still a challenge with specific findings for both locations. A peripheral location along the inner table of the skull and associated bone changes are typical for an extra-axial location as well as the presence of an interface between the lesion and the brain surface. This section provides a general overview of the practical aspects of MRI and the use of contrast media in brain tumor imaging, followed by a description of the MR appearances of the most common cerebral tumors. Functional MRI techniques which have impact on brain tumor imaging are described later on. 3.3.2  The WHO Classification of Brain Tumors In 1993 the WHO ratified a new comprehensive classification of neoplasms affecting the CNS (Kleihues et al. 1993; Lopes et al. 1993; WHO 1990). The classification of brain tumors is based on the premise that each type of tumor Table 3.3.1  World Health Organization list of brain tumors 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

Tumors of neuroepithelial tissue Astrocytic tumors Oligodendroglial tumors Ependymal tumors Mixed gliomas Choroid plexus tumors Neuroepithelial tumors of uncertain origin Neuronal and mixed neuronal-glial tumors Pineal parenchymal tumors Embryonal tumors

2 2.1 2.2 2.3

Tumors of the cranial and spinal nerves Schwannoma Neurofibroma Malignant peripheral nerve sheath tumor

3 3.1 3.2 3.3 3.4

Tumors of the meninges Tumors of meningothelial cells (meningiomas) Mesenchymal non-meningothelial tumors Primary melanocytic lesions Tumors of uncertain histogenesis

4

Lymphomas and hematopoietic neoplasms

5

Germ cell tumors

6

Cysts and tumor-like lesions

7

Tumors of sellar region

8

Local extensions from regional tumors

9

Metastatic tumors

10 Unclassified tumors

243

244

3  Brain, Head, and Neck

results from the abnormal growth of a specific cell type. To the extent that the behavior of a tumor correlates with basic cell type, tumor classification dictates the choice of therapy and predicts prognosis. The new WHO system is particularly useful in this regard, with only a few notable exceptions (for example all or almost all gemistocytic astrocytomas are actually anaplastic and hence grade III or even IV rather than grade II as designated by the WHO system). The WHO classification also provides a parallel grading system for each type of tumor. In this grading system, most named tumors are of a single defined grade. The new WHO classification provides the standard for communication between different centers in the United States and around the world. An outline of this classification is provided in Table 3.3.1. The WHO grading of CNS tumors also establishes a malignancy scale based on histologic features of the tumor (Kleihues et al. 1993). The histologic grades are as follows: • WHO grade I includes lesions with low proliferative potential, a frequently discrete nature, and the possibility of cure following surgical resection alone. • WHO grade II includes lesions that are generally infiltrating and low in mitotic activity but recur. Some tumor types tend to progress to higher grades of malignancy. • WHO grade III includes lesions with histologic evidence of malignancy, generally in the form of mitotic activity, clearly expressed infiltrative capabilities, and anaplasia. • WHO grade IV includes lesions that are mitotically active, necrosis-prone, and generally associated with a rapid preoperative and postoperative evolution of disease. 3.3.3  Practical Aspects of MR Imaging in Brain Tumors MRI is the method with the highest sensitivity for imaging brain tumors. This is true for both the detection and the description of the extent of disease. Although the method has been in use now for more than 15 years, the specificity is still limited. Using a standard MRI protocol for cerebral neoplasms allows high resolution imaging and characterization of lesions. Most tumors have low signal intensity on T1-weighted images and high signal intensity on T2weighted images (Byrne 1994). FLAIR imaging is a T2weighted sequence with suppression of the signal intensity of CSF or any fluid with the T1 of CSF. Initially used for imaging white matter changes, it has now become a part of most brain imaging protocols (Tsuchiya et al. 1996; Essig et al. 1998; De Coene et al. 1992; Rydeberg et al. 1994). It can be particularly useful in proving the cystic content of a lesion or in the delineation of tumor from CSF containing areas, e.g., the ventricles, surgical defects or cystic tumor parts. If a lesion is exactly isointense to CSF

on Tl-weighted, T2-­weighted, and FLAIR images, then one can state very confi­dently that the lesion is cystic, a pattern followed by arach­noid cysts and cysts associated with extra-axial masses. In patients with cerebral gliomas, FLAIR was found superior to conventional imaging in tumor delineation and in the differentiation between tumor and edema (Essig et al. 1998, 1999). Due to the suppression of the CSF signal and a reduced gray-to-white matter contrast on the FLAIR images, periventricular, cortical lesions, and callosal lesions could be better detected and delineated. Unfortunately for the radiologist, tumor cysts and cystic necrosis within neoplasms often contain blood byproducts or have a high protein content, which does not allow suppression of the cystic fluid on FLAIR imaging. Therefore, these regions show hyperintense to normal CSF on FLAIR with increasing signal parallel to increasing protein content. Fluid-debris intensity levels are a pathognomonic sign of cystic tissue and are of­ten quite striking and frequent in cases of cystic tumors. Hemor­rhage is uniquely depicted by MR imaging, because of the paramagnetic properties of many of the blood byproducts. The appearance of those products depends on the age and amount. On MR imaging, old hemorrhage is easily distinguished from other fluid (like CSF) because of the paramagnetic properties of methemoglobin, a marker of subacute to chronic intracra­nial hemorrhage (Forsting 1993). The tendency to bleed that is shown by certain primary intracranial neoplasms (e.g., glioblastoma, ependy­moma, and oligodendroglioma) and metastases (e.g., of melanoma, lung carcinoma, renal cell carcinoma, choriocar­cinoma), however, can be of great importance for a differential diagnosis and increases the sensitivity in tumor detection. Although it is important to discover hemorrhage, it is also critical to define its etiology, and MRI is of limited value for this. The signal intensity pattern of intratumoral hemorrhage differs from benign intracranial hematomas. Signal intensity presents heteroge­neously in tumor bleeding, while it is more homogeneous in parenchymal hemorrhage. The appearance of blood also differs between those entities. Blood may not evolve as rapidly if it is within tumor tissue (Forsting 1993) presenting different stages of blood, in compar­ison with the fast evolution of benign hematomas. Some tumors contain tissue that may mimic a hematoma. The most important of these are metastases from melanoma (Jeyapalan and Batchelor 2000; Bisese 1992). The melanin pigments shorten the T1 relaxation time substantially and the tumors appear hyperintense on the unenhanced T1 sequences. The effect on the T2 relaxation is small, and no signal changes are seen in most cases. Unfortu­nately, melanoma metastases are commonly both hemor­rhagic and melanotic, which in many subjects makes the imaging less specific. Calcifications can only be depicted if extensive. If there is a rim-shaped calcification, as in a cystic lesion, it may mimic chronic hemorrhage in the hemosiderin stage.

3.3  Intracranial Tumors

3.3.4  Blood–Brain Barrier and Tumor Enhancement: Mechanisms and Applications MR imaging of brain tumors requires a high CNS-to-lesion contrast, which depends on the signal intensity of the lesion relative to that of the surrounding normal tissue (Murofff and Runge 1995). Furthermore, detailed information on the internal morphology of the lesion is essential for differential diagnosis, grading, and for the selection and planning of therapy. For most diseases and for many of the currently available functional MR imaging methods, use of MR contrast media is mandatory. The standard dose employed for MR imaging of the CNS is 0.1 mmol/kg body weight, although numerous studies have shown that lesion detection may be improved with the use of higher doses and dedicated sequences (Erickson et al. 2002; Schneider et al. 2001). Contrastenhanced MRI also helps in distinguishing tumors from other pathologic processes, and depicting basic signs of tumor response to therapy, such as change in size, morphology, and degree of contrast material enhancement. The following section describes the fundamental features of the contrast mechanisms and their influence on brain tumor assessment. The different properties of currently available contrast media and their dosages and field dependencies are discussed. 3.3.4.1  Mechanisms of Contrast Enhancement in Brain Tumors Due to the presence of the blood–brain barrier (BBB), currently available MR contrast media do not leak into the brain tissue (Bart et al. 2000; Neuwelt 2004). The BBB consists of a complex of capillary endothelial cells, pericytes, and astroglial and perivascular macrophages and serves as an effective physical barrier to the entry of lipophobic substances into the brain (Neuwelt 2004). The BBB blocks all molecules except those that cross cell membranes by means of lipid solubility (such as oxygen, carbon dioxide, ethanol, and steroid hormones) and those that are allowed in by specific transport systems (such as sugars and some amino acids). Substances with a molecular weight higher than 180 Da, which include all available imaging contrast media, generally cannot cross the BBB (Gururangan and Friedman 2002). The integrity of the BBB can be altered by a variety of circumstances that increase the permeability, both for contrast media and drug delivery. A disruption of the BBB may be caused by osmotic means, e.g., steroids, biochemically by the use of vasoactive substances such as bradykinin, or even by localized exposure to focused ultrasound (Demeule et al. 2002; Kemper et al. 2004; Rautio and Chikhale 2004; Kroll and Neuwelt 1998).

In primary intraxial tumors, mainly gliomas, the BBB can be compromised by neovascularization and direct tumorous damage. Since non-neoplastic astrocytes are required to induce BBB features of cerebral endothelial cells, it is conceivable that malignant astrocytes have lost this ability due to dedifferentiation. Alternatively, glioma cells might actively degrade previously intact BBB tight junctions (Kido et al. 1991; Schneider et al. 2004). While the integrity of the barrier is often compromised within the tumor, this alteration in permeability is variable and dependent on the tumor type and size. Moreover, it is extremely heterogeneous in a given lesion (Earnest et al. 1988). Though the BBB is frequently leaky in the center of malignant brain tumors, the well-vascularized actively proliferating edge of the tumor, in the brain adjacent to tumor area, has been shown to have variable and complex barrier integrity. In secondary, metastatic intranaplastic astrocytomaxial tumors and extranaplastic astrocytomaxial tumors the vessels are different from normal cerebral vasculature and have no or strongly disturbed BBB (van den Bent 2004; Groothuis 2000). These entities normally have a strong enhancement pattern, with the whole tumor presenting as an enhancing mass. As brain edema is also thought to be due to breakdown of the BBB, one can expect a correlation between the degree of enhancement and the volume of the peritumoral edema. Holodny et al. (1999) and Pronin et al. (1997) studied this correlation in malignant gliomas as representative of primary intraaxial tumors and meningiomas as representative of extraaxial tumors. In their study, no correlation was found for meningiomas, which proved that the meningioma vessels have no BBB or no effect on the BBB in the surrounding brain tissue. For malignant gliomas a strong correlation was found, which provided evidence that the defect of the BBB is directly related to both the degree of lesion enhancement and the amount of edema. The interference may influence both the conventional contrast-enhanced MR imaging as well as some of the flow-dependent functional imaging techniques. The contrast enhancement patterns change substantially after corticosteroid treatment as first presented by Wilkinson’s group (2006). The enhancement is less both in intensity and size (Fig. 3.3.1). However, the concept of permeability across a disrupted or disturbed BBB in brain tumor patients has recently elicited interest for monitoring modern treatment strategies. First, changes in permeability may serve as a surrogate marker for other important physiologic processes in brain tumors, such as angiogenesis. Second, an understanding of permeability can elucidate the mechanisms by which therapeutic agents enter brain parenchyma. Third, an understanding of methods for increasing permeability can help in the development of methods to selectively alter the BBB to enhance drug delivery.

245

246

3  Brain, Head, and Neck

3.3.4.2  Contrast Media Dosage This topic is one of ongoing investigation in MRI and will be affected by further improvement of the hardware technology and changes in sequence design and more advanced contrast medium application strategies. For gadolinium (Gd)-DTPA, the pioneer agent now being used worldwide in more than 80 million applications (Knopp et al. 2006), the same minimum dose recommendation is printed in the package inserts for France, Germany, Italy, Japan, the United Kingdom, and the United States. However, different recommendations exist for the maximal dose. The maximal dose of 0.3 mmol/kg body weight did not gain approval in Japan or the United States. Nevertheless, it has been shown that a higher dose of Gd chelate-based contrast agents may help reveal more subtle disease states of the BBB and provide other improvements, such as an increase in the detection rate for metastatic lesions (Fig. 3.3.2) (Yu et al. 1994, 1995; Sze et al. 1998) or better tumor delineation (Fig. 3.3.3).

Several studies have been published on the detection and characterization of focal CNS lesions. Initial investigations by Yuh et al. (1994) demonstrated that a cumulative triple dose of 1 or 0.3 mmol/kg, as compared with a single dose of 0.05 or 0.1 mmol/kg, respectively, allowed detection of additional brain metastases 9 out of 29 patients. While they found significantly higher detection rates for small lesions, they found no difference in the detection rates for lesions larger than 10 mm. Detection was better with the application of a triple dose than with delayed imaging. Sze et al. (1998) compared single- and triple-dose contrast-enhanced MRI in screening for cerebral metastases in 92 patients. For all 70 patients with negative single-dose studies, the triple-dose studies depicted no additional metastases in terms of the standard of reference, which was panel review and long-term follow-up. No statistically significant difference was found between the results of the single- and triple-dose studies. For 5 (50%) of 10 patients with equivocal single-dose studies, the triple-dose study helped clarify the presence Fig. 3.3.1  Contrast-enhanced MRI before (a) and after (b) high-dose corticoid therapy. The intensity of contrast enhancement is substantially reduced after steroids. The rim-enhancing glioblastoma multiforme is also reduced in size

Fig. 3.3.2  Contrast-enhanced MRI in a patient with cerebral metastasis. Increasing the amount of contrast material from a single dose (0.1 mmol/kg) (a) to triple dose (0.3mmol/kg) (b) substantially increased the number of visible cerebral metastases

3.3  Intracranial Tumors

or absence of metastases. In 3 (25%) of 12 patients with a solitary metastasis seen on the single-dose study, the triple-dose study depicted additional metastases. In a separate analysis of interpretations by two blinded readers, use of triple-dose contrast significantly changed the distribution of results for only one reader, for whom the number of equivocal readings decreased while the number of false-positive readings increased. The authors therefore concluded that routine triple-dose contrast administration in all cases of suspected brain metastasis is not beneficial. The effect of higher dosage on lesion size in metastatic brain tumors was assessed in a study by Van et al. (1997). The contrast of brain metastases after cumulative doses of Gd chelate was quantified and compared in order to assess the clinical utility of high dosage in a series of 39 patients with metastatic brain tumors. The post–Gd MRI contrast doubled with dose escalation from 0.1 to 0.3 mmol/kg and also increased with lesion size, by a factor of 2.5 between metastases of 3 and 16 mm diameter. At 0.2 and 0.3 mmol/kg, the respective numbers of visible metastases increased by 15 and 43% compared with 0.1 mmol/kg (p < 0.0001). Image contrast figures differed significantly between doses (p = 0.018). Both the number of metastases detected and the image contrast are significantly higher when dose escalation is performed. It is likely that the number of detected metastases will increase further at Gd doses beyond 0.3 mmol/kg. Post–Gd MRI contrast increases with lesion size, to an extent that cannot be attributed to partial volume attenuation. Regarding the effect of field strength, no differences were described. In a study comparing 3 and 1.5T, cumulative triple-dose images of both field strengths were superior to standard field strengths. However, administration of a better contrast medium, such as gadodiamide, produces higher contrast between tumor and normal brain on 3 T than on 1.5 T, resulting in better detection of brain metastases and leptomeningeal involvement (Ba-Ssalamah et al. 2003).

Besides better detection, the high dose may also allow better characterization of gliomas. In suspected lowgrade tumors, a mild enhancement pattern might be better visualized, and a tumor recurrence may be better visualized as has been shown by Erickson et al (2002). In a prospective study protocol, they evaluated whether there exists a subset of brain tumors that demonstrate contrast enhancement with triple dose and magnetization transfer suppression (MTS) that do not enhance with standard imaging and standard contrast dose. In 15 patients with either newly diagnosed primary brain tumor or brain tumor that had been followed for more than 2 years T1-weighted MTS images without intravenous contrast, with 0.1 mmol/kg without MTS (single-dose [SD] images), and with additional 0.2 mmol/kg Gd and MTS (triple-dose [TD]/MTS) were obtained. None of the patients had enhancement on single dose imaging while six patients had areas of enhancement on triple dose MTS images. So far, it might be possible that those small areas of enhancement seen only with triple dose MTS might represent areas of higher-grade tumor that may benefit from a more intensive initial tumor therapy. 3.3.4.3  Contrast Agents Used at Different Field Strength Over the past decade, most clinical experience in the field of cerebral MR imaging has been with 1.5-T systems with a dose of 0.1 mmol/kg body weight of the conventional Gd chelates, as this combination seems to be an acceptable compromise between imaging expense and diagnostic sensitivity (Ba-Ssalamah et al. 2003; Haustein et al. 1992). The number of 3-T systems in clinical settings has been increasing over past few years, and systems operating at even higher field strengths are being used in clinical trials already. One of the main features of MR imaging at 3 T is the general gain in signal-to-noise ratio (SNR) compared Fig. 3.3.3  Contrast-enhanced MRI in a patient with malignant glioma at a single (a) and a triple (b) dose of contrast material. Note the stronger enhancement of the ring enhancing lesion. The volume appears to be increased and there is substantially more information about the intralesional heterogeneity

247

248

3  Brain, Head, and Neck Fig. 3.3.4  1.5- (a) and 3-T (b) contrast-enhanced MRI in a 58year-old patient with skull base meningioma. Both studies used an SE 3-mm transversal scan with identical acquisition parameters. The tumor enhancement is significantly stronger on the 3-T image, with a better background contrast

with that at lower field strengths. Therefore, one may anticipate that the increased SNR associated with a higher magnetic field will translate, at least to a certain degree, into a higher contrast-to-noise ratio (CNR) between enhancing and nonenhancing tissues. The increased CNR should improve the delineation of contrast agent–induced changes and, thus, could increase the sensitivity of detection of such signal intensity changes. In addition, the effectiveness of the T1-shortening effect of a Gd-based contrast agent depends on the baseline T1 relaxation time of local tissue. With the longer baseline T1 relaxation times brought about by a higher magnetic field strength, the T1-shortening effect of Gd-based contrast agents will be greater, as the relaxivity of such contrast agents changes only marginally between 1.5-T MR imaging and 3-T MR imaging (Elster 1997; Chang et al. 1994). Accordingly, the signal intensity changes caused by contrast enhancement observable in T1-weighted images should generally be stronger at 3 T than they are at 1.5 T (Fig. 3.3.4). 3.3.5  Intra-Axial Cerebral Tumors 3.3.5.1  Neuroepithelial Tumors 3.3.5.1.1  Astrocytic Tumors Astrocytic tumors account for up to 80% of glial neoplasms and refer to a diffuse infiltrating tumor originating from glial cells although a mixture with other neoplastic cells is not uncommon (Ohgaki et al. 2005; Behin et al. 2003; Cavenee et al. 2000). The tumor border on both imaging and histology is ill defined, with an infiltration that usually does not destroy the anatomic cerebral structures. The tumors often distort the structures with the appearance of cysts. The grading of astrocytomas is based on histologic features of the tumor as described above.

Under the recent WHO classification of primary intracranial tumors, low-grade gliomas would encompass grade I and grade II neuroepithelial tumors. The more common grade I tumors are pilocytic astrocytoma, dysembryoplastic neuroepithelial tumors (DNET), pleomorphic xantho-astrocytoma (PXA), neurocytoma, and ganglioglioma. The more common grade II tumors include astrocytoma, oligodendroglioma, and mixed oligoastrocytoma. This spectrum of discreet neuropathological entities is important since the grade I tumors generally can be cured by surgical excision and their symptoms very often alleviated (Burger et al. 2000). Conversely, the grade II tumors are generally incurable but have median survival times of >5 years (Cavenee et al. 2000; Steiber 2001). Some grade II gliomas are “diffuse” while others have relatively well-defined brain–tumor interfaces. Neuropathological diagnosis and tumor characteristics will therefore profoundly influence the impact of treatment strategies. Currently, even with the best MRI scanners, differentiation between grade I, II, and even III tumors is very difficult, therefore establishing tissue diagnosis can be important. Pilocytic Astrocytoma (WHO Grade I) Pilocytic astrocytoma (WHO grade I) is a grossly circumscribed, slow-growing, often cystic tumor that occurs primarily in children and young adults (Burger et al. 2000). The infratentorial form usually becomes manifest within the first two decades, while the supratentorial form peaks at the age of 20 (Salmon et al. 1994). They are rarely found in the fourth or fifth decades of life or later. Histologically, pilocytic astrocytomas are composed of varying proportions of compacted bipolar cells with Rosenthal fibers and loose-textured, multipolar cells with microcysts and granular bodies (Burger et al. 1994). This tumor is the most common glioma in children and represents 10% of cerebral and 85% of cerebellar astrocytic tumors. Occurring throughout the neuraxis, the preferred sites include

3.3  Intracranial Tumors Fig. 3.3.5  Typical appearance of an infratentorial pilocytic astrocytoma. The non-enhanced T1 (a) and T2 (b) present a partly solid and cystic tumor with surrounding edema and mass effect. The solid tumor parts present with a strong enhancement (c,d), accompanied by cystic non-enhancing areas

the optic nerve, optic chiasm/hypothalamus, thalamus and basal ganglia, cerebral hemispheres, cerebellum, and brain stem. Pilocytic astrocytoma is the principal CNS tumor associated with neurofibromatosis type 1 (NF1). Macroscopically, the tumor is a well-circumscribed mass that commonly has a large cyst and a focal mural nodule. The tumor can also be solid, with or without cystic degeneration. Microscopically, juvenile pilocytic astrocytoma demonstrates well-differentiated pilocytes with hair-like glial processes associated with microcysts that contain mucopolysaccharide material. The pilocytes are mixed with Rosenthal fibers, eosinophilic rod-shaped bodies and granular eosinophilic bodies, which are commonly found in indolent neoplasms. Capillary formation is usually present. On MRI, the tumors present with a different appearance upon their location (Lee et al. 1989; Mishima et al. 1992; Schneider et al. 1992; Takada et al. 1999) (Fig. 3.3.5). In the uncommon supratentorial location, the signal intensity is decreased on T1-weighted and increased on T2-weighted images (Fig. 3.3.6). Calcifications are hardly seen on MRI (Burger et al. 1996). The enhance-

ment patterns may vary from solid to rim-shaped enhancement, which makes it difficult to separate them from other astrocytic tumors (Burger et al. 1996; Strong et al. 1993: Beni-Adani et al. 2001). Lower-grade tumors tend to be homogeneous and well circumscribed. Peritumoral edema is mild, and no hemorrhage is present. Higher-grade tumors have more surrounding edema, are more heterogeneous in density, and may have areas of hemorrhage (Kuroiwa et al. 1999). Optic chiasm hypothalamic gliomas (Fig. 3.3.7) cannot be separated by their site of origin and are considered as a single entity (Kornreich et al. 2001; Alvord and Lofton 1988). On T1-weighted images, the signal intensity is low. On T2-weighted images, the signal intensity is generally increased. The T2 signal intensity increase may extend as far as the optic radiations, but it is not correlated directly with the presence of tumor. Enhancement is similar to that on CT scans. Fat-saturated T1-weighted postcontrast MRI of the intraorbital optic nerves is a sensitive method for demonstrating the tumor. Vermian pilocytic tumors are often associated with hydrocephalus and present with a mixture of solid and cystic masses (Schneider et

249

250

3  Brain, Head, and Neck

Fig. 3.3.6  Supratentorial pilocytic astrocytoma in a 35-year-old patient. Unenhanced T1 (a) and FLAIR (b) present a homogeneous mass lesion with centrally located cystic areas. Solid parts of the tumor present with enhancement (c). Histology confirmed a pilocytic tumor Fig. 3.3.7  Optical glioma in a 7-year-old patient. Note the hyperintensity on T2 in projection of the left optic nerve (a). Isointense to normal brain on FLAIR (b) and unenhanced T1 (c), the tumor shows marked enhancement within the orbits and the chiasm

3.3  Intracranial Tumors

al. 1992; Pencalet et al. 1999). The mainly solid tumors, which account often for less than 10% of subjects present with a strong homogeneous or heterogeneous enhancement. The more common mixed solid and cystic tumors show in approximately 50% simple cysts with a single strong enhancing mural nodule. However, multifocal localization is possible (Fig. 3.3.8). The cyst wall, which is histologically tumor free, does not enhance. Tumors presenting multiple solid parts and cysts are actually necrotic tumors with a cyst-like appearance. The periphery or cyst wall contains tumor and therefore enhances. Diffuse Astrocytoma (WHO Grade II) Diffuse astrocytoma (WHO grade II), also known as low-grade or fibrillary astrocytoma, is characterized by a slow growing and infiltrating mass lesion (Kitange et al. 2003a). Histologically, diffuse astrocytomas are composed of well-differentiated fibrillary or gemistocytic neoplastic astrocytes. This type of tumor typically affects young adults and has a tendency for malignant progression to anaplastic tumors and, ultimately, glioblastoma. Diffuse astrocytomas represent 35% of all astrocytic brain tumors (Kitange et al. 2003a). They may be located in any region of the CNS but most commonly develop in the cerebrum. Three histologic variants include fibrillary astrocytoma, gemistocytic as-

trocytoma, and protoplasmic astrocytoma (Pencalet et al. 1999). On MRI, grade II astrocytomas present as diffuse infiltrating T2-hyperintense mass lesions without low signal on T1 and missing contrast enhancement in the majority of cases (Fig. 3.3.9) (Earnest et al. 1988; Holland et al. 1985; Atlass et al. 1988; Marks et al. 1997; Phillipon et al. 1993). In the literature, about 20% of grade II tumors have been described as enhancing (Fig. 3.3.10). The tumors show a slow growth rate and in some cases, secondary lesions without a tumor bridge (Fig. 3.3.11). However, in up to 30% of what appear to be low-grade gliomas are in fact grade III or grade IV tumors presenting no or only faint enhancement patterns (Arienti et al. 2001; Bampoe et al. 1999). Spectroscopic analysis of the MRIs can help to clarify diagnosis but is certainly not definitive (Fig. 3.3.12). Radiological diagnosis of a “lowgrade glioma” can therefore pose a dilemma. Anaplastic Astrocytoma (WHO Grade III) WHO grade III tumors arise from a diffuse astrocytoma (secondary grade III tumor) or may arise de novo without indication of a less malignant precursor (primary anaplastic tumor) (Kleihues et al. 2000b). Histologically, this tumor shows increased cellularity, distinct nuclear atypia, and marked mitotic activity when compared with Fig. 3.3.8  Multifocal pilocytic astrocytoma. In this 12-yearold patient, multiple foci of the histologically confirmed tumor were present in the temporal lobe, the insula (arrow), and in the cerebellum (arrow). In this case the tumor pattern is a more solid one with a homogeneous enhancement of variable intensity

251

252

3  Brain, Head, and Neck

Fig. 3.3.9  Low-grade (WHO grade II) astrocytoma in a 26-year-old patient with a history of seizures. The MRI shows the typical low-grade appearance with diffuse infiltrating tumor masses of high signal intensity on the FLAIR (a) and T2-weighted (b) images. After contrast material application (c), no enhancement is visible Fig. 3.3.10  Low-grade (WHO grade II) astrocytoma in a 37-year-old patient with dysphasia. The tumor which was histologically confirmed as a low-grade tumor presents the typical appearance of astrocytoma on T1-weighted (a), T2-weighted (b), and FLAIR (c) images, while there is a mild enhancement visible on the post-contrast T1-weighted image (d)

3.3  Intracranial Tumors

Fig. 3.3.11  Secondary tumor manifestation of a low-grade astrocytoma in a 38-year-old patient. The patient had a recent resection of a frontoparietal grade II tumor (a) and presented with a secondary tumor manifestation at the border of the fourth

ventricle. The secondary lesion shows high signal intensity on T2-weighted and FLAIR (b) images and no enhancement (c). Histologic biopsy confirmed a low-grade astrocytoma (WHO grade II)

a diffuse astrocytoma. Anaplastic astrocytomas possess an intrinsic tendency to progress to glioblastoma. The mean age at biopsy is approximately 41 years. This tumor primarily affects the cerebral hemispheres. It has a high frequency of TP53 mutations, which is similar to that of diffuse astrocytoma. Chromosomal abnormalities are nonspecific. Many of the genetic alterations seen in anaplastic astrocytomas involve genes that regulate cell cycle progression (Kitange et al. 2003b). In neuroimaging, grade III tumors appear much more aggressive than grade II tumors. Heterogeneous signal patterns and a pathologic contrast enhancement are characteristic for anaplastic tumors (Graif et al. 1985; Kleihues and Ohgeki 2001; Muroff and Runge 1995; Provencale et al. 2006a; Pierallini et al. 1997). On T2 imaging the hyperintense tumor signal cannot be differentiated from vasogenic changes, therefore all of the T2 signal changes are defined as tumor tissue. Intratumoral heterogeneity may also be caused by a prominent hypervascularity of the tumors. Large cystic areas, if present, are an indication of a further malignization of the tumor toward glioblastoma multiforme. Calcifications are best seen on CT while intratumoral hemorrhage is better seen on MRI. Contrast enhancement can be focal or nodular (Figs. 3.3.13, 3.3.14) (Muroff and Runge 1995). Ring-like enhancement is suspicious for a further malignization.

tologically, this tumor is an anaplastic, cellular glioma composed of poorly differentiated, often pleomorphic astrocytic tumor cells with marked nuclear atypia and brisk mitotic activity. Secondary glioblastoma is the term used to describe a glioblastoma developed from a diffuse astrocytoma or an anaplastic astrocytoma. Glioblastoma is the most frequent brain tumor and accounts for approximately 12–15% of all brain tumors and 50–60% of all astrocytic tumors. The peak incidence occurs between the ages of 45 and 70 years. Glioblastoma primarily affects the cerebral hemispheres. Two histologic variants include giant cell glioblastoma and gliosarcoma. Glioblastomas have been associated with more specific genetic abnormalities than any other astrocytic neoplasm, but none are specific to it. Amplification of the epidermal growth factor receptor locus is found in approximately 40% of primary glioblastomas but is rarely found in secondary glioblastomas; mutations of the PTEN gene are observed in 45% of primary glioblastomas and are seen more frequently in primary glioblastomas than in secondary glioblastomas (Kitange et al. 2003b). Loss of heterozygosity (LOH) of chromosome 10 and loss of an entire copy of chromosome 10 are the most frequently observed chromosomal alterations. Glioblastomas are seen in mismatch repair–associated Turcot syndrome type 1. Glioblastomas are among the most aggressively malignant human neoplasms, with a mean total length of disease in patients with primary glioblastoma of less than 1 year. Mutation of the PTEN gene is associated with a poor prognosis in a subset of patients with gliomas. The usual appearance of a GBM or anaplastic astrocytoma on MRI is that of a contrast-enhancing lesion causing mass effect (Pronin et al. 1997; Hammond et

Glioblastoma Multiforme (WHO Grade IV) Malignant astrocytoma (WHO grade IV), also known as glioblastoma multiforme (GBM), may develop from a diffuse astrocytoma or an anaplastic astrocytoma but more commonly presents de novo without evidence of a less malignant precursor (Kleihues et al. 2000b). His-

253

254

3  Brain, Head, and Neck Fig. 3.3.12  MR spectroscopy in a 57-year-old patient with anaplastic astrocytoma. The tumor was diagnosed as a non-enhancing mass lesion suspicious for low-grade (WHO grade II) astrocytoma. Localization image for MR spectroscopy (a). The spectrum (b) shows a typical tumor pattern with elevated choline and reduced NAA. The presence of a lactate peak (arrow) is suspicious for an anaplastic or malignant transformation, which was histologically confirmed in this case

al. 1996). It is often very difficult to distinguish a GBM from a contrast-enhancing anaplastic astrocytoma lesion by MRI alone. GBMs characteristically spread along the white matter tracts of the corpus callosum to invade the contralateral hemisphere (“butterfly” appearance) (Fig. 3.3.15). GBMs usually exhibit heterogeneous signal intensity on both T1- and T2-weighted images caused by cysts, necrosis, and hemorrhage commonly seen with GBM (Fig. 3.3.16). Up to 95% of GBMs demonstrate contrast

enhancement, and they are usually associated with high signal on T2-weighted imaging. Anaplastic astrocytomas also exhibit heterogeneous signal intensity on noncontrast T1- and T2-weighted imaging, but the frequency of contrast enhancement is less than in GBM. Large GBM lesions may demonstrate prominent areas of central necrosis with peritumoral edema (Fig. 3.3.17), while others may show metastasis. In the diagnostic workup of GBM MRI also plays an important role for the planning of bi-

3.3  Intracranial Tumors Fig. 3.3.13  Anaplastic (grade III) astrocytoma in a 45-yearold patient. Anaplastic tumors present with a more heterogeneous signal pattern as seen on the T2-weighted (a) and FLAIR (b) images. Hemorrhage may be present as in this case and can be best seen on the unenhanced T1weighted image (c). After contrast media the enhancement patterns can be more focal, as in this case (d) or nodular as in the following case (Fig. 3.3.14)

Fig 3.3.14  Nodular enhancement patterns in a patient with anaplastic grade III astrocytoma. While there was no hemorrhage present in this case (a), there were multiple nodular-enhancing areas present (b)

opsy. The goal is to identify the most aggressive parts of the tumor and to differentiate between primary and secondary tumors. Pleomorphic Xanthoastrocytoma (WHO Grade I) Pleomorphic xanthoastrocytoma is a rare variant of an astrocytic tumor composed of pleomorphic and lipidized cells expressing glial fibrillary acidic protein (GFAP) (Kepes et al. 2000). This tumor accounts for less than 1%

of all astrocytic neoplasms, typically develops in children and young adults, and commonly involves the cerebrum and meninges. The tumor typically presents with a stable histology (grade I) and has a good prognosis. No specific cytogenetics or molecular genetics exist with this tumor. In histology, the tumor is well circumscribed with a low proliferation index in microscopy. Pleomorphic xanthoastrocytomas typically occur supratentorially in the cerebral hemispheres and present as hypointense

255

256

3  Brain, Head, and Neck Fig. 3.3.15  Glioblastoma multiforme in a 57-year-old patient with typical butterfly appearance. It presents as a mass lesion with spread over the corpus callosum into both frontal lobes (a) and spread along the midline into the insula (b). On FLAIR (c) imaging the extent of the tumor can be best described, especially those parts which grow along the ventricular walls. In this case no necrosis or hemorrhage is obvious (d). The tumor presents with a typical strong enhancement (e), which had grown substantially from a nodular lesion seen on MRI three months before (f)

on T1-weighted MRI, with the cystic tumor components having high signal intensity (Fig. 3.3.18) (Tien et al. 1992; Lipper et al. 1993; Pierallini et al. 1999). Subependymal Giant Cell Astrocytoma (SEGA) (WHO Grade I) Giant cell astrocytoma is a benign, slow-growing mass lesion typically arising in the wall of the lateral ventricles, close to the foramina of Monroe and composed of large ganglioid astrocytes (Wiestler et al. 2000a). SEGA occurs almost exclusively in patients with tuberous sclerosis complex (TSC); its incidence ranges from approximately 6–16% of patients with TSC. SEGA typically occurs during the first two decades of life. On imaging, the lesion

has a classic astrocytic appearance with common calcifications. On MRI, it can be identified by its typical location and the classical findings of tuberous sclerosis. The lesion is normally hyperintense on T2 with heterogeneous signal if calcification is present (Fig. 3.3.19). The hypo- to isointense lesion presents on T1-weighted imaging with a moderate enhancement (Tsuchida et al. 1984). 3.3.5.1.2  Oligodendroglial Tumors The most common genetic alteration in oligodendroglial tumors is LOH on the long arm of chromosome 19q, the incidence of which ranges from 50% to more than

3.3  Intracranial Tumors

Fig. 3.3.16  Large GBM in a 33-year-old patient presenting with a focal seizure. First diagnostic MRI presents a large mass lesion, with strong enhancement (a), and a heterogeneous signal pattern on both FLAIR (b) and T2-weighted MRI (c)

Fig. 3.3.17  Glioblastoma multiforme in a 68-year-old patient. Large, partly cystic tumor with mass effect on the lateral ventricles (a). After contrast administration, the tumor presents with a ring-shaped enhancement pattern, displaying the border of the macroscopic tumor parts (b)

Fig. 3.3.18  Xanthoastrocytoma in juvenile patients. The strong enhancing tumor with a subjacent cyst (a) is the typical MRI finding. The cystic components normally present with a high signal intensity (b) and can hardly be suppressed by FLAIR techniques because of the protein content

257

258

3  Brain, Head, and Neck

Fig. 3.3.19  Giant cell astrocytoma in a patient with tuberous sclerosis (Bourneville-Pringle disease). One can observe multiple hypointense lesions within the lateral left ventricle and a large mass of mixed intensity in the anterior aspect of the right ventricle (a–d). The hypointensity on both T2-weighted (a,b) and FLAIR (c, d) images represents calcification. After contrast media application (e–g), the giant cell tumor shows a moderate to marked enhancement (arrows), while the tuberous changes only present no or mild enhancement (arrows)

3.3  Intracranial Tumors

80% (Reifenberger et al. 2000a). Median postoperative survival times have been reported to range from 3 to 10 years for all histologic grades of oligodendroglial tumors (Kitange et al. 2001). Oligodendroglioma (WHO grade II) is a well-differentiated tumor, composed predominantly of cells morphologically resembling oligodendroglia, which grows diffusely in the cortex and white matter (Reifenberger et al. 2000a). This tumor is uncommon and accounts for approximately 50% of oligodendroglial tumors and between 5 and 18% of all gliomas. Most oligodendrogliomas occur in adults, with a peak incidence in the fifth and sixth decades of life, with the subcortical area as the most common location. Compared to patients with astrocytoma, patients with oligodendroglioma respond better to radiation therapy and chemotherapy. Temozolomide appears to have activity in low-grade oligodendrogliomas and oligoastrocytomas combined with a 1p allelic loss.

Clinical improvement was noted in 51% of patients, and the radiologic response rate was 31% (Hoang-Xuan et al. 2004). Anaplastic or malignant oligodendrogliomas may occur with an increased cellularity, cell pleomorphisms and fast growth rates, and a bad prognosis. In a clinical environment, one only differentiates between low grade and anaplastic oligodendroglioma. Imaging studies of oligodendrogliomas are useful for the differential diagnosis based on location, lesion characteristics, presence of calcification, and mass effect (Lee and Van Tassel 1989; Shaw et al. 1992). The typical imaging finding of oligodendroglioma is the calcification identified on unenhanced CT. The calcification is present in more than 85% of tumors. In MRI, oligoastrocytomas appear iso- to hypointense on T1 with very low intensity, representing the calcified areas (Fig. 3.3.20). On T2 or FLAIR, the tumors are typically hyperintense with a Fig. 3.3.20  Oligodendroglioma in a 39-year-old patient. The tumor typically presents as a heterogeneous mass lesion with cystic components and hypointensities both on T1- (a) and T2-weighted imaging (b), representing focal calcifications. Cortical thickening is best seen on T2-weighted and FLAIR imaging (c). Contrast enhancement may vary and appears in a cystic ring-shaped pattern in this case (d)

259

260

3  Brain, Head, and Neck Fig. 3.3.21  Anaplastic oligodendroglioma in a 35-year-old patient. On CT (a) the tumor presents as a CSF isodense lesion with multiple enhancing lesions and typical calcifications. The T2-weighted MRI (b) shows a lobulated cystic tumor with CSFlike signal. The areas of calcifications are hypointense. FLAIR (c) and proton density (d) images clearly delineate the hyperintense central solid tumor parts from the cystic CSF-filled components. The unenhanced T1-weighted image (e) does not show any hemorrhage, with a strong enhancement of the solid parts after contrast (f)

3.3  Intracranial Tumors

not-well delineated margin. Compared to astrocytomas, the lesions present with a lower signal on T2, which is related to the high cellular density. The most useful finding, however, is the typical cortical infiltration and marked cortical thickening. Small cystic lesions and hemorrhage is also a common finding. Enhancement is common and tends to be more intense in the anaplastic and malignant form (Fig. 3.3.21). 3.3.5.1.3  Mixed Gliomas Oligoastrocytoma (WHO grade II–IV) is composed of two distinct neoplastic cell types that morphologically resemble tumor cells in oligodendroglioma and diffuse astrocytoma (Reifenberger et al. 2000b). Estimates of incidence vary greatly. In one large US study, only 1.8% of gliomas were classified as mixed gliomas. The median age of patients is reported to range from 35 to 45 years. This tumor has a predilection for the cerebral hemispheres; the frontal lobes are most commonly affected, followed by the temporal lobes. These types of tumors contain no specific genetic alterations or chromosomal abnormalities; however, about 30% of oligoastrocytomas have genetic aberrations commonly found in astrocytic tumors. One study reported a median survival time of 6.3 years. Temozolomide appears to have activity in low-grade oligoastrocytomas and oligodendrogliomas combined with a 1p allelic loss. Clinical improvement was noted in 51% of patients, and the radiologic response rate was 31%. Oligoastrocytomas appear similar to oligodendrogliomas on both CT and MRI with common calcifications. Cystic changes are also common; hemorrhage and edema are rare. The signal intensities are hypointense on T1 and hyperintense on long TR/TE sequence types. Contrast enhancement is seen in all grades with a stronger enhancement in the anaplastic and malignant forms. 3.3.5.1.4  Gliomatosis Cerebri Gliomatosis cerebri is an uncommon low-grade primary infiltrative brain tumor manifested by diffuse overgrowth of neoplastic glial cells in at least two lobes of the brain with relative preservation of the underlying cytoarchitecture and sparing of the neurons (Dunn, Jr., and Kernohan 1956; Artigas et al. 1985). It represents a more extensive form of diffuse glioma and presents with an enlargement of the lobes and hemispheres without distortion of the normal brain anatomy. Diagnosis of this disease, which occurs in the second and third decades of life, is difficult because the symptoms are variable and nonspecific. Mental deterioration and seizures are the main clinical manifestations. The disease is usually slowly progressive, although a secondary transformation into glioblastoma multiforme can occur (Romero 1988).

Advances in imaging and pathologic evaluation have improved the detection and diagnosis of this disease, which in the past was diagnosed only at autopsy (Keene et al. 1999). MR is the imaging method of choice in gliomatosis cerebri because it reveals diffuse infiltration as increased signal intensity on T2-weighted images (Fig. 3.3.22) (Keene et al. 1999; Ponce et al. 1988). MRI, especially with the use of FLAIR, shows better than CT the continuity of abnormal tissue between the lesions and therefore enables a more correct estimation of disease extent, assisting accurate diagnosis. Enhancement is rare, but may occur in malignant transformation of the lesion. 3.3.5.1.5  Ependymal Tumors Ependymal tumors are neoplasms that are composed predominantly of neoplastic ependymal cells. They are common in children, accounting for about 10% of pediatric CNS neoplasms and 5% of all intra-axial tumors. The indolent variant myopapillary ependymoma, a grade I tumor, occurs along the filium terminale. Most ependymomas are found in the infratentorial compartment. Myxopapillary ependymoma (WHO grade I) is a slow-growing astrocytic tumor, histologically characterized by tumor cells arranged in a papillary pattern around vascularized mucoid stromal cores (Wiestler et al. 2000b). In a large series of cases of ependymal tumors, 13% were found to be of the myxopapillary type. The average age at presentation is approximately 36 years. This tumor almost exclusively occurs in the conus–cauda–filum terminate region of the spinal cord. Ependymoma is a WHO grade II slow-growing tumor of children and young adults that originates from the wall of the cerebral ventricles or from the spinal canal and is composed of neoplastic ependymal cells. These types of tumors account for 3–5% of all neuroepithelial tumors and for 30% of those in children younger than 3 years. Ependymomas are the most common neuroepithelial neoplasms in the spinal cord and make up 50–60% of spinal gliomas. These types of tumors occur at any site in the ventricular system and in the spinal canal; they develop most commonly in the posterior fossa and in the spinal cord, followed by the lateral ventricles and the third ventricle. Histologic variants include cellular ependymoma, papillary ependymoma, clear cell ependymoma, and tanycytic ependymoma. MR imaging shows the heterogeneity of the tumors, which reflects areas of necrosis, bleeding, and cysts (Spoto et al. 1990). In some cases, calcifications are present; however, those are small and hardly seen on MRI. On T2 the tumors are not as high in signal intensity as other gliotic tumors due to the high cellular density (Fig. 3.3.23). Hydrocephalus is a common finding due to the extension of the tumors in the ventricles and through the foramina.

261

262

3  Brain, Head, and Neck Fig. 3.3.22  Gliomatosis cerebri in 62- (a) and 58-year-old (b) patients. The tumor appears as a homogeneous mass lesion over multiple lobes on CT (a) and MRI (b). In the majority of cases there is no obvious enhancement; however, there are several reports of malignization into a glioblastoma multiforme. FLAIR imaging best shows the diffuse infiltration of this tumorous change into multiple lobes (c,d)

Subependymoma, another WHO grade I tumor is a slow-growing glial neoplasm that is typically attached to the ventricular wall (Wiestler et al. 2000c). In a large series of cases, this histologic type accounted for 8.3% of ependymal tumors. It is an unusual highly differentiated neoplasm that is considered a variant distinct from ependymoma. This tumor occurs most frequently in middle-aged and elderly males and is rarely found in children. Consistent cytogenetic abnormalities have not been found. Subependymoma carries a good prognosis; surgical removal is usually curative. A typical imaging finding in subependymoma is its location as an intraventricular lesion (Hoeffel et al. 1995) (Fig. 3.3.24). Differentiation from ependymomas is difficult especially in large tumors, which tend to be as heterogeneous as described above. Small masses, however, are homogeneously hyperintense on T2-weigthed images. Enhancement is uncommon and can be used for differentiation from other pathologies with intraventricular manifestation. Anaplastic ependymoma (WHO grade III) is a malignant glioma of ependymal origin with accelerated growth and an unfavorable outcome, particularly in children (Wiestler et al. 2000d). Incidence data vary con-

siderably. No specific genetic alterations for this tumor are known. Prognostic correlations between histology and clinical outcome have been inconsistent. In a large series, no correlation between survival times and classic histopathological findings of malignancy were observed. Based on imaging findings the anaplastic form presents with a stronger enhancement but cannot be differentiated based on its morphologic characteristics. 3.3.5.1.6  Neuroepithelial Tumors of Uncertain Origin and Mixed Neuronal–Glial Tumors These types of tumor are rare and include astroblastoma, choroid glioma, gliomatosis cerebri, and a list of mixed tumors (see Table 3.3.1). The imaging findings are unspecific in most of the entities despite the abovedescribed gliomatosis. The limited number of reports that exist describe signal characteristics as in glial tumors with high signal intensity on long TR/TE sequences and low signal on T1 (Naidich and Zimmerman 1984). Contrast enhancement varies according to the histological grade.

3.3  Intracranial Tumors Fig. 3.3.23  Ependymoma in a 16-year-old patient. The tumor appears as a large mass lesion (a,b) with rim enhancement and cystic necrosis (c,d). In this case the tumor is separate from the ventricular system and centers in the parenchyma. Ependymomas arise from ependymal cell rests in the parenchyma

3.3.5.1.7  Neuronal and Mixed Neuronal–Glial Tumors This is a heterogenous group of rare CNS tumors that originate from neuroepithelial cells with some glial components. Gangliocytoma (WHO grade I) and ganglioglioma (WHO grade I or II) are well-differentiated and well-circumscribed, slow-growing neuroepithelial tumors comprised of neoplastic, mature ganglion cells, either alone (gangliocytoma) or in combination with neoplastic glial cells (ganglioglioma) (Nelson et al. 2000a). Both types of tumors are relatively uncommon and generally have a favorable prognosis. Most tumors are supratentorial and the temporal lobe is the most common site. Infratentorial tumors can also be found. Anaplastic gangliogliomas (WHO grade III), i.e., gangliogliomas that show anaplastic features in their glial component, are sometimes seen; rare cases exhibit WHO grade IV (glioblastoma) changes in the glial component. These types of tumors account for 0.4% of all CNS tumors, 1.3% of all brain tumors, and can occur at any age. These types of tumors may occur throughout the CNS; most are supratentorial and involve the temporal lobe.

They present unspecific imaging findings with inhomogeneous hyperintensities on T2 and contrast enhancement in some cases (Fedi et al. 2004). Normally both cystic and solid components can be detected. The main differential diagnoses are pilocytic astrocytomas. The calcifications that are present in some of the cases might be helpful in this respect. Desmoplastic infantile astrocytoma (DIA) and desmoplastic infantile ganglioglioma (DIG) (WHO grade I) are large, cystic tumors of infants that involve the superficial cerebral cortex and leptomeninges, often attached to dura (Nikas et al. 2004). These rare neoplasms typically occur within the first 2 years of life. The lesions are typically partially cystic, and contain a variable neuronal component in addition to neoplastic astrocytes. The histology of the tumors is responsible for the heterogeneous appearance on MRI. Central neurocytoma (WHO grade I, II) is a recently described relatively benign intraventricular neoplasm that is composed of round cells with neuronal differentiation (Figarella-Branger et al. 2000). In a large surgical series, incidence ranged from 0.25 to 0.5% of all brain tumors. Almost 75% of these types of tumors are diagnosed between the ages of 20 and 40 years, with hydrocepha-

263

264

3  Brain, Head, and Neck Fig. 3.3.24  Subependymoma in a 55-year-old patient with a typical intraventricular location. Due to its high signal intensity the lesion is hardly seen on T2-weighted imaging (a) but is clearly visible on FLAIR (b) imaging. The lesion is isointense on the native T1-weighted image (c), and only a faint enhancement in the anterior part of the tumor became obvious

lus as the leading clinical finding. The clinical course of central neurocytoma is benign; the treatment of choice is complete surgical resection. On imaging studies, these tumors present typical characteristics with intraventricular location and attachment of the septum pellucidum (Wichman et al. 1991; Bolen et al. 1989). While calcifications can only be shown with CT, the mass lesions are homogeneous and isointense to gray matter on T2-weighted MRI (Fig. 3.3.25). Enhancement can be seen in most of the cases, which allows a differential diagnosis from heterotopic gray matter. Dysembryoplastic neuroepithelial tumor (WHO grade I) is a very uncommon, slow-growing benign, usually supratentorial, neuronal–glial neoplasm that occurs primarily in children and young adults with a long-standing history of partial seizures (Dauman-Duport et al. 2000). In one study, almost 90% of lesions associated with drug-resistant seizures were found to be dysembryoplastic neuroepithelial tumors. This tumor may develop in any part of the supratentorial cortex, but it has a predilection for the temporal lobe where a high intensity mass lesion on T2-weighted scans is the typical MRI finding (Fig.

3.3.26). Some dysembryoplastic neuroepithelial tumors may present cystic components and calcifications, which makes it difficult to distinguish them from oligodendrogliomas or oligoastrocytomas. 3.3.5.1.8  Nonglial Tumors Most nonglial tumors are considered intra-axial tumors. Nonglial tumors include a variety of rare entities including embryonal tumors and tumors of the pineal gland. Choroid plexus neoplasms are categorized as extra-axial tumors. 3.3.5.1.9  Embryonal Tumors Ependymoblastoma (WHO grade IV) or supratentorial primitive neuroectodermal tumor is a rare, malignant, embryonal brain tumor that occurs in neonates and young children (Becker et al. 2000). Ependymoblastomas are often large and supratentorial and generally relate

3.3  Intracranial Tumors Fig. 3.3.25  Central neurocytoma in a 16-year-old patient. The tumor appears to be isointense to gray matter on the T2-weighted images (a,b), which is a typical finding. After contrast a nodal enhancement within the large tumor mass becomes visible (c). The enhancement allows the differentiation form heterotopia

Fig. 3.3.26  Dysembryoplastic neuroepithelial tumor (DNET) in a 12-year-old patient. The tumor shows a mild hyperintensity both on T2-weighted (a) and FLAIR (b) images, with a small cystic component which is best displayed on FLAIR and T1-weighted (c) images. The tumor is in the temporal lobe, a typical location. After contrast media application (d), no enhancement could be detected

265

266

3  Brain, Head, and Neck

to the ventricles, although they do occur at other sites. These types of tumors grow rapidly, with craniospinal dissemination, and have a fatal outcome within 6 to 12 months of diagnosis. There are only a limited number of imaging studies available in this tumor. A recent study by Chawla et al. (2007) found that ependymoblastomas present as large, vascular, and hemorrhagic tumors. Due to the high cellular density, the tumors appear isointense to gray matter with a strong enhancement pattern. Medulloblastoma (WHO grade IV) is a malignant, invasive embryonal tumor of the cerebellum that occurs primarily in children, has a predominantly neuronal differentiation, and has a tendency to metastasize via CSF pathways (Roberts et al. 1991; Giangaspero et al. 2000). The annual incidence is 0.5 per 100,000 children younger than 15 years accounting for 20–25% of all intracranial tumors in children. In adulthood, 80% of medulloblastomas occur in people aged 21–40 years. These types of tumors rarely occur beyond the fifth decade of life. Medulloblastomas have been diagnosed in several familial cancer syndromes, including TP53 germline mutations, the nevoid basal cell carcinoma syndrome (NBCCS), and Turcot syndrome type 2. The tumor is the one who most likely tend to metastasize within and outside the neuroaxis, not uncommon to bone marrow or lymph nodes. Medulloblastomas normally occur in the midline, characteristically filling the fourth ventricle (Fig. 3.3.27). This is also the characteristic appearance on MRI: an intraventricular mass lesion in a midline or paramedian location with isointense signal to gray matter on T2-weighted imaging (Maleci et al. 1992; Koci et al. 1993; Bourgouin et al. 1992). The enhancement is intense with some heterogeneity and sometimes ring enhancement (Fig. 3.3.28). Subarachnoid seeding is common in medulloblastomas, occurring in up to 33% of all patients at the time of initial diagnosis (David et al. 1997) (Fig. 3.3.29). Some investigators believe that the prevalence of CSF seeding

may actually be much higher and perhaps present in all patients with the disease. Ventriculoperitoneal shunt involvement is common (20% of cases) and may lead to metastatic spread in the abdominal cavity. Numerous studies have shown that patients with evidence of CSF spread have a poorer prognosis compared with those in whom it is absent (Meyers et al. 2000). Therefore, its detection is crucial to optimal patient management, and those who review these imaging studies must be aware of its imaging manifestations. Supratentorial primitive neuroectodermal tumor (PNET or SPNET) (WHO grade IV) is an embryonal tumor in the cerebrum or suprasellar region that is composed of undifferentiated or poorly differentiated neuroepithelial cells, which have the capacity for differentiation along neuronal, astrocytic, ependymal, muscular, or melanocytic lines (Rorke et al. 2000). Synonyms include cere­­­ bral medulloblastoma (see above), cerebral neuroblastoma, cerebral ganglioneuroblastoma, blue tumor, and primitive neuroectodermal tumor. This is a rare tumor that occurs in children (mean age, 5.5 years); a precise incidence has not been determined. On MR imaging all PNETs were found to be either hypointense or isointense to normal white matter on T1-weighted images. In the study by Chawla et al. (2007), 10 of the 12 tumors were either isointense or hypointense on T2-weighted images, and 11 were isointense on FLAIR images. Patients with SPNETs had large, vascular, and hemorrhagic tumors. On DWI, all PNETs were hyperintense and had restricted apparent diffusion coefficient. MRS (two patients with medulloblastoma and one with a SPNET), showed elevated choline, decreased N-acetyl aspartate, and a small taurine peak in all three patients. Intraspinal tumor dissemination, visible as uniform or nodular enhancement coating the conus medullaris, was detected in 6 of 12 patients, 2 of whom also had intracranial dissemination.

Fig. 3.3.27  Medulloblastoma in a 6-year-old patient, in a typical midline location filling the fourth ventricle. The inhomogeneous mass is inhomogeneous on T1-weighted imaging (a) with a faint and inhomogeneous enhancement pattern (b,c)

3.3  Intracranial Tumors Fig. 3.3.28  Medulloblastoma in a 4-year-old patient. On T2weighted imaging, besides the typical findings (a,b) a heterogeneous enhancement pattern could be observed in this case (c,d)

3.3.5.1.10  Pineal Parenchymal Tumors and Cysts A diverse group of tumors may arise from this area, which comprises the pineal gland, the posterior third ventricle, the aqueduct and in some case also the basal ganglia. Vascular lesions may also be present in the pinealis region. Pineal parenchymal tumors arise from pineocytes or their precursors, and they are distinct from other pineal gland neoplasms such as astrocytic and germ cell tumors. Pineocytoma (WHO grade II) is a slow-growing pineal parenchymal neoplasm that primarily occurs in young adults (Mena et al. 2000a). Pineocytomas account for less than 1% of all brain tumors and comprise approximately 45% of all pineal parenchymal tumors. Adults aged 25–35 years are most frequently affected. No specific cytogenetic abnormalities or molecular genetics

exist with this tumor. The 5-year survival rate has been reported to be as great as 86%. Pineoblastoma (WHO grade IV) is a highly malignant primitive embryonal tumor of the pineal gland that manifests primarily in children (Mena et al. 2000b). Pineoblastomas are rare brain tumors that make up approximately 45% of all pineal parenchymal tumors. This tumor appears histopathologically similar to medulloblastoma, with focal areas of hemorrhage and necrosis. It tends to metastasize through the subarachnoid pathways into the meninges and subependymal space. The diagnosis of the exact histopathological type of pineal tumor is not possible in most cases; however, the main clinical question is the differentiation from cystic lesion in the pineal region. Mass lesions are normally solid with a marked tumor enhancement (Fig. 3.3.30). The signal is isointense to gray matter on T2-weighted imaging

267

268

3  Brain, Head, and Neck Fig. 3.3.29a,b  Metastatic spread in a 14-year-old patient with medulloblastoma. Note the small enhancing nodules along the cervical and thoracal myelon (b) (arrows)

and hypointense on the T1. Pineocytomas, with a high amount of cytoplasm are generally more intense on T2. Calcification may occur but the amount is normally too small to allow signal changes in MRI. The most common tumor is the benign pinealis cyst, which can be found in approximately 5% of people (Lee et al. 1987). Due to the high frequency and the often-unspecific clinical presentation, even in large cystic lesions, the main role of MRI is to distinguish between cysts and pineal neoplasms. Cysts are normally well circumscribed with a typical cystic appearance: high signal intensity on T2 and low CSF-like signal intensity on T1 (Fig. 3.3.31). Depending on the protein content of the cyst fluid, the signal can be also bright on FLAIR. The cyst wall is isointense to the normal brain tissue, and presents contrast enhancement that is normally weak and rim shaped. On late imaging, the cysts can fill with contrast media by diffusion of the contrast material. In such cases, the differential diagnosis from a malignant tumor can be difficult. Normally the cysts are small and have no mass effect. In larger cysts, an occlusion of the aqueduct can occur with resulting clinical symptoms of hydrocephalus (Oeckler and Feiden 1991). The main differential diagnosis of lesions in the pinealis region includes germinoma, teratoma, pineocytoma,

pineoblastoma, and glioma (Oeckler and Feiden 1991). Germinomas are homogeneous enhancing tumors with a signal appearance similar to white matter on the unenhanced scans. Due to the lack of a capsule, these tumors are normally not well circumscribed and invade the surrounding tissue. Metastasis has to be excluded by a whole CNS imaging procedure. The less differentiated teratomas are usually heterogeneous on imaging because they contain neoplastic tissue, calcifications, lipids, and fluid in varied proportions. 3.3.5.2  Lymphoma Primary CNS lymphoma is defined as lymphoma limited to the cranial–spinal axis without systemic disease and can be seen at any age (Fine and Mayer 1993). Although rare, an increasing incidence of this disease, up to 2%, has been seen among patients with acquired immunodeficiency syndrome (AIDS) and among other immunocompromised persons (Fine and Mayer 1993). The natural history of this disorder differs between patients with AIDS and those without AIDS. Both groups do equally poorly without therapy (1–3 month mean survival), but the overall survival for treated patients is much better for

3.3  Intracranial Tumors Fig. 3.3.30  Pinealoblastoma in a 2-year-old patient with hydrocephalus. The tumor presents as a mixed solid and cystic tumor causing a massive hydrocephalus. On FLAIR imaging (a), the cystic parts contain a large amount of proteins or blood, which does not allow suppression of the fluid. The solid tumor parts show a slight hyperintensity on T2-weighted imaging (b), with an inhomogeneous enhancement including the cystic walls (c,d)

patients without AIDS (18.9 months) than for those with AIDS (2.6 months) (Pollack et al. 1989). Central nervous systemic lymphomas are non-Hodgkin’s lymphoma in nearly all cases. In about 10%, a spread within the CNS was observed. When tumor progression occurs, it is usually confined to the CNS and/or the eye. Occult systemic disease can be excluded by staging with bone marrow biopsy and CT scans of the chest, abdomen, and pelvis (O’Neill et al. 1995; Abrey et al. 2005). Although more than 95% of patients with primary CNS lymphoma have lymphoma of B-cell origin, 45 patients with CNS lymphoma of T-cell origin showed no difference in presentation or outcome in a retrospective series with data collected from 12 cancer centers. The majority of primary CNS lymphomas have an intra-axial intraparenchymal location, whereas the secondary lymphomas mainly occur at the meninges (Scatliff et al. 1997; Thurner et al. 1997). The tumors are typically deep seated in the subependymal space and may present as a bilateral mass lesion. Most lesions are solitary but a multifocal and/or necrotic appearance can be present in AIDS patients. Imaging may therefore show ring enhancement in 50% of AIDS patients while patients without AIDS usually show only homogeneous enhancement (Thurner et al. 1997). The signal characteristics are

iso- to slightly hypointense on T1-weighted imaging and hyperintense on T2-weighted scans (Fig. 3.3.32). The extent of edema is normally less than observed in gliomas or metastases (Koeller et al. 2007). The enhancement is intense (Fig. 3.3.33), and the detection of enhancement along the perivascular spaces is typical for lymphomas, with sarcoidosis as the only differential diagnosis. Calcification and hemorrhage are uncommon in primary CNS lymphoma. Other differential diagnosis includes cerebral toxoplasmosis in the immunocompromised population with hemorrhage as a common finding. In the non-AIDS population, glioblastoma and metastases are the main differential diagnosis. Both can cross the corpus callosum and present comparable enhancement patterns. 3.3.5.3  Cerebral Metastases Metastasis to the brain is a frequent complication of systemic cancer and counts as one of the most common intracranial tumors after primary intra-axial neoplasms (gliomas) (Nelson et al. 2000b). Approximately 35–40% of intracranial neoplasms are metastatic. The incidence of cerebral metastases is rising with more sufficient treatment of primary tumors and therefore longer survival of

269

270

3  Brain, Head, and Neck Fig. 3.3.31  Pinealis cyst in a 36-year-old patient as an incidental finding. Classical appearance with isointensity to CSF on both T1-weighted (a) and T2-weighted (b,c) images. The cyst wall is of lower intensity on T2-weighted imaging and may show a mild enhancement on contrast enhanced T1-weighted imaging

patients (Patchell 2003). The primary tumors that most commonly metastasize to the brain include lung cancer, breast cancer, melanoma, renal cancer, and colon cancer. The majority of metastases (greater than 75%) are found supratentorially with the minority (less than 10%) found in the brainstem. Multiple metastases to the brain occur in more than 70% of cases, but solitary metastases are not uncommon (Nelson et al. 2000b). Most metastases occur at the corticomedullary junction of the cerebral hemispheres; however, they can be found in any location, with 15% being found in the cerebellum and 3% in the basal ganglia. As many as 40% to 50% of intramedullary spinal cord metastases originate from primary lung neoplasms. The most common primary cancers causing epidural spinal cord compression include breast cancer (22%), lung cancer (15%), prostate cancer (10%), and lymphoma (10%) (Patchell 2003). Leukemias, lymphomas, breast cancer, and gastrointestinal carcinomas are associated with diffuse infiltration of the leptomeninges. On MR imaging, cerebral metastases can mimic any cerebral tumor with iso- to hypointense signal on T1and hyperintensity on T2-weighted imaging (Yuh et al.

1995; Davis et al. 1991). In large lesions intensive edema is present, which is frequently absent in small metastatic lesions (Fig. 3.3.33). The metastatic foci are well separable from the surrounding edema because of the different signal characteristics that depend on the cellularity, hemorrhage, necrosis etc. Hemorrhage is frequent in metastatic CNS lesions, mainly in lesions from melanoma, lung cancer, hypernephroma, and choriocarcinoma. Hemorrhage can be seen as hyperintensity on unenhanced MRI, representing methemoglobin, but other blood byproducts, e.g., hemosiderin, may also be present (Fig. 3.3.34). The patients present with signs and symptoms referable to either increased intracranial pressure or focal pathology. In MRI, the signal characteristics of brain metastases vary: in most cases, the lesions appear hypointense on T1-weighted and hyperintense on T2-weighted imaging. A hyperintense signal on nonenhanced T1-weighted imaging represents bleeding into the metastasis or high melanin content of the lesion. In some rare cases, e.g., with gastrointestinal adenocarcinoma as the primary tumor, or in massive bleedings into the metastases, the lesions can appear hypointense on T2-weighted imaging.

3.3  Intracranial Tumors Fig. 3.3.32  Lymphoma in a 43-year-old patient with multiple cerebral lesions. The tumor presents as multiple hyperintense masses involving the corpus callosum, the lateral frontal brain surface, and the occipital lobe. There are more lesions detectable on the FLAIR image (a) than on the T2-weighted image (b). The tumor is slightly hypointense to brain tissue on the T1-weighted image (c). After contrast material administration, an intensive homogeneous enhancement can be observed in the lateral (d) lesion with a mild enhancement of the corpus callosum lesion. The lesion in the occipital lobe did not present any contrast enhancement

Fig. 3.3.33  A 60-year-old patient with known malignant melanoma and a large solitary metastasis. 3-T imaging using double dose of contrast media. Beside the large lesion a small lesion became obvious on T1-weighted contrast-enhanced imaging

(a, arrow). The lesion was not seen on either T2-weighted (b) or FLAIR (c) imaging because of the small size and the absence of perifocal edema

271

272

3  Brain, Head, and Neck Fig. 3.3.34  Patient with malignant melanoma and cerebral metastasis. The unenhanced T2-weighted and FLAIR sequences show a mass lesion with heterogeneous internal morphology including cystic parts and central hyperintensity (a,b). The hyperintense areas represent hemorrhage in the form of deoxyhemoglobin or hemosiderin. The hyperintensity on unenhanced T1-weighted imaging (c) represents methemoglobin. After contrast (d), a typical ring-shaped enhancement pattern is present

Fig. 3.3.35  Multiple metastasis of renal cell carcinoma. While few hyperintensities were visible on the standard SE imaging after a single dose of contrast media (a), multiple enhancing lesions became obvious after a magnetization transfer pulse sequence was applied (b)

Besides very small lesions, all cerebral metastases show enhancement. The degrees and patterns of enhancement may vary and can include homogeneous, inhomogeneous, and ring-shaped enhancement. In cases with hemorrhage, the enhancement can be hidden. The degree of enhancement also depends on the amount of contrast medium used and the time-point of imaging after contrast-media

application. There have been studies reporting a substantial increase in the detection of metastases with the use of a high dose of contrast media. An optimal dose may be a double dose (Colosimo et al. 2001; Schneider et al. 2001) (Fig. 3.3.32). The use of magnetization transfer technique was also found to be beneficial for detecting small or additional metastatic lesions (Fig. 3.3.35).

3.3  Intracranial Tumors

3.3.6  Extra-Axial Cerebral Tumors

Meningiomas (WHO grades I–III) are the most common primary nonglial intracranial tumors. They are typically slow-growing, benign, WHO grade I tumors attached to the dura mater and composed of neoplastic meningothelial (arachnoidal) cells (Black 1993). From large autopsy

series, meningiomas are estimated to comprise between 13 and 26% of primary brain tumors and have an annual incidence of approximately 6 per 100,000 persons. Meningiomas usually occur in adults, with a peak occurrence during the sixth and seventh decades of life. Women are affected more frequently than men, with a female-tomale ratio as high as 2:1. Most meningiomas arise within the intracranial, orbital, and intravertebral cavities. Spinal meningiomas are most common in the thoracic region; atypical and anaplastic meningiomas are more common in the falx cerebri and the lateral convexities. One can also differentiate low-grade or typical from atypical meningiomas (WHO grade II), which constitute 4.7–7.2% of meningiomas, while the anaplastic or malignant meningiomas (WHO grade III) account for 1.0–2.8% of meningioma cases. These higher-grade meningiomas may show a conspicuous predominance in males. Malignant behavior, including brain or bony invasion, may occur with any grade of meningioma. Multiple meningiomas are often associated with neurofibromatosis type 2 (NF2) and in subjects from other, non-NF2 families with a hereditary predisposition to meningioma. Although meningiomas can occur at any location it the brain they have a predilection for certain locations including the parasagittal side of the convexity with attachment to the sagittal sinus (Fig. 3.3.36) or the scull base

Fig. 3.3.36  Meningioma originating from the falx cerebri and adapted to the sagittal sinus in a 58-year-old patient. The meningioma is slightly hyperintense on T2-weighted imaging (a), hypointense on T1-weighted imaging (b) and presents with a

homogeneous intensive contrast enhancement (c–e). Imaging in multiple planes after contrast material administration allows assessment of the origin of the tumor and its relationship to the vascular structures

3.3.6.1  Meningeal Tumors Originating and in projection from the meninges can be found many tumor types. Most common are those arising from the meningothelial cells themselves, the meningiomas (Black 1993). Other mesenchymal, nonmeningothelial tumors also occur, but most are rare in the meninges and more commonly found at a different location. Hemangiopericytomas are listed in the WHO classification because they are more frequent and historically confused with meningiomas (Smirniotopoulos et al. 2007; Akiyama et al. 2004). There are also varieties of melanocytic lesions that can be found in the brain; these are rarely hemangioblastomas and are classified as of uncertain histogenesis. 3.3.6.1.1  Meningioma

273

274

3  Brain, Head, and Neck

attached to the cavernous sinus (Fig. 3.3.37) (Spagnoli et al. 1986; Nakano et al. 2002). Other common sites are the dura adjacent to the anterior Sylvian fissure and the olfactory grooves (Fig. 3.3.38). The orbits and the optic nerve are also common locations (Mafee et al. 1999). The tumors may arise from the optic nerve sheath directly or may extend into the orbits from the sellar region (Fig. 3.39). The therapeutic options for meningiomas include surgical resection or radiation therapy. After surgical resection, benign meningiomas (WHO grade I) recur in about 7–20% of cases, atypical meningiomas (WHO grade II) recur in 29–40% of cases, and anaplastic meningiomas

recur in about 50–78% of cases. Malignant histologic features correlate with shorter survival times; one series has reported a median survival of less than 2 years for patients with anaplastic meningiomas. Brain invasion indicates a greater likelihood of recurrence, regardless of histology. On imaging studies, meningiomas are usually broad based and attached to the adjacent dura. In some cases, if the tumor originates from pial meningeal cells, no dural attachment may be visible. This is also the case with intraventricular meningiomas that arise from the choroid plexus cells (Majos et al. 1999). The main differential diagnosis for the intraventricular localizaFig. 3.3.37  Meningioma attached to the cavernous sinus in a 62-year-old patient. In this case, imaged on a 3-T system, the tumor is isointense to gray matter on both T1- (a) and T2-weighted (b) images and shows the typical homogeneous enhancement pattern (c–e). At this location the patency of the carotid artery and the cavernous sinus is of importance

3.3  Intracranial Tumors Fig. 3.3.38  Small meningioma in the olfactory rim. The tumor, which is hardly seen on the unenhanced scans (a), appears as a thin layer of enhancement at the caudal border of the olfactory rim (arrow) (b). In some small tumors, the enhancing tumor is only visible in the coronal or sagittal orientation

Fig. 3.3.39  Opticus meningioma in a 48-year-old patient. The intraorbital tumor is hypointense on T2-weighted (a) and isointense on unenhanced T1-weighted imaging (b). After contrast the typical meningioma enhancement becomes visible (c,d).

The enhancement is best seen on fat-suppressed sequences, and for delineation from the orbital muscles, a coronary orientation is recommended

275

276

3  Brain, Head, and Neck

Fig. 3.3.40  Skull base meningioma in a 68-year-old patient. Note the induced hyper­ostosis and infiltration of the orbits in this huge meningioma with high signal intensity on T2-weighted imaging (a) and isointense signal on T1-weighted imaging (b). The contrast-enhanced scans (c,d) best display the extent of the tumor into the orbits and along the meninges. Along the en plaque parts at the meninges, the bone reaction exceeds the volume of the tumor itself (e)

3.3  Intracranial Tumors

Fig. 3.3.41  Anaplastic meningioma in a 56-year-old patient. The tumor presents with mass effect and perifocal edema on FLAIR imaging (a) and the typical homogeneous enhancement patterns (b). Angiography nicely displays the vascular supply (c)

tion is the choroid plexus papilloma, which appears as a nodular, heterogeneous mass in contrast to the smooth margin and oval configuration of meningioma. Calcification is common, and tumoral bleeding is rare. A typical finding is also the hyperostosis of the adjacent calvarium, which is commonly seen in lesions adjacent to the bone (Fig. 3.3.40). A penetration of the bone into the subcutaneous region was also reported. The degree of bone thickening is not related to the histological subtype or malignancy of the tumor; however, the bone is usually invaded by tumor cells. In some cases, e.g., en plaque meningiomas, the bony reaction far exceeds the volume of the tumor itself (Fig. 3.3.41). En plaque meningiomas consist of a thin layer of neoplasm that closely follows the contour of the inner table (Chabel et al. 1999). They present themselves as flat lesions with no or only minor mass effect. On MRI most meningiomas, in some series more than 85%, present with heterogeneous signal characteristics, both on T1- and T2-weighted sequences (Zimmerman et al. 1985; Spagnoli et al. 1986; Elster et al. 1989). This heterogeneity is related to the variety of histological subtypes of meningiomas, their vascularity, percentage of calcifi-

cation, and cystic and fatty components. In general, the T1 signal is more hypointense and the T2 signal more hyperintense (Figs. 3.3.36, 3.3.37, 3.3.38). Edema is a common finding in high-grade gliomas; however, it may also be present in large lesions with benign histology (Fig. 3.3.41). The contrast enhancement pattern is pronounced and homogeneous in most cases, even those that present with a large amount of calcifications. The enhancement can be central or ring-shaped and is more intense shortly after contrast media application, which reflects the high rate of vascularity and the fact that the meningioma capillaries have no BBB. The contrast enhanced scans allow a more precise definition of the tumoral boundaries and better detection of small meningiomas. The most typical finding is the so-called dural tail, a dural enhancement adjacent to the lesion (Fig. 3.3.34). This enhancement may or may not be associated with a dural thickening. In large series, most meningiomas present such an enhancement, but it does not automatically represent dural invasion by the tumor. Investigations have found both neoplastic invasion and reactive vascular changes. Dural enhancement may also be present in other diseases, e.g., metastases or invasive brain tumors, but the dural

277

278

3  Brain, Head, and Neck

tail sign is a very important diagnostic feature for the diagnosis of a meningeal tumor. Since meningiomas represent a group of highly vascularized tumors, MR angiography may play an important role in the diagnostic workup of these tumors. The primary vascular supply to a dural base meningioma originates from vessels supplying the normal dura in that region, which arise from the external cerebral artery. MRA allows the depiction of the vascular supply, which is of importance for the preoperative planning. Beyond that, MRA may display tumor involvement of arteries, veins, or dural sinuses. The arterial encasement is a common finding in skull base meningiomas. In most cases the arteries show no or only mild changes in their lumen; however, MRI is able to show a narrowing or irregularity of the lumen in cases of infiltration. The venous or sinus visualization in relation to the meningioma growth patterns is important with regard to tumor resectability and predicts the rate of recurrence.

3.3.6.1.2  Hemangiopericytoma Hemangiopericytoma of the CNS was long considered a meningioma, but it is now recognized as a mesenchymal, nonmeningothelial tumor histologically indistinguishable from hemangiopericytomas occurring in soft tissue and with a tendency to recur and to metastasize outside the CNS. A highly cellular and richly vascularized tumor, it is usually attached to the dura (Jääskeläinen et al. 2000). Histologic criteria for grading are not firmly established; however, these types of tumors appear to correspond histologically to WHO grade II or III. Meningeal hemangiopericytomas make up approximately 0.4% of all primary CNS tumors. These types of tumors tend to appear at a younger age than meningiomas, and they occur more often in men than in women (Jääskeläinen et al. 1985). In the series described in the literature, MR imaging findings and site manifestations were similar to those of meningiomas, but with slight differences (Ruscalleda et

Fig. 3.3.42  Recurrent hemangiopericytoma in a 52-year-old patient. In this case of a recurrent tumor there is not the typical multilobular appearance however we have the typical signal patterns with low signal on both T2weighted (a) and FLAIR (b) images and a homogeneous, strong enhancement (c,d)

3.3  Intracranial Tumors

al. 1994; Akiyama et al. 2004). Hemangiopericytomas had a more multilobular appearance with sometimes-irregular margins and strong enhancement, including the adjacent meningeal structures. Unlike meningiomas, they did not show hyperostosis or calcification. Their signal intensity has been described as various; in most cases the tumor was isointense to cortical gray matter on both T2 and unenhanced T1-weighted imaging. A common finding in hemangiopericytoma is a homogeneous enhancing, well-demarcated tumor with low signal on T2-weighted and FLAIR imaging (Fig. 3.3.42). On unenhanced T1-weighted imaging, the tumor was isointense to slightly hypointense compared with the gray matter. The low signal is not related to calcification but probably to a high vascular density within the lesion. In patients with hemangiopericytomas, imaging of the complete neuroaxis is mandatory to rule out metastases. In our case, an additional lesion was found in the cervical spine with strong dural thickening and enhancement.

3.3.6.1.3  Melanocytoma Melanocytic lesions are diffuse or circumscribed, benign, or malignant tumors arising from melanocytes of the leptomeninges (Jellinger et al. 2000). They include diffuse melanocytosis (diffuse melanosis) and neurocutaneous melanosis, melanocytoma, and malignant melanoma. Intermediate or mixed cases may occur. Melanocytoma accounts for 0.06–0.1% of brain tumors; the other melanocytic lesions are rarer. These lesions typically occur in the fifth decade of life with a femaleto-male ratio of 2:1. Diffuse melanocytosis involves the supratentorial and infratentorial leptomeninges; melanocytomas occur as solid masses in the cranial and spinal compartments. On both CT and MRI, the features of melanocytic lesions are similar to those of meningiomas (Fig. 3.3.43).

Fig. 3.3.43  Melanocytoma in a 7-year-old child. The lesions, initially misdiagnosed as meningiomas, present with an isointense signal on FLAIR (a) and hypointensity on T2-weighted (b) imaging. The main diagnostic finding is the hyperintensity on unenhanced T1-weighted imaging (c), with no evidence of contrast enhancement (d,e)

279

280

3  Brain, Head, and Neck

3.3.6.2  Choroid Plexus Tumors Choroid plexus papilloma (WHO grade I) and its malignant counterpart choroid plexus carcinoma (WHO grade III) are intraventricular papillary neoplasms derived from the choroid plexus epithelium (Aguzzi et al. 2000). These types of tumors are most commonly seen in the first 10 years of age and account for 0.4–0.6% of all cerebral tumors, 2–4% of brain tumors in children, and 10–20% of brain tumors manifesting in the first year of life. Papillomas outnumber carcinomas by a 10:1 ratio. These tumors are most commonly found in the lateral ventricles or the fourth ventricle but may also occur in the third ventricle and the cerebellopontine angle. Lateral ventricle tumors occur primarily in children; fourth ventricle tumors are evenly distributed among all age groups. An association between infection with simian virus 40

Fig. 3.3.44  Plexus papilloma in a 3-year-old patient with hydrocephalus. The tumor is isointense to white matter on FLAIR (a), T2-weighted (b,c) and unenhanced T1-weighted (d) images. Because of the intraventricular growth and the overproduction of CSF, hydrocephalus is present in most of the cases. After con-

(SV40) and choroid plexus tumors has been made. These types of tumors occasionally occur in patients with LiFraumeni syndrome. The main clinical finding is communicating hydrocephalus with diffuse ventricular dilatation, which can be caused either by an overproduction of CSF, a malresorption at the arachnoid granulations due to adhesions by proteinaceous material or a combination of both. Choroid plexus papilloma can be cured surgically and has a 5-year survival rate of as much as 100%. Choroid plexus carcinomas have a less favorable outcome and a 5-year survival rate of 40%. Besides the findings of hydrocephalus, the normally well-marginated, large lobular masses (Fig. 3.3.44) present with nonspecific CT and MRI characteristics (Martin et al. 1990; Jackson et al. 1992). On CT scan, the typical choroid plexus papilloma appears as a well-marginated, smooth, or lobulated isoor high-density mass protruding into the lumen of the

trast administration the enhancement pattern is heterogeneous with a lobulated aspect (e,f). The inhomogeneity is caused by the presence of calcification, hemorrhage, and high vascular density

3.3  Intracranial Tumors

ventricle with strong contrast enhancement. This marked homogeneous enhancement is related to the richly vascular nature of the tumor. Tumoral calcifications are uncommon in the pediatric age group. Choroid plexus papilloma and intraventricular meningioma cannot be differentiated by CT characteristics. MRI is the method of choice for the detection and especially delineation of those intraventricular lesions. The MRI characteristics are of intermediate signal intensity on T1-weighted and intermediate or increased signal intensity on T2-weighted images. There are areas of internal signal void, predominantly curvilinear, indicating enlarged intratumoral vessels. Both calcifications and hemorrhage may occur, and strong enhancement is typical because of the high vascularity of the lesions (Tasdemiroglu et al. 1996). 3.3.6.3  Tumors of the Sellar Region There are a variety of tumors that occur frequently in the sellar or parasellar region including pituitary tumors, craniopharyngiomas, granular cell tumors, and chordomas (Johnson et al. 1991). Other intrasellar masses include cysts, metastases, inflammatory lesions including abscesses, and vascular lesions, e.g., aneurysms that can mimic a solid tumor. In the parasellar region, meningiomas or nerve sheath tumors are the most common ones. MRI is the preferred imaging modality in this region because of the improved sensitivity and specificity of the method, based on the high tissue contrast. Because of its small structures, the radiologic imaging of the pituitary gland and the parasellar region is a challenge, with MRI being the method of choice for displaying the very small volume organ and its surrounding risk structures. A sellar imaging protocol should therefore include thin-section sagittal and coronal T1- and T2-weighted imaging. The contrast scans should be performed with a lower dose of contrast medium (0.05 mmol/kg of BW) with dynamic scanning in primarily coronal orientation. 3.3.6.3.1  Pituitary Tumors Depending on the cited literature, pituitary tumors represent from 10–25% of all intracranial neoplasms, and can be classified into three groups according to their biological behavior: benign, invasive adenoma, and carcinoma (Asa and Ezzat 1998; Ezzat et al. 2004). Pituitary adenomas compose the largest portion of pituitary neoplasms with an overall estimated prevalence of approximately 17%, but only a minority is symptomatic (Ezzat et al. 2004). The tumors arise from the pars distalis or anterior lobe of the pituitary gland, the adenohypophysis. Pituitary adenomas can be classified in three different ways: based on the anatomical or radiological approach, which classifies according to the size of the lesions on radiological imaging, by histological criteria us-

ing immunohistological characteristics or by functional criteria using the endocrine activities of the tumors. The radiological approach differentiates between microadenomas (i.e., the greatest diameter is <10 mm) and macroadenomas (i.e., the greatest diameter is ≥10 mm). Most pituitary adenomas are microadenomas. Historically, the most widely used radioanatomical classification was based primarily on a neuroradiological examination including skull X-rays, pneumoencephalography, polytomography, and carotid angiography (Hardy 1973) and subsequently validated by the application of more accurate CT and MRI (Elster 1993). Sagittal T1-weighted images, clearly displaying the anterior and posterior lobes and the stalk on the same plane, and coronal images, displaying the relation between the pituitary and cavernous sinuses, are optimal for identifying a pituitary adenoma. A 3-mm thin slice typically is used to obtain optimal resolution (Hall et al. 1994). Macroadenomas, which account for 70–80% of adenomas, are seen in all ages, most commonly in between the ages of 25 and 60 years, with a clear preference in females. Macroadenomas are usually endocrinologically inactive and present with clinical symptoms of mass effect or resulting hypofunction of the gland. Clinical important is its relation to the optical pathways, especially the deviation of the chiasm. The involvement of the chiasm, represented by a bitemporal hemianopia is an emergency that requires immediate treatment. The same is true for pituitary apoplexy, hydrocephalus, or cranial nerve involvement. In imaging studies, the tumor presents as a mass lesion in projection of the pituitary gland, in some cases with a typical figure eight appearance and a large suprasellar component (Fig. 3.3.45) (Cappaibanca et al. 1999; Schwartzberg 1992). Macroadenomas may extend into the neighboring areas such as the suprasellar cistern, cavernous sinus, sphenoid sinus, and nasopharynx in up to 70% of the cases and are best shown on MRI studies. They can present with cysts, areas of necrosis and hemorrhage. An involvement of the carotid sinus is described and may encase the carotid artery as well. On MRI studies, depending on the size and different components of the tumor, macroadenomas are usually isointense with cortical gray matter, and enhance intensely with Gd-containing contrast materials. Depending on the amount of cysts, bleeding, and hemorrhage the signal patterns can be quite inhomogeneous. On imaging studies and for treatment planning, e.g., surgical resection, the exact display of the size, the relation of the tumor to the optical pathways, and the infiltration of neighboring structures is essential and predicts the ability to completely remove the tumor via a transsphenoidal approach. The main differential diagnosis of a macroadenoma is a sellar or parasellar meningioma. These lesions are also isointense to brain parenchyma and present strong contrast enhancement. On T2 imaging, adenomas typically present with a higher signal intensity and do not show a

281

282

3  Brain, Head, and Neck

Fig. 3.3.45  Macroadenoma of the hypophysis in an 80-year-old patient with hemianopsia. The tumor is slightly hyperintense on both T2-weighted (a) and FLAIR (b) imaging and isointense on T1-weighted imaging (c). After contrast administration, the enhancement pattern is homogeneous (d). The tumor margins can best be differentiated on the enhanced imaging, including a series of highresolution coronal scans (e–g). Here the association to the optic pathways, especially the chiasm, can best be seen. In this case an elevation and impairment of the chiasm was verified

3.3  Intracranial Tumors

Fig. 3.3.46  Pituitary microadenoma in a 28-year-old patient with hormonal dysfunction. The dynamic scans (a–c) allow the differentiation of the adenoma (arrow) by a later and less intense

enhancement. A time resolution of 30 s for each scan with the use of a half-dose of contrast material (0.05 mmol/kg of body weight) is recommended

dural tail sign on enhanced scans. Bone erosion or hyperostosis is also uncommon for adenomas. Microadenomas less than 1 cm in diameter are best seen on thin-section coronal images and appear hypo- to isointense to the normal pituitary tissue. After contrast material application the microadenomas present with a less intense and later enhancement than the normal gland tissue, which can be nicely appreciated on the dynamic scans (Fig. 3.3.46) (Bartynski and Lin 1997; Rand et al. 2002). Using functional criteria, pituitary adenomas can be categorized into prolactin (PRL)-producing, ACTH-producing, growth hormone (GH)-producing, and the rare thyrotropin (TSH)-producing, also known as thyrotroph, tumors. The prolactin-producing adenomas are the largest group, accounting for about 30% of all adenomas. The second largest group is that of clinically nonfunctioning (i.e., endocrine-inactive) adenomas. This group (20–25% of adenomas) is composed predominantly of gonadotroph adenomas. They synthesize follicle-stimulating hormone (FSH) and/or luteinizing hormone (LH), or the alpha or beta subunits of these heterodimers. They are usually detected incidentally, or because of the presence of neurologic symptoms. Gonadotroph adenomas are inefficient secretors of the hormones they produce, so they rarely result in a clinically recognizable hormonal hypersecretion syndrome. In imaging studies, one cannot differentiate between hormone active and inactive adenomas. Invasive adenomas, which account for approximately 35% of all pituitary neoplasms, may invade into the dura mater, cranial bone, or sphenoid sinus (Scheithauer et al. 1986). Pituitary carcinomas account for 0.1 or 0.2% of all pituitary tumors (Ragel and Couldwell 2004). Both lesions can present with bone erosion and other malignant signs and are therefore barely distinguishable from other

malignant lesions, e.g., metastases or in some cases craniopharyngiomas or meningiomas. 3.3.6.3.2  Craniopharyngioma Craniopharyngioma (WHO grade I) is a benign, partly cystic epithelial tumor of the sellar region, presumably arising from squamous cell rests within the pars tuberalis of Rathke’s pouch (Atlas 1996). Two clinicopathological forms are distinguished: adamantinomatous and papillary. This type of tumor accounts for 1.2–4.6% of all brain tumors. The age incidence is bimodal; peaks are observed in children aged 5–14 years and in adults older than 50 years. The most common localization is suprasellar with an intrasellar component. Among these, 30% extend anteriorly, 23% extend into the middle fossa, and 20% extend into the retroclival area. In a large series, 60–93% of patients had a 10-year, recurrence-free survival. These tumors are benign in nature and grow slowly but they tend to recur. The most significant prognostic factor associated with tumor recurrence is the extent of surgical resection; lesions larger than 5 cm carry a worse prognosis. The recurrence rate is significantly higher after incomplete resection. On imaging, craniopharyngiomas present as heterogeneous mass lesions with solid and cystic components (Rennert and Doerfler 2007; Pisaneschi and Kapoor 2005; Chong and Newton 1993). Cystic components are present in 85% of cases. The cysts contain cholesterol crystals in varying concentrations as well as keratin debris. This may explain the variable signal intensities observed on MRI examinations (Fig. 3.3.47). Calcifications, which occur in 75% of cases, are better documented with CT than with MRI (Fig. 3.3.48). While on CT most cystic lesions present with the typical low-density appearance, the MRI signal characteristics may vary. Depending on the content

283

284

3  Brain, Head, and Neck Fig. 3.3.47  Craniopharyngioma in a 39-year-old patient. The tumor appears as a partly enhancing partly calcified lesion (a). The cystic part of the tumor is bright on both T2-weighted (b) and FLAIR (c) imaging, with the best differentiation on FLAIR. The bright signal is caused by high protein content in the cyst fluid. On non-enhanced T1-weighted imaging, the tumor is not clearly delineated (d), whereas after contrast material administration there is a ring-shaped enhancement (e)

3.3  Intracranial Tumors Fig. 3.3.48  Craniopharyngioma in a 55-year-old patient, initially misdiagnosed as a Rathke’s cleft cyst. CT (a) shows a small area of calcification that is not visible on MRI. The lesion has a cyst-like appearance with CSF signal on T2-weighted (b) and unenhanced T1-weighted (c) imaging. After contrast administration, there is only a slight enhancement of the cystic wall (d)

of the cysts, blood byproducts, cell debris, cholesterol or other crystals, the signal intensity can be increased or decreased both on T1- and T2-weighted imaging. 3.3.6.3.3  Rathke’s Cleft Cyst Rathke’s cleft cysts are uncommon benign cystic lesions that are derived from the remnants of the epithelium embryologically lining the craniopharyngeal duct (Naylor et al. 1995). Autopsy series have shown that Rathke’s cleft cysts can be found in up to 30% of asymptomatic subjects. Location is usually intrasellar in 50% of cases, supra­ sellar in 25% of cases, and both in 25% of cases. The cysts are usually simple, lined by a single epithelial layer. They may contain variable amounts of protein, mucopolysaccharide, cellular debris, and cholesterol. Depending on their contents, their signal intensity may be high, low, or intermediate on T1- and T2-weighted sequences. They usually do not show Gd enhancement, but they may rarely show a thin peripheral enhancement (Bonneville et al. 2006; Cohan et al. 2004).

Craniopharyngiomas and cystic pituitary adenomas should be considered in the differential diagnosis of Rathke’s cleft cyst. Craniopharyngiomas have common floccular calcification and usually have contrast enhancing solid or rim-like components. Cystic adenomas usually reveal peripheral contrast enhancement. Rathke’s cleft cysts are usually asymptomatic and are discovered incidentally. They may cause mass effect with headache, pituitary dysfunction or in very large cases visual disturbances. Most cases of Rathke’s cleft cyst are stable and do not need any treatment. 3.3.6.4  Tumors of Uncertain Histogenesis Capillary hemangioblastoma (WHO grade I) occurs sporadically and is associated with the familial tumor syndrome von Hippel-Lindau (VHL) disease (Böhling et al. 2000). VHL disease is inherited through an autosomal dominant trait and is characterized by the following: capillary hemangioblastomas of the CNS and retina, clearcell renal carcinoma, pheochromocytoma, pancreatic

285

286

3  Brain, Head, and Neck

tumors, and inner ear tumors. The syndrome is related to germline mutations of the VHL tumor suppressor gene, which is located on chromosome 3p25-26. VHL disease is estimated to occur at rates of 1:36,000–1:45,500 of the world population. Capillary hemangioblastomas typically occur in adults; the mean age of patients with VHL-associated tumors is 29 years. Capillary hemangioblastomas may occur in any part of the CNS; sporadic tumors occur primarily in the cerebellum. VHL patients often have multiple capillary hemangioblastomas at various sites, including the cerebellum, brain stem, and spinal cord. Because of advances in microsurgical techniques, mortality and morbidity are low for sporadic capillary hemangioblastomas. On imaging studies, the tumor is easily detectable on CT and MRI examination. On CT scan, the cystic component is generally sharply defined, with an attenuation value equal to, or slightly higher than cerebrospinal fluid. On an unenhanced scan, the mural nodule is isodense to brain tissue. After intravenous contrast injection, it enhances intensely and uniformly. Because of the high protein content of the cysts, they are slightly hyperintense to CSF on T1-weighted MR images (Fig. 3.3.49) (Kazner et al. 1989; Ho et al. 1992). They generally present without surrounding edema. The main differential diagnoses are pilocytic astrocytoma and hemangioblastoma, which are

indistinguishable. Simple arachnoid cyst, enlarged fourth ventricle secondary to obstruction of foramina of Magendie and Luschka, and cystic metastasis also have to be included in the differential diagnosis. 3.3.6.5  Tumors of Peripheral Nerves That Affect the CNS Schwannoma (WHO grade I), also known as neurilemoma and neurinoma, is usually an encapsulated benign tumor composed of differentiated neoplastic Schwann cells (Woodruff et al. 2000). This is a common tumor of the peripheral nerve sheath that accounts for an estimated 8% of cerebral tumors and 29% of primary spinal tumors. The term schwannoma best describes the underlying tumor pathology and should be used instead of the synonyms. Schwannomas occur frequently in patients with NF2. The peak incidence is in the fourth to sixth decades of life. Three histologic variants include cellular schwannoma, melanotic schwannoma, and plexiform schwannoma. Inactivating mutations of the NF2 gene on chromosome 22q12 have been detected in approximately 60% of schwannomas. Schwannomas are slow-growing benign tumors that grow at the periphery of the affected nerves and that only rarely undergo malignant changes.

Fig. 3.3.49  Capillary hemangioblastoma in a patient with von Hippel-Lindau (VHL) disease. The tumor presents as a large cystic mass (a), with incomplete suppression of the cyst content on FLAIR (b) imaging. After contrast administration (d,e), the solid tumor parts show intensive enhancement

3.3  Intracranial Tumors

Vestibulocochlear schwannomas, also known as acoustic neurinomas are the most common schwannomas, but they may affect all cranial nerves. MRI is the imaging method of choice in the work-up of patients with suspected schwannomas (Smirniotopoulos et al. 1996; Jager et al. 2001; Maroldi et al. 2001). It is

able to display even small peripheral tumors with a high sensitivity and specificity (Fig. 3.3.50). The typical signal characteristics are iso- to hyperintense on T1 and extremely hyperintense on T2 with a marked homogeneous or in some cases ring-shaped contrast enhancement. In large lesions, e.g., in those patients with NF2, cystic deFig. 3.3.50  Small intrameatal acoustic schwannoma in a 51year-old patient. On T2-weighted (a), FLAIR (b), and unenhanced T1-weighted (c) images, the tumor is isointense to normal tissue and appears as a thickening of the nerve in the inner acoustic canal. Strong and well-circumscribed enhancement (d,e) is pathognomonic. On CT (f), such small tumors are hardly seen and do not cause an enlargement of the inner acoustic canal

287

288

3  Brain, Head, and Neck

Fig. 3.3.51  Acoustic schwannoma in a 15-year-old patient with neurofibromatosis type 2. The large acoustic tumor has mass effect on T2-weighted imaging (a) and shows heterogeneous con-

trast enhancement (b,c). The non-enhancing tumor parts are not cystic. The present meningeal enhancement is caused by a ventricular shunt implantation

Fig. 3.3.52  Trigeminal schwannoma in a 64-year-old patient. The tumor is hardly seen on T2-weighted imaging, with hyperintensity in the course of the trigeminal nerve (a). After contrast

material administration, a homogeneous enhancing mass lesion becomes visible (b–d) in the area of the Meckel’s cavum

formation of the lesion with inhomogeneous enhancement patterns can be observed (Fig. 3.3.51). A dural tail as in meningiomas may be seen in some schwannomas adjacent to dural tissue. It can represent tumor infiltration but reflects only reactive changes in the majority of cases. In the most common type of nerve sheath tumor, the acoustic schwannoma, one differentiates between the intra- and extrameatal portions of the tumor. In large tumors, the differentiation between schwannoma and meningioma can be difficult. The presence of a significant heterogeneity with cystic or hemorrhagic areas is more typical for schwannoma.

The second most common schwannoma is that which originates from the trigeminal nerve (Fig. 3.3.52). It can involve Meckel’s cave with infiltration of the cavernous sinus with the characteristic signal changes seen in acoustic schwannoma. 3.3.7  Non-Tumorous Changes 3.3.7.1  Arachnoid Cysts Arachnoid cysts are cerebrospinal fluid filled cystic lesions covered by arachnoidal cells and collagen (Ariai

3.3  Intracranial Tumors

Fig. 3.3.53  Left temporal arachnoid cyst in a 59-year-old patient with right temporal anaplastic astrocytoma. While the tumor shows high signal on both long TR sequences, the signal intensity of the arachnoid cyst is equal to that of CSF on all

sequences. The tumor is hyperintense on T2-weighted imaging (a), hypointense on T1-weighted imaging, and shows no contrast enhancement (b). On FLAIR imaging the CSF in the cyst can be completely suppressed (c)

et al. 2005) that may develop between the surface of the brain and the cranial base or on the arachnoid membrane, one of the three membranes that cover the brain and the spinal cord. Arachnoid cysts account for approximately 1% of all intracranial neoplasms and are classified as a congenital disorder of unknown etiology (Gelabert-Gonzales 2004); most cases begin during infancy, but onset may be delayed until adolescence. The cysts contain CSF less complicated by hemorrhage or infection. They can be found in intracranial or spinal locations. The symptomatology of arachnoid cysts varies by size and location; however, small or even large cysts are found incidentally on routine imaging studies. The larger cysts present with symptoms of mass effect of the adjacent structures, a generalized increased pressure is rare. The imaging findings can be described as a mass lesion with the same signal characteristics as those of CSF-filled ventricles and no evidence of contrast enhancement or calcification (Fig. 3.3.53) (Weber and Knopf 2006). The thin membranes covering the CSF are not visible without contrast administration into the cyst fluid. In MRI, the FLAIR sequence with suppression of CSF is helpful to distinguish the lesion from more solid tumors (Fig. 3.3.53).

sule. They may grow slowly by progressive desquamation of the epithelial cells and conversion of this cell type into keratin and cholesterol. They can be located in intra- or extradural locations, and are often found in the cerebellopontine angle and the supra- or parasellar region. Other typical locations for these congenital lesions are the middle cranial fossa and the cisterna magna. On MRI, the epidermoid tumors typically present with a mild hypointensity on T1-weighted images and a marked hyperintensity on T2-weighted images (Hagen et al. 1994; Dechambre et al. 1999; Gao et al. 1992). The latter is more intense than the CSF signal, with substantial heterogeneity resulting from cellular debris and cholesterol crystals. In some cases a high-intensity rim can be seen, which represents a CSF cleft. In cases with high intensity on T1-weighted imaging, the epidermoid may contain a higher lipid content, which may cause a lowering of the signal on T2-weighted imaging. Although calcification or enhancement is extremely rare as in arachnoid cysts, these lesions can be easily differentiated by the use of FLAIR or diffusion-weighted imaging (Dechambre et al. 1999) (Fig. 3.3.54). While the CSF-containing arachnoid cysts are well suppressed on FLAIR, epidermoid cysts retain a bright signal. On DWI, epidermoid cysts typically show high signal intensity and are easily differentiated from CSF-containing cysts.

3.3.7.2  Epidermoid Cysts Epidermoid cysts, also referred to as epidermoid tumors or secondary cholesteatomas, are benign cysts developed out of ectodermal tissue and accounting for approximately 0.5–1.5% of brain tumors (Smirniotopoulos et al. 1993). Histologically, they are made of a thin layer of squamous epithelium covered by an external fibrous cap-

3.3.7.3  Dermoid Cysts A dermoid cyst is a teratoma that, similar to congenital epidermoid cysts, represents an ectodermal inclusion cyst that arises from inclusion of ectodermal elements at the

289

290

3  Brain, Head, and Neck Fig. 3.3.54  Epidermoid in a 35year-old patient with symptoms of headache. The MRI shows the typical findings, with inhomogeneous hyperintensity on T2weighted and FLAIR (a) imaging, hypointensity on T1-weighted imaging (b), and no contrast enhancement (c). The hyperintensity on DWI (d) is pathognomic

time of closure. Dermoid cysts contains developmentally mature skin, with hair follicles and sweat glands, sometimes luxuriant clumps of long hair, and often pockets of sebum, blood, fat, bone, nails, teeth, eyes, cartilage, and thyroid tissue (Smirniotopoulos and Chiechi 1995). Intracranially these are rare tumors that are always benign because they contain mature tissue. The rare malignant dermoid cyst usually develops squamous cell carcinoma. Dermoid tumors tend to occur near the midline with typical location supra- or parasellar, frontobasal, and in the posterior fossa. Due to the different components, the tumors are inhomogeneous on imaging studies with the tendency to hyperintensity on T1 due to their fatty content and hyper­ intensity on T2 (Fig. 3.3.55) (Smirniotopoulos and Chiechi 1995; Koeller et al. 1999). On both the signal can also be inverted, depending on the amount of the components. Calcification of the capsule may also be present but is hardly seen on MRI studies. In ruptured dermoid cysts the T1 signal is typically increased. Typically, there is no substantial contrast enhancement of the lesions. 3.3.7.4  Cerebral Lipomas Cerebral lipomas are developmental abnormal collections of fat in the subarachnoid spaces and arise from

abnormalities of the leptomeningeal membranes. They are most often found in the pericallosal cistern or in the corpus callosum. In some of these cases they are associated with a collosal agenesis or other brain malformations. They may also be found in other locations in the subarachnoid spaces, such as the quadrigemal, chiasmatic, cerebellopontine, or perimesencephalic cisterns. Lipomas are normally asymptomatic if small; large lesions can cause an obstructive hydrocephalus or mass effect. On MRI studies the lesions present with typical fat signal, which is high on T1 and can be suppressed by fat saturation techniques (Fig. 3.3.56) (Sari et al. 1998). Enhancement is uncommon. 3.3.8  Functional Imaging in Intracranial Tumors In the past few years a number of advanced, non-enhanced, and contrast-enhanced MR imaging techniques have been developed that provide new insights into the pathophysiology of brain tumors, mainly gliomas. These techniques include MR spectroscopy, perfusion MR imaging, dynamic contrast-enhanced MRI and diffusion tensor MRI. In the following review, the reader is provided with early results of the application of these techniques in brain tumor assessment. The technical considerations of the techniques have been described in earlier chapters.

3.3  Intracranial Tumors

Fig. 3.3.55  Dermoid tumor in a 48-year-old patient. The tumor in this case has a typical hyperintensity on all long TR sequences (a–c). The brightest signal is visible on proton density (PD) (b)

and FLAIR (c) imaging. On T1-weighted imaging, the tumor is also hyperintense with a brighter signal on SE (d) than on GRE (e) sequences. There was no enhancement

Fig. 3.3.56  Intracerebral lipoma as an incidental finding in a 50-year-old patient. The lesion at the scull base has identical signal characteristics as the scull fat and can be diagnosed by fat-suppressed sequences

291

292

3  Brain, Head, and Neck

Fig. 3.3.57  MR Spectroscopic finding in case of hamartoma (a) and in intra-axial tumor (b). In normal cerebral tissue, there is increasing MRS signal from choline over creatine to NAA. In tumor tissue (b), one can observe an increase of NAA, the

marker for membrane turnover, and a decrease of the neuronal marker NAA. The lactate peak indicates an anaplastic or malignant transformation

3.3  Intracranial Tumors

3.3.8.1  MR Spectroscopy Proton magnetic resonance spectroscopy (MRS) or chemical shift imaging (CSI) is becoming a common clinical tool because it can add to the diagnostic accuracy of MR imaging. Spectroscopic characterization of brain abnormalities has relied mostly on the calculation of ratios between the main proton spectrum metabolites, notably N-acetyl-aspartate (NAA), a neuronal marker, choline-containing compounds (Cho), a marker of membrane turnover, and creatine–phosphocreatine (Cr-PCr), and on the presence of lipids and lactate (Alger et al. 1990; Negendank et al. 1996; Meyerand et al. 1999; Dowling et al. 2001; Demaerel et al. 1991; Vuori et al. 2003). Brain tumors typically have a loss of NAA and an increase in the Cho content (Fig. 3.3.57). MR spectroscopy has also been used to differentiate non-tumoral lesions like hamartomas from gliomas (Jones et al. 2001; Castillo et al. 1995; Norfray et al. 1999). In several studies it has been reported that hamartomas did not differ significantly from the normal brain, while gliomas had lower NAA/ Cr-PCr, Cr-PCr/Cho, and NAA/Cho ratios (Fig. 3.3.57). In a study by Vuori et al. (2003), patients with seizures and a cortical brain lesion on MR images were studied with proton MR spectroscopy. A metabolite ratio analysis was performed, and the metabolite signals in the lesion core were compared with those in the contralateral centrum semiovale and in the corresponding brain sites of control subjects to separately obtain the changes in NAA, Ch), and Cr-PCr. In their study, 10 patients had a low-grade glioma (three, oligodendrogliomas; three, oligoastrocytomas; three, astrocytomas; and one, pilocytic astrocytoma), and 8 had focal cortical developmental malformations (five, focal cortical dysplasias and three, dysembryoplastic neuroepithelial tumors). They found that loss of NAA and increase of Cho were more pronounced in low-grade gliomas than in subjects with cortical developmental malformations. MR spectroscopy was also able to differentiate between subtypes of lowgrade gliomas. In a study of Law et al. (2004), who combined MR spectroscopy with the later described MR perfusion, both methods increased the sensitivity and the positive predictive value in the determination of the glioma grade when compared with conventional MR imaging. They were also able to provide thresholds for the metabolite ratios for the diagnosis of a high-grade glioma. A review of the literature, taking into account differences in MR spectroscopic technique such as the choice of TE and the method for determination of metabolite ratios, demonstrates that the mean maximal values obtained for Cho/Cr and Cho/ NAA and the mean minimum values obtained for NAA/ Cr-PCr ratios in their study were comparable to previously published data in differentiating between low- and high-grade gliomas. As in many functional techniques, one of the challenges is the standardized data acquisition

and postprocessing. For data acquisition the same echo time should be used and data from the contralateral unaffected white matter should be acquired. Studies comparing long and short TE MR spectroscopic studies found that a short TE provided a slightly better tumor classification (Delorme and Weber 2006; Galanaud et al. 2006). In a study by Majos et al. (2004), only meningiomas profited from long TE acquisitions. The use of modern scanner technology allows measurement of spectroscopic data from more than a single voxel. Two- or three-dimensional MR spectroscopy (2D or 3D CSI) enables the acquisition of multiple small voxels, which gives better information about the heterogeneity of a lesion (Fig. 3.3.58). The voxel information can be used to calculate metabolite ratios that can be colorcoded and overlaid on the anatomic images to better visualize, for example, hot spots within the tumor. Follow-up assessment of cerebral tumors is a promising field for MR spectroscopy. Increases in size and contrast enhancement are typical findings in tumor progression but also reflect therapy-induced changes (Delorme et al. 2002). The same is true for postoperative changes. Magnetic resonance spectroscopy (MRS) provides supplementary information about the possible extent and nature of changes on a routine MRI scan by permitting analysis of the presence and/or ratio of tissue metabolites such as NAA, Cr, Cho, and lactate (Fig. 3.3.59). The ratio of choline to normal creatine level usually is significantly higher in areas consistent with tumor compared with areas containing predominantly treatment effect. In fact, treatment effect is generally indicated by a marked depression of all the intracellular metabolite peaks from Cho, Cr, and N-acetyl compounds. However, MRS alone may not be helpful in instances where patients have mixed histologic findings comprised of necrosis and tumor. Because of this heterogeneity and as a result of low spatial resolution, MRS findings of Cho and NAA resonances below the normal range may indicate variable histologic findings ranging from radiation necrosis, astrogliosis, and macrophage infiltration to mixed tissues that contain some regions of tumor. The careful choice of voxel placement and interpretation of results in concordance with other imaging and clinical findings is critical in distinguishing between tumor and treatment-related changes. Furthermore, validation studies using image-guided acquisition of tissue need to be performed to correlate imaging with biology confidently. 3.3.8.2  Contrast-Enhanced Perfusion MR Imaging in Brain Tumors Perfusion-weighted imaging (PWI) in brain tumors has benefits for three major fields: differential diagnosis, biopsy planning, and treatment monitoring. In this section, we focus on the role of perfusion MR imaging in

293

294

3  Brain, Head, and Neck

Fig. 3.3.58  Spectroscopic imaging in a 57-year-old patient with anaplastic astrocytoma. a Localizer of the spectroscopic imaging sequence and the tumor coverage. With CSI a large area of the tumor can be assessed, which allows description of the heterogeneity of the tumor and depiction of hot spots of high

malignancy. The Cho-to-NAA ratio (b) shows high Cho ratios at the lateral part of the tumor, which were well correlated with the enhancement patterns. The Cho-to-Cr-PCr ratio (c) allows depiction of a hot spot in the anterior part of the tumor

Fig. 3.3.59  MR spectroscopy in a case with the parallel presence of radiotherapy-induced tissue changes and tumor progression. In the histologically confirmed necrosis on the patient’s right side the decrease of NAA was accompanied by an increase of the

lipid peak indicating necrosis. The lesion on the contralateral side, which was proved to represent tumor progression and malignization of the initially low-grade tumor, showed an increase of Cho parallel to the NAA decrease

improving diagnosis and monitoring brain tumors during therapy. For differential diagnosis, biopsy planning, and treatment monitoring of brain tumors, besides PWI other imaging methods such as diffusion-weighted imaging and spectroscopic imaging are beneficial (Provenzale et al. 2006b). Together with PWI, MRI now has the ability to provide quantitative cellular, hemodynamic, and metabolic information about brain tumor biology. PWI in neuro-oncology is mostly performed using first-pass

dynamic susceptibility-weighted contrast-enhanced (DSC) MR echo planar imaging approaches (Cha et al. 2002). Newer perfusion imaging approaches do not need extrinsic contrast media application and use the blood as an intrinsic contrast medium (Barbier et al. 2001). Since tumor specification is limited and sometimes conventional MRI cannot discriminate glioblastomas from solitary metastases, CNS lymphomas or other glioma grades, new methods like perfusion MRI are play-

3.3  Intracranial Tumors

Fig. 3.3.60  Perfusion MRI in a 39-year-old patient with suspected recurrent low-grade astrocytoma. The patient developed aphasia and was diagnosed for a recurrent grade II tumor with homogeneous hyperintensity on T2-weighted and FLAIR (a) imaging adjacent to the surgical resection cavity. With no presence of contrast enhancement (b) the patient was primarily rec-

ommended for radiotherapy. The rrCBV map of the perfusion MRI (c) showed a marked hyperperfusion in the central part of the tumor which was described by several groups as an indicator for a higher-grade tumor. Based on this information the patient underwent surgery and a grade III anaplastic astrocytoma was confirmed Fig. 3.3.61  rrCBF parameter maps in two different patients with non-enhancing glial tumors suspected to be grade II lesions. Patient in (a), whose tumor is in a right parieto-occipital location, shows the typical low-grade tumor perfusion with low values of rrCBF and rrCBV (arrows). In the second patient, independent from the morphological finding of a non-enhancing tumor, perfusion MRI (b) shows high values on the rrCBF parameter map (arrows), predicting an anaplastic tumor, which was confirmed by histology

ing an increasingly important role. The results of the available studies in literature, all with relatively limited patient numbers, indicate that DSC MRI is useful in the pretherapeutic workup of gliomas (Fig. 3.3.60), CNS lymphomas, and solitary metastases, as well as in the differentiation of these neoplastic lesions from infections and tumor-like manifestations of demyelinating disease (Cha et al. 2002; Hartmann et al. 2003; Law et al. 2003; Weber et al. 2006). In these applications, it delivers higher pre-

dictive values than conventional MRI. There was quite a large study on brain tumor differentiation that found compared to spectroscopic imaging and dynamic contrast-enhanced (DCE) imaging, PWI had superior diagnostic performance in predicting glioma grade (Fig. 3.3.61), providing follow-up information (Fig. 3.3.62), and differentiating glioblastoma from other tumor entities (metastases, meningiomas, and CNS lymphomas) (Weber et al. 2006).

295

296

3  Brain, Head, and Neck Fig. 3.3.62  Perfusion MRI in a patient after irradiation of a low-grade astrocytoma. On the 24-month control a pathologic contrast enhancement within the tumor became obvious (a). To differentiate tumor recurrence of progression from therapy-induced changes, perfusion MRI was performed. High rrCBF values (b) in conjunction with the new contrast enhancement confirmed the presence of a tumor malignization

Because of the shorter acquisition time and the better predictive values in differential diagnosis, we would favor the use of perfusion over spectroscopic imaging and DCE MRI on a 1.5-T MR unit. This may be the case in patients with diminished physical condition and compliance (often encountered in brain tumor patients), which result in a limited imaging time. Because of a significantly higher tumor perfusion in glioblastomas compared with CNS lymphomas, a thres­ hold value of 1.4 for relative CBV provided sensitivity, specificity, positive predictive value (PPV) and negative predicting value (NPV) of 100, 50, 90, and 100%, respectively. Although the conventional MRI characteristics of solitary metastases and primary high-grade gliomas may sometimes be similar, MR perfusion and spectroscopic imaging enable distinction between the two (Weber et al. 2006; Law et al. 2002). Although neither intratumoral metabolite ratios nor relative regional cerebral blood volume (rrCBV) values nor relative regional cerebral blood flow (rrCBF) values allow for discrimination between the two entities, analyzing the peritumoral T2-weighted, hyper­intense region enables discrimination between high-grade gliomas and metastases, since CBV is significantly higher in peritumoral non-enhancing T2 hyperintense regions of glioblastomas than in metastases. Thus, elevated perfusion in the peritumoral region of the lesion represents glioma with high specificity in the differentiation of glioma from metastasis or grade one meningioma. Hence, PWI allows us to readily appreciate tumor extension past obvious gross anatomic boundaries on conventional MRI. Correct grading of gliomas has significant clinical impact, because adjuvant therapy after surgery is usually ad-

ministered to high-grade but not low-grade gliomas. Histopathology as a gold standard based on biopsy samples is limited due to the inherent sampling error in these heterogeneous tumors. Several studies reported that highgrade gliomas had higher relative regional CBV (Weber et al. 2006; Law et al. 2002) and CBF than low-grade gliomas and that glioblastomas have the highest tumor perfusion among all other glioma grades (Fig. 3.3.35). However, there is a significant overlap of tumor perfusion between high- and low-grade gliomas, which may be explained by the inherent glioma heterogeneity and the sampling error of biopsy samples. This overlap leads to a low specificity, especially when differentiating grade 3 from grade 2 gliomas. Thus, PWI has limited utility for an individual patient in making a specific diagnosis, but it may be of great clinical value for biopsy guidance because of the potential glioma heterogeneity with highgrade components that may be interspersed among lowgrade components. Although PWI has a better diagnostic performance than do conventional MRI techniques in distinguishing different tumor entities, PWI cannot eliminate the need for a biopsy and histologic confirmation because modern treatment regimes also take genetic mutations of tumor cells into account. PWI also has a tremendous impact on treatment monitoring of low-grade gliomas, in addition to advantages in biopsy planning (Fig. 3.3.62). Because of the intact BBB, valid quantification of perfusion is possible in these entities. In case of a disrupted BBB, leakage of contrast agents from tumor vessels causes underestimation of tumor CBV in PWI (Uematsu and Maeda 2006). In lowgrade gliomas, determination of relative regional CBV measurements can be used to predict clinical response.

3.3  Intracranial Tumors

In a recent study, low-grade gliomas that had progressed more rapidly (mean time to progression of 245 days) had significantly higher CBV than those with stable tumor volumes at follow-up (mean time to progression of 4,620 days). The authors proposed a threshold value of relative regional CBV (>1.75) to indicate a propensity for malignant transformation (Law et al. 2006). The reason for this finding is presumably—as in biopsy planning—that PWI depicts focal anaplastic areas in low-grade gliomas that have not yet led to a disruption of the BBB and therefore to a contrast-enhancement on conventional MRI. The same applies for low-grade gliomas after radiotherapy. PWI also detects a subset of patients with higher tumor CBV and shorter progression-free survival (Fuss et al. 2001). Thus, PWI enables a better prediction of prognosis after radiotherapy than conventional MRI. But also after antiangiogenic chemotherapy of gliomas, PWI has shown its potential to better predict treatment outcome than tumor volume determined on conventional MRI (Cha et al. 2000). For other intra-axial lesions, such as brain metastases (Essig et al. 2003; Weber et al. 2004), PWI has also shown its potential to better predict treatment outcome than tumor volume. In this context, a reduction of CBV was highly predictive of treatment response while an increase in CBV was a hint for non-response. In extra-axial lesions, such as meningiomas, DCE MRI might be a good alternative to DSC MRI for treatment monitoring, because the technique is not as susceptible as DSC MRI to susceptibility artifacts arising from bone and air. In summary, PWI delivers higher predictive values than conventional MRI by providing maps of the regional variations in cerebral microvasculature of normal and diseased brains. PWI can easily be incorporated as part of the routine clinical evaluation of intracranial mass lesions due to the relatively short imaging and data processing times and the use of a standard dose of contrast agent. Thus, PWI together with conventional MRI should be regarded as the test of choice to diagnose and monitor brain tumors before, during, and after therapy. 3.3.8.3  Dynamic Contrast-Enhanced MR Imaging Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is the acquisition of serial images before, during, and after the administration of extracellular low-molecular-weighted MR contrast media. The resulting signal intensity measurements of the tumor reflect a composite of tumor perfusion, vessel permeability, and the extravascular–extracellular space (Brix et al. 1991; Tofts and Kermode 1991). DCE-MRI has been investigated for a range of clinical oncologic applications including cancer detection, diagnosis, staging, and assessment of treatment response (Padhani and Husband 2001; Choyke et al. 2003; Knopp

et al. 2001). DCE-MRI allows measurement of permeability and its aberrations, while microvascular density (MVD) measures only a histopathologically partial picture of the tissue microvasculature. Furthermore, MVD is a heterogeneous property of tumors and is limited by histopathological sampling. Tumor microvascular meas­ urements by DCE-MRI have been found to correlate with prognostic factors such as tumor grade, microvessel density (MVD), and vascular endothelial growth factor (VEGF) expression and with recurrence and survival outcomes (de Lussanet et al. 2005). In addition, changes of DCE-MRI in follow-up studies during therapeutic intervention have been shown to correlate with outcome (Giesel et al. 2006), suggesting a role for DCE-MRI as a predictive marker. In contrast to conventional (static post-contrast T1-weighted) enhanced MRI, which simply presents a snapshot of enhancement at one time point, DCE-MRI permits a fuller depiction of the wash-in and washout contrast kinetics within tumors, and this provides insight into the nature of the bulk tissue properties on its microvascular level. With the strong demand in drug development (especially with the introduction of anti-VEGF trials) to identify a biomarker that can assess tumor microvascular properties non-invasively in animal as well in human studies (Hylton 2006), this technique seems to be most appealing as a possible imaging biomarker. DCE-MRI also allows further non-invasive characterization of brain lesions and their microcirculatory properties (Fig. 3.3.63). This information can be used for improved biopsy and treatment planning, as well as for monitoring therapeutic interventions, e.g., chemo- and/ or radiotherapy. The dynamic information reflects the angiogenic profile and heterogeneity of a tumor (M. Essig, 2007, unpublished data) (Fig. 3.3.64). Hawighorst et al. (1988) reported that prior and during radiotherapy DCE-MRI is helpful to characterize lesion changes prior to structural changes and predict therapeutic response. Furthermore DCE-MRI in combination with DSCMRI is helpful to assess functional tumor response, which is in concordance to the glioma WHO grading and additionally is helpful for differentiating gliomas from other brain tumors (Ludemann et al. 2001). The method also proved of value in the assessment of meningiomas as the most common nonglial primary tumor and the most common intracranial extranaplastic astrocytomaxial neoplasm. Almost all meningiomas are characterized by a rapid and intense contrast enhancement after the application of contrast material (Hawighorst et al. 1997). However, the intensity of enhancement does not allow the differentiation of typical from atypical tumors. Yang et al. (2003) reported that the exchange rate parameter from dynamic contrast-enhanced studies allowed atypical meningiomas to be distinguished from

297

298

3  Brain, Head, and Neck

Fig. 3.3.63  DCE-MRI in a 64-year-old patient with glioblastoma. The tumor shows a typical strong and ring-shaped enhancement on contrast-enhanced scans (a). The signal intensity time curve of the lesion (b) describes the time-resolved enhancement characteristics with a fast enhancement and a later plateau phase. This is a typical malignant enhancement pattern. The calculated color-coded parameter maps (c) enable visualization of the heterogeneity of these tumors, which enhanced homogenously on late imaging, based on the regional differences in the enhancement characteristics. This information proved to be helpful in biopsy planning and treatment selection

typical meningiomas, independent of the contrast behavior. Other assessed parameters have not been able to show a statistical difference. Hawighorst et al. (1997) assessed the follow-up after radiotherapy with DCE-MRI. Though meningiomas are usually slow-growing lesions, their response to radiotherapy is often difficult to assess and DCE-MRI techniques showed promising results during radiotherapy (Fig. 3.3.10). Changes in tumor volume are easily measured after therapy and reflect possible therapy-induced changes, but do not provide information about the tumor microcirculation. This additional information could be assessed and evaluated by pharmacokinetic analysis on the basis of DCE-MRI technique. 3.3.8.4  Diffusion-Weighted and Diffusion Tensor MRI Diffusion-weighted MRI is used routinely in the assessment of cerebral infarction and infectious diseases. Both DWI and diffusion tensor imaging (DTI) also play an important role in the diagnostic workup and monitoring of patients with cerebral tumors (Stadnik et al. 2002). The

signal intensity of gliomas varies from hyper- to iso- to hypointense. The diffusion signal mainly depends on the cellularity of the lesion with some influence from the amount of necrosis, water content, and hemorrhage. The “normal” glioma appears hyperintense on DWI with a reduced apparent diffusion coefficient (ADC). ADC values cannot be used in individual cases to differentiate glioma types reliably (Kono et al. 2001; Stadnik et al. 2001; Sugahara et al. 1999). However, in the study of Kono et al. (2001), the combination of routine image interpretation and ADC values had a higher predictive value. In the study of Gauvain et al. (2001), there was a clear distinction between the low-grade gliomas and the embryonal tumors. A differentiation between glioma and peritumoral edema based on ADC measurements was also not possible. In patients with metastases the signal intensity of the non-necrotic components was highly variable on DW images (Krabbe et al. 1997). The necrotic components of metastases show marked signal suppression on DWI and increased ADC values, which may be related to increased free water. Increased signal intensity on DW images and a low ADC value are unusual but possible in lesions with hemorrhage (Hartmann et al. 2001).

3.3  Intracranial Tumors

Fig. 3.3.64  DCE MRI in a 48-year-old patient with glioblastoma multiforme prior to serial biopsy planned on the basis of the DCE parameter map. The strong and early enhancing parts of the lesion present with a higher vascular density and higher values of the angiogenic parameter VEGF

In other cerebral tumors the signal intensities on DWI were also found to be very variable and did not allow differentiation with respect to malignancy. Some studies, however, have reported more intense signal in atypical or malignant meningiomas than in typical tumors (Filippi et al. 2001). Lymphomas typically have a hyperintense signal on DWI (Johnson et al. 1991). However, due to the high variability in other tumor entities, differentiation of lymphoma from metastases or glial tumors is not possible. A clear role, however, exists for DWI in the assessment of epidermoid tumors (Fig. 3.3.65). Since these tumors are difficult to exactly detect and delineate based on conventional imaging techniques, they present with a very bright signal on DWI and are easily differentiated from other lesions, especially from arachnoid cysts (Chen et al. 2001). The differential diagnosis of epidermoid and arachnoid cyst is straightforward on DW images. The epidermoid cyst is bright, while the arachnoid cyst is dark. As diffusion is truly a three-dimensional process, molecular mobility in tissues may be anisotropic, as in brain white matter. With DTI, diffusion anisotropy effects can be fully extracted, characterized, and exploited, provid-

Fig. 3.3.65  Three-dimensional reconstruction of the cortical fibers in a normal volunteer. Displayed are the corona radiate and the superior cerebellar peduncle. The starting point for reconstruction was set in the cerebral peduncle

299

300

3  Brain, Head, and Neck Fig. 3.3.66  DTI-based fiber tracking in projection on a 3D volumetric dataset in a patient with glioblastoma multiforme. The study was performed for surgical planning. The tumor causes a medial shift or infiltration of the motoric fibers towards medial. The contralateral side shows normal motor tracts

ing even more exquisite details on tissue microstructure. The most advanced application is certainly that of fiber tracking in the brain (Fig. 3.3.65), which, in combination with other functional MRI techniques, might open a window on the important issue of connectivity. DTI has also been used to demonstrate subtle abnormalities in a variety of diseases (including stroke, multiple sclerosis, dyslexia, and schizophrenia) and is currently becoming part of many routine clinical protocols. DTI has recently proved to be a valuable diagnostic tool in the assessment of intracranial neoplasm (Lu et al. 2004; Stieltjes et al. 2006) (Fig. 3.3.66). In a recent study by Lu et al. (2004), peritumoral diffusion tensor metrics could not be used to distinguish intra-axial from extra-axial lesions or to determine the grade of gliomas preoperatively. However, peritumoral DTI values were reported to be helpful in distinguishing solitary intraaxial metastatic lesions from gliomas. In addition, the method enables one to distinguish presumed tumor-infiltrated edema from vasogenic edema composed purely of extracellular water. These capabilities of diffusion tensor

MR imaging are helpful in current diagnostic scenarios and conceivably will be useful for broader applications in the future. Stieltjes et al. (2006) used a model based on probabilistic voxel classification for a user-independent analysis of DTI-derived parameters. The proposed quantification method proved to be highly reproducible both in healthy controls and patients. Fiber integrity in the corpus callosum (CC) was measured using this quantification method, and the profiles of fractional anisotropy (FA) provided additional information about the possible extent of infiltration of primary brain tumors when compared with conventional imaging (Fig. 3.3.67). This yielded additional information on the nature of ambiguous contralateral lesions in patients with primary brain tumors. The results showed that DTI-derived parameters can be determined reproducibly and may have a strong impact on evaluation of contralateral extent of primary brain tumors. A recent study also proved that the method enables prediction of the patterns of glioma recurrence (Price et al. 2007).

3.3  Intracranial Tumors

Fig. 3.3.67a–e  Diffusion tensor imaging and fractional anisotropy mapping in a patient with anaplastic astrocytoma. In the treatment planning exam (a,b), the large tumor had a mass effect but no definite infiltration of the contralateral side over the corpus callosum. A contralateral lesion was seen (arrow); however, no differentiation from a physiologic subependymal gliosis was possible. The fractional anisotropy (FA) map (c) and the quantitative FA values showed reduced values at the frontal section of the corpus callosum indicating an infiltration not obvious on conventional imaging. After resection and on follow up (e), the lesion demonstrated growth and was confirmed as a tumor manifestation without a clear bridge from the original tumor site

301

302

3  Brain, Head, and Neck

References 1.

2.

3.

4.

5.

6. 7.

8.

9.

10. 11. 12.

13.

14.

15.

16.

17.

Abrey LE, Batchelor TT, Ferreri AJ et al (2005) Report of an international workshop to standardize baseline evaluation and response criteria for primary CNS lymphoma. J Clin Oncol 23:5034–5043 Aguzzi A, Brandner S, Paulus W (2000) Choroid plexus tumours. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 84–86 Akiyama M, Sakai H, Onoue H, Miyazaki Y, AbeT (2004) Imaging intracranial haemangiopericytoma: study of seven cases. Neuroradiology 46:194–197 Alger JR, Frank JA, Bizzi A et al (1990) Metabolism of human gliomas: assessment with H-1 MR spectroscopy and F-18 fluorodeoxyglucose PET. Radiology 177:633–641 Alvord EC, Lofton S (1988) Gliomas of the optic nerve or chiasm: outcome bypatients sage, tumor site, and treatment. Neurosurgery 68:85–98 American Cancer Society (2007) Cancer facts and figs. 2007. American Cancer Society, Atlanta Ariai S, Koerbel A, Bornemann A, Morgala M, Tatagiba M (2005) Cerebellopontine angle arachnoid cyst harbouring ectopic neuroglia. Pediatr Neurosurg 41:220–223 Arienti VM, Botturi A, Boiardi A, Broggi G, Collice M. Fariselli L, Zanni D, Botturi M (2001) Adult brain lowgrade astrocytomas: suvival after surgery and radiotherapy. Neurol Sci 22:233–238 Artigas J, Cervos-Navarro J, Iglesias JR, Ebhardt G (1985) Gliomatosis cerebri: clinical and histological findings. Clin Neuropathol 4:135–148 Asa SL, Ezzat S (1998) The cytogenesis and pathogenesis of pituitary adenomas. Endocr Rev 19:798–827 Ashkan K, Rose P, Walter P (2002) Cystic meningioma: challenges in the diagnosis. Br J Neurosurg 16:72–73 Atlas SW (ed) (1996) Magnetic resonance of the brain and spine, 2nd edn. Lippincott-Raven, Philadelphia, pp 894–898 Atlas SW, Grossman RI, Hackney DB et al (1988) Ca1cified intracraniallesions: detection with gradient-echoacquisition rapid MR imaging. AINR Am I Neuroradiol 9:253–259 Bampoe J, Bauman G, Cairncross J (1999) Adult low-grade gliomas: natural history, prognostic factors and timing of treatment. In: Rock J, Rosenblum M, Shaw E, Cairncross J (eds) The practical management of low-grade primary brain tumors. Lippincott Williams & Wilkins, Philadelphia, pp 135–148 Barbier EL, Lamalle L, Decorps M (2001) Methodology of brain perfusion imaging. J Magn Reson Imaging 13:496–520 Bart J, Groen HJ, Hendrikse NH et al (2000) The blood– brain barrier and oncology: new insights into function and modulation. Cancer Treat Rev 26:449–462 Bartynski WS, Lin L (1997) Dynamic and conventional spin-echo MR of pituitary microlesions. Am J Neuroradiol 18:965–972

18. Ba-Ssalamah A, Nobauer-Huhmann IM, Pinker K et al (2003) Effect of contrast dose and field strength in the magnetic resonance detection of brain metastases. Invest Radiol 38:415–422 19. Becker LE, Cruz-Sanchez FF (2000) Ependymoblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 127–128 20. Behin A, Hoang-Xuan K, Carpentier AF, Delattre JY (2003) Primary brain tumours in adults. Lancet 361:323–331 21. Beni-Adani L, Gomori M, Spektor S, Constantini S (2000) Cyst wall enhancement in pilocytic astrocytoma: neoplastic or reactive phenomenon. Pediatr Neurosurg 32:234–249 22. Bent MJ van den (2004) Management of metastatic (parenchymal, leptomeningeal, and epidural) lesions. Curr Opin Oncol 16:309–313 23. Bisese J (1992) MRI of cranial metastasis. Sem Ultrasound CT MR 13:473–483 24. Black PM (1993) Meningiomas. Neurosurgery 32:643–657 25. Böhling T, Plate KH, Haltia MJ et al (2000) Von Hippel-Lindau disease and capillary haemangioblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 223–226 26. Bolen J, Lipper M, Caccamo D (1989) Intraventricular central neurocytoma: CT and MR findings. J Comput Assist Tomogr 13:495–497 27. Bonneville F, Cattin F, Marsot-Dupuch K, Dormont D, Bonneville JF, Chiras J (2006) T1 signal hyperintensity in the sellar region: spectrum of findings. Radiographics 26:93–113 28. Bourgouin PM, Tampieri D, Grahovac SZ, Leger C, Del Carpio R, Melancon D (1992) CT and MR imaging findings in adults with cerebellar medulloblastoma: comparison with findings in children. AJR Am J Roentgenol 159:609–612 29. Brant-Zawadzki M, Norman D, Newton TH et al (1984) Magnetic resonance of the brain: the optimal screening technique. Radiology 152:71–77 30. Brix G, Semmler W, Port R et al (1991) Pharmacokinetic parameters in CNS Gd-DTPA enhanced MR imaging. J Comput Assist Tomogr 15:621–628 31. Burger PC, Scheithauer BW (1994) Atlas of tumor pathology, 3rd series, fascicle 10: tumors of the central nervous system. Armed Forces Institute of Pathology, Washington, D.C. 32. Burger PC, Breiter SN, Fisher PG (1996) Pilocytic and fibrillary astrocytomas of the brain stem: a comparative clinical, radiological, and pathological study (Abstr). J Neuropathol Exp Neurol 55:640 33. Burger PC, Scheithauer BW, Paulus W et al (2000) Pilocytic astrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 45–51 34. Byrne TN (1994) Imaging of gliomas. Semin Oncol 21:162–171

3.3  Intracranial Tumors 35. Cappabianca P, Cirillo S, Alfieri A, D’Amico A, Maiuri F, Mariniello G, Caranci F, de Divitiis E (1999) Pituitary macroadenoma and diaphragma sellae meningioma: differential diagnosis on MRI. Neuroradiology 41:22–26 36. Castillo M, Green C, Kwock L et al (1995) Proton MR spectroscopy in patients with neurofibromatosis type 1: evaluation of hamartomas and clinical correlation. AJNR Am J Neuroradiol 16:141–147 37. Cavenee WK, Furnari FB, Nagane M et al (2000) Diffusely infiltrating astrocytomas. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 10–21 38. Cha S, Knopp EA, Johnson G et al (2000) Dynamic contrast-enhanced T2-weighted MR imaging of recurrent malignant gliomas treated with thalidomide and carboplatin. AJNR Am J Neuroradiol 21:881–890 39. Cha S, Knopp EA, Johnson G et al (2002) Intracranial mass lesions: dynamic contrast-enhanced susceptibilityweighted echo-planar perfusion MR imaging. Radiology 223:11–29 40. Chang KH, Ra DG, Han MH et al (1994) Contrast enhancement of brain tumors at different MR field strengths: comparison of 0.5 T and 2.0 T. AJNR Am J Neuroradiol 15:1413–1419 41. Charbel FT, Hyun H, Misra M, Gueryikian S, Mafee RF (1999) Juxtaorbital en plaque meningiomas. Report of four cases and review of literature. Radiol Clin North Am 37:89–100 42. Chawla A, Emmanuel JV, Seow WT, Lou J, Teo HE, Lim CC (2007) Pediatric PNET: pre-surgical MRI features. Clin Radiol 62:43–52 43. Chen S, Ikawa F, Kurisu K, Arita K, Takaba J, Kanou Y (2001) Quantitative MR evaluation of intracranial epidermoid tumors by fast fluid-attenuated inversion recovery imaging and echo-planar diffusion-weighted imaging. AJNR Am J Neuroradiol 22:1089–1096 44. Chong BW, Newton TH (1993) Hypothalamic and pituitary pathology. Radiol Clin North Am 31:1147–1153 45. Choyke PL, Dwyer AJ, Knopp MV (2003) Functional tumor imaging with dynamic contrast-enhanced magnetic resonance imaging. J Magn Reson Imaging 17:509–520 46. Cohan P, Foulad A, Esposito F, Martin NA, Kelly DF (2004) Symptomatic Rathke’s cleft cysts: a report of 24 cases. J Endocrinol Invest 27:943–948 47. Colosimo C, Ruscalleda J, Korves M et al (2001) Detection of intracranial metastases: a multi-center, intra-patient comparison of gadobenate dimeglumine-enhanced MRI with routinely used contrast agents at equal dose. Invest Radiol 36:72–81 48. Daumas-Duport C, Pietsch T, Lantos PL (2000) Dysembryoplastic neuroepithelial tumour. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 103–106

49. David KF, Casey ATH, Hayward RD, Harkness WFJ, Phipps K, Wade AM (1997) Medulloblastoma: is the 5-year survival rate improving? A review of 80 cases from a single institution. J Neurosurg 86:13–21 50. Davis PC, Hudgin PA, Peterman SB, Hoffman JC Jr (1991) Diagnosis of cerebral metastases: double-dose delayed CT vs contrast enhanced MR imaging. Am J Neuroradiol 12:293–300 51. De Coene B, Hajnal JV, Gatehouse P et al 1992MR of the brain using fluid-attenuated inversion recovery (FLAIR) pulse sequences. Am J Neuroradiol 13:1555–1564 52. Dechambre S, Duprez T, Lecouvet F, Raftopoulos C, Gosnard G (1999) Diffusion-weighted MRI postoperative assessment of an epidermoid tumor in the cerebellopontine angle. Neuroradiology 41:829–831 53. Delorme S, Weber MA (2006) Applications of MRS in the evaluation of focal malignant brain lesions. Cancer Imaging 6:95–99 54. Demaerel P, Johannik K, van Hecke P et al (1991) Localized 1H NMR spectroscopy in fifty cases of newly diagnosed intracranial tumors. J Comput Assist Tomogr 15:67–76 55. Demeule M, Regina A, Jodoin J et al (2002) Drug transport to the brain: key roles for the efflux pump P-glycoprotein in the blood–brain barrier. Vascul Pharmacol 38:339–348 56. Dowling C, Bollen AW, Noworolski SM et al (2001) Preoperative proton MR spectroscopic imaging of brain tumors: correlation with histopathologic analysis of resection specimens. AJNR Am J Neuroradiol 22:604–612 57. Dunn J Jr, Kernohan JW (1956) Gliomatosis cerebri. Arch Pathol 64:82–91 58. Earnest F IV, Kelly PJ, Scheithauer BW et al (1988) Cerebral astrocytomas: histopathologic correlation of MR imaging and CT contrast enhance­ment with stereotactic biopsy. Radiology 166:823–827 59. Elster AD (1993) Modern imaging of the pituitary. Radiology 187:1–14 60. Elster AD (1997) How much contrast is enough? Dependence of enhancement on field strength and MR pulse sequence. Eur Radiol 7:276–280 61. Elster AD, Challa VR, Gilbert TH, Richardson DN, Contento JC (1989) Meningiomas: MR and histopathologic features. Radiology 170:857–862 62. Erickson BJ, Campeau NG, Schreiner SA et al (2002) Tripledose contrast/magnetization transfer suppressed imaging of “non-enhancing” brain gliomas. J Neurooncol 60:25–29 63. Essig M, Hawighorst H, Schönberg SO et al (1998) Fast fluid-attenuated inversion-recovery (FLAIR) MR imaging in the assessment of intra-axial brain tumors. J Magn Reson Imag 8:789–798 64. Essig M, Schönberg SO, Hawighorst H et al (1999) Cerebral gliomas and metastases: Assessment with contrast enhanced fast fluid-attenuated inversion-recovery MR imaging. Radiology 210:551–557 65. Essig M, Waschkies M, Wenz F et al (2003) Assessment of brain metastases with dynamic susceptibility-weighted contrast-enhanced MR imaging: initial results. Radiology 228:193–199

303

304

3  Brain, Head, and Neck 66. Ezzat S, Asa SL, Couldwell WT et al (2004) The prevalence of pituitary adenomas: a systematic review. Cancer 101:613–619 67. Fedi M, Anne Mitchell L, Kalnins RM et al (2004) Glioneuronal tumours in neurofibromatosis type 1: MRI-pathological study. J Clin Neurosci 11:745–747 68. Figarella-Branger D, Soylemezoglu F, Kleihues P et al (2000) Central neurocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 107–109 69. Filippi CG, Edgar MA, Ulu AM et al (2001) Appearance of meningiomas on diffusion-weighted images: correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 22:65–72 70. Fine HA, Mayer RJ (1993) Primary central nervous system lymphoma. Ann Intern Med 119:1093–104 71. Forsting M, Albert FK, Kunze S, Adams HP, Zenner D, Sartor K (1993) Extirpation of glioblastomas: MR and CT follow-up of residual tumor and regrowth patterns. Am J Neuroradiol 14:77–87 72. Fuss M, Wenz F, Essig M et al (2001) Tumor angiogenesis of low-grade astrocytomas measured by dynamic susceptibility contrast-enhanced MRI (DSC-MRI) is predictive of local tumor control after radiation therapy. Int J Radiat Oncol Biol Phys 51:478–482 73. Galanaud D, Nicoli F, Chinot O, Confort-Gouny S, Figarella-Branger D, Roche P, Fuentes S, Le Fur Y, Ranjeva JP, Cozzone PJ (2006) Noninvasive diagnostic assessment of brain tumors using combined in vivo MR imaging and spectroscopy. Magn Reson Med 55:1236–1245 74. Gao PY, Osborn AG, Smirniotopoulos JG, Harris CP (1992) Radiologic-pathologic correlation. Epidermoid tumor of the cerebellopontine angle. AJNR Am J Neuroradiol 13:863–872 75. Gauvain KM, McKinstry RC, Mukherjee P et al (2001) Evaluating pediatric brain tumor cellularity with diffusiontensor imaging. AJR Am J Roentgenol 177:449–454 76. Gelabert-Gonzalez M (2004) Intracranial arachnoid cysts. Rev Neurol 39:1161–1166 77. Giangaspero F, Bigner SH, Kleihues P, Pietsch T, Trojanowski JQ (2000) Medulloblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 129–137 78. Giesel FL, Bischoff H, von Tengg-Kobligk H et al (2006) Dynamic contrast-enhanced MRI of malignant pleural mesothelioma: a feasibility study of noninvasive assessment, therapeutic follow-up, and possible predictor of improved outcome. Chest 129:1570–1576 79. Graif M, Bydder GM, Steiner RE et al (1985) Contrast-enhanced MR imag­ing of malignant brain tumors. AINR Am I Neuroradiol 6:855–862 80. Groothuis DR (2000) The blood–brain and blood–tumor barriers: a review of strategies for increasing drug delivery. Neurooncol 2:45–59

81. Gururangan S, Friedman HS (2002) Innovations in design and delivery of chemotherapy for brain tumors. Neuroimaging Clin N Am 12:583–597 82. Hagen T, Kujat C, Donauer E, Piepgras U (1994) Neuroradiological aspects of intracranial epidermoids. Radiologe 34:639–647 83. Hall WA, Luciano MG, Doppman JL et al (1994) Pituitary magnetic resonance imaging in normal human volunteers: occult adenomas in the general population. Ann Intern Med 120:817–820 84. Hammound MA, Sawaya R, Shi W, Thall PF, Leeds NE (1996) Prognostic significance of preoperative MRI scans in glioblastoma multiforme. J Neurooncol 27:65–73 85. Hardy J (1973) Transsphenoidal surgery of hypersecreting pituitary tumors. In: Kohler PO, Ross GT (eds) Diagnosis and treatment of pituitary tumors: proceedings of a conference sponsored jointly by the National Institute of Child Health and Human Development and the National Cancer Institute, 15–17 January 1973, Bethesda, Md. Excerpta medica, Amsterdam, pp 179–198 86. Hartmann M, Jansen O, Heiland S, Sommer C, Mönkel K, Sartor K (2001) Restricted diffusion within ring enhancement is not pathognomonic for brain abscess. AJNR Am J Neuroradiol 22:1738–1742 87. Hartmann M, Heiland S, Harting I et al (2003) Distinguishing of primary cerebral lymphoma from high-grade glioma with perfusion-weighted magnetic resonance imaging. Neurosci Lett 338:119–122 88. Haustein J, Laniado M, Niendorf HP et al (1992) Administration of gadopentetate dimeglumine in MR imaging of intracranial tumors: dosage and field strength. AJNR Am J Neuroradiol 13:1199–1206 89. Hawighorst H, Engenhart R, Knopp MV et al (1997) Intracranial meningiomas: time- and dose-dependent effects of irradiation on tumor microcirculation monitored by dynamic MR imaging. Magn Reson Imaging 15:423–432 90. Hawighorst H, Knopp MV, Debus J et al (1998) Pharmacokinetic MRI for assessment of malignant glioma response to stereotactic radiotherapy: initial results. J Magn Reson Imaging 8:783–788 91. Ho VB, Smirniotopoulos JG, Murphy FM et al (1992) Radiologic-pathologic correlation: hemangioblastoma. AJNR Am J Neuroradiol 13:1343–1352 92. Hoang-Xuan K, Capelle L, Kujas M et al (2004) Temozolomide as initial treatment for adults with low-grade oligodendrogliomas or oligoastrocytomas and correlation with chromosome 1p deletions. J Clin Oncol 22:3133–3138 93. Hoeffel C, Boukobza M, Polivka M et al (1995) MR manifestations of subependymomas. AJNR Am J Neuroradiol 16:2121–2129 94. Holland BA, Kucharcyzk W, Brant-Zawadzki M et al (1985) MR imaging of ca1cified intracraniallesions. Radiology 157:353–356 95. Holodny AI, Nusbaum AO, Festa S et al (1999) Correlation between the degree of contrast enhancement and the volume of peritumoral edema in meningiomas and malignant gliomas. Neuroradiology 41:820–825

3.3  Intracranial Tumors 96. Hylton N (2006) Dynamic contrast-enhanced magnetic resonance imaging as an imaging biomarker. J Clin Oncol 24:3293–3298 97. Jääskeläinen J, Serve A, Haltia M, Wahlstrom T, Valtronen S (1985) Intracranial hemangiopericytoma: radiology, surgery, radiotherapy and outcome in 21 patients. Surg Neurol 23:227–236 98. Jääskeläinen J, Louis DN, Paulus W et al (2000) Haemangiopericytoma. I In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 190–192 99. Jackson A, Panizza B, Hughes D, Reid H (1992) Primary choroid plexus papilloma of the cerebellopontine angle: magnetic resonance imaging, computed tomographic and angiographic appearances. Br J Radiol 65:754–757 100. Jager L, Reiser M (2001) CT and MR imaging of the normal and pathologic conditions of the facial nerve. Eur J Radiol Nov 40:133–146 101. Jellinger K, Chou P, Paulus W (2000) Melanocytic lesions. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 193–195 102. Jeyapalan S, Batchelor T (2000) Diagnostic evaluation of neurologic metastases. Cancer Invest 18:381–394 103. Johnsen De, Woodruff WW, Allen IS, Cera PJ, Funkhouser GR, Coleman LL (1991) MR imaging of the sellar and juxtasellar regions. Radiographics 11:727–758 104. Johnson BA, Fram EK, Johnson PC, Jacobowitz R (1997) The variable MR appearance of primary lymphoma of the central nervous system: comparison with histopathologic features. AJNR Am J Neuroradiol 18:563–572 105. Johnson JD, Young B (1996) Demographics of brain metastasis. Neurosurg Clin North Am 7:337–344 106. Jones AP, Gunawardena WJ, Coutinho CM (2001) 1H MR spectroscopy evidence for the varied nature of asymptotic focal brain lesions in neurofibromatosis type 1. Neuroradiology 43:62–67 107. Kazner E, Wende S, Grumme T et al (1989) Computed tomography and magnetic resonance tomography of intracranial tumours, 2nd edn. Springer, Berlin Heidelberg New York 108. Keene DL, Jimenez C, Hsu-E (1999) MRI diagnosis of gliomatosis cerebri. Pediatr Neurol 20:148–151 109. Kemper EM, Boogerd W, Thuis I et al (2004) Modulation of the blood–brain barrier in oncology: therapeutic opportunities for the treatment of brain tumours? Cancer Treat Rev 30:415–423 110. Kepes JJ, Louis DN, Giannini C et al (2000) Pleomorphic xanthoastrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 52–54 111. Kido G, Wright JL, Merchant RE (1991) Acute effects of human recombinant tumor necrosis factor-alpha on the cerebral vasculature of the rat in both normal brain and in an experimental glioma model. J Neurooncol 10:95–109

112. Kitange GJ, Smith JS, Jenkins RB (2001) Genetic alterations and chemotherapeutic response in human diffuse gliomas. Expert Rev Anticancer Ther 1:595–605 113. Kitange GJ, Templeton KL, Jenkins RB (2003a) Recent advances in the molecular genetics of primary gliomas. Curr Opin Oncol 15:197–203 114. Kitange GJ, Templeton KL, Jenkins RB (2003b) Recent advances in the molecular genetics of primary gliomas. Curr Opin Oncol 15:197–203 115. Kleihues P, Burger PC, Scheithauer BW (1993) The new WHO classification of brain tumours. Brain Pathology 3:255–268 116. Kleihues P, Burger PC, Collins VP et al (2000a) Glioblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 29–39 117. Kleihues P, Davis RL, Coons SW et al (2000b) Anaplastic astrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 27–28 118. Kleihues P, Davis RL, Ohgaki H et al (2000c) Diffuse astrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 22–26 119. Kleihues P, Ohgeki H (2001) Primary and secondary glioblastomas: from concept to clinical diagnosis. Neuro Oncol 52:181–188 120. Knopp MV, Giesel FL, Marcos H et al (2001) Dynamic contrast-enhanced magnetic resonance imaging in oncology. Top Magn Reson Imaging 12:301–308 121. Knopp MV, Balzer T, Esser M et al (2006) Assessment of utilization and pharmacovigilance based on spontaneous adverse event reporting of gadopentetate dimeglumine as a magnetic resonance contrast agent after 45 million administrations and 15 years of clinical use. Invest Radiol 41:491–499 122. Koci TM, Chiang F, Mehringer CM et al (1993) Adult cerebellar medulloblastoma: imaging features with emphasis on MR findings. AJNR Am J Neuroradiol 14:929–939 123. Koeller KK, Alamo L, Adair CF, Smirniotopoulos JG (1999) Congenital cystic masses of the neck: radiologic-pathologic correlation. Radiographics 19:121–146 124. Koeller KR, Smirniotopoulos JG, Jones RV (2007) Primary central nervous system lymphoma: Radiologogic-pathologic correlation. Radiographics 17:1497–1526 125. Kono K, Inoue Y, Nakayama K et al (2001) The role of diffusion-weighted imaging in patients with brain tumors. AJNR Am J Neuroradiol 22:1081–1088 126. Kornreich L, Blaser S, Schwarz M et al (2001) Optic pathway glioma: correlation of imaging findings with the presence of neurofibromatosis. AJNR Am J Neuroradiol 22:1963–1969 127. Krabbe K, Gideon P, Wagn P, Hansen U, Thomsen C, Madsen F (1997) MR diffusion imaging of human intracranial tumours. Neuroradiology 39:483–489

305

306

3  Brain, Head, and Neck 128. Kroll RA, Neuwelt EA (1998) Outwitting the blood–brain barrier for therapeutic purposes: osmotic opening and other means. Neurosurgery 42:1083–1099 129. Kuroiwa T, Ohta T, Tsutsumi A (1999) Malignant pilocytic astrocytoma in the medulla oblongata: case report. Brain Tumor Pathol 16:81–85 130. Law M (2004) MR spectroscopy of brain tumors. Top Magn Reson Imaging Oct 15:291–313 131. Law M, Cha S, Knopp EA et al (2002) High-grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 222:715–721 132. Law M, Yang S, Wang H et al (2003) Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 24:1989–1998 133. Law M, Oh S, Babb JS et al (2006) Low-grade gliomas: dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging—prediction of patient clinical response. Radiology 238:658–667 134. Lee DH, Norman D, Newton TH (1987) MR imaging of pineal cysts. J Comput Assist Tomogr 1:586–590 135. Lee Y, Van Tassel P (1989) Intracranial oligodendrogliomas: imaging find­ings in 35 untreated cases. AINR Am I Neuroradiol 10:119–127 136. Lee Y, Van Tassel P, Bruner JM et al (1989) Juvenile pilocytic astr, tomas: CT and MR imaging characteri,stics. AJNR Am J Neuroradiology 10:363–370 137. Levin VA, Leibel SA, Gutin PH (2001) Neoplasms of the central nervous system. In: DeVita VT Jr, Hellman S, Rosenberg SA (eds) Cancer: principles and practice of oncology, 6th edn. Lippincott Williams & Wilkins, Philadelphia, pp 2100–2160 138. Lipper MH, Eberhard DA, Phillips CD, Vezina LG, Cail WS (1993) Pleomorphic xanthoastrocytoma, a distinctive astroglial tumor: neuroradiologic and pathologic features. AJNR Am J Neuroradiol 14:1397–1404 139. Lopes MBS, VandenBerg SR, Scheithauer BW (1993) The World Health Organization classification of nervous system tumors in experimental neuro-oncology. In: Levine AJ, Schmidek HH (eds) Molecular genetics of nervous system tumors. Wiley-Liss, New York, pp 1–36 140. Lu S, Ahn D, Johnson G, Law M, Zagzag D, Grossman RI (2004) Diffusion-tensor MR imaging of intracranial neoplasia and associated peritumoral edema: introduction of the tumor infiltration index. Radiology 232:221–228 141. Ludemann L, Grieger W, Wurm R et al (2001) Comparison of dynamic contrast-enhanced MRI with WHO tumor grading for gliomas. Eur Radiol 11:1231–1241 142. Lussanet QG de, Langereis S, Beets-Tan RG et al (2005) Dynamic contrast-enhanced MR imaging kinetic parameters and molecular weight of dendritic contrast agents in tumor angiogenesis in mice. Radiology 235:65–72 143. Mafee MF, Goodwin J, Dorodi S (1999) Optic nerve sheath meningiomas. Role of MR imaging. Radiol Clin North Am 37:37–58

144. Majos C, Cucurella G, Aguilera C, Coll S, Pons L (1999) Intraventricular meningiomas: MR imaging and MR spectroscopic findings in two cases. AJNR Am J Neuroradiol 20:882–885 145. Majos C, Julia-Sape M, Alonso J, Serrallonga M, Aguilera C, Acebes JJ, Arus C, Gili J (2004) Brain tumor classification by proton MR spectroscopy: comparison of diagnostic accuracy at short and long TE. AJNR Am J Neuroradiol 25:1696–1704 146. Maleci A, Cervoni L, Delfini R (1992) Medulloblastoma in children and in adults: a comparative study. Acta Neurochir (Wien) 119:62–67 147. Marks JE, Gado M (1977) Serial computed tomography of primary brain tumors following surgery, irradiation, and chemotherapy. Radiology 125:119–125 148. Maroldi R, Farina D, Palvarini L, Marconi A, Gadola E, Menni K, Battaglia G (2001) Computed tomography and magnetic resonance imaging of pathologic conditions of the middle ear. Eur J Radiol 40:78–93 149. Martin N, Pierot L, Sterkers O, Mompoint D, Nahum H (1990) Primary choroid plexus papilloma of the cerebellopontine angle: MR imaging. Neuroradiology 31:541–543 150. Mena H, Nakazato Y, Jouvet A et al (2000) Pineocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 118–120 151. Mena H, Nakazato Y, Jouvet A et al (2000) Pineoblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 116–118 152. Meyerand ME, Pipes JM, Mamourian A, Tosteson TD, Dunn JF (1999) Classification of biopsy-confirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol 20:117–123 153. Meyers SP, Wildenhain SL, Chang JK et al (2000) Postoperative evaluation for disseminated medulloblastoma involving the spine: contrast-enhanced MR findings, CSF cytologic analysis, timing of disease occurrence, and patient outcomes. AJNR Am J Neuroradiol 21:1757–1765 154. Mishima K, Nakamura M, Nakamura H, Nakamura O, Funata N, Shitara N (1992) Leptomeningeal dissemination of cerebellar pilocytic astrocytoma: case report. J Neurosurg 77:788–791 155. Muroff LR, Runge VM (1995) The use of MR contrast in neoplastic disease of the brain. Top Magn Reson Imaging 7:137–157 156. Naidich TP, Zimmerman RA (1984) Primary brain tumors in children. Semin Roentgenol 19:100–114 157. Nakano T, Asano K, Miura, Itoh S, Suzuki S (2002) Meningiomas with brain edema: radiological characteristics on MRI and review of the literature. Clin Imaging 26:243–249 158. Naylor MF, Scheithauer BW, Forbes GS, Tomlinson FH, Young WF (1995) Rathke cleft cyst: CT, MR, and pathology of 23 cases. J Comput Assist Tomogr 19:853–859 159. Negendank WG, Sauter R, Brown TR et al (1996) Proton magnetic resonance spectroscopy in patients with glial tumors: a multicenter study. J Neurosurg 84:449–458

3.3  Intracranial Tumors 160. Nelson JS, Bruner JM, Wiestler OD et al (2000a) Ganglioglioma and gangliocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 96–98 161. Nelson JS, Von Deimling A, Petersen I et al (2000b) Metastatic tumours of the CNS. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 250–253 162. Neuwelt EA (2004) Mechanisms of disease: the blood– brain barrier. Neurosurgery 54:131–140 163. Nikas I, Anagnostara A, Theophanopoulou M, Stefanski K, Michail A, Hdjegeorgi Ch (2004) Desmoplastic infantile ganglioglioma: MRI and histological findings case report. Neuroradiology. Dec 46:1039–1043 164. Norfray JF, Darling C, Byrd S et al (1999) Short TE proton MRS and neurofibromatosis type 1 intracranial lesions. J Comput Assist Tomogr 23:994–1003 165. Oeckler R, Feiden W (1991) Benign symptomatic lesions of the pineal gland. Report of seven cases treated surgically. Acta Neurochirurg 108:40–44 166. Ohgaki H, Kleihues P (2005) Population-based studies on incidence, survival rates, and genetic alterations in astrocytic and oligodendroglial gliomas. J Neuropathol Exp Neurol 64:479–489 167. O’Neill BP, Dinapoli RP, Kurtin PJ et al (1995) Occult systemic non-Hodgkin’s lymphoma (NHL) in patients initially diagnosed as primary central nervous system lymphoma (PCNSL): how much staging is enough? J Neurooncol 25: 67–71 168. Padhani AR, Husband JE (2001) Dynamic contrast-enhanced MRI studies in oncology with an emphasis on quantification, validation and human studies. Clin Radiol 56:607–620 169. Patchell RA (2003) The management of brain metastases. Cancer Treat Rev 29:533–540 170. Pencalet P, Maixner W, Sainte-Rose C et al (1999) Benign cerebellar astrocytomas in children. J Neurosurg 90:265–273 171. Phillipon JH, Clemenceau SH, Fauchon FH et al (1993) Supratentorial low­grade astrocytomas inadults. Neurosurgery 32:554–559 172. Pierallini A, Bonamini M, Bozzao A et al (1997) Supratentorial diffuse astrocytic tumours: proposal of an MRI classification. Eur Radiol 7:395–399 173. Pierallini A, Bonamini M, Di Stefano D, Siciliano P, Bozzao L (1999) Pleomorphic xanthoastrocytoma with CT and MRI appearance of meningioma. Neuroradiology 41:30–34 174. Pisaneschi M, Kapoor G (2005) Imaging the sellar and parasellar region. Neuroimaging Clin N Am 15:203–219 175. Pollack IF, Lunsford LD, Flickinger JC et al (1989) Prognostic factors in the diagnosis and treatment of primary central nervous system lymphoma. Cancer 63:939–947

176. Ponce P, Alvarez-Santullano MV, Otermin E, Santana MA, Garcia-Ludena, MV (1998) Gliomatosi cerebri: findings with computed tomography and magnetic resonance imaging. Eur J Radiol 28:226–229 177. Price SJ, Jena R, Burnet NG, Carpenter TA, Pickard JD, Gillard JH (2007) Predicting patterns of glioma recurrence using diffusion tensor imaging. Eur Radiol (in press) 178. Pronin IN, Holodny AI, Petraikin AV (1997) MRI of highgrade glial tumors: correlation between the degree of contrast enhancement and the volume of surrounding edema. Neuroradiology 39:348–350 179. Provenzale JM, York G, Moya MG, Parks L, Choma M, Kealey S, Cole P, Serajuddin H (2006a) Correlation of relative permeability and relative cerebral blood volume in high-grade cerebral neoplasms. AJR Am J Roentgenol 187:1036–1042 180. Provenzale JM, Mukundan S, Barboriak DP (2006b) Diffusion-weighted and perfusion MR imaging for brain tumor characterization and assessment of treatment response. Radiology 239:632–649 181. Ragel BT, Couldwell WT (2004) Pituitary carcinoma: a review of the literature. Neurosurg Focus 16:E7 182. Rand T, Lippitz P, Kink E, Huber H, Schneider B, Imhof H (2002) Evaluation of pituitary microadenoma with dynamic MR imaging. Eur J Radiol 41:131–135 183. Rautioa J, Chikhale PJ (2004) Drug delivery systems for brain tumor therapy. Curr Pharm Des 10:1341–1353 184. Reifenberger G, Kros JM, Burger PC et al (2000a) Oligodendroglioma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 56–61 185. Reifenberger G, Kros JM, Burger PC et al (2000b) Oligoastrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 65–67 186. Rennert J, Doerfler A (2007) Imaging of sellar and parasellar lesions. Clin Neurol Neurosurg 109:111–124 187. Ries LAG, Eisner MP, Kosary CL et al (2003)Trends in SEER incidence and US mortality using the joinpoint regression program 1975–2000 with up to three joinpoints by race and sex. In: Ries LAG, Eisner MP, Kosary CL et al (eds) SEER Cancer Statistics Review, 1975–2000. Section 3: Brain and other nervous system cancer (invasive), Table III-1. National Cancer Institute, Bethesda, Md. 188. Roberts RO, Lynch CF, Jones MP, Hart MN (1991) Medulloblastoma: a population-based study of 532 cases. J Neuropathol Exp Neurol 50:134–144 189. Romero FJ, Ortega A, Titus F, Ibarra B, Navarro C, Rovira M (1988) Gliomatosis cerebri with formation of a glioblastoma multiform. Study and follow-up by magnetic resonance and computed tomography. J Comput Assist Tomogr 12:253–257 190. Rorke LB, Hart MN, McLendon RE (2000) Supratentorial primitive neuroectodermal tumour (PNET). In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 141–144

307

308

3  Brain, Head, and Neck 191. Ruscalleda J, Feliciani M, Avila A, Castaner E, Guardia E, de Juan M (1994) Neuroradiological features of intracranial and intraorbital meningeal haemangiopericytomas. Neuroradiology 36:440–445 192. Rydberg JN, Hammond CA, Grimm RC et al (1994) Initial clinical experience in MR imaging of the brain with a fast fluid-attenuated inversion-recovery pulse sequence. Radiology 193:173–180 193. Salmon I, Dewitte 0, Pasteeis JL et al (1994) Prognostic scoring in adult as­trocytic tumors using patient age, histopathologic grade, and DNA histogram type. I Neurosurg 80:877–883 194. Sari A, Dinc H, Gumele HR (1998) Interhemispheric lipoma associated with subcutaneous lipoma. Eur Radiol 8:628–630 195. Scatliff JH, Kwock L, Chancellor K, Bouldin TW, Kapoor CC, Castillo M (1997) Postmortem MR imaging of the brains of patients with AIDS. Neuroimaging Clin N Am 7:297–320 196. Scheithauer BW, Kovacs KT, Laws ER Jr et al (1986) Pathology of invasive pituitary tumours with special reference to functional classification. J Neurosurg 65:733–744 197. Schneider G, Kirchin MA, Pirovano G et al (2001) Gadobenate dimeglumine-enhanced magnetic resonance imaging of intracranial metastases: effect of dose on lesion detection and delineation. J Magn Reson Imaging 14:525–539 198. Schneider JH, Jr Raffel C, McComb JG (1992) Benign cerebellar astrocytomas of childhood. Neurosurgery 30:58–63 199. Schneider SW, Ludwig T, Tatenhorst L et al (2004) Glioblastoma cells release factors that disrupt blood–brain barrier features. Acta Neuropathol (Berl) 107:272–276 200. Schwartzberg DG (1992) Imaging of pituitary tumors. Semin Ultrasound CT MR Jun 13:207–223 201. Shaw EG, Scheithauer BW, O’Fallon I et al (1992) Oligodendrogliomas: the Mayo Clinic experience. I Neurosurg 76:428–434 202. Smirniotopoulos JG, Rushing EJ, Mena H (1992) Pineal region massess: differential diagnosis. Radiographics 12:577 203. Smirniotopoulos JG, Yue NC, Rushing EJ (1993) Cerebellopontine angle masses: radiologic–pathologic correlation. Radiographics 13:1131–1147 204. Smirniotopoulos JG, Chiechi MV (1995) Teratomas, dermoids, and epidermoids of the head and neck. Radiographics 15:1437–1455 205. Smirniotopoulos JG, Yue NC, Rushing EJ (1996) Cerebellopontine angle masses: radiologic-pathologic correlation. Radiographics 16:1131–1143 206. Smirniotopoulos JG, Murphy FM, Rushing EJ, Rees JH, Schroeder JW (2007) From the archives of the AFIP: patterns of contrast enhancement in the brain and meninges. Radiographics 27:525–551 207. Spagnoli MV, Goldberg HI, Grossman RI, Bilaniuk LT, Gomori JM, Hackney DB, Zimmerman RA (1986) Intracranial meningiomas: high-field MR imaging. Radiology 161:369–375

208. Spagnoli MV, Goldberg HI, Grossman RI et al (1986) Intracranial meningiomas: high-field MR imaging. Radiology 161:369–375 209. Spoto G, Press G, Hesselink J, Solomon M (1990) Intracranial ependymoma and subependymoma: MR manifestations. AJNR Am J Neuroradiol 11:83–91 210. Stadnik TW, Chaskis C, Michotte A et al (2001) Diffusionweighted MR imaging of intracerebral masses: comparison with conventional MR imaging and histologic findings. AJNR Am J Neuroradiol 22:969–976 211. Stadnik TW, Demaerel P, Luypaert RR et al (2002) Imaging Tutorial: Differential diagnosis of bright lesion on diffusion-weighted MR images. Radiographics doi:10.1148/radiol.e7 212. Stieber VW (2001) Low-grade gliomas. Curr Treat Options Oncol 2:495–506 213. Stieltjes B, Schluter M, Didinger B, Weber MA, Hahn HK, Parzer P, Rexilius J, Konrad-Verse O, Peitgen HO, Essig M (2006) Diffusion tensor imaging in primary brain tumors: reproducible quantitative analysis of corpus callosum infiltration and contralateral involvement using a probabilistic mixture model. Neuroimage 31:531–542 214. Strong JA, Hatten HP Jr, Brown MT et al (1993) Pilocytic astrocytoma: correlation between the initial imaging features and clinical aggressiveness. AJR Am J Roentgenol 161:369–372 215. Sugahara T, Korogi Y, Kochi M et al (1999) Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 9:53–60 216. Sze G, Johnson C, Kawamura Y et al (1998) Comparison of single- and triple-dose contrast material in the MR screening of brain metastases. AJNR Am J Neuroradiol 19:821–828 217. Takada Y, Ohno K, Tamaki M, Hirakawa K (1999) Cerebellopontine angle pilocytic astrocytoma mimicking acoustic schwannoma. Neuroradiology 41:949–950 218. Tasdemiroglu E, Awh M, Walsh J (1996) MRI of cerebellopontine angle choroid plexus papilloma. Neuroradiology 38:38–40 219. Thurnher MM, Thurnher SA, Schindler E (1997) CNS involvement in AIDS: spectrum of CT and MR findings. Eur Radiol 7:1091–1097 220. Tien RD, Cardenas CA, Rajagopalan S (1992) Pleomorphic xanthoastrocytoma of the brain: MR findings in six patients. AJR Am J Roentgenol 159:1287–1290 221. Tofts PS, Kermode AG (1991) Measurement of the blood– brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 17:357–367 222. Tsuchida T, Kamata K, Kawamata M et al (1984) Brain tumors in tuberous sclerosis: reports of 4 cases. Childs Brain 8:271–283 223. Tsuchiya K, Mizutani Y, Hachiya J (1996) Preliminary evaluation of fluid-attenuated inversion-recovery MR in the diagnosis of intracranial tumors. Am J Neuroradiol 17:1081–1086

3.3  Intracranial Tumors 224. Uematsu H, Maeda M (2006) Double-echo perfusionweighted MR imaging: basic concepts and application in brain tumors for the assessment of tumor blood volume and vascular permeability. Eur Radiol 16:180–186 225. Van DP, Sijens PE, Schmitz PI et al (1997) Gd-enhanced MR imaging of brain metastases: contrast as a function of dose and lesion size. Magn Reson Imaging 15:535–541 226. Vuori K, Kankaanranta L, Hakkinen AM et al (2004) Lowgrade gliomas and focal cortical developmental malformations: differentiation with proton MR spectroscopy. Radiology 230:703–708 227. Weber F, Knopf H (2006) Incidental findings in magnetic resonance imaging of the brains of healthy young men, J Neurol Sci 240:81–84 228. Weber MA, Thilmann C, Lichy MP et al (2004) Assessment of irradiated brain metastases by means of arterial spin-labeling and dynamic susceptibility-weighted contrast-enhanced perfusion MRI: initial results. Invest Radiol 39:277–287 229. Weber MA, Zoubaa S, Schlieter M et al (2006) Diagnostic performance of spectroscopic and perfusion MRI for distinction of brain tumors. Neurology 66:1899–1906 230. Wichmann W, Schubiger O, von DA, Schenker C, Valavanis A (1991) Neuroradiology of central neurocytoma. Neuroradiology 33:143–148 231. Wiestler OD, Lopes BS, Green AJ et al (2000a) Tuberous sclerosis complex and subependymal giant cell astrocytoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 227–230 232. Wiestler OD, Schiffer D, Coons SW et al (2000b) Myxopapillary ependymoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 78–79

233. Wiestler OD, Schiffer D (2000c) Subependymoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 80–81 234. Wiestler OD, Schiffer D, Coons SW et al (2000d) Ependymoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 72–76 235. Wilkinson ID, Jellineck DA, Levy D et al (2006) Dexamethasone and enhancing solitary cerebral mass lesions: alterations in perfusion and blood-tumor barrier kinetics shown by magnetic resonance imaging. Neurosurgery 58:640–646 236. Woodruff JM, Kourea HP, Louis DN et al (2000) Schwannoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, pp 164–166 237. World Health Organization (1990) Classification of brain tumors. World Health Organization, Zurich 238. Yang S, Law M, Zagzag D, Wu HH, Cha S, Golfinos JG, Knopp EA, Johnson G (2003) Dynamic contrast-enhanced perfusion MR imaging measurements of endothelial permeability: differentiation between atypical and typical meningiomas. Am J Neuroradiol 24:1554–1559 239. Yuh WT, Nguyen HD, Tali ET et al (1994) Delineation of gliomas with various doses of MR contrast material. AJNR Am J Neuroradiol 15:983–989 240. Yuh WTC, Tali ET, Nguyen HD, Simonson TM, Mayr NA, Fisher DJ (1995) The effect of contrast dose, imaging time and lesion size in the MR detection of intracerebral metastasis. Am J Neuroradiol 16:373–380 241. Zimmerman RD, Fleming CA, Saint-Louis La, Lee BC, Manning JJ, Deck MD (1985) Magnetic resonance imaging of meningiomas. AJNR Am J Neuroradiol 6:149–157

309

310

3  Brain, Head, and Neck

3.4  Cerebrovascular Disease D.C. Bergen, J.M. Fagnou, and R.J. Sevick 3.4.1  Introduction Non-invasive imaging techniques play a critical role in the investigation of patients with cerebrovascular disease and intracranial hemorrhage. At most institutions, nonenhanced CT continues to play a central role in the initial identification of hemorrhage. Development of new stroke therapies mandates rapid investigation and identification of ischemic change, vascular occlusion, and perfusion changes. Recent advances in MR technology have been paralleled to a large extent by CT. In the patient with acute ischemic stroke both modalities have advantages and disadvantages, and selection will depend to some extent on the availability of resources and personal preference. MR investigation of intracerebral hematomas has become routine. This section focuses on MR findings in acute ischemic stroke and intracerebral hemorrhage. 3.4.2  MR Technique As in all MR applications, evaluation of patients with cerebrovascular disease requires a balance of acquisition time, signal-to-noise ratio (SNR), and resolution for a given amount of signal. In the patient with acute ischemic stroke, time is of the essence, and technique will be tailored to minimize acquisition time while providing adequate SNR and resolution. Choice of sequences is limited to those providing essential information for diagno-

sis and treatment planning. Time is clearly not as great a consideration in the patient presenting with intracerebral hemorrhage. Added signal, and the time, resolution, and SNR benefits associated, can be achieved by imaging at a higher field strength (3 T) (Frayne et al. 2003). Optimally, all patients are imaged using a matrix of head and neck phased-array coils. This allows for cranial imaging as well as imaging of the neck vessels when required, with no need to switch coils or remove the patient from the magnet. In the acute ischemic stroke patient, neck imaging is not routinely performed; however, we have found that it can be valuable from a treatment planning perspective to have knowledge of anatomy and pathology (stenosis, dissection, and occlusion) of neck vasculature prior to any endovascular interventions. In the non-acute ischemic stroke patient, imaging of the neck vasculature is routinely performed. Imaging is confined to the head when evaluating most cases of intracranial hemorrhage. As mentioned above, evaluation of the hyperacute ischemic stroke patient must be accomplished rapidly. MR safety, screening, and consent issues are often the most significant impediment to rapid imaging. Key information for diagnosis and treatment planning is obtained from sequences including diffusion-weighted imaging (DWI), gradient-recalled echo images (GRE), dynamic contrast enhanced susceptibility weighted perfusion imaging (DSC-PWI) and pre/post-contrast 3D time-offlight MR angiography (3D TOF MRA). These techniques and their application in cases of stroke are discussed in detail later in the chapter. Table 3.4.1 details a typical MR imaging protocol (1.5 T) for routine evaluation of cerebrovascular disease in the head and neck. Sequences include standard mul-

Table 3.4.1  Cerebrovascular MR imaging technique Sequence

Plane

TR/TE/TI (ms)

Slice thickness/gap (mm)

FOV

Matrix

T1

Sagittal

400/16

5/1.5

220

192 × 256

T2

Axial

4,090/109

5/2.5

220

204 × 512

FLAIR

Axial

9,080/94/2400

5/2.5

220

184 × 256

DWI

Axial

2,400/70

5/2.5

230

90 × 128

GRE

Axial

650/23

5/2.5

220

168 × 256

3D TOF MRA

Axial

26/3.35

1 (4 slabs, 32 slices/slab), 16-mm overlap

220

256 × 256

3D TOF MRA: neck

Coronal

3.23/1.22

0.9/0.18 (88 slices/slab)

300

176 × 512

Contrast

10 ml at 2 ml/s (1 M)

3.4  Cerebrovascular Disease

tiplanar anatomical imaging; DWI, 3D TOF MRA of the intracranial circulation (posterior fossa and circle of Willis), and dynamic contrast-enhanced 3D TOF MRA of the neck. For DSC-PWI and post-contrast intracranial 3D TOF MRA, we routinely use a 0.5 M gadolinium chelate. Recently we have been performing all contrast enhanced 3D TOF MRA examinations of the neck with a 1 M gadolinium chelate. This contrast material is still being investigated, but it appears to provide superior image quality. In this situation, a shorter bolus of more concentrated contrast is used, so the timing of the acquisition becomes more critical. 3.4.3  Acute Ischemic Stroke Acute ischemic stroke continues to be a significant medical challenge around the world. In developed countries, the estimated stroke incidence ranges from 4 to 12 cases per 1,000 persons yearly, with an overall prevalence of 42–72 per 1000 people. These numbers are expected to rise, as evidenced by a trend of increasing incidence and a continually aging population. Acute strokes are associated with a mortality rate of roughly 23% (within 1 month of stroke onset) and are considered to be the second leading cause of death worldwide (Feigin et al. 2003). In terms of morbidity, the estimated lifetime cost of an acute ischemic stroke ranges from US $41,000 to 105,000, which underscores the large economic burden (Palmer et al. 2005). Ischemic stroke is defined as the prolonged deprivation of brain cells from oxygen and metabolic substrates, typically in the form of decreased blood flow. The clinical presentation of stroke is variable, but will primarily manifest itself in the form of acute neurological deficits. These acute deficits commonly include: paralysis, visual disturbances, speech problems, gait difficulties, and altered level of consciousness. Acute ischemic strokes are considered a medical emergency and recent studies support the notion that “time is brain.” If selected stroke patients can be accurately diagnosed and treated with intravenous thrombolytic within 3 h of onset, it has been demonstrated that functional measures and neurological outcomes are improved (Hacke et al. 1995; Hill et al. 2000). Thus, an accurate and timely radiologic investigation is paramount to patient care in this setting. Current standards of care generally utilize CT in the initial assessment of the acute stroke patient. The benefits of CT relative to MRI are predominantly related to its accessibility, lower cost, faster scan times, and its ability to detect intracranial hemorrhage. However, because tissue contrast on CT is dependent on X-ray beam attenuation (i.e., density), it is limited in its ability to detect the earliest

signs of cerebral ischemia. This is especially true during the first 6 h from onset, which includes the therapeutic window for thrombolytic treatment (Kalafut et al. 2000; Mullins et al. 2002). However, using modern CT angiography (CTA) techniques, critical information regarding vascular patency can be obtained rapidly. In many cases, CT and CTA provide sufficient information to guide selection of initial emergent therapy. At our institution, MR is used as a problem solving tool in patients who are fluctuating clinically, and to assess the extent of ischemia in patients with suspected vertebrobasilar disease. Other causes of acute neurological deficits that can mimic acute ischemic stroke include subdural and intracerebral hemorrhage. Less common mimics include cerebritis in the setting of infection or inflammation, or focal seizures. Hemiplegic or hemisensory migraines can imitate a stroke but have no radiological findings. Their exclusion would rely primarily on clinical history. MRI is ideally suited to the detection of acute ischemic stroke. The earliest changes of ischemia manifest as cellular dysfunction and cytotoxic edema: As the ischemic cells are deprived of oxygen and substrates, metabolism declines and ATP stores are depleted. This leads to the failure of ATP ion-exchange pumps in the cell membrane and an ionic imbalance occurs, which allows the influx of extracellular water into the cell. This phenomenon results in diffusion abnormalities which can be detected with MR within minutes of onset. Another ischemic pathological process relevant to MR is vasogenic edema. It is the subacute process (occurring ~6 h after stroke onset) whereby the blood–brain barrier begins to break down and become more permeable. The eventual result is a shift of water from within blood vessels to the extravascular space, effectively increasing the water content of the lesion (Loubinoux et al. 1997). With conventional anatomical imaging techniques, and their assessment of T1, T2, and proton density, MRI is able to capture a great deal of information about these water-related abnormalities, thereby enabling detection of some of the early events of ischemia. Coupled with the vascular and hemodynamic assessments provided by perfusion weighted imaging and MR angiography, it is perhaps not surprising that MRI has been reported to have a higher sensitivity and specificity than CT for the detection of acute ischemic stroke, especially within the first few h of acute stroke onset (Mullins et al. 2002). 3.4.3.1  MR Findings in Acute Ischemic Stroke 3.4.3.1.1  T1-Weighted Imaging On T1-weighted images, one of the earliest signs of a vascular occlusion is the absence of normal flow-related signal void. In a patent artery, intraluminal flow-related sig-

311

312

3  Brain, Head, and Neck

nal void is usually observed, reflecting fast-flowing blood (causing time of flight effects) or turbulence. This can also be visualized on T2 and FLAIR imaging, and it is best observed with T2-weighted imaging. Another early vascular finding on T1-weighted imaging is the presence of arterial enhancement within the suspected territory of ischemia following administration of contrast. This will appear as enhancing curvilinear structures and is representative of slowed circulation within cortical vessels or leptomeningeal collaterals (Essig et al. 1996; Sato et al. 1991). Either the absence of flow-related signal void or the presence of intraluminal enhancement can be indicative of a proximal occlusion slowing down or arresting flow distally. It should therefore be noted that these intraluminal findings may not necessarily represent the actual site of the occlusion, but could represent its downstream effects. In early acute stroke, there are general structural findings that are analogous to some of the CT findings. The morphological changes are shared and are caused by the mass effect of edema on the surrounding tissue. This mass effect can present as a spectrum of findings from the more subtle findings of sulcal effacement and gyral thickening to more overt findings of midline shift, compression of the ventricles and brain herniation. These findings are best assessed with T1-weighted imaging, can occur within the first few hours of ischemia, and will usually diminish after 72 h (Baker et al. 1991: Yang et al. 2002). A relatively late finding on T1-weighted imaging is

that of parenchymal signal hypointensity, which will typically appear around 16 h after the initial ischemic insult (Provenzale 2003: Yu et al. 1991). This is due to the cumulative effects of vasogenic and cytotoxic edema present within the infarcted tissue. 3.4.3.1.2  T2-Weighted Imaging On T2-weighted imaging, ischemic parenchyma will begin to appear hyperintense 6–12 h after stroke onset. This high signal intensity is a reflection of the increased water content of the lesion caused by cytotoxic and vasogenic edema. The increased water content leads to prolonged T2 relaxation, and thus higher signal intensity on T2-weighted images. This can present as a loss of gray– white differentiation within the zone of ischemia with increased signal in the white matter. Time course of evolution of signal changes on T2-weighted images (as well as DWI/ADC) are illustrated in Fig. 3.4.1. 3.4.3.1.3  Fluid-Attenuated Inversion Recovery Imaging The same mechanisms of cytotoxic and vasogenic edema also yield early signal changes on FLAIR images in acute stroke. Increased parenchymal signal can be detected

Fig. 3.4.1  Evolution of an acute stroke. Evolutionary signal changes in a large right MCA territory infarct are demonstrated on T2, DWI, and ADC images at <6 h, 24 h, 7 days, and 1 month post-onset of symptoms. At <6 h, T2 images are essentially normal, with patchy increased signal on DWI, and restricted diffusion evident on ADC map. At 24 h, infarction has evolved and appears mildly hyperintense on T2, with marked signal hyperintensity on DWI and corresponding restricted diffusion evident on the ADC map. At 7 days, the extent of T2 signal hyperintensity is greater. DWI continues to show hyperintensity while the ADC map indicates normalization of diffusion in some areas of the infarct. Signal hyperintensity on DWI is therefore due to prolonged T2 relaxation (T2 shine-through) and restricted diffusion. At 1 month, T2 signal hyperintensity is marked. DWI shows a heterogeneous appearance due to effects of increased diffusion (evident on the ADC map) and prolonged T2 relaxation

3.4  Cerebrovascular Disease

within 3 h of onset (Noguchi et al. 1997). This finding alone has a reported sensitivity of 46% and specificity of 92% (Cosnard et al. 1999). One of the advantages of FLAIR imaging over conventional T2 imaging is in its innate CSF signal suppression. This allows superior assessment of gray matter that is adjacent to CSF. Furthermore, a longer echo time (TE) can be used in the FLAIR sequence resulting in heavier T2 weighting. This confers a greater sensitivity to edema, allowing for improved distinction between normal and abnormal parenchyma. Another finding that has been reported on FLAIR imaging is the hyperintense vessel sign (HVS), which is the finding of increased signal intensity within cerebral blood vessels. It most likely represents a combination of slow flow or stasis in small arteries, veins, or collateral vessels supplying the ischemic territory (Schellinger et al. 2005). Again, similarly to the arterial enhancement pattern seen on T1-weighted images, this may represent the site of occlusion or the downstream effects of reduced flow. Although both of these findings can be found within 3 h of an acute cerebrovascular event (about 85% of cases) and correlate well with angiographic findings and perfusion abnormalities, they can also be found in a smaller percentage of asymptomatic patients with chronic cerebrovascular disease. HVS in a non-acute patient is likely indicative of a high risk for future infarction (Wolf 2001). The finding of an HVS has a sensitivity of 66% and a specificity of 92%. Moreover, when found in conjunction with hyperintense parenchymal signal, the sensitivity and specificity increases to 85% and 92% respectively (Fig. 3.4.2). In terms of prognosis, the presence of an HVS is not independently useful for determining patient outcomes, response to therapy, recanalization, or hemorrhagic transformation (Schellinger et al. 2005). An interesting observation however, is that in patients with a positive HVS, there is an association with larger final infarct volumes. 3.4.3.1.4  Gradient-Recalled Echo Imaging On T2* susceptibility-weighted GRE imaging, a sign known as the susceptibility vessel sign (SVS) (Fig. 3.4.3) is encountered and is similar to the aforementioned HVS. It is defined as an intraluminal signal intensity loss of diameter larger than the normal diameter of the contralateral vessel. This hypointense signal is predominantly found within the middle cerebral artery (MCA) and distal internal carotid artery (ICA) (Rovira et al. 2004), and is likely due the susceptibility effects of deoxyhemoglobin containing thrombus. This is attributable to the fact that deoxyhemoglobin is a paramagnetic substance and as such will cause local rapid dephasing of proton spins, ultimately resulting in a loss of signal on T2*-weighted imaging. The greater the content of deoxyhemoglobin, the greater the signal loss. This is relevant because de-

oxyhemoglobin content will increase as the thrombus ages. Thus the SVS is dependent on thrombus evolution time. This is represented in the scientific literature with reported sensitivity of 34% and specificity of 75% for the <3 h of the stroke-onset window (Schellinger et al. 2005), with a subsequent rise to 83 and 97% respectively, in the under 6 h window (Rovira et al. 2004). There is also further evidence that the sensitivity is dependent on the location of the thrombus, as evidenced by the SVS being much more sensitive for occlusions proximal to the MCA trunk (97%) than occlusions distal to the MCA bifurcation (38%) (Rovira et al. 2004). Prognostic significance of the SVS has yet to be determined. 3.4.3.1.5  Magnetic Resonance Angiography In MRA, there are two principle non-enhanced techniques employed, TOF, and phase contrast and TOF. 3D TOF MRA TOF techniques rely on the inflow of non-saturated blood into a plane already saturated by a train of short TR interval excitation pulses. The relatively short T1 of inflowing blood yields higher signal intensity than that of the surrounding stationary (saturated) background tissue. In the setting of acute cerebral ischemia, an intracranial 3D TOF angiographic assessment of the major vessels can be performed to evaluate for flow abnormalities such as occlusions, dissections, or high-grade stenosis. Such vascular flow changes will present on 3D TOF as a loss of the expected flow-related enhancement within the vessel (Sohn et al. 2003). This finding has an accuracy of 68% (Cosnard et al. 1999), and its presence is significantly linked to poorer neurological outcomes (Derex et al. 2004). The problems encountered with the 3D TOF technique are inherent in its reliance on the blood remaining unsaturated. If the flow of blood is not perpendicular to the plane (such as in the distal intracranial branch vessels), or its flow is slow, it may reside in the plane long enough to become saturated by the frequent excitation pulses. If the blood becomes saturated, there is a loss of distinction between its signal and that of the background tissues. Moreover, vessel narrowing, junctions, stenoses and occlusions can lead to complex, non-laminar blood flow (generically termed disturbed flow) (Underwood and Mohiaddin 1993). This complex, disturbed flow can lead to intra-voxel dephasing, where the spins within a voxel become randomly oriented such that the net signal is significantly reduced. In such cases, loss of flow-related enhancement is not an indication of slow-flowing blood or stationary tissue, but rather an artifact caused by the localized, complex blood flow patterns. Saturation effects and disturbed flow create the potential for a false positive

313

314

3  Brain, Head, and Neck

Fig. 3.4.2  Right vertebral artery thrombosis with associated medullary infarct. a On the coronal MIP image from a CE MRA of the neck, there is absent flow-related enhancement secondary to complete occlusion of the distal right vertebral artery. b T2 and c FLAIR signal hyperintensity in the right dorsolateral medulla indicates secondary infarction in this location. Note the absence of flow-related signal void in the right vertebral artery on the T2 image and corresponding hyperintense vessel sign in the same location on the FLAIR image

3.4  Cerebrovascular Disease

Fig. 3.4.3  Susceptibility vessel sign. a DWI demonstrates a focal infarction of the right posterior insular cortex. The GRE image (b) demonstrates the susceptibility vessel sign due to thrombus in a posterior Sylvian branch of the right MCA

interpretation, commonly a subtotal occlusion mistakenly being considered a total occlusion (Fig. 3.4.4). This can be partially mitigated by administration of IV contrast. Nonetheless, 3D TOF is recommended and commonly performed in the assessment of intracranial stenoocclusive disease (Özsarlak et al. 1998). Phase-Contrast MRA Phase-contrast (PC) techniques assess velocity of moving tissue (in this case blood) by comparing two acquisitions of a single axis, one acquisition with a velocity-encoding gradient, and one without. Stationary tissues (zero velocity) yield no signal, while moving blood will have a higher signal. The result is a sequence that can provide information about the speed and direction of flow, but takes longer than other techniques because of the requirement to obtain two acquisitions per axis. It also shares many of the same problems of TOF techniques, including intra-voxel dephasing, and signal loss for in plane flow. In comparison, it does have some advantages over TOF, as it can distinguish flow directions and detect slower flow in smaller vessels. These advantages tend to be limited to the smaller intracranial vessels, however, and contribute

less to the assessment of the major intracranial vessels as compared to TOF techniques. In general, TOF MRA has a greater sensitivity and specificity than PC MRA in the assessment of intracranial occlusive disease (Oelerich et al. 1998). Contrast-Enhanced MRA In contrast-enhanced MRA (CE MRA), a paramagnetic contrast agent such as gadolinium is administered intravenously (Fig. 3.4.5). As a consequence, the T1 relaxation time of contrast enhanced blood is significantly reduced relative to the background tissues. Using a short TR sequence, blood will emit high signal relative to saturated stationary background tissue. With CE MRA, there is a lesser reliance on inflow effects for signal. Consequently many of the artifacts encountered using TOF acquisitions are eliminated or minimized, and a higher signal to noise ratio is achieved. This is particularly advantageous in areas of slow flow, as CE MRA can distinguish with greater certainty whether there is a true occlusion or critical stenosis; the presence or absence of which may dictate further management. In particular, the observed signal loss on the pre-contrast TOF images can be com-

315

316

3  Brain, Head, and Neck

Fig. 3.4.4  MRA M1 branch occlusion. a Collapsed and b coronal re-projected MIP reconstructions from a TOF MRA show apparent complete occlusion of the left M1 segment. In contrast,

c axial source and d collapsed MIP CE MRA images of the same patient demonstrate a non-occlusive thrombus with diminished distal flow

3.4  Cerebrovascular Disease

Fig. 3.4.5  Normal CE MRA neck. Coronal MIP reconstruction of a normal CE MRA of the neck. 1 M Gd chelate, 10 ml total injected at a rate of 2 ml/s. Excellent depiction of aortic arch, great vessel origins, and extracranial vasculature with minimal venous contamination

pared to the CE MRA findings: a commensurate absence of signal on the post-contrast CE MRA image confirms an occluded vessel, whereas normal CE MRA signal suggests that flow in the vessel is slow or disturbed and that the pre contrast TOF signal loss is artifactual. Another advantage of CE MRA is in its superior ability to visualize smaller, distal and out of plane vessels, allowing more confidence in interpreting signal abnormalities found in these locations. A third advantage of CE MRA is in its significantly shorter acquisition time relative to the non contrast enhanced techniques. This reduces the risk of motion by the patient, and expedites the diagnostic process. Prognostically, CE MRA may also have a role in assessing the degree of collateral circulation, which is an independent predictor of favorable outcome (Kucinski et al. 2003). The major disadvantage of CE MRA lies in the requirement for accurate timing of the acquisition relative to the contrast bolus as it travels through the circulatory system. Ideally, the acquisition should be captured during the arterial phase and before the venous phase. If the acquisition is too early, arterial anatomy will be missed. If it is too late, the superim-

position of venous structures may obscure arterial anatomy. This is particularly true in the cavernous sinus, as it will render assessment of the adjacent ICA and M1 segments difficult. Parenchymal enhancement is also possible if there is break down of the blood brain barrier, typically seen 5–6 days after stroke onset. There are several methods for administering contrast in CE MRA. The first is simple post-contrast CE MRA, whereby the contrast is administered without attempting to capture the arterial phase on MRA resulting in non-selective enhancement of both arterial and venous structures. In the setting of acute stroke, the presence of contrast following perfusion-weighted imaging studies is often a convenient source of this. In spite of the venous enhancement, with proper placement of the 3D acquisition slab (oblique–axial and just superior to the cavernous sinus), the post-contrast findings are considered useful in the assessment of intracranial circulation, especially in comparison with pre contrast TOF (Yang et al. 2002). As such, in the setting of acute ischemic stroke, when used in conjunction with 3D TOF, post-contrast CE MRA provides a more accurate assessment of the patency of intracranial vasculature, and allows better interpretation of the mechanism of disrupted flow whether from dissection, stenosis, or occlusion (Yang et al. 2002). Dynamic CE MRA is the administration of contrast followed by an attempt to capture the arterial phase of enhancement. This synchronization is accomplished by either (1) directly measuring the delay between contrast injection and arterial arrival time using a bolus timing scan, followed by the CE MRA acquisition incorporating this delay time, or (2) utilizing fluoroscopic imaging, which automatically senses or displays to the operator the arrival of contrast (which obviates the need for a timing bolus). With regard to imaging the intracranial circulation, the arterial enhancement window of the intracranial arteries is considered too brief to image appropriately without sacrificing adequate spatial resolution. Although this limitation will likely be overcome in the near future, dynamic CE MRA in the assessment of acute intracranial occlusive disease is not yet recommended (Summers et al. 2001). Dynamic CE MRA is useful in imaging the extracranial circulation during the arterial passage of the contrast agent. Recent advances in this area have yielded elliptic centric CE MRA acquisitions, which collect the center of k-space first, then radiate outward, and can significantly improve the resolution and minimize venous contamination. Still, one must ensure that the acquisition of central k-space data coincide with the passage of the MR contrast agent through the arterial vessels of interest. The current gold standard for extracranial assessment is digital subtraction angiography (DSA). However, because of its inherent procedural risks and resource costs, non-invasive alternatives are being sought. TOF and PC techniques are used in the assessment of extracranial vas-

317

318

3  Brain, Head, and Neck

cular disease, but dynamic CE MRA using elliptic centric acquisitions is emerging as the preferred technique. This is because it is much less susceptible to the artifacts of PC and TOF, and is thus better suited to assessing areas of slow flow and turbulence such as artery origins, stenosis, dissections, and ulcerations (Cloft et al. 1996; Phan et al. 2001). There is ongoing debate whether CE MRA is acceptable alone, or should be used in conjunction with duplex ultrasonography (DUS) or CT angiography (CTA), and whether it can replace DSA. When compared with DSA results, CE MRA alone has a reported sensitivity of 93% and specificity of 81% in the detection of severe carotid stenosis, and 97 and 99% respectively, in the detection of complete occlusions (U-King-Im et al. 2004). In our practice we rarely perform DSA for assessment of extracranial vascular disease, and two concordant noninvasive tests (MRA, CTA, DUS) are required prior to surgical or endovascular intervention. 3.4.3.1.6  Diffusion-Weighted Imaging Diffusion is the spontaneous and random movement of molecules in solution due to their thermal motion. In a normal physiologic setting, the molecular movement is restricted to an extent by the presence of cellular structures such as cell membranes and organelles. Other factors that influence diffusion include temperature, molecular mass, and pH. A quantitative measure of diffusion is the apparent diffusion coefficient (ADC), which is essentially a mathematical coefficient representative of the molecular mobility. The diffusion coefficient is specific to an area of interest, is an average of both the intracellular and extracellular space, and areas of higher diffusion will have a larger diffusion coefficient, while in areas of limited diffusion it will be smaller. In normal brain tissue, the average measured diffusion coefficient is about 0.79 × 10–3 mm /s (Pereira et al. 2002). In ischemic stroke, the cascading effect from normal to ischemic to necrotic tissue alters diffusion. In the acute setting, cytotoxic edema occurs resulting in a shift of extracellular water to the intracellular space. The intracellular space contains many more barriers than the extracellular space, and thus there is a decrease in diffusion. Furthermore, the cells begin to swell, effectively decreasing the extracellular area, and restricting diffusion in that compartment. Acute ischemic tissues will have a decreased ADC compared to normal tissue (Fig. 3.4.6). The average ADC of acutely ischemic areas is about 0.46 × 10–3 mm /s (Pereira et al. 2002). As the infarcted tissue evolves into necrotic tissue, cell lysis occurs, and much of the barriers restricting diffusion are removed. Therefore, in chronic ischemic areas there is actually an increase in diffusion relative to normal tissue, and therefore the ADC is higher.

In MR imaging, diffusion sensitization is typically accomplished using a spin-echo approach (Stejskal and Tanner 1965) whereby two large diffusion sensitizing gradient pulses are located on either side of the 180° refocusing radiofrequency pulse. Each gradient lobe induces a phase change of the water signal. A static molecule experiences no net phase change (i.e., the phase accrual of the second gradient cancels that of the first), whereas the random motion of water at the sub cellular level produces a non-zero phase, the net result of which is exponential signal loss. Analogous to how the TE determines the amount of T2 weighting in the image, the b value (measured in seconds per millimeter squared) controls the sensitivity of diffusion weighting. The degree of diffusion sensitization is directly proportional to the b value. In acute stroke, b values are typically within 1,000 to 1,500 s/mm2 to ensure adequate signal to noise and keep TE reasonably short (≤100 ms) (Fig. 3.4.7). The signal intensity on diffusion weighted images is determined by both T2 relaxation and diffusion (see Fig. 3.4.1). In acute ischemic stroke, the diffusion coefficient decreases and on DWI signal intensity increases. In other words, the relatively greater diffusion in normal brain causes more signal loss than on DWI, rendering zones of acute ischemia relatively hyperintense. By comparison, subacute strokes (i.e., those beyond 1–2 weeks) have an increased diffusion coefficient, and one would expect signal intensity to decrease. However, the ensuing vasogenic edema increases T2 significantly, and a hyperintense lesion is seen on DWI. This is due to increased T2, an effect called “T2 shine-through.” Thus, it is necessary to interpret DWI images in combination with T2-weighted images, and maps of the diffusion coefficient itself. These are the ADC maps, and estimate the average diffusion of intra- and extra-cellular water in each voxel. Generally, spin-echo diffusion-weighted EPI is used due to its speed, which reduces overall imaging time and motion artifacts. It is, however, very sensitive to off-resonance effects (e.g., shim offsets, B0 inhomogeneity, and magnetic susceptibility) and timing errors, each of which may create phase discrepancies and produce distortion and/or ghosting effects. In acute stroke, DWI has become one of the first line diagnostic imaging methods of choice. When compared to CT and conventional MR imaging, it has been consistently shown to be superior for demonstrating the presence, extent, and location of early cerebral ischemia. It is able to display some of the earliest findings, as increased signal intensity and decreased ADC have been reported within minutes of ischemia in animal models, and within 30 min in humans (Burdette et al. 1998). It is a robust technique with reported sensitivities, specificities, and accuracies within the 90–100% range (Mullins et al. 2002). False negatives have been reported (rarely) on

3.4  Cerebrovascular Disease

Fig. 3.4.6  Acute lacunar infarction. a DWI sequence and b ADC map demonstrate restricted diffusion in an acute lacunar infarction in the left corona radiata. Mild corresponding

signal hyperintensity is noted on the c FLAIR and d T2 images. The acute lesion is easily distinguished from adjacent chronic ischemic change on DWI

319

320

3  Brain, Head, and Neck

Fig. 3.4.7  Acute midbrain stroke. Patient with acute onset right third nerve palsy. a Initial T2-weighted image shows no midbrain abnormality. However, on DWI there is increased signal in the right midbrain consistent with acute infarction. The ab-

normality is subtle, with a b value of 1,000 s/mm2 (b), but more obvious with a b value of 1,500 s/mm2 (c). However, at the higher b value, SNR is decreased. Follow-up d T2 image shows right midbrain infarct

3.4  Cerebrovascular Disease

DWI for small brainstem and lacunar infarcts (Gonzalez et al. 1999), and occasional false positives have also been reported due to T2 shine-through, cerebral abscesses, and tumors (Baird and Warch 1998). However, when appropriately viewed in conjunction with conventional MR, these false findings should be minimized (Schaefer et al. 2000). On a pathological level, the area of restricted diffusion identified on DWI has traditionally been thought to represent the “infarcted core.” Centered within a larger area of ischemia not visible on DWI, the infarcted core is considered to be an area of irreversible cell damage. This is supported by the observation that many of the latent findings on CT and conventional MR ultimately evolve and correspond to the same area of initial DWI involvement, even if an occluded artery was recanalized within the 3-h interval. This theory of irreversibility has been challenged with observations that initial DWI lesions are sometimes larger than final infarct size, and some reversibility is possible with early reperfusion. Thus, while the area represented by DWI lesions is primarily irreversible, this may not be absolute. 3.4.3.1.7  Perfusion-Weighted Imaging Perfusion is the passage of fluid through the capillaries in a tissue to supply nutrients and/or oxygen, and to remove metabolic by products. In the brain, disruption of this process leads to ischemic stroke, and as such its effects can be detected within minutes of stroke onset. For MR, there are three general classes of perfusion-weighted imaging: (1) susceptibility-based techniques that require an intravenous bolus injection of an exogenous paramagnetic contrast agent (e.g., chelated compounds such as gadolinium diethylenetriamine penta-acetic acid, or Gd-DTPA), (2) blood oxygen level–dependent (BOLD) acquisitions that detect the changes of endogenous deoxyhemoglobin (a paramagnetic substance), and (3) arterial spin labeling methods that selectively and magnetically “tag” the flowing blood and image the resulting changes in the tissue of interest. The most common PWI approach for acute stroke is the first method, which is commonly called dynamic-susceptibility contrast perfusion-weighted imaging (DSC-PWI). The contrast agent shortens the longitudinal (T1) and transverse (T2, T2*) relaxation times. When the blood–brain barrier is intact, gadolinium compounds remain intravascular and the T1 shortening effect produces signal enhancement primarily within the blood volume (as previously described in the MRA section). The paramagnetic nature of the intravascular contrast agent also causes magnetic susceptibility differences between the lumen of the vessel and the surrounding tissue; this induces a magnetic field gradient that extends beyond the blood vessel, which in turn in-

creases spin dephasing, decreases the T2* relaxation time, and causes localized signal loss. This is the basis of the DSC perfusion measurements. PWI is able to assess the hemodynamics of tissue by comparing the temporal and volumetric aspects of perfusion. A series of sequential images for each slice are analyzed, noting differences in the arrival time and/or degree of T2*-related signal loss. These values are then compared with the contralateral tissue to obtain a relative comparison of the signals. Assessment of the data in this manner gives relative information about blood flow, and is not quantitative. This information is then processed and the data is interpreted in the form of mean transit time, time to peak, and cerebral blood volume maps. Relative time-to-peak (rTTP) is the amount of time elapsed for the lowest signal intensity to be obtained. It essentially represents the time taken for the contrast to have attained its maximum (peak) concentration in that area. It is one of the easier parameters to interpret because of its high contrast-to-noise ratio. Its timing is useful in the assessment of delayed flow, and its amplitude can be useful in the assessment of blood volume. Ischemic regions will often show either a delayed and/or diminished response relative to healthy tissue; the delayed response is an indication that collateral vessels may be supplying blood to the tissue, whereas a diminished response suggests that less contrast agent is present and that the cerebral blood volume (CBV) in this area of the brain is reduced. The second metric is the relative mean transit time (rMTT), which gives an average for the arrival time of blood in the tissue. It is meaningful when comparing the degree of transit delay (increased time) between normal and infarcted tissue. A third important measure is the relative cerebral blood volume (rCBV); this is an estimate of the volume of blood in a particular voxel. Larger blood volumes contain more MR contrast agent and thus exhibit greater T2* signal loss. Again, only relative statements can be made about CBV. A decreased volume is consistent with hypoperfusion. An increased volume seen in the early stages of ischemia may represent compensatory vasodilatation or collateralization in response to a proximal occlusion, and if seen in the later stages, may represent possible recanalization. In terms of detection, all three parameters are useful, and should be interpreted in conjunction with each other. Quantitative measures of cerebral blood flow are under investigation, but are not currently in routine clinical use. PWI may demonstrate an area of hypoperfusion, which in the absence of reperfusion, is at risk of progressing to infarction. Previous studies demonstrate that in acute ischemic stroke, untreated PWI lesions usually progress to infarct, with follow up findings ultimately matching the volume of the PWI abnormality on initial presentation. Conversely, in patients with prompt recanalization following an identified PWI lesion, the fi-

321

322

3  Brain, Head, and Neck

nal infarct size generally regresses to a smaller volume than the initial PWI lesion volume. This has led many to believe that the area of PWI abnormalities in excess of the infarcted core represents areas of potentially salvageable tissue, and it has been labeled the ischemic penumbra. Similar to the DWI infarcted core theory, however, it is likely more complex than this. Recent studies have demonstrated that not all of a PWI lesion will necessarily progress to infarcted tissue if left untreated. The revision to this theory then is that the PWI lesion does include area at risk, but it also includes a peripheral area of benign oligemia, which while underperfused at the time of imaging, is not necessarily at risk of progressing to infarction. This carries important treatment decision implications. 3.4.3.1.8  Diffusion-Weighted Imaging and Perfusion-Weighted Imaging In hyperacute stroke, a comparison is often made between the DWI and PWI lesions. As previously mentioned, the DWI lesion represents the infarcted core and the PWI lesion represents the infarcted core plus ischemic penumbra. These lesions invariably overlap usually with the DWI lesion at the center of the PWI lesion. Since the infarcted core represents the eventual effect of a perfusion deficit, it is intuitive to think that during the early stages, the DWI lesion area would be smaller than the PWI lesion (Fig. 3.4.8). This is referred to as a DWI–PWI mismatch, and is seen in approximately 75% of cases (Parsons et al. 2001). The importance of this finding is that it can provide the clinician with an approximation of how much salvageable tissue is still present, and whether or not the patient would likely benefit from reperfusion therapy. A small mismatch for instance, would indicate that the patient has less salvageable tissue, and the risks of reperfusion therapy might outweigh whatever small benefit could be obtained. Although a mismatch is defined operationally when the difference between the two volumes is larger than 20% (Baird and Warach 1998), the interpretation of such findings requires caution. The reliability is only as accurate as the subjective eye of the observer, and the DWI and PWI images on which the mismatch is based. Not only are there areas of debate on what exactly PWI and DWI represent, but the degree of mismatch depends on what parameters are used in the assessment, how much time has elapsed, and what tissues are involved. The findings of diffusion abnormalities and perfusion abnormalities progress in a non-linear relationship (Hamon et al. 2005) and thus the degree of mismatch is dependent on the time at which the study was obtained. The irreversible damage manifested by the DWI lesion, will occur below a critical level of perfusion, but this critical

level is specific to the tissue and the time of exposure. Because the perfusion parameters are relative, there is currently no reliable method of measuring an absolute threshold identifying the true ischemic penumbra. Furthermore, in PWI imaging, rMTT and rTTP maps tend to overestimate final infarct size, while rCBV has been shown to both overestimate and underestimate the final infarct size (Parsons et al. 2001). As such, the degree of mismatch and area of the ischemic penumbra depends significantly on which parameter is used. At present there are no exact standards or thresholds to reliably distinguish salvageable tissue from non-salvageable tissue. Until a valid method is established and the areas are better defined, the presence of a DWI–PWI mismatch should be interpreted as an indication that salvageable tissue may be present. In addition to DWI–PWI mismatch, there are two other possible patterns. If the areas of diffusion and perfusion abnormalities are equal, this is considered a match (Fig. 3.4.9), and if the diffusion lesion is greater than the perfusion lesion it is considered a reverse match. The main determinant of presentation of these patterns appears to be time. In the early phase of acute stroke, mismatch is the predominant pattern for reasons already explained. As the ischemic area evolves into either reperfusion or infarction, the borders of the PWI and DWI lesion converge, leading to a match. As the stroke progresses further still, the perfusion abnormality can regress to an area smaller than the diffusion abnormality, and a reverse match may be seen. Other factors that may alter the DWI–PWI presentation include the degree of collateralization and varying stroke etiologies. Recent studies have demonstrated that mismatches are often associated with suspected large artery atherosclerotic strokes, and reverse mismatches are being associated with cryptogenic strokes (Restrepo et al. 2005). 3.4.3.2  Stroke Etiology The radiological identification of stroke mechanism is useful in the acute setting, particularly with regard to prognosis and therapeutic management. For example, an acute stroke caused by an occlusive thrombus would likely be treated with thrombolytics, while a stroke attributed to an intracerebral hemorrhage would be an absolute contraindication. There are many proposed mechanisms of acute stroke and their identification relies on the clinical presentation, patient history, and radiological findings. From a radiological perspective, pertinent areas to assess include the time elapsed between onset and acquisition, the size, shape, location and number of infarcts, and the status of associated vasculature.

3.4  Cerebrovascular Disease

Fig. 3.4.8  ACA territory hypoperfusion with subsequent infarct. a Initial DWI sequence demonstrates no lesion but there is prolonged rMTT (b) in the left anterior cerebral artery territory

indicating a DWI–PWI mismatch. Follow-up DWI (c) shows a small infarct in the region of previous perfusion abnormality

3.4.3.2.1  Vascular Territories

3.4.3.2.2  Embolism

The current convention of identifying cerebral vascular supply begins proximally and proceeds in the cephalad direction. There are two main circulations, the anterior circulation, derived from the internal carotid arteries, supplying the majority of both cerebral hemispheres, and the posterior circulation, derived from vertebral arteries, supplying posterior fossa structures, midbrain, thalamus, and posterior temporal/occipital lobes. Approximately 80–90% of acute ischemic strokes will involve the anterior circulation (Yang et al. 2002). Areas located between the vascular territories are referred to as the border zone or watershed areas, and are considered the most distal regions of cerebrovascular blood supply. Although vascular territories provide a common framework to use, they are generalized across the population. As such, they do not necessarily reflect the true vascular anatomy, because they do not account for the large anatomic variance between patients or the asymmetry found within each patient (van der Zwan et al. 1993). In the etiological determination of acute stroke, identifying infarcts within these territories can be useful (when interpreted in conjunction with other findings), but none are certain for any one etiology. For instance, infarcts located in the border zone/watershed areas were previously considered highly suggestive of a hemodynamic compromise, but recent studies have proposed that they are not as specific as once thought. Not only is the identification of the border zone often inaccurate, but similar findings can also be generated by embolic mechanisms (Hennerici et al. 1998). MR assessment of the acute stroke can identify stroke mechanisms including embolism, hemodynamic compromise, dissection, and dural sinus occlusion.

Embolic pathogenesis of stroke is by far the most common etiology of acute stroke. General MR findings of embolic stroke include a typical wedge-shaped, peripherally located area of infarction (Provenzale 2003). The lesions can be multiple, involving multiple vascular territories, and when compared with other etiologies, are more likely to undergo hemorrhagic transformation following reperfusion therapy (Baker et al. 1991). Proximal origins of emboli account for 30% of all strokes, and typically include cardiac sources such as atrial and ventricular thrombi, myxomas, endocarditis, and occasionally paradoxical emboli. If cardioembolic sources are radiologically suspected, it is important to perform the appropriate cardiac investigations such as an echo cardiogram and ECG. More distal sources of emboli originate from other arteries, the most common being atherosclerotic plaque in the ICA (Fig. 3.4.10) (Sohn et al. 2003). When comparing the two embolic origins, it is important to consider the anatomy of where the emboli originate relative to the vasculature. The more proximal the embolus, the greater is the potential to break into multiple fragments with involvement of multiple arterial branch points. Thus, cardiogenic emboli are more likely to involve multiple vascular territories. This has been demonstrated in MR studies of stroke comparing emboli from distal artery sources versus cardiogenic sources. The appearance of simultaneous infarcts in both the anterior and posterior circulation is significantly more frequent in the latter (Park et al. 2000). Furthermore, cardiogenic emboli more commonly yield lesions in both the superficial and deep territories, whereas infarcts from distal artery emboli tend to be limited to superficial territories (Tei et al. 1999).

323

324

3  Brain, Head, and Neck

Fig. 3.4.9  Matched PWI–DWI abnormality. In the posterior right frontal lobe, there is restricted diffusion on the a DWI sequence and b ADC map consistent with infarction. c Decreased

rCBV and d prolonged rMTT are identified in the same location, a DWI–PWI match. e Collapsed MIP from a 3D TOF MRA shows occlusion of the right ICA and an M2 branch

3.4.3.2.3  Hemodynamic Compromise

tion (Liu et al. 2004), and are best visualized on CE MRA. On PWI, areas of hemodynamic ischemia tend to yield larger abnormalities than other stroke mechanisms (Restrepo et al. 2005), a finding which is expected given the mechanism of ischemia. Additionally, infarcts of this nature can present with multiple, deep border-zone lesions on the ipsilateral side, often giving the appearance of a string of beads within the centrum semiovale (Knapf et al. 1998).

Hemodynamic compromise implies a general decrease in blood flow to an area of the brain, usually the downstream effects of a more proximal event. Sudden events capable of eliciting an acute stroke would include acute states of shock such as cardiac failure, sepsis, anaphylaxis, or hypovolemia. MR findings of global ischemia usually demonstrate diffuse bilateral lesions, often with selective necrosis of cerebral neocortex, hippocampus, and the striatum (Fig. 3.4.11) (Baker et al. 1991). Although critical stenosis of the large arteries is a predominantly chronic process, a hemodynamic acute stroke in the setting of an isolated critical stenosis can occur. Often the patient will have a preceding history of TIAs. Critical stenoses are primarily extracranial; most commonly at the bifurcation of the common carotid, or at the vertebrobasilar junction in the posterior circula-

3.4.3.2.4  Dissection Acute strokes caused by an arterial dissection are relatively rare, accounting for approximately 2% of all stroke mechanisms (Provenzale 1995), although it has been shown in some studies to be the most common cause of stroke (20%) within otherwise healthy young adults (Provenzale

3.4  Cerebrovascular Disease

Fig. 3.4.10  ICA stenosis. a Sagittal MIP from CE MRA of the neck shows moderate left ICA stenosis. b DWI demonstrates resultant infarct in the left hemisphere

1997). Patients at risk are generally those with a history of vasculopathy, collagen–vascular diseases, or trauma. The majority of dissections will occur extra-cranially in the cervical portion of the internal carotid, emphasizing the importance of including MRA studies of the neck, preferably contrast enhanced. The second most common location is the C1–C2 portion of the vertebral artery, where approximately 66% of vertebral artery dissections occur. Intra-cranially, the most common site of carotid dissection is the supraclinoid segment (Provenzale 1997), and these tend to have worse outcomes than extracranial dissections. The exact sequence of events leading to the acute stroke is still in question, but it is likely a combination of local arterial occlusion and associated generation of emboli causing distal infarcts. Current studies demonstrate that at least 15% of all dissections will have associated MRI findings of distal embolization (Provenzale 1997). In practice, it is not uncommon to observe findings of

both hemodynamic compromise and embolism within the distal brain parenchyma. At the site of dissection, multiple MR findings may be seen (Fig. 3.4.12). On T1- and T2-weighted MRI, a thin periarterial crescent or circumferential line of hyperintense signal is often visualized. Associated with this, the overall diameter of the artery may appear increased. Within the lumen, loss of normal flow-related signal void is also possible, seen as narrowing or absence. On MRA, evidence of intramural hemorrhage can be visualized as the double lumen sign. This is represented by a narrowed central column of high intensity (flow in the true lumen) surrounded by or adjacent to a layer of abnormal increased signal intensity. Often the signal intensity of this intramural hemorrhage is higher than that of the surrounding tissue, but lower than the high signal intensity of flowing blood (Provenzale 1995).

325

326

3  Brain, Head, and Neck

Fig. 3.4.11  Diffuse cerebral anoxia. a DWI and b ADC images demonstrate diffuse bilateral ischemia of the cortical gray matter. Subtle edema and cortical hyperintensity are present on the c FLAIR and d T2 sequences

3.4  Cerebrovascular Disease

3.4.3.4.5  Dural Sinus Occlusion

3.4.4  Intracerebral Hemorrhage

Also known as dural sinus thrombosis (DST), this mechanism of acute stroke and intracerebral hemorrhage is relatively infrequent but increasingly recognized due to improved detection methods. It typically affects younger to middle aged adults, with a slight predisposition to females. Risk factors include hypercoagulable states, infection, or the compressive effects of a mass (Provenzale 1997). The mechanism of infarct stems from venous occlusion with subsequent venous congestion and propagation of thrombus. The most common location for DST is in the superior sagittal sinus, often with extension of thrombus into the transverse or sigmoid sinus (Peter et al. 1996). MRI in conjunction with MRV is currently the diagnostic investigation of choice for detecting DST (Fig. 3.4.13), combining the ability to assess vascular patency with high sensitivity to parenchymal changes such as edema, infarction and hemorrhage. Unlike arterial infarcts, venous infarcts do not conform to arterial vascular territories, the infarcts are usually subcortical in location, and they are often hemorrhagic. Further, infarcted areas of parenchyma often demonstrate slightly reduced or even increased ADC values after 48 h (Selim et al. 2002). Within the dural sinuses, conventional MR sequences demonstrate the loss of flow-related signal void, and replacement with abnormal signal intensity. This abnormal signal is typically thrombus, with its intensity dependent on the thrombus age. On T1-weighted imaging, the thrombus will initially appear isointense, becoming increasingly hyperintense after 4–5 days. T2-weighted images initially show hypointense thrombus, with progressive conversion to hyperintense signal after approximately 5 days (Provenzale 1997). After 2 weeks, both T1 and T2 images are hypointense presumably due to recanalization. The findings on GRE imaging are much more striking, with intravenous thrombus and any associated intraparenchymal hemorrhage appearing hypointense due to the susceptibility effects of deoxyhemoglobin (Fig. 3.4.14) (Selim et al. 2002). Of the MR venography sequences available, 2D TOF MRA is the study of choice in the detection of DST because of its ability to cover a large volume with superior spatial resolution and its relatively fast acquisition time (Wasay and Azeemuddin 2005). On MRV the occlusion is visualized as an intraluminal absence of flow-related high signal, which is often surrounded by a collar of increased signal. This is similar in appearance to the “empty delta sign” visualized on contrast-enhanced CT. If the sinus has recently recanalized, the high flow signal will be present but with “frayed” margins (Wasay and Azeemuddin 2005). An additional finding suggestive of DST includes the presence of dilated venous collaterals. These usually include the deep medullary and emissary veins. Potential false positives include slow flowing blood (TOF artifacts), anatomical sinus hypoplasia/aplasia, and prominent arachnoid granulations.

Intracerebral hemorrhage is defined as the presence of blood or its breakdown products outside of the normal cerebral vasculature, either spontaneously or secondary to trauma or an underlying lesion. It carries an approximate incidence of 20 cases per 100,000 people, with a distribution skewed toward increasing age (Hanal et al. 2002). Specific races are at increased risk, with the incidence in African-Americans and Japanese being double that of Caucasians. Generally, the disease burden is relatively high with a 25% mortality rate within the first month, leaving 33% of survivors severely disabled (MacWalter et al. 2001). The appearance of intracerebral hemorrhage on MR depends on time-related factors such as the oxidative state of hemoglobin, degree of local oxygenation, and integrity of cell membranes, as well as anatomical location of the bleed and the precipitating etiology. 3.4.4.1  Parenchymal Hematoma: Temporal Evolution of MR Findings Elapsed time from onset to scanning plays a major role in the MR appearance of blood products, particularly with respect to the progressive oxidation of hemoglobin. When hemoglobin is exposed to O2, it binds an O2 molecule to become oxyhemoglobin. Eventually it will release the O2 in the capillaries to become deoxyhemoglobin. In the normal physiological circulation, hemoglobin will alternate between these two forms. When exposure to O2 is limited however, deoxyhemoglobin will undergo reversible oxidative denaturation into methemoglobin, which contains a heme iron in oxidized ferric form. Given more time in an anoxic environment, methemoglobin will continue to oxidize and denature further into hemichromes, and will ultimately be broken down by macrophages into ferritin (Bradley 1993): Oxyhemoglobin ↔ Deoxyhemoglobin ↔ Methemoglobin ↔ Hemichromes Generally, as hemoglobin progresses through the oxidative process from oxyhemoglobin (diamagnetic), to increasingly paramagnetic substances there is an increase in magnetic susceptibility and subsequent ability to induce a magnetic field around the hemoglobin. While the red blood cell membranes remain intact, a magnetic gradient is generated between the intracellular and extracellular compartments. As water molecules diffuse across the cell membrane, they are subjected to this magnetic gradient which results in dephasing and T2 shortening (Bradley 1993). This interplay between the oxidative state of hemoglobin and the presence or absence of cell membranes is a key component in determining its appearance on MRI. It should be noted that the evolutionary stages of hemorrhage are generalized, and do not occur in exclusivity

327

328

3  Brain, Head, and Neck

Fig. 3.4.12  Vertebral artery dissection with PICA territory and thalamic infarction. a,b Axial FLAIR images show posterior inferior cerebellar artery (PICA) and thalamic infarction. c CE

MRA of the neck shows luminal narrowing indicating dissection of the left vertebral artery (arrow), confirmed by catheter angiogram (d)

3.4  Cerebrovascular Disease

Fig. 3.4.13  Acute sagittal and transverse sinus thrombosis. a Coronal T1-weighted image demonstrates isointense acute thrombus in the superior sagittal sinus (arrow) and right transverse sinus (arrowhead). On b axial T2 images, thrombus simu-

lates flow-related signal void. c Coronal MIP reconstruction from 2D TOF MRA shows flow in the straight sinus (arrow) as well as left transverse and sigmoid sinuses. d Sagittal MIP shows complete occlusion of the superior sagittal sinus

329

330

3  Brain, Head, and Neck

Fig. 3.4.14  Cortical vein thrombosis with venous infarction and hemorrhage. a FLAIR and b T2 images demonstrate heterogeneous signal intensity in a left frontoparietal venous infarct. On GRE images, c a tubular focus of hypointensity represents magnetic susceptibility effect of acute thrombus in a cortical vein. Parenchymal hemorrhage is more conspicuous on the GRE image (d)

of each other. Depending on various factors involved, it is not uncommon to observe variable rates of progression, or the findings of one stage coexisting with another (Fig. 3.4.15). By convention, the age of a hematoma is determined by the most mature form of hemoglobin present within the hemorrhage. 3.4.4.1.1  Hyperacute Phase The hyperacute phase is usually defined as being within the first 24 h. At this stage, the inner core of the hematoma exists predominantly as oxyhemoglobin (due to shielding from the outer core) and to a large extent the MR signal characteristics are related to the outer core and surrounding tissues. The local tissue surrounding the hematoma shows vasogenic edema, with hyperintense signal on T2-weighted imaging, and corresponding hypointense signal on T1-weighted images. Within the outer core, deoxyhemoglobin is present in sufficient quantity to induce magnetic gradients, thereby appearing

hypointense on T2-weighted, and especially T2*-weighted imaging. T2*-weighted sequences such as GRE are very sensitive to magnetic susceptibility effects, and the acquisition of images for this intent is known as susceptibility-weighted imaging (SWI). SWI is able to demonstrate some of the earliest findings of hemorrhage, with the earliest reported findings being within 23 min of onset (Linfante et al. 1999). The signal loss over areas of magnetic susceptibility is marked, rendering SWI capable of identifying very small areas of hemorrhage within the first couple of hours (Fiebach 2004a). Possible false positives in the form of susceptibility artifacts include poorly oxygenated venous blood and areas adjacent to air filled sinuses (Wycliffe et al. 2004). Although deoxyhemoglobin is considered a paramagnetic substance, it does not generate T1 shortening due to very short spin–spin relaxation times, and its molecular structure preventing the heme iron from interacting with water molecules. Paradoxically, the T1-weighted signal of the affected area often appears hypointense relative to normal brain due to increased water content.

3.4  Cerebrovascular Disease

3.4.4.1.2  Acute Phase

3.4.4.1.3  Early Subacute Phase

This stage of evolution generally refers to the period of time from 24 to 72 h after onset. As a hematoma evolves, the central collection of oxyhemoglobin progressively converts to deoxyhemoglobin from the outside-inward. The MR features of the outer core will progressively overtake those of the inner core. Further T2 shortening occurs at this stage due to thrombus formation and hemoconcentration. T1-weighted images become more diffusely hypointense relative to normal tissue, and SWI demonstrates significant signal loss due to deoxyhemoglobin.

Within 3 to 7 days, there is further oxidation of deoxyhemoglobin to methemoglobin. Because O2 is more readily available at the periphery and is required for oxidation, the conversion begins at the outer core and progresses inward over the following days. Methemoglobin is strongly paramagnetic, and causes marked T1 shortening, generating hyperintense signal on T1-weighted imaging. T1-weighted imaging is superior in distinguishing the acute stage from the subacute stage. The methemoglobin is still contained within the red blood cells at this stage, perpetuating a magnetic gradient between the extracel-

Fig. 3.4.15  Intracerebral hematoma, early subacute stage. a Sagittal T1 image shows a large hematoma with predominant signal hyperintensity with corresponding signal hypointensity on b T2 and c GRE images

331

332

3  Brain, Head, and Neck

lular and intracellular space. This contributes to significant T2 shortening, and continued signal hypointensity on T2-weighted images. Again, SWI will show significant signal loss in areas of deoxyhemoglobin and intracellular methemoglobin. 3.4.4.1.4  Late Subacute Phase The late subacute phase typically occurs between 1 and 2 weeks post-hemorrhage. It is distinguished from the early subacute phase by the breakdown of the red blood cell membrane. This allows equalization of methemoglobin levels between the compartments, effectively removing the local magnetic gradient. The loss of magnetic gradient removes the T2 shortening effects, and increased water content from cell lysis causes T2 prolongation. In contrast to the early subacute phase, this results in hyperintense signal on T2-weighted imaging. The T1 signal generated, which is independent of cellular compartmentalization, remains hyperintense. As such, T2-weighted imaging is best suited to distinguish between the early and late subacute phases of hemorrhage. 3.4.4.1.5  Chronic Phase After approximately 2 weeks have elapsed, hemoglobin is degraded into soluble ferritin and insoluble hemosiderin within macrophages. As hemosiderin is strongly paramagnetic and intracellular, it is seen on T2-weighted imaging as a hypointense ring surrounding the resolving hematoma. This ring often surrounds an inner ring of hyperintense T2-weighted signal, which represents extracellular methemoglobin that the macrophages have yet to phagocytize. As the hematoma evolves further, the hypointense ring will constrict and eventually collapse to form a hemosiderin lined slit (Bradley 1993). 3.4.4.1.6  CT versus MRI in the Detection of Intracerebral Hemorrhage CT is widely considered to be necessary in the exclusion of intracerebral hemorrhage, particularly in the assessment of acute stroke patients (NINDS rt-Pa Study Group 1995). Despite never being formally assessed in this capacity, CT is still considered the non-invasive gold standard in the detection of hyperacute intracerebral hemorrhage. In view of MR being implemented in the primary diagnosis of acute ischemic strokes, and continuing MR research, there has been significant focus on the sole utilization of MR for detection of intracerebral hemorrhage. This is beneficial in terms of efficiency and in expediting the diagnostic process. Many studies have focused on SWI, with its exquisite sensitivity in detecting paramagnetic

substances, and its consequent ability to detect the earliest signs of hemorrhage. Two studies have reported susceptibility changes in hyperacute ICH where the patients were examined from 2–6 h after the onset of symptoms (Patel et al. 1996; Schellinger et al. 1999). T2*-weighted EPI detected all hemorrhages in both studies. MR imaging was performed immediately after CT, and it was found that T2*-weighted EPI was as sensitive as CT in detecting hemorrhage. In another recent study, five patients with ICH (verified by CT) were examined within two h using an acute MR imaging protocol. They concluded that T2*weighted GRE-EPI was the most sensitive MRI modality in detecting hyperacute ICH and suggested that it may be of diagnostic value provided that larger prospective studies can be initiated (Lifante et al. 1999). The Hemorrhage Early MRI Evaluation (HEME) study (Kidwell et al. 2004) evaluated whether conventional GRE and GRE-EPI susceptibility-weighted MR imaging are as sensitive as CT for the detection of acute ICH. This trial was a non-randomized, collaborative, blinded, two-center paradigm of 200 adult patients presenting within 6 h of acute stroke symptoms, and showed that there was no statistical difference between the MR and CT techniques for the diagnosis of acute hemorrhage; in fact, in 49 patients MR imaging identified chronic hemorrhage that CT did not detect. Most recently, a prospective multicenter trial analyzing DWI, T2- and T2*-weighted images, determined that MR alone was able to identify hyperacute ICH with a sensitivity and overall accuracy of 100% (Fiebach et al. 2004a). In summary, these results are encouraging and suggest that robust and reliable MR hemorrhage detection techniques are forthcoming. It is likely that our collective understanding of the role and pitfalls associated with T2*-weighted fast gradient-recalled echo (and GRE-EPI) techniques will improve, and that in the near future these approaches can be relied on for hemorrhage detection. 3.4.4.2  Anatomical Location of Hemorrhage and Its Appearance on MR Temporal evolution of MR findings in parenchymal hemorrhage has been described. As a template, most intracranial hemorrhages will generally follow this progression and yield the expected MR findings. However there are some unique features and distinctive alterations to this process that are dependent on the location of hemorrhage. 3.4.4.2.1  Subarachnoid Hemorrhage Subarachnoid hemorrhage accounts for 5% of all strokes, most commonly arising from a ruptured aneurysm (about 60%) (Licata and Turazzi 2003; Mohamed et al. 2004). Its initial appearance and progression are relatively similar to that of intraparenchymal hemorrhage, but with some key

3.4  Cerebrovascular Disease

differences. CSF has a higher O2 tension relative to brain parenchyma. Due to the increased O2 availability from the surrounding environment, the natural evolution of hemorrhage within this space is often prolonged (Rumboldt et al. 2003). This is perceptible on MR as a relative delay in the expected T1 and T2-weighted changes, where both are often isointense to CSF during the hyperacute phase. As such, T1 and T2-weighted imaging are both considered unreliable in the detection of hyperacute subarachnoid hemorrhage (Wiesmann et al. 2002). For patients who present in the subacute stage, FLAIR images have been shown to be superior to CT in the detection of subarachnoid hemorrhage (Fig. 3.4.16) (Noguchi et al. 1997). T2*-weighted findings are also commonly delayed, but to a lesser degree because of the high sensitivity to trace amounts of deoxyhemoglobin. In the chronic-recurrent setting, repeated hemosiderin staining of the leptomeninges may be seen as a consistent T2* loss of signal. This is known as superficial siderosis. T2*-weighted sequences suffer from the susceptibility artifacts generated by structures adjacent to the subarachnoid space such as dense bone or air filled sinuses. These can make it very hard or impossible to assess certain areas of the subarachnoid space, particularly the posterior fossa and pontine cisterns (Fiebach et al. 2004b). Within the first few hours, FLAIR and proton density (PD) images are considered very sensitive and equal, if not superior to CT in the detection of subarachnoid hemorrhage (Fiebach et al. 2004b; Mohamed et al. 2004; Wiesmann et al. 2002). The increased protein content (from blood) within the CSF interferes with the suppression of fluid, and generates high signal intensity in contrast to hypointense CSF. Both FLAIR and PD hyperintensity are often visible within 3 h of onset, and play an essential role

in the acute MR assessment of a suspected subarachnoid hemorrhage (Wiesmann et al. 2002). Possible false positives include infectious meningitis, leptomeningeal carcinomatosis and CSF flow artifacts. In the assessment of a suspected hyperacute subarachnoid hemorrhage, most centers still consider CT first line because of its availability, and high sensitivity within the first 24 h. MRI can play a supplementary role in cases where CT findings are negative/equivocal, or if a specific etiology is sought in which the MR will provide a superior depiction. If all radiological findings are negative after 12 h and there is continued suspicion of a subarachnoid hemorrhage an LP should be performed and analyzed for xanthochromia (Liebengerg et al. 2005).

Fig. 3.4.16  Subacute subarachnoid hemorrhage. Patient presented with left hemiparesis several days after onset of acute headache. a Axial FLAIR image shows hyperintense subarachnoid blood in the right Sylvian Fissure. b Axial DWI shows sig-

nal hyperintensity from infarction in the right MCA territory. c Angiography demonstrates severe vasospasm and multilobular right MCA aneurysm

3.4.4.2.2  Subdural Hemorrhage Hemorrhage into the potential space between the arachnoid and dura is termed subdural hemorrhage. This form of hemorrhage is more common in the elderly, and is typically the result of a deceleration trauma generating a shearing force on cortical vessels, such as a bridging vein. Occurring less frequently are spontaneous subdural hemorrhages, typically seen in the anticoagulated patient. In the elderly, hemorrhage is frequently asymptomatic during the early stages, and will often present in the later stages of evolution. Rarer causes of subdural hematomas include dural arteriovenous malformations and ruptured arachnoid cysts (Bradley 1993). Distinguishing features include the distribution of the hematoma. It follows its anatomical space, yielding a characteristically medially concave or crescent shape. Because this space is not restricted by dural suture lines,

333

334

3  Brain, Head, and Neck

blood is able to spread throughout this space along the hemispheres and fissures. Its limits are often in accordance with the dural reflections such as the tentorium cerebelli and falx cerebri. In addition, it is common to identify bilateral subdural hematomas, especially in the elderly population (Fig. 3.4.17) (Bradley 1993). In terms of temporal progression, the dura matter is well vascularized and thus O2 is supplied in greater abundance relative to brain parenchyma. Consequently, the temporal progression of subdural hematomas may be slower than that expected in an intraparenchymal bleed. Subdural blood in the hyperacute setting will often show distinct signal relative to brain. T1-weighted images demonstrate hypointense signal, whereas T2-weighted images show hyperintense signal (Wasenko et al. 2002). In the chronic setting, there is very little hemosiderin deposition due the absence of macrophages within the subdural space. Methemoglobin continues its oxidative denaturation nonetheless, but in contrast to intra-axial bleeds, it yields non-paramagnetic hemichromes (Bradley 1993). As such, susceptibility effects due to hemosiderin are not usually present, and T1-weighted signal is more hypointense. Furthermore, it is not uncommon for chronic stages to demonstrate heterogeneous signal intensity. This finding is likely an indicator of recurrent bleeding in several stages of progression (Wasenko et al. 2002). The heterogeneous signal may also appear in layers, effectively representing blood products at different stages of evolution. 3.4.4.2.3  Epidural Hemorrhage Epidural hemorrhages occur in the space between the dura and the inner table of the calvarium. This form of hemorrhage frequently has a preceding history of head trauma, with an associated skull fracture and subsequent disruption of dural arteries. The most common site of an epidural hematoma is in the temporal-parietal region due to a laceration of the middle meningeal artery. Bleeding may also be venous in origin, usually when a dural venous sinus is lacerated. Accounting for about a third of cases, these are less common than arterial sources, and they tend to present less acutely or even incidentally. This is due to the lower pressure of venous hemorrhage, and its slower evolution and presentation (Domenicucci et al. 1995). Distinguishing features are based primarily on the location and appearance of the blood. The epidural spaces are confined by the cranial suture lines where the dural layer is firmly adhered to the skull. In conjunction with a high pressure arterial bleed; these confined spaces generate a “bulging” of the dura away from the skull. This results in a medially convex, or lentiform shape. Often it is possible to identify the dural layer on T1-weighted imaging as a hypointense band between the hematoma and brain. Another distinguishing finding is an adjacent skull

fracture, which may be visualized on MR if fluid is present within the fracture line. The epidural space itself is a very similar environment to the subdural space, and thus an epidural hematoma will share the same features of a subdural hematoma such as a delayed temporal progression and a paucity of hemosiderin deposition (Bradley 1993). 3.4.5  Intracerebral Hemorrhage Etiology The differential diagnosis of intracerebral hemorrhage is broad, and is grossly broken down into two categories: traumatic and non-traumatic. The traumatic injuries present very acutely and are relatively obvious based on history and physical findings. Their radiological presentation is highly variable depending on the mechanism of injury. Non-traumatic injuries can present acutely (especially in the presence of hemorrhage), and it is not uncommon to identify evidence of previous hemorrhage. This category includes primary spontaneous hemorrhages due to hypertension and amyloidosis, and secondary hemorrhages due to vascular malformations, tumors, aneurysms, stroke transformation, sinus thrombosis (covered in the stroke section), and other less common causes. 3.4.5.1  Vascular Malformations 3.4.5.1.1  Pial Arteriovenous Malformations Arteriovenous malformations (AVM) are a congenital vascular lesion. Arising during the embryonic stage due to capillary dysmorphogenesis, an AVM is typically comprised of one or more arteries feeding a nidus of abnormal vessels (in place of a normal capillary bed) and drained by one or more veins. This is essentially a high-flow shunt, and is the commonest of symptomatic vascular malformations. AVMs are predominantly supratentorial in location, and supplied by the intracranial circulation. AVMs carry a 2–3% yearly risk of hemorrhage, and are estimated to account for 1–2% of all strokes (Ko et al. 2003). Identified factors that increase the risk of hemorrhage include previous bleed, infratentorial location, solitary feeding artery, solitary draining vein, small feeding artery diameter (<1 mm), small nidus size (<2 cm) and deep venous drainage (Kubalek et al. 2003). The hemorrhage itself is usually intraparenchymal, but it may also include the subarachnoid or intraventricular spaces. On MR, the most important finding of an AVM is non-anatomical curvilinear areas of flow-related signal void within the parenchyma, best visualized on T2 images. These represent high flow feeding arteries, or the nidus itself (Fig. 3.4.18). Other T2 findings may include faint perilesional hyperintensity due to venous congestion or a fibrofatty matrix (Connor et al. 2005). Conversely, hypointense T2 signal may also be present due to

3.4  Cerebrovascular Disease

Fig. 3.4.17  Bilateral subdural hematomas of differing age. a Coronal T1 and b axial T2 images show bilateral hematomas which are both hyperintense on T1. The right-sided lesion, however, shows slight relative signal hypointensity on T2, indicating

the presence of intracellular methemoglobin. The more chronic left-sided hematoma shows multiple septations and predominantly hyperintense signal on T2 (breakdown of RBC membrane, extracellular methemoglobin)

calcification of intervening brain tissue. Very strong T2 signal loss is also seen if hemosiderin is present due to previous silent hemorrhages. Apart from depiction of vascular flow void, T1-weighted images are not particularly useful. TOF MRA images are able to identify abnormally dilated feeding and draining channels, and provide a general overview of AVM architecture. However, they are not as sensitive in the precise identification of a small nidus or feeding/draining vessels. CE MRA is superior in this aspect, as it is able to delineate the feeding arteries and nidus, and is satisfactory in depicting draining veins. DSA currently remains the gold standard in the detection of AVMs, and is required for surgical/endovascular treatment planning (Suzuki et al. 2003).

of hemorrhage as an AVM. DAVFs are considered to be an acquired malformation, and are thought to arise from previous trauma or veno-occlusive diseases (Connor et al. 2005). When an abnormal cluster of flow-related signal void is observed adjacent to a venous sinus, a DAVF should be suspected (Fig. 3.4.19). Other findings on conventional MR include perivenous white matter T2 hyperintensity, due to venous hypertension from the high-flow shunt. Dilated leptomeningeal or medullary vessels may also be visualized as a result of high venous pressure, although these are better assessed on MRA. MRA is considered more sensitive for DAVFs as it allows direct visualization of the fistula, and the absence of a nidus aids in its distinction from an AVM. The presence of venous flow–related enhancement within the affected sinus is fairly specific, representing arterialized flow (Kwon et al. 2005). Other important aspects of DAVFs can also be assessed with MRA, specifically the presence of retrograde leptomeningeal venous drainage, which indicates an increased risk of hemorrhage and poorer prognosis (Willinksy et al. 1999). Although MRA results are encouraging, DSA of the external carotids remains the gold standard in the diagnosis of DAVFs.

3.4.5.1.2  Dural Arteriovenous Fistulas Dural arteriovenous fistulas (DAVF) are also high-flow shunts, and are made up of an artery connecting directly to a dural venous sinus. In contrast to AVMs, the blood supply is extracranial in origin, typically from a dural/ meningeal artery with fistulization to the transverse, sigmoid, or cavernous sinus. DAVFs are predominantly infratentorial in location, but carry the same general risk

335

336

3  Brain, Head, and Neck

Fig. 3.4.18  Right frontal AVM with acute hemorrhage. a Sagittal T1 and b axial GRE images show acute intraventricular (arrowhead) and corpus callosum hemorrhage (arrows). c Axial T2 shows nidal vessels in the right medial frontal lobe and d coro-

nal T2 demonstrates intranidal aneurysm (arrow). e MIP reconstruction from 3D TOF MRA and f catheter angiogram show AVM arising from pericallosal artery branches

3.4  Cerebrovascular Disease

Fig. 3.4.19  Posterior fossa DAVF. a Axial T2 image shows edema in the vermis and left-cerebellar hemisphere as well as numerous dilated cortical and medullary vessels. b Post-contrast axial T1-weighted image demonstrates multiple enhancing cortical and deep medullary veins. c GRE image shows small bi-

lateral cerebellar hemorrhages. d Sagittal MIP from CE MRA of the neck and external carotid injection from catheter angiogram (e) demonstrate enlarged occipital artery (arrowhead) supplying dural fistula with venous reflux (arrows)

3.4.5.1.3  Developmental Venous Anomalies

T2-weighted images, tubular flow-related signal voids are seen within the large trunk and some of its venules (Rigamonti et al. 1988). Smaller radiating veins are best demonstrated on contrast enhanced T1-weighted images. Given the stable nature of the lesion, there are very rarely any perilesional findings. MRA yields similar findings, but is usually not necessary for diagnosis. The MR findings are considered very specific for DVAs, and thus DSA is rarely required unless another pathology must be excluded (Brown et al. 2005).

Also known as venous angiomas, developmental venous anomalies (DVA), are the most common of cerebral vascular malformations. Typically an incidental finding, they are very rarely hemorrhagic unless spontaneous thrombosis occurs or another malformation is present, most commonly a cavernous malformation (Lee et al. 1997). DVAs are congenital in origin; the focal result of aberrant venous development generating thin walled vascular channels amongst normal arteries, capillaries, and brain parenchyma. Because the abnormality is limited to the venous side, DVAs are low flow malformations and they are not shunts (Vilanova et al. 2004). DVAs are usually solitary, can occur in any area of the brain, and commonly drain into a dural sinus. DVAs have a very characteristic caput medusa appearance: a large, dilated venous trunk extending through the cortex, associated with smaller venules converging radially on the trunk like spokes (Fig. 3.4.20). On T1- and

3.4.5.1.4  Cavernous Malformations Also know as cavernous angiomas, angiomas, cavernomas, cavernous hemangiomas or cryptic vascular malformations, cavernous malformations (CM) are composed of a spongiform network of multiple thin walled vascular cavities located within the intercapillary spaces. These are not shunts, are slow flowing, and are informally thought

337

338

3  Brain, Head, and Neck

Fig. 3.4.20  Developmental venous anomaly. a,b Axial T2 and sagittal T1 images show abnormal tubular signal voids within the right posterior temporal and occipital lobes (arrows). c–f Post-

contrast T1 images demonstrate a large DVA with multiple small veins radiating toward larger branches which converge medially e–f see next page

3.4  Cerebrovascular Disease

Fig. 3.4.20  (continued) e–f Post-contrast T1 images demonstrate a large DVA with multiple small veins radiating toward larger branches which converge medially

of as a “blood sponge.” There are two types, the sporadic congenital form, and the familial form. Congenital CMs are typically solitary, while the familial form consists of multiple CMs, and de novo formations are possible (Brunereau et al. 2002). CMs are the third most prevalent vascular malformation, and carry a yearly risk of symptomatic hemorrhage around 1% (Dillon 1997). Despite a relatively low rate of symptomatic hemorrhage there are likely repeated minor occult bleeds or thrombosis occurring, few of which are symptomatic. CMs tend to have a distinctive appearance with a core and periphery (Fig. 3.4.21). The core is typically a well circumscribed lesion of heterogeneous intensity, consisting of a septal-reticulated pattern often described as a “bunch of grapes.” Sinusoids and caverns of varying diameters are characteristic features of CMs. The heterogeneity of the T1- and T2-weighted signal is due to the frequent occurrence of minor hemorrhages or thrombosis, many of which are in different stages of resolution. Although flow-related signal voids are not typical of CMs, round areas of signal void may be visualized due to the presence of phleboliths or other calcifications (Connor et al. 2005). A peripheral rim of signal hypointensity is characteristic on T2- and T2*-weighted images due to the susceptibility effect from hemosiderin deposition. Local effects on

surrounding tissues such as mild edema are also possible, particularly in association with acute hemorrhage. MR is the superior imaging modality for CMs because it is more sensitive than is CT, and findings on DSA are characteristically inconspicuous (Brunereau 2002). 3.4.5.1.5  Capillary Telangiectasia A Capillary Telangiectasia refers to a congenital collection of ectatic capillaries. There is some speculation that they may be spectrally related to DVAs. They are often seen in association with DVAs, and share similar traits such as being low flow, non-shunting malformations that very rarely hemorrhage unless in association with another malformation (commonly a CM). They are typically asymptomatic and carry a general prevalence of roughly 0.1% (Brown et al. 2005). Capillary telangiectasias can be located anywhere within the central nervous system, but demonstrate a predilection for the pons (Vilanova et al. 2004). They are frequently undetectable on standard T1- and T2-weighted imaging, although mild hyperintensity on T2 may be identified. The scarcity of findings is partially attributable to the stability of these lesions; there is a lack of edema,

339

340

3  Brain, Head, and Neck

Fig. 3.4.21  Cavernoma. a Axial T2 image shows a heterogeneously hyperintense lesion with hypointense rim in the mesial temporal lobe, the hypointense rim “blooms” on the GRE image (b). The lesion is iso-intense on pre-contrast T1 images (c) and shows mild central enhancement (d)

3.4  Cerebrovascular Disease

gliosis, calcification, or hemosiderin deposition. However, there is frequently mild T2* hypointensity, purportedly due to static blood within the ectatic vessels promoting deoxyhemoglobin formation. Furthermore, there is often subtle gadolinium enhancement of these lesions which is accompanied by a brush like border. The combination of isolated T2* hypointensity and gadolinium enhancement is key in the diagnosis of capillary telangiectasia (Lee et al. 1997). CT is not useful in their detection, and DSA rarely demonstrates any specific findings. 3.4.5.2  Tumors Neoplasms account for approximately 5% of all intracerebral hemorrhages (Licata and Turazzi 2003). Generally, tumoral hemorrhage is secondary to malignant neoplasms although benign tumors can also bleed and should not be excluded from the differential. About half of all hemorrhagic neoplasms are high grade gliomas such as glioblastomas (33%) and anaplastic gliomas (12%), while highly vascular metastases account for about a fifth of all cases (Licata and Turazzi 2003). The most common primary tumors associated with hemorrhagic brain metastases are melanoma and carcinomas of the breast, kidney, chorion, thyroid, and lungs (Bradley 1993). Hemorrhagic neoplasms can be difficult to distinguish from other causes of intracerebral hemorrhage. Although features of specific neoplasms are beyond the scope of this chapter and there are no pathognomonic findings, there are some general MR characteristics of neoplasms that can aid in differentiation. The presence of a solid, enhancing mass strongly favors tumor-related hemorrhage. Generally, tumors will demonstrate heterogeneous T1 and T2 signal intensity within the core, which can represent tumor matrix, necrotic tissue, evolving blood products or calcifications; however, the exact composition is dependent on the neoplasm type. Most neoplastic hemorrhages are intratumoral in location where blood products evolve slower than intraparenchymal hematomas. A hemosiderin rim may be present, but in contrast to other etiologies, they will often appear to have burst, owing to the outward expansion of a growing neoplasm. Additionally, vasogenic edema of the surrounding tissues is often much more prominent with neoplasms than vascular malformations (Sze et al. 1981). In highly vascularized tumors, a serpentine pattern of flow-related signal void is often visualized. A distinguishing trait from vascular malformations is the presence of tumor matrix (possibly enhancing) adjacent to the abnormal vessels (Connor et al. 2005). 3.4.5.3  Saccular Aneurysms Intracranial aneurysms are classified into 4 types: saccular, fusiform, dissecting, and mycotic types. The saccular

form, also known as a berry aneurysm, accounts for 90% of aneurysms. Saccular aneurysms are predominantly located within the circle of Willis or its branching arteries, with two thirds split evenly between the anterior communicating artery and the internal carotid artery (Gasparotti and Liserre 2005). It is estimated that approximately 5% of the general population harbor an intracranial aneurysm, and that the average aneurysm carries a 0.5% yearly risk of rupture. That risk of rupture is further increased with increasing size (>10 mm), posterior circulation location and a previous subarachnoid hemorrhage (SAH) (Wardlaw and White 2000). On conventional MRI, the findings of T1- and T2weighted imaging are relatively insensitive and nonspecific. It is possible to see flow-related signal void out of keeping with the expected anatomy of the vasculature. Furthermore, previous sentinel bleeds may also be evinced, in the form of leptomeningeal or intraparenchymal hemosiderin staining. These will appear as adjacent areas of hypointensity on T2- and T2*-weighted images, the latter being more sensitive. MRA findings are basic, demonstrating an unexpected deviation of flow signal enhancement from the artery into an adjoining balloon or dome shaped structure. Often it is possible to identify a neck and dome of the aneurysm, the measurements of which are important for treatment planning. Saccular aneurysms typically occur at bifurcations of vessels, and multiple aneurysms are not uncommon. Within the aneurysm itself, there may not always be complete flow enhancement due to intra aneurysmal thrombosis. Furthermore, if SAH is present and contains methemoglobin, it may obscure an aneurysm on T1-weighted spin-echo images and MRA. Challenges encountered with non-contrast MRA include the loss of flow enhancement signal from turbulence, or saturation effects from slow flow and in-plane flow (Gasparotti and Liserre 2005; White et al. 2001). CE MRA is theoretically superior to non-enhanced MRA in this regard; however, recent studies suggest equivalence, except in the depiction of giant aneurysms (Fig. 3.4.22) (Gasparotti and Liserre 2005). Comparing PC and TOF MRA for aneurysm assessment, TOF is currently regarded as the superior modality (White et al. 2001). Overall the sensitivity of MRA is highly dependent on several factors. Aneurysms located along the MCA bifurcation or ICA are notoriously difficult to visualize due to vessel overlap and adjacent bone. The size of the aneurysm is also crucial with reported sensitivities ranging from 86% to 92% for aneurysms over 5 mm, but a discouraging 25–35% for aneurysms <5 mm (White et al. 2001). CTA and all forms of MRA have similar accuracies of about 90% and comparable deficiencies with small-diameter aneurysms. Thus in the detection of intracerebral aneurysms, DSA remains the gold standard (studies suggest equivalence, except in the depiction of giant aneurysms (Fig. 3.4.22) (Gasparotti and Liserre 2005; Wardlaw and White 2000; White et al. 2001).

341

342

3  Brain, Head, and Neck

Fig. 3.4.22  Giant aneurysm. a Sagittal T1 image shows a parasellar lesion with iso- to hypointense signal and flow-related artifact. b The lesion shows heterogeneous signal on FLAIR and flow-related signal void on T2 (c). d On non-enhanced 3D TOF

MRA the lesion appears isointense but shows conspicuous flow related enhancement on CE MRA (e). f Catheter angiogram confirms the presence of a giant posterior communicating artery aneurysm

3.4.5.3.1  Aneurysm Screening

do not appear to benefit despite increased risk, include those with Marfan’s syndrome, Ehlers-Danlos type 4, and neurofibromatosis and acquired risk factors. In patients with one first-degree relative who has had a ruptured aneurysm, screening is also not currently recommended. The identified patient groups who may be appropriate for screening are those between the ages of 20 and 70 with: (1) autosomal dominant polycystic kidney disease (2) two or more first or second degree relatives who have had a ruptured aneurysm and (3) a previous history of aneurysmal SAH (Bederson et al. 2000; Rinkel 2005).

Aneurysm screening remains a controversial issue. DSA has poor sensitivity, but it is invasive and carries a complication rate of 0.5–1% (Bederson et al. 2000). Conversely, MRA and CTA have excellent sensitivities for aneurysms >5 mm and a negligible complication rate. Although MRA and CTA have poor sensitivities for aneurysms <5 mm, such small aneurysms have a very low risk of rupture, and are therefore much less important from a screening perspective. As such, both MRA and CTA are employed as screening modalities, with 3D TOF MRA being most common. The question of who to screen also remains to be defined, although there is some general consensus. Asymptomatic patients without risk factors do not benefit from screening (Bederson et al. 2000). Other patient groups that

3.4.5.4  Spontaneous Hemorrhages Non-traumatic spontaneous hemorrhage in the absence of an obvious etiology is most frequently due to hyper-

3.4  Cerebrovascular Disease

The hematoma can be of variable size and extend into the subarachnoid space. 3.4.5.4.1  Hypertension Hypertensive hemorrhage accounts for 90% of all spontaneous hemorrhage (Greenburg 1998). It is typically located in the supratentorial deep gray matter (putamen, thalamus) as well as the cerebellum and brainstem (Fig. 3.4.24). In addition to the acute bleed, it is not uncommon to identify multiple chronic microbleeds in these locations. It is important to note that hypertensive hemorrhage is not limited to these areas, and although considered atypical, it can occur in the lobar regions (Kim et al. 1994; Rosand 2004). Intraventricular hemorrhage may accompany hypertensive bleeds. Blood in the lateral or third ventricles can originate from hemorrhage in the thalamus, putamen, and caudate, while blood found in the fourth ventricle can arise from cerebellar hemorrhage. A hypertensive hemorrhage should be suspected when these findings are presented and other etiologies are not identifiable. Fig. 3.4.23  Microbleeds. GRE image demonstrates small, multifocal areas of magnetic susceptibility effect secondary to prior microbleeds.

tension or amyloid disease. Hypertensive hemorrhage can occur in the elderly in the setting of prolonged hypertension and vasculopathy, or in the young due to a bleeding diathesis or hypertensive crisis. Cerebral amyloid angiopathy (CAA) is implicated to a lesser extent in spontaneous hemorrhage, and it tends to affect those with Alzheimer’s disease as well as otherwise healthy elderly patients. Both entities share common MRI findings, in particular microbleeds (Fig. 3.4.23). On T2- and particularly T2*-weighted imaging, these appear as multiple small areas of focal signal loss, typically in the absence of any other related findings on MR. In the context of acute intracerebral hemorrhage these foci are frequently not directly involved with the area of hemorrhage. However, their presence is considered a significant indicator of CAA or hypertension. These discrete areas are thought to represent hemosiderin deposition following the resolution of previous minor, asymptomatic intraparenchymal bleeds, hemorrhages that were likely precipitated by the friable vessels of chronic CAA or hypertension. This is based on the significant correlation between multiple microbleeds and spontaneous intracerebral hemorrhages (Lee et al. 2004). In terms of acute hemorrhage, both share non-specific T1- and T2-weighted findings of intraparenchymal hemorrhage and mild peri-lesional edema.

3.4.5.4.2  Cerebral Amyloid Angiopathy CAA hemorrhage is commonly referred to as lobar hemorrhage, in reference to its affinity for the lobar regions of the brain. CAA related hemorrhage is accompanied by numerous microbleeds typically found in the cortical and sub cortical regions, most frequently involving the occipital lobe (Yamada 2000). Microbleeds are rarely found within the territories typical of hypertensive hemorrhage, except for the cerebellum. Although this is a relatively distinctive pattern, the presence of lobar hemorrhage is not diagnostic of CAA because atypical lobar hypertensive hemorrhage is actually more common in the general population (Itoh et al. 1993; Yamada 2000). Nonetheless, in normotensive elderly patients with lobar hemorrhage and findings of multiple microbleeds, in the absence of other etiologies, a presumptive diagnosis of CAA can be made. In terms of prognosis, there is evidence that the greater the number of microbleeds identified, the greater the recurrence rate and severity of future bleeds (Greenburg 1998; Greenburg et al. 1999). 3.4.5.5  Hemorrhagic Transformation of Ischemic Stroke Approximately 40–50% of all ischemic strokes will develop some degree of hemorrhage within hours to months of stroke onset (Tong et al. 2000). Many are asymptomatic in the sense that they do not exacerbate existing neurological symptoms any further, but they can be identified on follow up images. In general, patients are at increased

343

344

3  Brain, Head, and Neck

Fig. 3.4.24  Hypertensive hemorrhage. a Acute hemorrhage centered within the left putamen is hypointense on T2 with surrounding hyperintense edema. b DWI image is dominated by the susceptibility effect of the blood with surrounding hyperin-

tense signal. c ADC map demonstrates evidence of both ischemia (hypointense, restricted diffusion) and vasogenic edema (hyperintense, increased diffusion)

risk of hemorrhagic transformation (HT) in the setting of a prolonged ischemic period followed by sudden reperfusion (Zaheer et al. 2000). Reperfusion can occur physiologically or iatrogenically. With respect to acute stroke intervention, the increased risk of hemorrhagic transformation with time is one of the primary limitations of the therapeutic stroke window. HT should be suspected in all stroke patients who show clinical deterioration following the acute event. There are no MR findings specific to HT aside from the hemorrhage occurring within the previously identified ischemic area. MRI demonstrates the expected signal intensities of intraparenchymal hemorrhage, which evolve in a similar manner. T2*-weighted sequences are generally regarded as the most sensitive, the development of hypointense signal being one of the earliest findings of HT (Arnould et al. 2004). Extent of hemorrhage can range from gyral petechial hemorrhage at the margins of the infarct, to large mass producing hematomas. Much research has focused on predictors of HT in the acute ischemic stroke setting, with its potential to guide management and extend the therapeutic window. A strong association has been observed in ischemic strokes of embolic origin, as 95% of these infarct lesions are predicted to demonstrate hemorrhage at some point (Tong et al. 2000). Furthermore, the presence of early T1 parenchymal enhancement following Gadolinium administration is a strong predictor of subsequent HT (Vo et al. 2003). Moderate predictors of future HT include large infarct size, microbleeds, and very low ADC values over numer-

ous voxels (Selim et al. 2002; Tong et al. 2000; Zaheer et al. 2000). A more recently proposed predictor is the (abnormal) visualization of transcerebral veins. This is seen as hypointense signal tracing the course of veins around the area of infarct, and it is best visualized on T2*-weighted images. When this finding is seen during the initial stages of ischemic stroke, it may be a good prognostic indicator of HT (Hermier et al. 2003). All of these findings predicting HT deserve some consideration, although their role in screening and selection of candidates for thrombolysis is still under investigation. References 1.

2. 3.

4.

Arnould M, Grandin C, Cosnard G, Duprez T (2004) Comparison of CT and three MR sequences for detecting and categorizing early (48 h) hemorrhagic transformation in hyperacute ischemic stroke. AJNR 25:939–944 Baird AE, Warach S (1998) Magnetic resonance imaging of acute stroke. J Cereb Blood Flow Metab 18:583–609 Baker LL, Kucharczyk J, Sevick RJ, Mintorovitch J, Moseley ME (1991) Recent advances in MR imaging/spectroscopy of cerebral ischemia. AJR Am J Roentgenol 156:1133–1143 Bederson J et al (2000) Recommendations for the management of patients with unruptured intracranial aneurysms: a statement for healthcare professionals from the Stroke Council of the American Heart Association. Stroke 31:2742–2750

3.4  Cerebrovascular Disease 5. 6.

7.

8.

9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Bradley W (1993) MR appearance of hemorrhage in the brain. Radiology 189:15–26 Brown R et al (2005) Natural history, evaluation, and management of intracranial vascular malformations. Mayo Clin Proc 80:269–281 Brunereau L, Labauge P, Lasserve E, Laberge S, Levy C, Houtteville J (2002) Familial form of intracranial cavernous angioma: MR imaging findings in 51 families. Radiology 214:209–216 Burdette JH, Ricci PE, Petitti N, Elster AD (1998) Cerebral infarction: time course of signal intensity changes on diffusion-weighted MR images. Am J Roentgenol 171:791–795 Cloft HJ, Murphy HJ, Prince MR, Brunberg JA (1996) 3D gadolinium-enhanced MR angiography of the carotid arteries. Magn Reson Imaging 14:593–600 Connor S, Flis C, Langdon J (2005) Vascular masses of the head and neck. Clin Radiol 60:856–868 Cosnard G, Duprez T, Grandin C, Smith A, Munier T, Peeters A (1999) Fast FLAIR sequence for detecting major vascular abnormalities during the hyperacute phase of stroke: a comparison with MR angiography. Neuroradiology 41:342–346 Derex L, Nighoghossian N, Hermier M, Adeleine P, Philippeau F, Honnorat J, Yilmaz H, Dardel P, Froment JC, Trouillas P (2004) Thrombolysis for ischemic stroke in patients with old microbleeds on pretreatment MRI. Cerebrovasc Dis 17:238–241 Dillon W (1997) Cryptic vascular malformations: controversies in terminology, diagnosis, pathology and treatment. AJNR 18:1839–1846 Domenicucci M, Signorini P, Strzelecki J, Delfini R (1995) Delayed post-traumatic epidural hematoma. A review. Neurosurg Rev 18:109–122 Essig M, Kummer R, Egelhof T, Winter R, Sartor K (1996) Vascular MR contrast enhancement in cerebrovascular disease. AJNR 17:887–894 Feigin VL, Lawes CM, Bennett DA, Anderson C S (2003) Stroke epidemiology: a review of population based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol 2:43–53 Fiebach JB, Schellinger PD, Gass A, Kucinski T, Siebler M, Villringer A, Olkers P, Hirsch JG, Heiland S, Wilde P, Jansen O, Rother J, Hacke W, Sartor K (2004a) Stroke magnetic resonance imaging is accurate in hyperacute intracerebral hemorrhage: a multicenter study on the validity of stroke imaging. Stroke 35:502–506 Fiebach JB, Schellinger PD, Hacke W, Geletneky K, Wilde P, Meyer M, Sartor K (2004b) MRI in acute subarachnoid haemorrhage; findings with a standardised stroke protocol. Neuroradiology 46:44–48 Frayne R et al (2003) Magnetic resonance imaging at 3.0 Tesla: challenges and advantages in clinical neurological imaging. Invest Radiol 38:385–402 Gasparotti R, Liserre R (2005) Intracranial aneurysms. Eur Radiol 15:441–447

21. Gonzalez RG, Schaefer PW, Buonanno FS, Schwamm LH, Budzik RF, Rordorf G, Wang B, Sorensen AG, Koroshetz W (1999) Diffusion-weighted MR imaging: diagnostic accuracy in patients imaged within 6 h of stroke symptom onset. Radiology 210:155–162 22. Greenberg S (1998) Cerebral amyloid angiopathy: prospects for clinical diagnosis and treatment. Neurology 51:690–694 23. Greenberg S, O’Donnell H, Schaefer P, Kraft E (1999) MRI detection of new hemorrhages: Potential marker of progression in cerebral amyloid angiopathy. Neurology 53:1135–1138 24. Hacke W, Kaste M, Fieschi C et al (1995) Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA 274:1017–1025 25. Hamon M, Marie M, Clochon P, Coskun O, Constans M, Viader F, Courtheoux P, Baron C (2005) Quantitative relationships between ADC and perfusion changes in acute ischemic stroke using combined diffusion-weighted imaging and perfusion MR (DWI/PMR). J Neuroradiol 32:118–124 26. Hanel RA, Xaxier AR, Mohammad Y, Kirmani JF, Yahia AB, Qureshi AI (2002) Outcome following intracerebral hemorrhage and subarachnoid hemorrhage. Neurol Res 24:S58–S62 27. Hennerici M, Daffertschofer M, Jakobs L (1998) Failure to identify cerebral infarct mechanisms from topography of vascular territory lesions. AJNR 19:1067–1074 28. Hermier M et al (2003) Hypointense transcerebral veins at T2*-weighted MRI: a marker of hemorrhagic transformation risk in patients treated with intravenous tissue plasminogen activator. J Cereb Blood Flow Metab 23:1362–1370 29. Hill MD, Barber PA, Demchuk AM et al (2000) Building a “brain attack” team to administer thrombolytic therapy for acute ischemic stroke. CMAJ 162:1589–1593 30. Itoh Y et al (1993) Cerebral amyloid angiopathy: a significant cause of cerebellar as well as lobar cerebral hemorrhage in the elderly. J Neuro Sci 116:135–141 31. Kalafut MA, Schriger DL, Saver JL, Starkman S (2000) Detection of early CT signs of >1/3 middle cerebral artery infarctions. Stroke 31:1667–1671 32. Kidwell CS et al (2004) Comparison of MRI and CT for detection of acute intracerebral hemorrhage. JAMA 292:1823–1830 33. Kim S, Lee H, Lee C (1994) Small primary intracerebral hemorrhage: clinical presentation of 28 cases. Stroke 25:1500–1506 34. Ko H, Johnston S, Young W, Singh V, Klatsky A (2003) Distinguishing intracerebral hemorrhages caused by arteriovenous malformations. Cerebrovasc Dis 15:206–209 35. Krapf H, Widder B, Skalej M (1998) Small rosary-like infarctions in the centrum ovale suggest hemodynamic failure. AJNR 19:1479–1484

345

346

3  Brain, Head, and Neck 36. Kubalek R, Moghtaderi A, Klisch J, Berlis A, Quiske A, Schumacher M (2003) Cerebral arteriovenous malformations: influence of angioarchitecture on bleeding risk. Acta Neurochir 145:1045–1052 37. Kucinski T, Koch C, Eckert B, Becker V, Kro H, Heesen C, Grzyska U, Freitag J, Röther J, Zeumer H (2003) Collateral circulation is an independent radiological predictor of outcome after thrombolysis in acute ischaemic stroke. Neuroradiology 45:11–18 38. Kwon B, Han M, Kang H, Chang K (2005) MR imaging findings of intracranail dural arteriovenous fistulas: relations with venous drainage patterns. AJNR 26:2500–2507 39. Lee R, Becher M, Benson M, Rigamonti D (1997) Brain capillary telangiectasia: MR imaging appearance and clinicohistopathologic findings. Radiology 205:797–805 40. Lee SH, Bae HJ, Kwon SJ, Kim H, Kim YH, Yoon BW, Roh JK (2004) Cerebral microbleeds are regionally associated with intracerebral hemorrhage. Neurology 62:72–76 41. Licata C, Turazzi S (2003) Bleeding cerebral neoplasms with symptomatic hematoma. J Neurosurg Sci 47:201–210 42. Liebengerg WA, Worth R, Firth GB, Olney J, Norris JS (2005) Aneurysmal subarachnoid haemorrhage: guidance in making the correct diagnosis. Postgrad Med J 81:470–473 43. Linfante I, Llinas RH, Caplan LR, Warach S (1999) MRI features of intracerebral hemorrhage within 2 h from symptom onset. Stroke 30:2263–2267 44. Liu LH, Chen CC, Chang MH (2004) Clinical findings of intracranial vertebral artery disease using magnetic resonance angiography. Acta Neurol Taiwan 13:120–125 45. Loubinoux I, Volk A, Borredon J, Guirimand S, Tiffon B, Seylaz J, Meric P (1997) Spreading of vasogenic edema and cytotoxic edema assessed by quantitative diffusion and T2 magnetic resonance imaging. Stroke 28:419–427 46. MacWalter RS, Ersoy Y, Wolfson DR (2001) Cerebral haemorrhage: parenchymal intracranial haemorrhage. Gerontology 47:119–130 47. Mohamed M, Heasely C, Yagmurlu B, Yousem DM (2004) Fluid-attenuated inversion recovery MR imaging and subarachnoid hemorrhage: not a panacea. AJNR 25:545–550 48. Mullins ME, Schaefer PW, Sorensen GA, Halpern EF, Ay H, He J, Koroshetz WJ, Gonzalez RG (2002) CT and conventional and diffusion-weighted MR imaging in acute stroke: study in 691 patients at presentation to the emergency department. Radiology 224:353–360 49. NINDS rt-PA Study Group The (1995) Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 333:1581–1587 50. Noguchi K et al (1997) Subacute and chronic subarachnoid hemorrhage: diagnosis with fluid-attenuated inversion-recovery MR imaging. Radiology 203:257–262 51. Noguchi K, Ogawa T, Inugami A, Fujita H, Hatazawa J, Shimosegawa E, Okudera T, Uemura K, Seto H (1997) MRI of acute cerebral infarction: a comparison of FLAIR and T2-weighted fast spin-echo imaging. Neuroradiology 39:406–410

52. Oelerich M, Lentschig G, Zunker P, Reimer P, Rummeny J, Schuierer G (1998) Intracranial vascular stenosis and occlusion: comparison of 3D time-of-flight and 3D phasecontrast MR angiography. Neuroradiology 40:567–573 53. Özsarlak O, Van Goethem W, Maes M, Parizel M (2004) MR angiography of the intracranial vessels: technical aspects and clinical applications. Neuroradiology 46:955–972 54. Palmer AJ, Valentine WJ, Roze S, Lammert M, Spiesser J, Gabriel S (2005) Overview of costs of stroke from published, incidence-based studies spanning 16 industrialized countries. Curr Med Res Opin 21:19–26 55. Park DC, Nam HS, Lim SR, Lee PH, Heo JH, Lee BI, Kim DI (2000) MRI features of infarcts with potential cardiac source of embolism in the Yonsei Stroke Registry. Yonsei Medical Journal 41:431–435 56. Parsons MW, Yang Q, Barber PA, Darby DG, Desmond PM, Gerraty RP, Tress BM, Davis SM (2001) Perfusion magnetic resonance imaging maps in hyperacute stroke relative cerebral blood flow most accurately identifies tissue destined to infarct. Stroke 32:1581–1587 57. Patel MR, Edelman RR, Warach S (1996) Detection of hyperacute primary intraparenchymal hemorrhage by magnetic resonance imaging. Stroke 27:2321–2324 58. Pereira RS, Harris AD, Sevick RJ et al (2002) Effect of b value on contrast during diffusion-weighted magnetic resonance imaging assessment of acute ischemic stroke. J Magn Reson Imaging 15:591–596 59. Phan T, Huston J, Bernstein A, Riederer J, Brown D (2001) Contrast-Enhanced magnetic resonance angiography of the cervical vessels. Stroke 32:2282–2286 60. Preter M, Tzourio C, Ameri A, Bousser G (1996) Longterm prognosis in cerebral venous thrombosis. Stroke 27:243–246 61. Provenzale JM (1995) Dissection of the internal carotid and vertebral arteries: imaging features. AJR Am J Roentgenol 165(5):1099–1104 62. Provenzale JM (1997) Brain infarction in young adults: etiology and imaging. AJR Am J Roentgenol 169:1161–1168 63. Provenzale JM, Jahan R, Naidich TP, Fox AJ (2003) Assessment of the patient with hyperacute stroke: imaging and therapy. Radiology 229:347–359 64. Restrepo L, Jacobs A, Barker B, Beauchamp J, Skolasky L, Keswani C, Wityk J (2005) Etiology of perfusion-diffusion magnetic resonance imaging mismatch patterns. J Neuroimaging 15:254–260 65. Rigamonti D et al (1988) Appearance of venous malformations on magnetic resonance imaging. J Neurosurg 69:535–539 66. Rinkel G (2005) Intracranial aneurysm screening: indications and advice for practice. Lancet Neurol 4:122–128 67. Rosand J (2004) Hypertension and the brain. Neurology 63:6–7 68. Rovira A, Orellana P, Alvarez-Sabin J, Arenillas F, Aymerich X, Grive E, Molina C, Rovira-Gols A (2004) Hyperacute ischemic stroke: middle cerebral artery susceptibility sign at echo-planar gradient-echo MR imaging. Radiology 232:466–473

3.4  Cerebrovascular Disease 69. Rumboldt Z, Kalousek M, Castillo M (2003) Hyperacute subarachnoid hemorrhage on T2-weighted MR images. AJNR 24:472–475 70. Sato A, Takahashi S, Soma Y, Ishii K, Kikuchi Y, Watanabe T, Sakamoto K (1991) Cerebral infarction: early detection by means of contrast-enhanced cerebral arteries at MR imaging. Radiology 178:433–439 71. Schaefer PW, Grant PE, Gonzalez RG (2000) Diffusion-weighted MR imaging of the brain. Radiology 217:331–345 72. Schellinger PD, Chalela JA, Kang D, Latour LL, Warach S (2005) Diagnostic and prognostic value of early MR imaging vessel signs in hyperacute stroke patients imaged <3 h and treated with recombinant tissue plasminogen activator. AJNR Am J Neuroradiol 26:618–624 73. Schellinger PD, Jansen O, Fiebach JB et al (1999) A standardized MRI stroke protocol: comparison with CT in hyperacute intracerebral hemorrhage. Stroke 30:765–768 74. Selim M et al (2002) Predictors of hemorrhagic transformation after intravenous recombinant tissue plasminogen activator. Stroke 33:2047–2052 75. Selim M, Fink J, Linfante I, Kumar S, Schlaug G, Caplan R (2002) Diagnosis of cerebral venous thrombosis with echoplanar T2*-weighted magnetic resonance imaging. Arch Neurol 59:1021–1026 76. Sohn CH, Sevick RJ, Frayne R (2003) Contrast-enhanced MR angiography of the intracranial circulation. Magn Reson Imaging Clin N Am 11:599–614 77. Stejskal EO, Tanner JE (1965) Spin diffusion measurements: spin-echoes in the presence of a time-dependent field gradient. J Chem Phys 42:288–292 78. Summers PE, Jarosz JM, Markus H (2001) MR angiography in cerebrovascular disease. Clinical Radiology 56:437–546 79. Suzuki M, Matsui O, Kobayashi K, Ueda F, Saitoh C, Katagiri A, Sanada J, Tawara M, Terayama N, Kawashima H, Kida S, Yamashita J (2003) Contrast-enhanced MRA for investigation of cerebral arteriovenous malformations. Neuroradiology 45:231–235 80. Sze G, Krol Ge, Olsen W, Harper P, Galicich J Heier L, Zimmerman R, Deck M (1987) Hemorrhagic neoplasms: MR mimics of occult vascular malformations. AJR Am J Roentgenol 149:1223–1230 81. Tei H, Uchiyama S, Koshimizu K, Kobayashi M, Ohara K (1999) Correlation between symptomatic radiological and etiological diagnosis in acute ischemic stroke. Acta Neurol Scand 99:192–195 82. Tong D, Adami A, Moseley M, Marks M (2000) Relationship between apparent diffusion coefficient and subsequent hemorrhagic transformation following acute ischemic stroke. Stroke 31:2378–2384 83. U-King-Im JM, Trivedi RA, Graves MJ, Higgins NJ, Cross JJ, Tom BD, Hollingworth W, Eales H, Warburton EA, Kirkpatrick PJ, Antoun NM, Gillard JH (2004) Contrastenhanced MR angiography for carotid disease. Neurology 62:1282–1290

84. Underwood RS, Mohiaddin RH (1993) Magnetic resonance imaging of atherosclerotic vascular disease. Am J Hypertens 6:335S–339S 85. Vilanova JC et al (2004) Hemangioma from head to toe: MR imaging with pathologic correlation. Radiographics 24(2):367–385 86. Vo K, Santiago F, Lin W, Hsu C, Lee Y, Lee J (2003) MR imaging enhancement patterns as predictors of hemorrhagic transformation in acute ischemic stroke. AJNR 24:674–679 87. Wardlaw J, White P (2000) The detection and management of unruptured intracranial aneurysms. Brain 123:205–221 88. Wasay M, Azeemuddin M (2005) Neuroimaging of cerebral venous thrombosis. J Neuroimaging 15:118–128 89. Wasenko J, Lieberman KA, Rodziewicz GS, Holsapple J (2002) Magnetic resonance imaging characteristics of hyperacute hemorrhage in the brain and spine. Clin Imaging 26:330–337 90. White P, Teasdale E, Wardlaw J, Easton V (2001) Intracranial aneurysms: CT angiography and MR angiography for detection - prospective blinded comparison in a large patient cohort. Radiology 219:739–749 91. Wiesmann M, Mayer TE, Yousry I, Medele R, Hamann GF, Bruckmann H (2002) Detection of hyperacute subarachnoid hemorrhage of the brain by using magnetic resonance imaging. J Neurosurg 96:684–689 92. Willinsky R, Goyal M, Brugge K, Montanera W (1999) Tortuous, engorged pial veins in intracranial dural arteriovenous fistulas: correlation with presentation, location and MR findings in 122 patients. AJNR 20:1031–1036 93. Wolf R (2001) Intraarterial signal on fluid-attenuated inversion recovery images: a measure of hemodynamic stress? AJNR 22:1015–1017 94. Wycliffe ND, Choe J, Holshouser B, Oyoyo UE, Haacke M, Kido DK (2004) Reliability in detection of hemorrhage in acute stroke by a new three-dimensional gradient recalled echo susceptibility-weighted imaging technique compared to computed tomography: a retrospective study. J Magn Reson Imaging 20:372–377 95. Yamada M (2000) Cerebral amyloid angiopathy: An overview. Neuropathology 1:8–22 96. Yang JJ, Hill MD, Morrish WF, Hudon ME, Barber PA, Demchuk AM, Sevick RJ, Frayne R (2002) Comparison of pre- and postcontrast 3d time-of-flight MR angiography for the evaluation of distal intracranial branch occlusions in acute ischemic stroke. AJNR Am J Neuroradiol 23:557–567 97. Yuh WT, Crain MR, Loes DJ, Greene GM, Ryals TJ, Sato Y (1991) MR imaging of cerebral ischemia: findings in the first 24 h. AJNR Am J Neuroradiol 12:621–629 98. Zaheer A, Ozsunar T, Schaefer P, (2000) Magnetic resonance imaging of cerebral hemorrhagic stroke. Top Magn Reson Imaging 11:288–299 99. Zwan A van der, Hillen B, Tulleken C, Dujovny M (1993) A quantitative investigation of the variability of the major cerebral arterial territories. Stroke 24:1951–1959

347

348

3  Brain, Head, and Neck

3.5  Intracranial Infections E. Turgut Tali and Serap Gültekin Infections of the central nervous system (CNS) frequently present diagnostic and therapeutic challenges. Prompt diagnosis and aggressive treatment often are necessary to allow the best chance of recovery without sequelae. The prognosis depends on rapid identification, the site of the inflammation, and the pathogen. Improvements in diagnostic imaging, and particularly in magnetic resonance imaging (MRI), have greatly facilitated the diagnosis, treatment, and treatment monitoring of intracranial infections. MRI is the most sensitive modality for imaging CNS infections because of its superior soft tissue detail, multiplanar imaging capabilities, and vascular, metabolic, and physiological imaging techniques. Advanced MRI techniques aid in the investigation of physiological properties of tissues. Diffusion-weighted imaging (DWI) acquires images in which the signal intensities of tissues depend on the speed of water motion within the tissues. Diffusion tensor imaging (DTI) represents mean longitudinal direction of axons in the white matter tracts, and depicts the normal course, the displacement, or interruption of white matter tracts around a pathological entity. MR spectroscopy (MRS) provides in vivo biochemical assessment of intracranial lesions. Several of the major metabolite peaks of the brain are localized with the MRS technique. MR angiography (MRA) provides essential information to detect early and subtle vascular abnormalities such as vasculitis, vascular spasm, stenosis, aneurysm and occlusion, dural sinus thrombosis and accompanying infections. There are four principal routes of infection to the brain: 1. Hematogenous spread 2. Direct spread from an adjacent focus of infection (i.e., sinonasal cavities, otorhinologic, birth canal) 3. Direct inoculation from penetrating injury, surgery, etc. 4. Ascending through the peripheral nerves, etc. into the CNS. This section describes the basic and recent findings in neuroimaging of CNS infections. 3.5.1  Meningitis The meninges consist of three layers: the outermost membrane, the dura mater, is also referred as the pachymeninx; both the pia mater and the arachnoid mater are referred to as the leptomeninges. Meningitis is a pathological process involving diffuse inflammation of the membranes (pia-arachnoid mater and cerebrospinal fluid [CSF] or dura-arachnoid mater or both) surrounding the

brain and spinal cord. Meningitis is more common in children. Despite the widespread and early use of antibiotics, up to one third of children may develop sequelae from the disease (Harris and Edwards 1992). Meningitis as the most common form of CNS infections is divided into the following categories: 1. Bacterial meningitis 2. Viral meningitis 3. Meningitis caused by specific organisms 3.5.1.1  Bacterial Meningitis Nonepidemic meningitis is most common in neonates, infants, and children. Epidemic meningitis can occur at any age. New antibiotics have markedly diminished the incidence of bacterial meningitis and its complications (Osborn 1994). The etiology varies according to the ages of the patients and the status of their immune systems. Gram-negative bacilli (particularly Escherichia coli), group B streptococci, and other enteric bacilli are the major causative agents during the neonatal period. Haemophilus influenzae and Neisseria meningitidis are the major agents among children beyond one month of age. In adults, the most common agents include streptococci and N. meningitidis, although disease caused by aerobic gram-negative bacilli is increasing in frequency, especially in elderly. Immunocompromised patients are prone to infections caused by E. coli, Klebsiella, Pseudomonas, and fungi. Iatrogenic infections are usually the result of gram-negative microorganisms. N. meningitidis is the only major cause of epidemics of bacterial meningitis. Pathologically, bacteria that reach into the subarachnoid space stimulate the production of cytokines and other inflammatory products. Neutrophils migrate into the subarachnoid space and produce purulent exudates, which distend the subarachnoid space and also extend into the perivascular spaces. Frequently, basal cisterns are filled with pus. However, pneumococcal pus commonly accumulates over the convexities. The inflammation may extend to the walls of arteries and veins, resulting in vasculitis and thrombosis that may cause hemorrhage and cortical infarction. Coagulopathy caused by septicemia (especially meningococcal) may also cause hemorrhage. Degenerated neutrophils are removed by the macrophages in the later stage, and lymphocytes and fibroblasts proliferate in the exudate. Accumulation of the exudate and inflammatory debris may cause obstruction of CSF flow, and also may interfere with resorption of the CSF over the convexities or at the arachnoid villi. Communicating and noncommunicating hydrocephalus may result. Increased intracranial pressure, deteriorated cerebral perfusion, and loss of autoregulation of the cerebral blood flow may complicate the pathologic changes and result in severe clinical status such as coma or death. Involvement

3.5  Intercranial Infections

of cranial nerves results in neuropathies. Extension into the brain produces cerebritis that may progress to an abscess (Castillo 2004; Anslow 2004). Peripheral blood studies are nonspecific, peripheral leukocytosis with a shift to the left, and elevation of the erythrocyte sedimentation rate (ESR) are frequently present. Blood cultures are positive in up to 75% of patients. The definitive diagnosis rests on identification of the organism from culture. Lumbar puncture with examination of the CSF represents the cornerstone of the diagnosis in suspected cases. Although lumbar puncture can probably be safely performed without neuroimaging studies in most patients, CT or MRI should be performed routinely to exclude the presence of a mass lesion, especially in patients with focal neurological signs. However, if bacterial meningitis is strongly suspected, antibiotic treatment should never be delayed while waiting for an imaging study. In such cases, blood cultures should be obtained, and empiric antibiotics should be started promptly (Choi and Akins 2002; Kastrup et al. 2005). There are two views on the imaging of meningitis: (1) imaging is not indicated and is reserved to look for complications; and (2) imaging is indicated if the clinical diagnosis is unclear, neurologic deterioration occurs or recovery is slow. Unenhanced MR scans of patients with uncomplicated acute bacterial meningitis, particularly at the early stage, may be unremarkable. Negative neuroimaging never rules out meningitis. Only 50% of patients with meningitis show subarachnoid space obliteration (Castillo 2004). In addition to the distension of subarachnoid space (with widening of interhemispheric fissure), post-contrast MR studies can demonstrate leptomeningeal enhancement (Fig. 3.5.1). Bacterial meningitis is usually associated with marked meningeal enhancement (pia-subarachnoid or dura-subarachnoid type), which is best seen on coronal T1-weighted post-contrast images (T1-weighted images) (Figs. 3.5.1, 3.5.2). However, abnormal meningeal enhancement is not pathognomic of meningitis and can be associated with any condition causing inflammation of meninges. Meningeal enhancement can be seen in postoperative areas, after infarction, and in other inflammatory conditions, such as after radiation, trauma, and with sarcoidosis (Anslow 2004; Whiteman et al. 2002; Smith and Caldemeyer 1999). Also, in normal patients, there may be insignificant meningeal enhancement—perhaps a small, focal area of thin, linear enhancement not seen on consecutive slices after contrast (Osborn 1994; Smith and Caldemeyer 1999). Double/triple dose, higher relaxivity, or higher molar contrast agents provide better enhancement. Recently, introduction and widespread use of new imaging techniques have contributed significantly to the rapid and more specific diagnosis of meningitis and its complications. The introduction of fluid attenuated inversion recovery (FLAIR), DWI, and MRS has led to a bet-

ter understanding of meningitis. FLAIR images show the CSF within the subarachnoid spaces to be of higher signal intensity in meningitis (Figs. 3.5.1, 3.5.2). This is due to prolongation of the relaxation time, which is caused by the presence of excess protein in the CSF (Tsuchiya et al. 1997). The subcortical white matter may be hypointense in regions underlying areas of intense meningeal inflammation on FLAIR images (Lee et al. 2002). In addition, post-contrast FLAIR imaging shows the abnormal meningeal enhancement better than the post-contrast T1-weighted (Ercan et al. 2004) (Fig. 3.5.2). Magnetization transfer (MT) techniques also may aid the visualization of enhancing meninges as well as the identification of infection-related intracerebral lesions (Mehta et al. 1995; Kamra et al. 2004). Quantification of MT ratios in cases of meningitis with different etiologies may be of value in differentiating between etiologic agents. MT ratios of meningitis were found to be significantly higher in pyogenic and fungal meningitis than in tuberculosis (TB), and significantly higher in TB than in viral meningitis (Gupta et al. 1999). On MRS, elevation of lactate, large amounts of amino acids, acetoacetate, reduced glucose with normal N-acetyl-l-aspartate (NAA), creatine (Cr), and choline (Cho) have been reported (Shawle 1995; Roy et al. 1996). Decreased brain pH, without differences in cellular energy metabolites with CSF lactic acidosis and hypoglycemia have also been shown by using phosphorus MRS (Matthews et al. 1989). 3.5.1.2  Viral Meningitis Aseptic meningitis is usually caused by viral pathogens. Non-polio enteroviruses (echovirus and coxsackie viruses), mumps virus, arboviruses and Epstein-Barr virus are common causes of aseptic meningitis. Parasitic, para-meningeal, neoplastic, chemical, immune-mediated, vascular, and toxic factors may also cause aseptic meningitis. Viral meningitis classically manifest as a febrile illness accompanied by the typical symptoms and signs indicative of meningeal irritation, which are collectively known as meningismus. The patients most commonly present with headache, neck stiffness, fever, and chills. The condition usually follows a subacute course. In some cases it is biphasic, with an initial viral prodromal of fever and myalgias followed by resolution of symptoms and then reappearance of fever accompanied by headache and stiff neck (Patterson and Ling 2002). For patients in whom meningitis is suspected (regardless of the cause), a lumbar puncture should be performed without delay. The laboratory diagnosis of viral meningitis is based on CSF analysis. Identification of specific viral agent may be difficult (Rotbart 1997). Imaging findings in patients with viral meningitis are frequently normal. Viral meningitis generally does not

349

350

3  Brain, Head, and Neck

Fig. 3.5.1  Uncomplicated pia-subarachnoid type meningitis. Axial pre-contrast T1-weighted image (a) and T2-weighted image (b) show no prominent changes. However, axial FLAIR image (c) demonstrates high signal intensity CSF in the cortical sulci compared with normal dark CSF signal from ventricles.

Post-contrast axial (d) and coronal (e) T1-weighted images show pia-subarachnoid type enhancement in the cortical sulci. Post-contrast coronal (f) and axial (g) T1-weighted images from another patient exhibit diffuse and nodular leptomeningeal enhancement at the right temporal lobe

3.5  Intercranial Infections

Fig. 3.5.2  Dura-subarachnoid type meningitis. Axial pre-contrast T1-weighted image (a) and T2-weighted image (b) show only suspicious meningeal thickening with hydrocephalus. However, axial FLAIR image (c) demonstrates diffuse high signal intensity of the dura mater over the cerebral convexities.

Post-contrast axial (d) and coronal (e) T1-weighted images demonstrate abnormal dural thickening and enhancement over the convexities as well as the falx and tentorium. Diffuse dural enhancement is more obvious on post-contrast FLAIR image (f)

produce the meningeal enhancement seen on gadolinium-enhanced MRI in cases of bacterial meningitis. Unless an accompanying encephalitis is present, the brain parenchyma is also normal, and even with some viral encephalitides, such as that caused by enterovirus, the MR appearance still remains normal (Osborn 1994).

diffuse form (e.g., basal exudative leptomeningitis) or in a localized form (e.g., tuberculoma, abscess) (Whiteman et al. 2002; Arbealáez et al. 2004; Bernaerts et al. 2003). Brain imaging studies allow for early diagnosis of CNS TB, and contrast-enhanced MRI is the method of choice for diagnosis and follow-up. The infection disseminates hematogenously from a distant focal point, and the meningeal compromise is considered to occur due to rupture of microscopic cortical, subependymal, and/or sub-pial lesions called “Rich’s foci” into the subarachnoid space (Arbealáez et al. 2004; Bernaerts et al. 2003).The meningeal TB can also be secondary to the rupture of a tuberculoma into a blood vessel, or can be acquired from TB mastoiditis (Cannard 2002). Clinically, TB meningitis most often presents with fever, headache, a decreased level of consciousness, and

3.5.1.3  Meningitis Caused by Specific Organisms (Tuberculous Meningitis) Tuberculosis (TB) can affect any organ system. Meningeal tuberculosis is the most common presentation of CNS TB and is most frequently seen in children and adolescents. Coexisting pulmonary TB is often present (25– 83%). CNS TB can present in different forms, either in a

351

352

3  Brain, Head, and Neck

meningeal signs. The CSF shows lymphocytic pleocytosis, a marked increase in proteins, and decreased glucose levels (Zuger and Lowy 1997). Polymerase chain reaction (PCR) for CNS TB can also be obtained and can confirm the diagnosis rapidly, well before culture results are available. Overall mortality for tuberculous meningitis is over 25% and even higher in children. Early detection and appropriate long-term antituberculous therapy together with corticosteroids may reduce mortality and morbidity in patients with TB meningitis (Sütlaş et al. 2003). The common triad of imaging findings in TB meningitis is basal meningeal enhancement, hydrocephalus,

and infarctions in the supratentorial brain parenchyma and brain stem. Basal meningeal enhancement is the most consistent feature, caused by the “leaky” inflammatory neovessels (Fig. 3.5.3). Occasionally, meningeal enhancement is seen over the cerebral convexities, Sylvian fissures, and the tentorium (Arbealáez et al. 2004; Bernaerts et al. 2003). In patients with AIDS, there may be minimal or absent meningeal enhancement, supposedly due to an impaired immunological response (Harisinghani et al. 2000). High-resolution proton MRS of the CSF has been shown to be a valuable method of confirmation (Roy et al. 1996). It shows large resonance from

Fig. 3.5.3  Tuberculous meningitis. Axial pre-contrast T1weighted image (a) and T2-weighted image (b) demonstrate no abnormal finding. However, axial FLAIR image (c) of the same plane shows abnormal linear high signal surrounding the brain stem in the basal cisterns. Post-contrast axial T1-weighted images (d,e) demonstrate abnormal enhancement in basal cisterns including perimesencephalic and ambient cisterns. Slight dilata-

tion of the temporal horn of right lateral ventricle is seen. Postcontrast sagittal T1-weighted image (f) of a different patient depicts basal exudative tuberculous meningitis with intense diffuse and nodular enhancement in the basal subarachnoid cisterns, meninges, and third and forth ventricles (Courtesy of B. Diren, MD)

3.5  Intercranial Infections

lactate, acetate and sugars with cyclopropyl rings (–0.5 to +0.5 ppm) and phenolic glycolipids (7.1 ppm) (Roy et al. 1996). Extension of the inflammatory response to the ventricular system may result in choroid plexitis (Choi et al. 1998). Hydrocephalus is the most frequent complication of TB meningitis and develops mainly in children. It is secondary to blockage of the basal cisterns by the inflammatory exudate. Vasculitis of the small and medium-sized vessels is the second most common complication that causes cerebral infarctions. Vascular occlusions may be caused either by the infection itself or by the fibrosis of the exudate around the vessels. MRA is a useful technique to demonstrate the vascular complications. The differential diagnosis of TB meningitis includes other infections caused by viruses, fungi, and parasites and also noninfectious causes, such as rheumatoid disease, sarcoidosis, and primary or secondary neoplasias compromising meningeal surfaces. The suspicion of fungal meningitis is always increased in immunocompromised patients. Radiologically, healing may be recognized by the absence of basal meningovascular enhancement. After treatment, meningeal enhancement may even persist in some cases of TB meningitis (Jinkins 1991). 3.5.1.4  Complications of Meningitis Not infrequently, patients with bacterial meningitis develop neurological sequelae as a result of their illness. CNS complications develop in 50% of adult patients with bacterial meningitis and can arise during or after the onset of meningitis (Osborn 1994). The development of seizure activity and/or focal neurological symptoms and signs during the course of meningitis are usually caused by complications such as cerebritis, brain abscess, cortical infarction, enlarging subdural effusions, or empyema. 3.5.1.4.1  Hydrocephalus Hydrocephalus is a complication of meningitis seen more frequently in children and may be communicating or non-communicating (obstructive) in nature. Communicating hydrocephalus is more common. It is frequently seen 2–3 weeks after the onset of infection. A mild degree of hydrocephalus is found in most patients with bacterial meningitis, but, in most of them, this is a transient phenomenon. In some patients, the ventricular size never returns to normal but becomes stable, and there is no need for therapy. Only a small number of patients will require CSF-shunting procedures (Castillo 2004; Smith and Caldemeyer 1999). Periventricular CSF accumulation secondary to ventricular obstruction is seen on FLAIR, proton density, and T2-weighted images as areas of hyperintense signal surrounding the ventricular sys-

tem. CSF flow imaging and analysis are helpful to evaluate the CSF dynamics, CSF flow and flow disturbances. 3.5.1.4.2  Subdural Effusions and Empyema The common causes of prolonged fever in patients with bacterial meningitis are subdural effusions, drug fever, and concomitant arteritis or pneumonia (Roos et al. 1997). Extra-axial collections generally occur during the acute infection (Whiteman et al. 2002). If the subdural effusions are infected, subdural empyema will result. Up to 15% of all subdural effusions will become empyema. The cerebral convexities and interhemispheric fissures are common sites for empyema. Bilateral collection adjacent to the frontal and parietal lobes are usual in children having H. influenza meningitis. Most subdural effusions resolve spontaneously and do not require treatment, whereas empyemas need to be drained (Castillo 2004; Smith and Caldemeyer 1999). For this reason, accurate diagnosis of empyema is essential for proper treatment. Usually, empyema causes an associated inflammation of the adjacent brain, since the pia-arachnoid is a poor barrier (Osborn 1994). It is difficult to differentiate subdural collections from dilated subarachnoid spaces. If veins are seen coursing through collections, then the fluid is located in the subarachnoid space (Fig. 3.5.4). Subdural effusions can often be distinguished from empyema, which has higher protein content, and are often of greater signal intensity than CSF on T1-weighted images, FLAIR and proton density images (Fig. 3.5.4). A surrounding membrane that enhances intensely and uniformly following contrast administration is typically identified in empyema, which also tends to have internal septations (Fig. 3.5.4). The most helpful imaging technique to identify empyema is DWI. Effusions often demonstrate low intensity on DWI (Castillo 2004; Smith and Caldemeyer 1999). 3.5.1.4.3  Cranial Nerve Involvement Cranial nerve dysfunction occurs in 10–20% of patients, especially with infections caused by N. meningitidis. Cranial nerve VIII is the most commonly affected nerve followed, by VI, III, and II (Choi and Akins 2002). Some degree of sensorineural hearing loss is one of the most common sequelae of pyogenic meningitis occurring in up to 30% of patients. When deafness occurs, it is almost always permanent. Unfortunately, bilateral involvement tends to be the rule. T2-weighted images (particularly CISS FIESTA sequences, etc.) may show lack of the normal high signal intensity of the endo- and perilymph within the labyrinth (Castillo 2004). In the chronic phase, CT may show labyrinthitis ossificans involving the inner ear structures.

353

354

3  Brain, Head, and Neck

Fig. 3.5.4  Subdural empyema complicating meningitis. Coronal and axial T2-weighted images (a,b), and FLAIR image (c) show subdural empyema in the left frontal region. The fluid has higher signal intensity than CSF. Internal septations are also seen clearly. Right parietal parenchymal nodular lesion is also seen on T2-weighted images. Bilateral high signal along the cor-

tical sulci on FLAIR image is evident due to the corresponding meningitis. Post-contrast T1-weighted image (d) shows marked enhancement of neighboring meninges and also empyema mildly. There is also dilatation of subarachnoid space in the right frontoparietal region that has coursing vessels through it as a distinguishing feature from empyema

3.5.1.4.4  Cerebritis and Abscess

in hyperintense areas in sub-pial cortex and underlying white matter, which do not conform to an expected arterial distribution (Whiteman et al. 2002). Dural sinus thrombosis will manifest as a loss of the usual flow void in the affected sinus. The signal intensity within the sinus depends on the age of the clot and the nature of the hemoglobin present. Thrombosis and loss of normal venous patency can be demonstrated by MR venography particularly with gradient-echo (GRE) images.

Approximately 10% of patients with meningitis show parenchymal abnormalities in imaging studies. Cerebritis is seen in approximately 25% of cases. Cerebritis can evolve to frank abscess formation (Osborn 1994). This subject will be discussed later in this section. 3.5.1.4.5  Venous Thrombosis In children, dehydration associated with meningitis may contribute to the development of venous thrombosis and results in cerebral infarctions in up to 30% of affected patients (Castillo 2004). Thrombosis may be seen in the cortical venous sinuses and cortical veins, and results

3.5.1.4.6  Vasculopathy Meningitis may result in spasm or in direct infection/ inflammation of arterial walls (Kanamalla et al. 2000). Catheter angiography shows narrowing and spasm of

3.5  Intercranial Infections

the larger caliber arteries, irregularities of medium size arteries, and occlusion of distal branches (Smith and Caldemeyer 1999). Occlusion of small perforating arteries results in focal infarcts of the basal ganglia whereas spasm of anterior or middle cerebral arteries may lead to massive infarctions (Whiteman et al. 2002). MRA is helpful to demonstrate the vascular abnormalities. 3.5.2  Empyema 3.5.2.1  Epidural Empyema Cranial epidural empyema is defined as a suppurative infection of the epidural space, which is the space between the dura mater and the inner table of the skull. Cranial epidural empyema is a rare intracranial infection accounting for only 2–5% of all cases of cranial suppuration. It most often occurs as a complication of neurosurgery. In contrast to subdural empyema, epidural empyema is rare in young children, reported more commonly in people in the sixth decade of life (Ackerman and Traynelis 2002). They generally have fulminant clinical course and require prompt diagnosis and urgent treatment. Epidural empyema usually occurs as a result of infection caused by Staphylococcus aureus, Staphylococcus epidermidis, enteric gram-negative bacilli (especially E. coli), Pseudomonas species, Bacteroides species, and other anaerobes. Usually aerobic and microaerophilic streptococci, and rarely, Salmonella and Mucor species are responsible for infection that has spread from the paranasal sinuses, or infrequently secondary infection of epidural hematomas/collections (Gellin et al. 1997). In approximately 10% of cases, epidural empyema is associated with subdural empyema. At autopsy, 81% of patients with epidural empyema are found to have infections extending into the subdural space. Autopsy evidence of meningitis is present in 35% of patients with epidural empyema and evidence of brain abscess is present in 17% (Ackerman and Traynelis 2002). Epidural empyema often manifests more insidiously than subdural empyema. It enlarges too slowly to produce sudden major deficits, and most often patients present with focal findings. Neck stiffness, nausea, vomiting, lethargy, and hemiparesis do not develop until the infection has reached the subdural space. Usually, the patient presents with headache that is either diffuse or localized to one side with scalp tenderness. Seizures might as well be the first presenting symptom in some cases (Ackerman and Traynelis 2002; Gellin et al. 1997). Surgical exploration, decompression, and debridement, along with antibiotic therapy, are the mainstays of treatment of epidural empyema. Imaging studies in epidural empyema show hypoor isointense crescentic extra-axial fluid collections on T1-weighted images, which is mildly hyperintense to

CSF on T2-weighted images. The cerebral convexities and interhemispheric fissures are common sites. Epidural empyema does not necessarily respect the midline whereas a subdural empyema is usually confined to one hemisphere by the anatomy of the subdural space (Anslow 2004). Unlike subdural empyema, epidural empyema shows a hypointense rim representing displaced dura depicted at the interface between the lesion and the brain (Kastrup et al. 2005). A surrounding membrane that enhances strongly after contrast administration is typically identified. The enhancing membrane is usually thicker and more irregular than that seen in subdural empyema (Ackerman and Traynelis 2002). Cortical vein thrombosis with venous infarction, and cerebritis and abscess formation may ensue. 3.5.2.2  Subdural Empyema Subdural empyema (SDE) can occur at any age, but about two thirds of patients are aged between 10 and 40 years. The most common cause is extension from paranasal sinusitis, especially from the frontal and ethmoidal sinuses followed by otitis media, mastoiditis. Common causative organisms are anaerobes, aerobic streptococci, staphylococci, H. influenzae, Streptococcus pneumoniae, and other gram-negative bacilli (Ackerman and Traynelis 2001; Helfgott et al. 1997). Subdural empyema is a primarily intracranial infection located between the dura mater and the arachnoid mater. It has a tendency to spread rapidly through the subdural space until limited by specific boundaries (e.g., falx cerebri, tentorium cerebelli, base of the brain, foramen magnum). With progression, SDE has a tendency to behave like an expanding mass lesion with associated increased intracranial pressure (ICP) and cerebral intraparenchymal penetration. Cerebral edema and hydrocephalus may be present secondary to disruption of blood flow or CSF flow caused by the increased ICP. Cerebral infarction may be present from thrombosis of the cortical veins, cavernous sinuses or from septic venous thrombosis of contiguous veins in the area of the SDE. In infants and young children, SDE most often occurs as a complication of meningitis. In such cases, SDE should be differentiated from reactive subdural effusion (i.e., sterile collection of fluid due to increased efflux of intravascular fluids from increased capillary wall fenestrations into the subdural space). When making the diagnosis of SDE, one must pay close attention to the patient’s history. Any patient presenting with a history of sinusitis and neurologic symptoms should raise the index of suspicion for subdural empyema. Invariably, the patient with subdural empyema will have fever, but the clinical triad of sinusitis, fever, and neurologic deficit remains a classic presentation for subdural empyema, which should be considered as an

355

356

3  Brain, Head, and Neck

Fig. 3.5.5  Subdural empyema. Axial T1-weighted image (a), T2weighted image (b), and proton-weighted image (c) show right parafalcine subdural empyema with higher signal on all the images to the CSF. Thickened meninges along with falx are also seen. Post-contrast axial T1-weighted images (d) demonstrates enhancement of the outer and inner margins of empyema. e Corresponding trace DWI shows the empyema with bright signal due to the restricted water diffusion inside the empyema.

f On the ADC map, the lesion is seen isointense. Coronal proton-density- and T2-weighted images (g) demonstrate other example of parafalcine subdural empyema that has higher signal intensity than CSF does in subarachnoid spaces in a different patient. Post-contrast coronal T1-weighted image (h) depicts tentorial subdural empyemas with a ring enhancement in another patient

3.5  Intercranial Infections Fig. 3.5.5  (continued) Subdural empyema. Coronal T2-weighted (i) and post-contrast T1-weighted (j) images of the different patient show bilateral frontoparietal subdural empyemas. Right-sided collection is larger owing to greater mass effect, with marked peripheral meningeal enhancement

emergency. Because the infection involves the leptomeninges, patients may present with headache (usually frontal), meningeal signs, focal findings, seizures, and signs of increased intracranial pressure (Ackerman and Traynelis 2001; Helfgott et al. 1997). MRI demonstrates the subdural collections that have generally higher signal than CSF on T1-weighted images, FLAIR, and T2-weighted images (Figs. 3.5.5, 3.5.6). Thickened meninges can be seen as hypointense linear borders of the SDE. Post-contrast T1-weighted images typically demonstrate the collections with marked peripheral enhancement of the thickened meninges (Figs. 3.5.5, 3.5.6). The associated cortical edema, parenchymal infection, infarction, vascular occlusion, vasculopathies, or hydrocephalus, if present, show up well on MRI. Differential diagnosis of SDE includes chronic subdural hematoma, subdural hygroma, subdural effusion, and epidural empyema. Subdural hemorrhage has a varying signal according to the blood degradation products on both T1-weighted images and T2-weighted images. Epidural empyema may cross the midline, and is lentiform with displaced hypointense dura. 3.5.3  Cerebritis and Abscess 3.5.3.1  Pyogenic Cerebritis and Abscess Cerebritis represents the earliest phase of parenchymal infection of the brain. Cerebritis might progress to abscess with delayed diagnosis and inappropriate treatment. Brain abscess is a focal intracerebral infection that begins as a localized area of cerebritis and develops into a collection of pus surrounded by a well-vascularized capsule. The likely microorganisms responsible for brain abscesses depend on the predisposing condition for abscess formation. Cerebritis/brain abscesses can be bacterial, fungal, or parasitic. Streptococcus is the most common causative organism (30–50%). Less frequently, gram negative or-

ganisms such as E. coli, Pseudomonas, or H. influenza may cause brain abscesses. Certain bacteria, fungi, and protozoa have been observed with increasing frequency in immunocompromised patients (Zeidman 2002). Congenital heart disease is a common cause of sepsis in children, while bronchiectasis and bacterial endocarditis are the most common sources of infection in adults. Hematogenous abscesses tend to be multiple rather than solitary, to start at the gray–white matter interface or in white matter, and are seen most commonly in the distribution of the middle cerebral artery and rarely in basal ganglia (Anslow 2004). Solitary pyogenic abscesses of the brain stem are rare occurrences that constitute less than 1% of all intracranial abscesses (Suzer et al. 2005). Clinically, patients present with the cardinal features of fever, epilepsy, headache, nausea, and vomiting. Depending on the part of the brain affected, there may be focal neurological signs and depressed level of consciousness. It is very important to stress that, while the majority of patients present with one of the cardinal symptoms, occasionally patients may have none of these symptoms. Therefore, in case of a ring-enhancing mass in an afebrile patient, abscess still needs to be considered (Anslow 2004). Abscess formation progresses through four stages: early cerebritis, late cerebritis, early capsule, and late capsule. The cerebritis phase typically lasts from 10 to 14 days. Pathologically, cerebritis is described as a poorly demarcated area of parenchymal softening with scattered necrosis, edema, vascular congestion, and petechial hemorrhage without tissue liquefaction. The focus of cerebritis progresses to abscess when the central zone of necrosis becomes liquefied, better defined, and encircled by a collagen capsule, which itself is surrounded by a prominent zone of gliosis (Whiteman et al. 2002). Formation of a collagen capsule is the single most important response in limiting the spread of infection in the brain. Brain abscess is the only condition in which fibroblasts proliferate within the brain parenchyma. A well-developed layer of fibroblasts is more prominent on the cortical side than

357

358

3  Brain, Head, and Neck Fig. 3.5.6  Cerebritis. Axial T1-weighted image (a), T2weighted image (b), and FLAIR image (c) show an area of abnormal signal intensity in the right frontal lobe, with sulcal effacement and mass effect. Post-contrast axial T1-weighted image (d) demonstrates abnormal heterogenous focal parenchymal enhancement. The central area of the lesion shows mild enhancement then the periphery, which may be consistent with the late stage of the cerebritis. Associated mild pia-subarachnoid enhancement, consistent with meningitis, is also evident

on the ventricular side with reticulin deposition. A brain abscess combines features of inflammation and fibrosis. Histologically, the four layers of a mature brain abscess are as follows (from within): a necrotic center (pus), a layer of inflammatory cells, a collagen capsule, and a gliotic margin (Tunkel 2000; Nguyen et al. 2004). Pathological analysis of cerebritis and abscess can be correlated with CT and MRI appearances. In early cerebritis, an area of increased signal intensity on T2 and FLAIR sequences with or without ill-defined, patchy, irregular enhancement in the depth of a sulcus is seen on imaging studies (Fig. 3.5.6). As the focus of early cerebritis matures into the late cerebritis stage, the level of contrast enhancement is increased (Fig. 3.5.6). Gyral swelling and mass effect may also be observed on T1-weighted images. FLAIR imaging has been shown to be well suited for the detection of these early inflammatory changes (Fig. 3.5.6). The differentiation of cerebritis from infarction is important since both appear as increased signal intensity on T2-weighted images. DWI is useful in that issue. Acute infarction appears hyperintense on the DWI with a matching area of decreased signal on the appar-

ent diffusion coefficient (ADC) map (Whiteman et al. 2002). However, cerebritis shows high signal on DWI and ADC due to the T2 shine-through effect. After the second week, the abscess starts to form and is characterized by an incomplete rim of enhancement (Fig. 3.5.7). At this time, the capsule is incomplete (early capsule formation) and true pus is lacking. By the end of the third week, a mature abscess forms. In a typical abscess with central liquefactive necrosis, the center of the cavity is slightly hyperintense to CSF, whereas the surrounding edematous brain is slightly hypointense to normal brain parenchyma on T1-weighted images (Fig. 3.5.8). Signal intensities are quite variable depending on the TE chosen and the contents (protein composition, hemorrhage, etc.) and viscosity of the material in the central cavity on T2-weighted images (Fig. 3.5.8). With a long TE (>100), the differences in signal intensity between CSF, cavity fluid and edema usually diminish. Peripheral edema is nearly always present. Edema surrounding an abscess may be greater in volume than the abscess itself (Figs. 3.5.7, 3.5.8) (Whiteman et al. 2002). A mature abscess has a capsule that is isointense or slightly hyperintense to

3.5  Intercranial Infections

Fig. 3.5.7  Cerebral abscess with early capsule formation. Axial T2-weighted (a) image shows increased signal in the right frontal lobe with sulcal effacement. Post-contrast axial T1-weighted image (b) demonstrates a rim enhancement of the lesion with hypointense surrounding edema. Multi-voxel MRS (TE 35 ms) (c) of the ring enhancing lesion demonstrates the presence of glutamate (2.1 ppm), and amino acids (AA) (0.9 ppm) in the lesion. Due to the small size of the lesion, there is contamination of the spectra from normal surrounding parenchyma and prevents the demonstration of the decrease in NAA and creatine levels

white matter on T1-weighted images and is hypointense on T2-weighted images (Fig. 3.5.8). The signal properties of the capsule have been attributed to the collagen, hemorrhage, or paramagnetic free radicals, which are heterogeneously distributed in the periphery of the abscess (Whiteman et al. 2002). The abscess capsule shows ring enhancement pattern on post-contrast images (Figs. 3.5.7, 3.5.8). If there has not yet been frank necrosis in the abscess cavity, on delayed imaging, then the central area of the ring-enhancing mass will fill with contrast. However, the abscess with frank necrosis does not fill on delayed scanning (Fig. 3.5.7) (Smith and Caldemeyer 1999; Hunter and Morriss 2003). The ring is usually smooth and thin walled (approximately 5 mm) and is often thinner along the medial margin, a reflection of the better blood supply laterally (Fig. 3.5.8). Neoplasm and tumefactive demyelinating plaques often have the opposite pattern, with thicker enhancement along the ependymal (medial

border) (Hunter and Morriss 2003). Daughter abscesses may be seen as smaller enhancing rings, frequently adjacent to the medial margin (Whiteman et al. 2002). Immunocompromised patients and patients receiving steroid therapy or antibiotic treatment may show a significant reduction in the degree of ring enhancement associated with an abscess. Contrast enhancement of the capsule, edema, and mass effect resolve with successful treatment. Gliosis and/or calcification may remain. It is also important to evaluate for the complications such as herniations, intraventricular rupture, or subarachnoid extension. Ring-enhancing cavities on imaging studies are not specific for cerebral abscess. Primary brain tumor (highgrade astrocytoma), metastasis, infarction, resolving hematoma, thrombosed aneurysm, radiation necrosis, and other inflammatory conditions (e.g., demyelinating disease) may mimic brain abscesses (Whiteman et al. 2002).

359

360

3  Brain, Head, and Neck

Fig. 3.5.8  Mature cerebral abscess. Axial T1-weighted image (a), T2-weighted image (b), and FLAIR image (c) show a lesion in the left posterior parietal lobe, with marked surrounding edema and mass effect. The capsule is hyperintense on T1-weighted images whereas hypointense on the T2-weighted images. The central cavity of the lesion is hyperintense to CSF fluid due to the debris and proteinaceous material content on T1-weighted im-

ages. Post-contrast axial (d) and coronal T1-weighted images (e) demonstrate the regular, thin enhancing capsule. Corresponding trace DWI (f) shows that the center of the abscess is very bright. On the ADC map (g), the lesion is hypointense due to restricted water diffusion in the purulent abscess. Bright signal around the abscess is due to unrestricted vasogenic edema in the white matter

3.5  Intercranial Infections

DWI is helpful in differentiating abscess from necrotic tumor, which may be indistinguishable with conventional MR sequences. On DWI, an abscess demonstrates high signal intensity with a corresponding reduction in the ADC (Fig. 3.5.8). Increased signal on DWI of intracranial abscesses is related to water molecule movement restriction from marked cellularity and viscosity of pyogenic abscess. The pattern on DWI for an abscess is similar to the pattern for an acute cerebral infarction. However, calculated ADC values for abscess and acute infarct are substantially different (ADC values of abscesses are lower than those of infarcts (Kastrup et al. 2005; Whiteman et al. 2002; Nguyen et al. 2004; Leuthardt et al. 2002). MRS adds another layer of certainty by identifying the metabolites that are specific for pyogenic brain abscesses. MRS reveals end-products of bacterial breakdown and the degradations of proteolytic enzymes (lactate 1.3 ppm, acetate 1.92 ppm, succinate 2.4 ppm, and amino acid peaks [leucine, isoleucine–valine 0.9 ppm, alanine 1.5 ppm, and others]), which can be found in all abscesses, but not in necrotic/cystic tumors. Acetate and succinate are not seen in association with necrotic tumors and are therefore specific markers for pyogenic abscesses. However, these two resonances are not consistently identified in all abscess cavities. The cystic portion of tumors often reveals only a lactate resonance peak (Whiteman et al. 2002; Nguyen et al. 2004). MRS is also a potential tool for monitoring the treatment. Some authors have noted the disappearance of succinate, acetate, alanine, and amino acid peaks, with the persistence only of a lactate peak (Burtscher and Holtas 1999). A decrease in the lactate/amino acid ratio has also been shown to be helpful for the effectiveness of the treatment (Dev et al. 1998). 3.5.3.2  Tuberculosis Despite all the measures taken to combat them, TB infections including CNS TB infections are increasing worldwide. CNS TB can be seen in different forms, including meningitis, cerebritis, tuberculoma, abscess and TB encephalopathy, 3.5.3.2.1  Parenchymal Tuberculosis Parenchymal TB can manifest as tuberculous granuloma (tuberculoma), focal areas of cerebritis, cerebral abscess, or rarely as TB encephalopathy and may be isolated or associated with basal meningitis. Parenchymal TB is more common in human immunodeficiency virus (HIV)-infected patients. The most common parenchymal form of CNS TB is tuberculoma, which is multiple in 10–34% of cases. These lesions originate as a conglomerate of microgranulomata in an area of TB cerebritis that join to form a mature non-caseating tuberculoma. In most cases subse-

quent solid, central caseous necrosis will develop, which may eventually liquefy. Outside the capsule, there is parenchymal edema and astrocytic proliferation (Arbealáez et al. 2004; Bernaerts et al. 2003; Jinkins et al. 1995). The clinical manifestations are nonspecific. The common presenting symptoms and signs are headache, intracranial hypertension, seizures, focal neurological signs, and papilledema. Fever may be present. The CSF culture is usually negative. Neuroimaging is essential for the diagnosis. Tuberculomas are located at the corticomedullary junctions; most are infratentorial in children while most are supratentorial and affect the frontal and parietal lobes in adults (Arbealáez et al. 2004; Bernaerts et al. 2003). The radiological presentation depends on whether the tuberculoma is non-caseating, caseating with a solid center, or caseating with a liquid center (Fig. 3.5.9). The degree of surrounding edema is inversely proportional to the maturity of the lesion (Bernaerts et al. 2003; Jinkins et al. 1995; McGuinness 2000). One or more lesions with different characteristics and stages and with various sizes may be seen in the same patient (Fig. 3.5.9). Non-caseating tuberculomas are hypointense on T1-weighted images and usually hyperintense but occasionally hypointense on T2-weighted images to white matter (Fig. 3.5.9). On contrast-enhanced studies, homogenous enhancement is seen (Fig. 3.5.9). Solid caseating tuberculomas are iso- or hypointense on T1 and T2 sequences with an iso- or hyperintense rim (Fig. 3.5.9). Shortening of T2 signal may be due to the presence of free oxygen radicals in macrophages, fibrosis/gliosis, or macrophage infiltration (Wilson and Castillo 1994). Solid caseating tuberculomas show central heterogeneous enhancement with a capsule presenting a ring enhancing pattern (Fig. 3.5.9). This ring enhancement is usually of uniform thickness. A tuberculoma with central liquefaction of caseous material demonstrates a central hypointense signal on T1-weighted images and hyperintense signal on T2-weighted images, with variable hypointense signal of the rim (Fig. 3.5.9). An intense rim enhancement is seen on post-contrast T1-weighted images (Fig. 3.5.9). In this stage, lesions may be indistinguishable from true tuberculous or pyogenic abscess on MRI (Arbealáez et al. 2004; McGuinness 2000). In general, edema is less than that surrounding a pyogenic abscess of comparable size (Whiteman et al. 2002). Miliary CNS TB is characterized by multiple 2- to 30-mm tuberculomas. Multiple tiny hyperintense/hypointense lesions are seen on T2-weighted images. The contrast-enhanced T1-weighted images may show numerous enhancing foci that are hyperintense on T2-weighted images (Bernaerts et al. 2003; Gee et al. 1992). MT-SE imaging has been shown to improve assessment of the disease load in CNS TB owing to better lesion detection and is also helpful for treatment monitoring (the MT ratio in white matter reverts to normal with treatment) (Gupta et al. 1999).

361

362

3  Brain, Head, and Neck

3.5  Intercranial Infections 9 Fig. 3.5.9  Parenchymal tuberculoma. Axial FLAIR (a) and post-contrast T1-weighted (b) images show enhancing nodular lesion in right frontal lobe at the gray-white matter interface representing a small non-caseating granuloma. Axial T2weighted images (c), FLAIR (d) and post-contrast T1-weighted (e) images of the same patient from different level demonstrate two ring enhancing solid caseating granulomas in the posterior regions of right frontal and left parietal lobes with surrounding vasogenic edema. The center of the lesions are markedly hypointense on T2-weighted images corresponding to caseous necrosis, and do not enhance. The wall of the lesions have high T2 signal, therefore cannot be distinguished from perilesional edema. DWI shows high signal on trace images (f) and low signal on

the corresponding ADC map (g) reflecting mild restriction. Although Cho:NAA ratio seems to be preserved on multi-voxel MRS (TE 35 ms) (h), the lactate elevation reflects the central necrosis. In a different case, axial post-contrast T1-weighted image (i) and T2-weighted image (j) show a ring enhancing lesion with surrounding edema in the right lentiform nucleus, representing caseating granuloma with central liquefaction. Note that the centrum of the lesion is hyperintense on T2-weighted images, in contrast to solid caseating granuloma. Corresponding trace DWI (k) shows that the center of the lesion is very bright. The lesion is hypointense due to restricted water diffusion on the ADC map. There are also two non-caseating granulomas anterior to that lesion (Courtesy of N. Bulakbasi, MD)

MR spectroscopy may show prominent lipid peaks (at 0.9, 1.3, 2.0, 2.8, and 3.7 ppm) due to high lipid content in the caseous material (Whiteman et al. 2002; Arbealáez et al. 2004; Gupta et al. 1996). The presence of serine has been accepted as a distinct feature of tuberculomas. Phenolic lipids also represent biochemical fingerprints of Mycobacterium tuberculosis in a granuloma. Full resolution of CNS tuberculoma requires months to years of medical therapy. Treatment monitoring may show a decrease of edema and of the size of the lesion, and also contrast enhancement. Lesions may disappear, calcify or results in gliosis, encephalomalacia. Healed tuberculomas may calcify in up to 23% of cases (Whiteman et al. 2002). Atrophy is frequently a long-term sequelae of CNS TB. Occasionally, newly developing or enlarging intracranial tuberculomas may be observed despite antibiotic therapy. Patients who develop signs of raised intracranial pressure or new neurological signs should have urgent neuroimaging to exclude the development of new lesions or enlargement of existing granulomas (Bernaerts et al. 2003). The differential diagnosis for parenchymal TB includes mainly other granulomatous processes (e.g., sarcoidosis, fungal lesions, parasitic disease such as cysticercosis and toxoplasmosis), multicentric primary neoplasm, as well as metastatic neoplasms. T2 shortening, which is not found in most other space-occupying lesions, may be a useful sign (Bernaerts et al. 2003).

larger than tuberculomas and pyogenic abscesses (often >3 cm), and have a more accelerated clinical course. On imaging studies, they show multiloculation, and varying types of walls, from thin and smooth to irregular and thick with marked enhancement on post-contrast T1-weighted images (Fig. 3.5.10). TB abscesses may mimic all the ring-like lesions.

3.5.3.2.2  Tuberculous Abscess Tuberculous abscesses occur in less than 10% of patients with CNS TB, and are more common in the elderly or in immunocompromised patients (Arbealáez et al. 2004). In contrast to tuberculomas, in which solid caseating is often seen, tuberculous abscesses are formed by semiliquid material that is rich in tubercle bacilli (Whiteman et al. 2002). The abscesses are similar to pyogenic ones, although they are often multiloculated. TB abscesses are

3.5.3.2.3  Tuberculous Encephalopathy Tuberculous encephalopathy is typically found in young children and presents diffuse compromise of white matter with no masses or meningeal involvement. Extensive damage to the white matter seems to be due to a type IV hypersensitivity mechanism including cell-mediated immunity to tuberculoprotein. Neuroimaging studies show uni- or bilateral cerebral edema, perivascular demyelination or hemorrhagic leukoencephalopathy. Death often occurs within 1 to 2 months of onset of neurological illness despite antituberculous medical treatment (Arbealáez et al. 2004; Bernaerts et al. 2003). 3.5.3.3  Fungal Infections Fungal infections of the CNS are far less common than bacterial infections are, although their clinical presentations are strikingly similar. Fungi that are involved in CNS infections are either pathogenic fungi (Coccidioidomyces, Blastomyces, Histoplasma, etc.), which can infect any host, or saprophytic fungi (Candida, Mucor, Aspergillus, etc.), which generally infect immunocompromised hosts. There are an increased number of CNS fungal infections in immunocompromised patients. The pathophysiology of mycotic lesions in the CNS varies with the type of fungal form. Fungi that grow in infected tissues as yeast cells (Cryptococcus, Histoplasma) spread hematogenously and access meningeal circulation, causing leptomeningitis or less commonly granulomas. Fungi that grow in infected

363

364

3  Brain, Head, and Neck

tissue as hyphae only (Aspergillus, Mucor) or as pseudohyphea (Candida) tend to involve parenchyma rather than meninges, because larger hyphal forms have limited access to the meningeal circulation. Generally, granulomatous reactions occur with variable degrees of suppuration. Hyphal forms are capable of invading arteries, leading to infarction and cerebritis. Pseudohyphal forms are larger than individual cells but smaller than true hyphae. Thus, they produce scattered parenchymal granulomatous microabscesses secondary to small vessel occlusion (Whiteman et al. 2002; Perdiago et al. 2004). Definitive diagnosis of fungal infection of the CNS is made by isolating the fungus from the tissue specimen or CSF with the exception of Cryptococcus neoformans, which typically does not grow in culture from CSF specimens. Urgent diagnostic approaches are essential because of the high morbidity and mortality rates from these infections. Neuroimaging with contrast agents helps identify the infection but rarely establishes its cause (Perdiago et al. 2004). 3.5.3.3.1  Cryptococcosis

Fig. 3.5.10  Tuberculous abscess. Axial T2-weighted image (a) and post-contrast T1-weighted image (b) demonstrate multiloculated enhancing lesion in the left frontal lobe with extensive surrounding edema and mass effect. Note the large size of the lesion (Courtesy of B. Diren, MD)

Cryptococcosis is the most common fungal infection of the CNS. It is rare in immunocompetent individuals and is clinically evident in 6% to 7% of AIDS patients. CNS cryptococcosis occurs most commonly in young or middle-aged adults, with male dominancy. Headache is the most common symptom. Cryptococcal CNS infections may show different manifestations as most commonly meningitis, parenchymal mass lesions (also known as cryptococcomas), meningoencephalitis, pseudocyst in dilated Virchow-Robin spaces, plexitis, and a mixed pattern. The pseudocysts, dilated perivascular, and/or Virchow-Robin spaces are distended with fungus and gelatinous mucoid material, most evident in basal ganglia, and midbrain and may be bilateral. Dilated perivascular spaces (Virchow-Robin spaces) and cystic lesions in HIV-positive patients suggest cryptococcal infection of the CNS. MRI has limited sensitivity in the detection of meningitis from cryptococcal infection. Meningeal enhancement may be seen in immunocompetent patients whereas immunocompromised/immunosuppressed patients rarely demonstrate meningeal enhancement. Hydrocephalus is not rare. Parenchymal lesions have higher signal intensity on T2-weighted images and show marked enhancement. Gelatinous pseudocysts and dilated perivascular spaces are isointense to CSF on T1- and T2-weighted images and do not demonstrate edema. A hyperintense lesion on T2-weighted images with marked contrast enhancement is typical for plexitis. There are reports for the use of MT imaging to differentiate cryptococcal infections from other infections (Gupta et al. 1999). No specific finding has been stated for MRS of CNS cryptococcosis.

3.5  Intercranial Infections

3.5.3.3.2  Aspergillosis

3.5.3.3.4  Mucormycosis

CNS aspergillosis is one of the common fungal infections and mostly caused by Aspergillus fumigatus and Aspergillus flavus. Aspergillosis involves the CNS by direct extension from nasal cavity/paranasal sinus infection or, more commonly, by hematogenous dissemination. A broad spectrum of immunosuppressed patients, including not only those with hematologic malignancies and transplant recipients, but also patients receiving steroids or broadspectrum antibiotics are at risk for aspergillosis. AIDS patients may develop invasive cerebral aspergillosis, and the majority of them will have pulmonary disease. Ten to 25% of immunocompromised patients with pulmonary aspergillosis will develop CNS involvement (Perdiago et al. 2004). Cerebral aspergillosis carries a mortality rate close to 100%, and diagnosis is often made after death. In hematogenous spread, Aspergillus hyphae lodge in the blood vessel, grow, and cause occlusion, producing hemorrhagic infarction. Initial sterile infarct converts to septic infarction with associated cerebritis and abscess formation after the erosion of the vessel wall. Vascular invasion may result in a mycotic aneurysm and fungal vasculitis. Heterogeneous high signal is seen with low signal due to hemorrhage on MRI (Whiteman et al. 2002; Perdiago et al. 2004). In the aggressive forms, no contrast enhancement is seen; this lack of contrast enhancement could be explained by rapid progression and/or immunocompromization. If the organism is isolated by the host defense system, granuloma or abscess formation can be seen with a nodular or ring enhancement (Fig. 3.5.11). The ring enhancing wall is often thicker and more irregular than the pyogenic abscess is (Erdogan et al. 2002). Patients with aspergillosis may also show multiple nonspecific enhancing brain lesions. Evidence of vasculitis may be seen on MRA. The presence of hemorrhage is often a clue to the underlying diagnosis of aspergillosis (Whiteman et al. 2002; Perdiago et al. 2004). Based on imaging studies, the differential diagnosis of aspergillosis granuloma includes abscess, tuberculoma, meningioma, and glioma (Dubey et al. 2005).

Mucormycosis is an aggressive disease with a high morbidity (up to 90%) that is most commonly caused by genus Rhizopus. The clinical presentations induced by these organisms are virtually identical. While the rhinocerebral form is most common, hematogenous spread or direct invasion can cause an acute, fulminant CNS infection. The only definite way of diagnosing mucormycosis is by detection of hyphae in tissue specimens or by growth of the fungus in culture. Mucor tend to spread along perivascular or perineural channels through cribriform plate into the frontal lobe or through orbital apex into the cavernous sinus. Intracranial Mucor can also form a fungal abscess or invade blood vessels and cause cerebral infarction (Osborn 1994). Basal ganglia are frequently involved. Lesions show nonspecific hypointense signal on T1-weighted images, hyperintense signal on T2-weighted images with contrast enhancement.

3.5.3.3.3  Candidiasis The brain is the second most common organ involved with Candida infection, following the kidneys. Candida infections include meningitis, meningoencephalitis, vasculitis with mycotic aneurysm, microabscesses with granuloma formation, frank abscess formation, and ependymitis. On T2-weighted images, the Candida abscess appears as a well-demarcated area with hypointense signal surrounded by edema, called a “target appearance” (Whiteman et al. 2002; Perdiago et al. 2004). Ring enhancement is seen on post-contrast images.

3.5.3.4  Parasitic Infections Many parasites can cause CNS infections. A wider variety and number of exotic parasites are being seen with the increasing mobility of populations. Neurocysticercosis and echinococcosis are common CNS parasitic infections. 3.5.3.4.1  Neurocysticercosis Neurocysticercosis is the most common CNS parasitic infection worldwide. Cysticercosis is caused by the parasite Taenia solium. Hematogenous spread to neural, muscular, and, ocular tissue occurs. Intracranially, the oncospheres, the primary larval form of the T. solium may burrow into brain parenchyma, meninges, ependyma, and choroid plexus. Oncospheres develop into a secondary larval form called cysticerci. The parenchymal type (Cysticercus cellulosae) is the most common one among the subarachnoid, intraventricular (Cysticercus racemose), and mixed types. The corticomedullary junction is the primary location of the parenchymal form. Intraventricular cysticercosis can be seen in 20–50% of cases, and the fourth ventricle is the most common site. More than one anatomical site is often involved. The mature parenchymal cysts are present 2–3 months after the ingestion of ova. This stage is generally asymptomatic but may result in seizures. Each cyst measures 3–18 mm in diameter and contains a scolex. A dying cyst causes an intense inflammatory reaction, leading to seizures or some focal neurologic signs (Osborn 1994; Whiteman et al. 2002). Intraparenchymal lesions elicit a higher immune response from the host and are easily

365

366

3  Brain, Head, and Neck Fig. 3.5.11  Aspergillosis. Postcontrast T1-weighted image (a) in the axial plane shows a ring enhancing lesion at the left middle frontal lobe. Bilateral multiple millimetric-enhancing foci at the corticomedullary junction is also seen. DWI (b) at the same level demonstrates a target lesion, with high signal intensities reflecting diffusion restriction in the central part of the lesion, consistent with abscess formation. Multi-voxel MRS (c) (TE 35 ms) of the ring enhancing lesion demonstrates the presence of glutamate (2.1 ppm), lactate (1.3 ppm) and amino acids (AA) (0.9 ppm) at the central portion of the lesion, reflecting end products of fungal metabolism. Due to the small size of the lesion, there is contamination of the spectra from normal surrounding parenchyma and prevents the demonstration of the decrease in NAA and creatine levels. 20-day-follow-up post-contrast T1-weighted image (d) in the axial plane during the antifungal therapy shows decrease in size of the ring-enhancing lesion in the left frontal lobe. Please note the complete resolution of bilateral multiple enhancing foci at the corticomedullary junction. Multi-voxel MRS (TE 35 ms) (e) of the ring enhancing lesion demonstrates the presence of lactate (1.3 ppm) with near complete regression of the amino acid peak (0.9 ppm), reflecting a good response to antifungal therapy

3.5  Intercranial Infections

reached by antiparasitic medications, and thus these patients have a good prognosis (Castillo 2004b). Parenchymal cysticercosis has been classified into four stages. 1 Vesicular stage: Cysticercosis consists of a thin capsule that surrounds the viable larva and its fluid-containing bladder. On imaging studies, the larva appears as a cyst containing a small, mural nodule that represents its scolex located usually near the gray–white matter junction. Edema and contrast enhancement are rare, but the scolex may show some enhancement. The mural nodule is best seen on proton density or FLAIR images (Osborn 1994; Castillo 2004b). 2 Colloidal vesicular stage: When the larva dies, the fluid becomes turbid, and its capsule thickens. Edema and contrast enhancement are seen in this stage Degenerating cysts may be hyperintense on both T1- and T2-weighted images due to their contents. Ring-like enhancement is seen in two third of cases. 3 Granular nodular stage: As the parasite dies, the cyst begins to collapse, and surrounding edema develops, the capsule thickens, and the scolex calcifies. There is intense contrast enhancement of the cyst walls. The cyst is isointense compared to brain on T1-weighted images. A target or bull’s-eye appearance is seen with the calcified scolex in the center of the mass (Osborn 1994; Castillo 2004b). 4 Nodular calcified stage: In this stage the granulomatous lesion is completely mineralized. There is no active immune response from the host. A small, calcified lesion measuring between 2 and 10 mm is typical. Usually, there is no mass effect or contrast enhancement. However, contrast enhancement or perilesional edema is also identified in some lesions, probably related to recent seizure activity originating from chronic regions (Osborn 1994; Castillo 2004b). Subarachnoid–intraventricular cysticercosis appears as a large cyst or multiple cysts (5–9 cm) with a “bunch-of-grapes” appearance (racemose form). The scolex is absent in this form. The cerebellopontine angle and suprasellar cisterns are the most common locations. These cysts are difficult to identify because they are isointense to CSF, and subarachnoid lesions do not enhance. The intraventricular cyst may enhance. Hydrocephalus is often present especially at the fourth ventricular cysticercosis. High resolution higher T2weighted sequences (CISS, etc.) may be more helpful than spin-echo sequences are (Castillo 2004b). The lesions may be difficult to distinguish (particlarly if solitary) during the vesicular stage from Echinococcus (usually a large, solitary cyst in the region of the distribution of the middle cerebral artery), sparganosis (seen in patients from Southeast Asia, multiple cystic lesions), coenurosis (a very rare disease caused by the cestode M. multiceps), cystic metastases, and multiple abscesses. During the colloidal and granular stages,

both solitary and multiple cysts may be difficult to distinguish from metastases. Since the MRS findings in cysticercosis and pyogenic abscesses may be similar, DWI may be helpful to differentiate between them. Pyogenic abscesses are nearly always bright on DWI, while cysticercal cysts are darker (Castillo 2004b). 3.5.3.4.2  Echinococcosis (Hydatid Disease) The pathogen responsible for CNS echinococcosis/hydatid disease is Echinococcus, specifically Echinococcus granulosus and Echinococcus multilocularis. The larval stage is known as the hydatid cyst. Clinical presentation is due to mass effect by the cyst. On imaging studies, a single, thin-walled spherical cyst with CSF intensity is typical. The cyst may be unilocular, unilocular with septations, unilocular with daughter cysts or multilocular (Fig. 3.5.12). The wall of the hydatid cyst has three layers. The first, inner layer of the cyst is a thick, nucleated membrane called a germinative membrane, which is the active layer, rich in collagen, granular in appearance, syncytial (endocyst). The endocyst may be seen with hyperintense signal on T1-weighted images and PDWI, and with hypointense signal on T2-weighted images (Fig. 3.5.12). The second layer of the cyst is secreted by the endocyst and is made up of an acellular, laminated, keratinous, mucopolysaccharide membrane of variable thickness (ectocyst) (1 mm). The third layer, which is caused by host response, is made up of granulation tissue of variable intensity with fibroblasts, giant cells, and eosinophils surrounding the hydatid cyst (pericyst). Ectocyst and pericyst may show hyperintense signal on PD and isointense signal on T1- and T2-weighted images (Fig. 3.5.12). They usually do not show edema or enhancement if there is no obvious interaction with the host. However, degenerated cysts may show edema and enhancement (Fig. 3.5.12). Calcification may be seen in the cyst wall. The free brood capsules, free protoscolices, scolices, and daughter cysts that lie within the fluid of the cyst are collectively referred to as hydatid sand, an important feature both radiologically and pathologically (Fig. 3.5.12). MRS shows dominant resonances of acetate (1.92 ppm), succinate (2.4 ppm), lactate (1.33 ppm) and glycine (3.56 ppm) and small contributions from alanine and isoleucine, leucine, and valine (Kohli et al. 1995). The imaging features of echinococcosis by E. multilocularis mimic those of tumors and abscesses. A multilocular mass lesion with heterogeneous hypointense signal on T1-weighted images and hyperintense signal on T2-weighted images with peripheral edema and heterogeneous enhancement is seen.

367

368

3  Brain, Head, and Neck

Fig. 3.5.12  Hydatid disease. Axial T1-weighted image (a), T2-weighted image (b), and FLAIR image (c) show a unilocular cyst that is isointense to CSF, in the left posterior parietal lobe. The lesion has only local mass effect without any perilesional vasogenic edema. Post-contrast T1-weighted image (d) in the axial plane shows no enhancement of the wall of the cyst. In a different case, a unilocular cystic lesion located in the right

posterior parietal lobe is seen to be isointense on (e) DWI and (f) the ADC map, demonstrating that there is no restricted diffusion inside the cyst due to its clear content. Pre-contrast sagittal T1-weighted image of another patient (g) show multiple cysts that are isointense to CSF. (Courtesy of A. Alkan, MD) Axial T1-weighted image (h) and T2-weighted image (i) see next page

3.5  Intercranial Infections

Fig. 3.5.12  (continued) (h) and T2-weighted image (i) of another case show a unilocular cyst with multiple daughter cysts in the left frontal lobe. The lesion has local mass effect, without perilesional edema. The rim of the cyst is hypointense on T2-

weighted images. Pre-, and post-contrast coronal T1-weighted images (j) of another case demonstrate an unilocular cyst with detached membrane. There is no enhancement of the wall of the cyst. In a different case, T2-weighted image (k–l) see next page

369

370

3  Brain, Head, and Neck

Fig. 3.5.12  (continued) (k) and proton-weighted image (l) also show the lesion with detached membrane and its associated slight perilesional edema. The wall of the cyst has higher signal compared with other uncomplicated cyst in previous examples

3.5.3.4.3  Amebiasis Several amebic microorganisms may involve the CNS. Entamoeba histolytica, Naegleria fowleri, and Acanthamoeba are the most frequently encountered microorganisms. E. histolytica is the etiologic agent in cerebral amebiasis, and a well-known human intestinal pathogen. It is a relatively common infection in the endemic parts of the world. The infection is generally confined to the large bowel. Involvement of CNS is more unusual and the incidence was reported to be between 0.7 and 8 % in various studies. Most patients are between second and fourth decade, with male dominancy. CNS involvement of E. histolytica is almost always fatal with current treatment. The infection tends to be located at the gray–white matter junction, making bloodstream spread highly probable. Pathologically, the infection is a focal destructive process due to the histotoxic and phagocytic nature of the pathogen. Although the lesions are called abscesses, the histologic properties are relatively different from classical bacterial abscesses of the brain (Celik et al. 2005). Instead of a pus-filled central cavity and a well-formed fibrous capsule, the central portion of these lesions contains necrotic tissue, and the capsule is not a prominent finding and clearly delineated (Celik et al. 2005). Accurate radiological and clinical data supplied to the pathologist is of utmost importance for correct diagnosis because it is

not always easy to discern amoebae in the surgical specimens. It is also surprising that the inflammatory reaction of the brain is relatively low compared to other infectious diseases of the brain (Celik et al. 2005). Single or multiple ring enhancing lesions are seen on MRI (Fig. 3.5.13). Whereas the typical brain abscess has a homogenous, bright signal on DWI associated with decreased ADC values, the central portion of an entamoebal brain abscess is dark on DWI. This darkness may be attributed to necrosis of the central portion of the lesion, which differs from the central pus accumulation of brain abscesses (Fig. 3.5.13). High ADC values are also consistent with central necrosis, reflecting the histopathological property of the entamoebal brain abscess (Celik et al. 2005). MRS may be useful for differentiating abscesses from neoplastic disorders. 3.5.3.4.4  Toxoplasmosis Toxoplasma encephalitis is caused by the intracellular protozoan Toxoplasma gondii and has a worldwide distribution. Toxoplasma encephalitis represents a recurrence of latent infection in most cases. It is the most common opportunistic brain infection seen in AIDS patients (in 10– 34% of adult AIDS autopsies) presenting with altered mental status, fever, seizure, and/or focal neurological deficit.

3.5  Intercranial Infections

Fig. 3.5.13  Entamoebal brain abscess. Axial T2-weighted image (a) shows a left frontal hyperintense lesion containing scattered small hypointense foci indicating hemorrhage. Post-contrast axial T1-weighted image (b) reveals left frontal mass lesion having a thick, rim-like contrast enhancement with relatively preserved gyral pattern. ADC map (c) demonstrates increased diffusion at the lesion site particularly in the central necrotic portion. Relative cerebral blood volume (rCBV) map of perfusion MRI (d) generated from time–signal intensity curve, with red areas representing high and blue areas representing low perfusion, shows decreased rCBV at the involved sites at both hemispheres. MRS (e) (TE 144 ms) demonstrates the presence of choline (3.2 ppm), lipid (1.2 ppm), and negative lactate (1.3 ppm) in the central portion of the lesion

371

372

3  Brain, Head, and Neck

Fig. 3.5.14  Toxoplasmosis. Axial T1-weighted (a) and coronal proton-weighted and T2-weighted images (b) show cystic lesions in the right temporal lobe. The lesions are isointense to CSF. Coronal post-contrast T1-weighted images (c) demonstrate nodular enhancement of the upper two lesions, and peripheral enhancement in the lower lesion

3.5  Intercranial Infections

Pathologically, toxoplasmosis lesions are characterized by three zones. The innermost zone consists of coagulative necrosis with few organisms. The intermediate zone is hypervascular and contains numerous inflammatory cells mixed with tachyzoites and encysted organisms. The peripheral zone is mostly composed of encysted organisms. Vasogenic edema surrounds the mass. Toxoplasma lesions do not have a capsule (Osborn 1994). The radiological appearance correlates well with the pathological findings, with the central hypointensity on T2-weighted images corresponding to the region of avascular coagulative necrosis. The enhancing ring corresponds to the region of intense inflammation, and the peripheral zone may appear as edema. Toxoplasma is commonly seen in basal ganglia, and lesions are usually multiple. On T2-weighted images, active lesions show variable intensity, with an appearance of hyperintensity or hypointensity called a “target sign” that is rather nonspecific. Post-gadolinium studies reveal ring or nodular enhancement (Fig. 3.5.14). Hemorrhage in lesions is rare (Whiteman et al. 2002). With treatment, the size and the number of lesions decrease. Treated lesions may appear normal, or may demonstrate calcification. The mineralized lesions may have foci of increased or decreased signal on MRI. In AIDS patients, all lesions are needed to be followed to resolution, because multiple pathologies may coexist in the same patient (Whiteman et al. 2002). Primary Toxoplasma encephalitis is often difficult to distinguish from CNS lymphoma. Toxoplasma encephalitis is often multicentric, whereas primary CNS lymphoma is usually solitary. Thallium-201 brain SPECT will be positive for CNS lymphoma in an HIV-seropositive patient, whereas negative studies are presumed to be due to an infectious agent. MRS in toxoplasmosis reveals elevated lipid and lactate peaks, consistent with an anaerobic acellular environment within an abscess. In contrast, primary CNS lymphoma shows a mild to moderate increase in lactate and lipids, a markedly elevated choline peak, and preservation of some normal metabolites with variably decreased levels of NAA and Cr (Whiteman et al. 2002). Toxoplasmosis is also a common cause of congenital infection. Clinically significant abnormalities occur when the fetus is infected prior to 26 weeks of gestational age. Neuroimaging studies reveal hydrocephalus and intracranial calcifications (more diffuse than those seen in cytomegalovirus [CMV]), most commonly in the cortex and basal ganglia (Osborn 1994; Whiteman et al. 2002). 3.5.4  Encephalitis Encephalitis refers to any diffuse inflammatory process involving the brain that can be caused by a broad spectrum of agents. The most common encephalitides are viral. The resistance of the brain to viral attack is believed

to due to blood–brain barrier, which acts as a mechanical, anatomical, and physiological filter. The spectrum of brain involvement and the outcome of the disease are dependent on the specific pathogen, the immunological state of the host, and a range of environmental factors. In adulthood, the virus most frequently responsible for acute viral encephalitis is herpes simplex virus type 1 (HSV-1), and in the neonatal period it is herpes simplex virus type 2 (HSV-2). Viral encephalitides in immunocompromised patients are usually caused by HIV and CMV. Typical non-viral causes in these patients include T. gondii, A. fumigatus, and Listeria monocytogenes (Osborn 1994). 3.5.4.1  Herpes Encephalitis Herpes viruses are a large group of double-stranded DNA viruses. HSV-1 is the causative agent in 95% of herpetic encephalitis and the most common cause of fatal sporadic encephalitis. HSV-2 virus may result in eye/ skin/mouth disease (43%), disseminated disease (23%), and encephalitis (34%) (Buff et al. 1994). The virus gains access to neuronal tissue either by diffusion through the blood–brain barrier or by infecting endothelial cells of intracranial blood vessels. In adults, the infection usually arises in those individuals with preexisting antibodies and thus represents viral reactivation (latent in the trigeminal Gasserian ganglion in 50% of cases) (Osborn 1994; Whiteman et al. 2002). For HSV-2, the primary route of infection is via the maternal birth canal, either ascending or by contact during the delivery. In non-AIDS patients, HSV-1 may result in necrotizing encephalitis of the temporal lobes and the orbital surfaces of frontal lobes. The insular cortex, cerebral convexity, and the posterior occipital cortex and, later in the course of the disease, the cingulate gyrus may become involved. Disease is usually bilateral with sparing of the basal ganglia (Whiteman et al. 2002). On rare occasions, the infection may be restricted to the brainstem (Jereb et al. 2005). However, neonatal HSV-2 is a diffuse, nonfocal infection. The gross appearance of brain in adults with herpes encephalitis initially shows acute inflammation, congestion and/or hemorrhage, and softening. The meninges overlying the infected area may appear clouded and congested. After approximately 2 weeks, these changes proceed to frank necrosis and liquefaction. Vascular changes that have been reported in the area of infection include hemorrhagic necrosis and perivascular cuffing. Glial nodules are also evident after the second week. The microscopic appearance becomes dominated by the evidence of necrosis and eventually, inflammation. Although found in 50% of patients, the presence of intranuclear inclusions (Cowdry type A inclusions) supports the diagnosis of viral infection (Whiteley 1997).

373

374

3  Brain, Head, and Neck

Fig. 3.5.15  HSV-1 encephalitis. Axial (a) T2-weighted and (b) FLAIR images show diffuse hyperintense signal involving both gray and white matter primarily in the right temporal lobe, but also affecting left temporal lobe in a lesser degree.

Post-contrast sagittal (c) and axial T1-weighted images (d) reveal gyriform enhancement in the right and left temporal lobes. (e–h) see next page

Patients with HSV-1 encephalitis may present fever, headache, nausea and vomiting, and neck stiffness with altered mental status. No specific symptoms are seen for the HSV-2 encephalitis, and they generally begin at about the second or third weeks of life (Castillo and Thurner 2004). The mainstay in diagnosis is the detection of HSV DNA in the CSF by using PCR. However,

this laboratory test might produce false-negative results (Maschke et al. 2004). MRI examination is more sensitive and demonstrates the early findings of edematous changes of HSV-1 encephalitis. These are increased signal seen in particularly temporal and inferior frontal lobes on FLAIR and on T2-weighted images (Fig. 3.5.15). This hyperintense signal

3.5  Intercranial Infections

Fig. 3.5.15  (continued) (e) DWI and ADC map (f) demonstrate high signal in the affected region compared with the contralateral unaffected area, consistent with vasogenic edema. MRS (g) shows elevated choline, low NAA due to loss of structural neuronal activity. On DTI (h), due to the intense edema fiber tracts cannot be readily appreciated on the right temporal lobe

375

376

3  Brain, Head, and Neck

may be seen as early as 48 h after onset of signs and symptoms and may involve both cortex and white matter (Fig. 3.5.15). Areas of involvement frequently coalesce, and there is associated mass effect. The signal abnormalities often extend into the contralateral temporal lobe, insular cortex, and cingulate gyri and spare the basal ganglia. Involvement of cingulate gyrus and contralateral temporal lobe is highly suggestive of HSV-1 encephalitis (Steiner et al. 2005). Enhancement is often absent in the early stages, but gyriform enhancement may be seen with progression (Fig. 3.5.15). Although HSV-1 results in necrotizing and hemorrhagic encephalitis, frank blood is not commonly seen on imaging studies. Acute hemorrhage on MRI may appear as a focus of moderate to marked hypointensity on gradient-echo images or on T2-weighted images. The late sequelae of HSV-1 are atrophy, encephalomalacia, and dystrophic calcifications (Jordan and Enzmann 1991). For HSV-2 encephalitis, MRI shows patchy areas of increased T2 signal particularly in periventricular white matter. After a few days, these areas show contrast enhancement. Neonatal infections do not have the temporal lobe predilection seen in HSV-1 encephalitis, and are less frequently hemorrhagic. Encephalomalacia, which develops rapidly, may be visible within the first 2–3 weeks. Dystrophic calcification occurs in the basal ganglia and periventricular white matter as a late sequela. In the acute stages of herpes encephalitis, dedicated sequences, such as FLAIR and DWI, may show more subtle alterations (Hunter and Morriss 2003; Castillo and Thurner 2004; Andreula 2004). In patients with herpes encephalitis, two distinct types of findings on DWI (b = 1,000s/mm2) and ADC maps were noted: lesions similar to cytotoxic edema, and lesions similar to vasogenic edema. The patients with the former type of lesions have fulminating disease and are in a severe clinical condition. Those with the latter represent early cases, and they are in fairly good clinical condition and are likely to have a good outcome with prompt therapy. DWI appears to be a promising sequence for monitoring the changes in the brain tissue in herpes encephalitis, and in other infections as well with respect to restriction of movement (cytotoxic edema) or relatively high-motion (vasogenic edema) of water molecules, providing data on the severity of the disease (Sener 2001). MR spectroscopy shows elevated choline, low NAA, and the presence of lipids and lactate, reflecting necrosis. Due to loss of structural or functional neuronal activity, the NAA:Cr ratio may be decreased. Occasionally, HSV-1 infects the eighth cranial nerve, resulting in acute sensorineural hearing loss. In these patients, MRI shows marked contrast enhancement of the intracanalicular portion of the eighth nerve (Castillo and Thurner 2004). Differential diagnosis of herpes encephalitis includes infiltrative neoplasms, ischemia, and post-infection encephalitis.

3.5.4.2  HIV Encephalitis HIV belongs to the retrovirus family of RNA viruses, with a reverse transcriptase that transcribes viral RNA into provirus DNA, which is then integrated into the host-cell genome (Andreuala 2004). HIV infects mainly the brain in all patients and its most common clinical manifestation is encephalopathy (28% of all AIDS autopsies). HIV encephalopathy is a progressive, subcortical dementing illness that was recognized shortly after the initial description of AIDS; it is caused by CNS infection with the HIV virus itself, and is not an opportunistic infection (Whiteman et al. 2002). It is a form of subacute encephalitis. The virus is moved to brain by macrophages and monocytes. The infection damages the blood–brain barrier and induces the production of cytokines. Initially, there is neuronal death, leading to atrophy. Afterward, the virus causes astrogliosis, myelin pallor, and the presence of multinucleated cells in the white matter and less commonly in the gray matter (Castillo and Thurner 2004). This causes gradual demyelination and perivascular nodules of microglial cells, monohistiocytes, and macrophages (Andreula 2004). Mood swings, nervousness, leg weakness, loss of coordination, and loss of bowel or bladder control are clinical signs and symptoms of HIV encephalopathy. When the infection becomes clinically symptomatic, the most common MR findings are diffuse brain atrophy and multifocal or diffuse lesions with increased T2 signal intensity predominantly in the periventricular white matter on T2-weighted images and particularly on FLAIR images (Fig. 3.5.16). Cortical atrophy is a distinguishing feature since non-HIV–mediated cerebral atrophy usually starts centrally. White matter lesions corresponding to the areas of myelin pallor on pathology are usually bilateral (predominantly the frontal lobes), and may be symmetric or asymmetric. Gray matter is generally spared and mass effect is absent. Occasionally, the deep gray matter nuclei (particularly basal ganglia) may be affected. Lesions usually do not enhance following contrast administration. The extent of disease roughly parallels the clinical deterioration. The nodules in white matter do not change in size or number in HIV-positive patients who remain neurologically stable. The lesions may reduce in size when treated (Whiteman et al. 2002; Castillo and Thurner 2004; Andreula 2004). MRS has been used to evaluate patients with HIV infection. MRS shows low NAA. Additionally, the level of choline peak may elevated, and the elevated myoinositol may be seen in these patients when short echo time MRS is used. MRS abnormalities may be detected before the onset of encephalopathy and may be used to monitor the response to treatment (Castillo and Thurner 2004). The differential diagnosis of non-enhancing white matter abnormalities in AIDS patients includes mainly progressive multifocal leukoencephalopathy (PML) and

3.5  Intercranial Infections Fig. 3.5.16  HIV encephalitis. Coronal T2-weighted image (a) and axial FLAIR image (b) show patchy areas of hyperintensity in periventricular white matter, extending to subcortical region in the left frontal lobe. The hyperintense areas do not have mass effect (Courtesy of M. Thurnher, MD)

377

378

3  Brain, Head, and Neck

CMV infection. The lesions of HIV demyelination tend to be somewhat more symmetric and central, while PML is more often multifocal and asymmetric, with scalloping and a greater predilection for the subcortical white matter. Although both entities appear hyperintense on T2weighted images, the lesions of HIV demyelination are usually isointense on T1-weighted images, whereas those of PML are often hypointense and well-demarcated on T1-weighted images. Clinical correlation is also extremely helpful since HIV encephalitis often manifests as a global encephalopathy, whereas PML is usually associated with a focal neurological deficit (Whiteman et al. 2002; Castillo and Thurner 2004).

References 1.

2. 3. 4. 5.

3.5.4.3  Cytomegalovirus Encephalitis CMV is a type of herpes virus that produces infection in babies and in adults. Approximately 1% of newborn babies are infected, but progression to systemic disease occurs in only 10% of cases (Andreula 2004). In adults, CMV is a frequent pathogen in the AIDS population and in other immunocompromised patients, occurring not only in the CNS but throughout the body. The virus exists in a latent form in most of the normal population, with nearly 90% of adults having antibodies to CMV. In some immunodeficient patients, reactivation usually results in disseminated infection and/or necrotizing meningoencephalitis (Whiteman et al. 2002). Histologically, CMV produces atrophy, periventricular necrosis, neuronal loss, and accumulation of enlarged microglial nodules (Castillo and Thurner 2004). CMV is a markedly neurotrophic virus, and any part of the brain may be affected but characteristically involvement of ventricles to the subcortical regions is seen. Early infection in utero can cause congenital malformation, with interference in neuronal migration producing migrational disorders ranging from mild micropolygyria to severe agyria/pachygyria with subsequent microcephaly and a small skull (Andreula 2004). MRI findings include migrational anomalies, encephalomalacia with non-specific ventricular enlargement, prominent sulci, delayed myelination, subependymal paraventricular cysts, and calcification (Osborn 1994). Widespread parenchymal calcifications are most commonly seen at the periventricular region. In addition to atrophy, MRI may demonstrate increased signal in the periventricular white matter, which may be patchy and less confluent on T2-weighted images (Whiteman et al. 2002). Infrequently, subependymal enhancement (30–45% of cases) is evident, and if present, is a valuable diagnostic clue (Griffiths 2004). HIV encephalopathy should be included in the differential diagnosis. The centrum semiovale is most commonly involved in HIV leukoencephalopathy, and HIV encephalopathy may result in infarctions.

6.

7.

8.

9. 10. 11. 12.

13. 14.

15.

16.

17.

18.

Ackerman LL, Traynelis VC (2002) Dural space infections: cranial subdural empyema and cranial epidural abscess. In Osenbach RK, Zeidman SM (eds) Infections in neurological surgery: diagnosis and management, 1st edn. Lippincott-Raven, Philadelphia, pp 85–99 Andreula C (2004) Cranial viral infections in the adult. Eur Radiol 14:E132–E144 Anslow P (2004) Cranial bacterial infection. Eur Radiol 14: E415–E154 Arbealáez A, Medina E, Restrepo F, Castillo M (2004) Cerebral tuberculosis. Semin Roentgenol 474–480 Bernaerts A, Vanhoenacker FM, Parizel PM, Van Goethem JWM, Altena RV, Laridon A, Roeck JD, Coeman V, De Schepper AM (2003) Tuberculosis of the central nervous system: overview of neuroradiological findings. Eur Radiol 13:1876–1890 Buff BL, Mathews VP, Elster AD (1994) Bacterial and viral parenchymal infections of the brain. Top Magn Reson Imaging 6:11–21 Burtscher IM, Holtas S (1999) In vivo proton MR spectroscopy of untreated and treated brain abscesses. AJNR Am J Neuroradiol 20:1049–1053 Cannard KR (2002) Tuberculous meningitis and tuberculoma. In Osenbach RK, Zeidman SM (eds) Infections in neurological surgery: diagnosis and management, 1st edn. Lippincott-Raven, Philadelphia, pp 33–40 Castillo M (2004a) Imaging of meningitis. Semin Roentgenol 39:458–464 Castillo M (2004b) Imaging of neurocysticercosis. Semin Roentgenol 39:465–473 Castillo M, Thurner M (2004) Imaging viral and prion infections. Semin Roentgenol 482–484 Celik H, Karaosmanoglu AD, Gultekin S, Tokgoz N, Tali ET (2005) Cerebral amebiasis: MRI, DWI, perfusion and MRS features. Riv Neuroradiol 18:559–563 Choi IC, Chang KH, Kim YH et al (1998) MRI features of choroids plexitis. Neuroradiology 40:303–307 Choi JY, Akins PT (2002) Bacterial meningitis. In Osenbach RK, Zeidman SM (eds) Infections in neurological surgery: diagnosis and management, 1st edn. LippincottRaven, Philadelphia, pp 3–12 Dev R, Gupta RK, Poptani H, Roy R, Sharma S, Husain M (1998) Role of in vivo proton MRS in the diagnosis and management of brain abscess. Neurosurgery 42:37–43 Dubey A, Patwardhan RV, Sampth S, Santosh V, Kolluri S, Nanda A (2005) Intracranial fungal granuloma: analysis of 40 patients and review of the literature. Surg Neurol 63:254–260 Ercan N, Gultekin S, Celik H, Tali ET, Oner AY, Erbas G (2004) Diagnostic value of contrast-enhanced fluid attenuated inversion recovery (FLAIR) sequence in intracranial metastases. AJNR Am J Neuroradiol 25:761–765 Erdogan E, Beyzadeoglu M, Arpacı F, Celasun B (2002) Cerebellar aspergillosis: a case report and literature review. Neurosurgery 50:874–876

3.5  Intercranial Infections 19. Gee GT, Bazin C, Jinkins JR (1992) Miliary tuberculosis involving the brain: MR findings. Am J Roentgenol 159:1075–1076 20. Gellin BG, Weingarten K, Gamache FW, Hartman BJ (1997) Epidural abscess. In Scheld WM, Whitley RJ, Durack DT (eds) Infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, pp507–522 21. Griffiths P (2004) Cytomegalovirus infection of the central nervous system. Herpes 11(Suppl 2):95A–99A 22. Gupta RK, Kathuria MK, Pradhan S (1999) Magnetization MR imaging in CNS tuberculosis. AJNR 20:867–875 23. Gupta RK, Roy R, Dev R (1996) Fingerprinting of Mycobacterium tuberculosis in patients with intracranial tuberculomas by using in vivo, ex vivo, and in vitro MRS. Magn Reson Med 36:829–833 24. Harisinghani M. McLoud TC, Shepard JA, Ko JP, Shroff MM, Mueller PR (2000) Tuberculosis from head to toe. Radiographics 20:449–470 25. Harris TM, Edwards MK (1992) Meningitis. Neuroimaging Clin N Am 1:39–55 26. Helfgott DC, Weingarten K, Hartman BJ (1997) Subdural Empyema. In: Scheld WM, Whitley RJ, Durack DT (eds) Infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, pp 495–505 27. Hunter JV, Morriss MC (2003) Neuroimaging of central nervous infections. Semin Pediatr Infect Dis 14:140–164 28. Jereb M, Lainscak M, Marin J, Popovic M (2005) Herpes simplex virus infection limited to the brainstem. Middle Eur J Med 117:495–499 29. Jinkins JR (1991) Computed tomography of intracranial tuberculosis. Neuroradiology 33:126–135 30. Jinkins JR, Gupta R, Chang KH, Rodriguez-Carvajal J (1995) MR imaging of central nervous system tuberculosis. Radiol Clin N Am 33:771–786 31. Jordan J, Enzmann DR (1991) Encephalitis. Neuroimag Clin N Am 1:17–38 32. Kamra P, Azad R, Prasad KN, Jha S, Pradhan S, Gupta RK (2004) Infectious meningitis: prospective evaluation with magnetization transfer MRI. Br J Radiol 77:387–394 33. Kanamalla US, Ibarra R, Jinkins JR (2000) Imaging of cranial meningitis and ventriculitis. Neuroimaging Clin N Am 10:309–329 34. Kastrup O, Wanke I, Maschke M (2005) Neuroimaging of infections. NeuroRx 2:324–332 35. Kohli A, Gupta RK, Poptani H, Roy R (1995) In vivo proton MRS in a case of intracranial hydatid cyst. Neurology 45:562–564 36. Lee JH, Na DG, Choi KH et al (2002) Subcortical low intensity on MR imaging of meningitis, viral encephalitis, and leptomeningeal metastasis. AJNR Am J Neuroradiol 23:1369–1377 37. Leuthardt EC, Wippold FJ, Oswood MC, Rich KM (2002) Diffusion-weighted MR imaging in the preoperative assessment of brain abscesses. Surg Neurol 58:395–402 38. Maschke M, Kastrup O, Forsting M, Diener HC (2004) Update on neuroimaging in infectious central nervous systems disease. Curr Opin Neurol 17:475–480

39. Mathews PM, Shoubridge E, Arnold DL (1989) Brain phosphorus MRS in acute bacterial meningitis. Arch Neurol 46:994 40. McGuinness FE (2000) Intracranial tuberculosis. In: Clinical imaging in non-pulmonary tuberculosis. Springer, Berlin Heidelberg New York, pp 5–25 41. Mehta RC, Pike GB, Haros SP, Enzmann DR (1995) Central nervous system tumor, infection, and infarction: detection with gadolinium-enhanced magnetization transfer MR imaging. Radiology 195:41–46 42. Nguyen JB, Black BR, Leimkuehler MM, Halder V, Nguyen JV, Ahktar N (2004) Intracranial pyogenic abscess: imaging diagnosis utilizing recent advances in computed tomography and magnetic resonance imaging. Crit Rev Comput Tomogr 45:181–224 43. Osborn AG (1994) Infections of the brain and its linings. In: Osborn A (ed) Diagnostic neuroradiology. Mosby, St. Louis, pp 673–715 44. Patterson FA, Ling GSF (2002)Viral meningitis. In: Osenbach RK, Zeidman SM (eds) Infections in neurological surgery: diagnosis and management, 1st edn. LippincottRaven, Philadelphia, pp 13–22 45. Perdiago J, Rojas R, Verzelli LF, Castillo M (2004) Fungal infections of the central nervous system. Semin Roentgenol 39:505–518 46. Roos KL, Tunkel AR, Scheld MS (1997) Acute bacterial meningitis in children and adults. In: Scheld WM, Whitley RJ, Durack DT (eds) Infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, 335–416 47. Rotbart HA (1997) Viral meningitis and the aseptic meningitis syndrome. In Scheld WM, Whitley RJ, Durack DT (eds) Infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, pp 23–46 48. Roy R, Gupta RK, Kishore J, Taparia S, Poptani H, Bhakuni V (1996) High-resolution proton MR spectroscopy of the cerebral fluid from children with tuberculous meningitis. Proceedings of Annual Meeting of ISMRM, p 1157 49. Sener RN (2001) Herpes simplex ensefalitis: diffusion MR imaging findings. Comput Med Imag Graph 25:391–397 50. Shawl S (1995) Neurologic evaluation of patient with acute bacterial meningitis. Neurol Clin 13:549–577 51. Smith RR, Caldemeyer KS (1999) Neuroradiologic review of intracranial infection. Curr Probl Diagn Radiol 28:1–26 52. Steiner I, Budka H, Chaudhuri A, Koskiniemi M, Sainio K, Salonen O, Kennedy PGE (2005) Viral encephalitis: a review of diagnostic methods and guidelines for management. Eur J Neurol 12:331–343 53. Sütlaş PN, Ünal A, Forta H, Şenol S, Kırbaş D (2003) Tuberculous meningitis in adults: review of 61 cases. Infection 6:387–391 54. Suzer T, Coskun, Cirak B, Yagci B, Tahta K (2005) Brain stem abscesses in childhood. Childs Nerv Syst 21:27–31 55. Tsuchiya K, Inaoka S, Mizutana Y, Hachiya J (1997) Fast fluid-attenuated inversion-recovery MR of intracranial infections. AJNR Am J Neuroradiol 18:909–913 56. Tunkel AR (2000) Brain abscess. Curr Treat Options Infect Dis 2:449–460

379

380

3  Brain, Head, and Neck 57. Whiteman MLH, Bowen BC, Post MJD, Bell M (2002) Intracranial infection. In: Atlas SW (ed) Magnetic resonance imaging of the brain and spine, 3rd edn. Lippincott-Raven, Philadelphia, pp 1099–1175 58. Whitley RS (1997) HSV virus. In: Scheld WM, Whitley RJ, Durack DT (eds) Infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, pp 73–89 59. Wilson JD, Castillo M (1994) Magnetic resonance imaging of granulomatous inflammations: sarcoidosis and tuberculosis. Top Magn Reson Imag 6:32–40

60. Zeidman SM (2002) Pyogenic brain abscess. In: Osenbach RK, Zeidman SM (eds) Infections in neurological surgery: diagnosis and management, 1st edn. Lippincott-Raven, Philadelphia, pp 101–122 61. Zuger A, Lowy FD (1997) Tuberculosis. In: Scheld WM, Whitley RJ, Durack DT (eds) infections of the central nervous system, 2nd edn. Lippincott-Raven Philadelphia, pp 417–443

3.6  Neurodegenerative Disorders

3.6  Neurodegenerative Disorders S. Karimi and A.I. Holodny 3.6.1  Introduction The radiological evaluation of neurodegenerative disorders such as dementias, hydrocephalus, movement disorders, and temporal lobe epilepsy has markedly improved because of technical advances in magnetic resonance imaging (MRI). Diagnosing neurodegenerative disorders and differentiating them from one another is a difficult clinical challenge which is further complicated by the ongoing normal aging process of the brain. The capability of MRI to depict the brain in multiple planes and show its function and metabolic signature is a significant aid in the diagnostic evaluation of neurodegenerative disorders. Understanding and recognizing the utility of MRI in the management of patients with neurodegenerative disorders, is particularly important as the population ages and new therapies are introduced. 3.6.1.1  Normal Aging Brain The ventricles and sulci become dilated with normal ageing due to cerebral parenchymal atrophy (Fig. 3.6.1). Cerebral atrophy is variable and can range from minimal to severe. Sulcal dilatation is typically seen with advancing age even if ventricular dilatation is not appreciable. The metabolism of the brain does not diminish as age-related atrophic changes occur on FDG PET. At times people over the age of 60 do demonstrate ventricular dilatation and may have only mild or no sulcal widening. Quantification studies using MRI have demonstrated that the sulcal-to-ventricular-volume ratio approximates unity among normal young individuals, and that this ratio increases after age 60 in normal subjects (Rusinek et al. 1991). MR imaging has revealed similar results for the volume of the amygdale, hippocampus, and temporal horn in healthy individuals. These structures remain relatively stable in volume till age 60 and then undergo atrophy with age (Mu et al 1999). 3.6.2  Dementia 3.6.2.1  Alzheimer’s Disease Alzheimer’s disease (AD) is the most common dementing disorder among the elderly. It is considered one of the most significant health problems of the century. The prevalence of AD increases with advancing age. It has been estimated that AD afflicts 10% of people over 65 years of age and 50% of individuals over the age of 85. AD patients typically present with impairment of memory,

Fig. 3.6.1  Axial CT image of a 58-year-old healthy individual shows mild prominence of the sulci due to generalized parenchymal atrophy

judgment, abstract thought, and other higher cortical functions. Abnormalities of the personality and behavior are also common among AD patients. These features are not specific to AD and can be seen with other dementing disorders, and AD often overlaps with other clinical entities. The earliest pathologic examinations showed a generalized pattern of cortical atrophy and did not establish diagnostic criteria. However, they demonstrated that atrophy varied in extent and in some cases was absent. It should be noted that atrophy of the brain is not specific for AD and can be seen with other degenerative conditions as well as with normal aging. In as many as 40% of the patients with AD, genetic factors have been implicated. Three different genes have been identified which when mutated can cause AD. Sporadic cases can account for as many as 20% of AD cases. Inherited cases can be transmitted by autosomal dominant means. The definite diagnosis of AD can only be made by histopathological examination (McKhann 1984). The histopathological hallmarks of AD include neurofibrillary tangles and senile plaques that accumulate in the medial temporal and the temporoparietal lobes. Studies correlating clinical and pathological diagnoses have reported

381

382

3  Brain, Head, and Neck

clinical diagnostic accuracy rates of 60–90% (Joachim et al. 1988; Morris et al. 1988). The clinical diagnosis of AD is not uncommonly certain, particularly among patients early in the disease with mild cognitive changes. Many investigators have focused on imaging structural changes to help make the diagnosis of AD since biopsy is not a practical option for most patients. In the early 1980s, the hippocampal formation was the first specific region of the brain that was implicated in the pathogenesis of AD. The histopathological changes in the early cases of AD are restricted to the hippocampus and parahippocampal gyrus (Prince et al. 1991). Later it was shown that other observations described previously were secondary both in temporal evolution and importance to the changes that occur in the hippocampal formation. The hippocampi have reciprocal connections with structures which are crucial to memory and cognition such as the basal forebrain, thalamus, and hypothalamus. Damage to these neurons in the hippocampus correlates well with the neuropsychological impairment in patients with AD. The pattern of degeneration leads to “isolation” of the hippocampus from the neocortical association areas in AD (Hyman et al. 1984). In 1985, the central role of the hippocampus in the development of AD was recognized and hippocampal dementia was proposed as a new definition for AD (Kemper 1984). This led to the efforts of a number of investigators to focus on the hippocampal formation and to evaluate imaging criteria for the diagnosis of AD. Pathological evidence is also supportive of the concept of AD as a hippocampal dementia (Ball et al. 1985). Severe cell loss and neurofibrillary tangle formation has been shown histologically in the subiculum and entorhi-

nal cortex in patients with AD (Narkiewicz et al. 1993). Neuropathology in this region is thought to be responsible for memory impairment, which is typical in AD. The pathological hallmarks in AD again include senile plaques, neurofibrillary tangles, and granulovacuolar degeneration with progressive neuronal loss and atrophy (Ball et al. 1985; Hyman et al. 1984; Kemper 1984). The mesial temporal lobe structures are where the earliest atrophic changes can be detected (Fig. 3.6.2). As the disease progresses atrophy extends to other parts of the temporal and the parietal lobes. As pathologic attention was directed to the temporal lobes and CT imaging improved, enlargement of the cerebrospinal fluid (CSF) spaces in the peri hippocampal region was noted as a sign of hippocampus and parahippocampal gyrus atrophy. A number of studies showed that it was possible to differentiate AD patients from age-matched controls by assessing mesial temporal lobe structures (Kido et al. 1989; LeMay et al. 1986; Sandor et al. 1988). Reverse-angle CT imaging parallel to the long axis of the temporal lobes was found to have sensitivity of 82%, specificity of 75%, and an overall accuracy of 80% in differentiating AD patients from age-matched controls (George et al. 1990). It was also found that these observations might be of value in predicting AD in patients who have only mild cognitive impairment (de Leon et al. 1989). MR imaging has significantly improved evaluation of the mesial temporal lobes particularly on coronal imaging. Most patients with AD have some degree of medial temporal lobe atrophy with secondary widening of the adjacent CSF spaces. These structural changes can be helpful and sensitive in identifying patients with AD

Fig. 3.6.2  Axial CT images (a,b) of a 65-year-old patient with memory loss and mild cognitive decline demonstrate mild atrophy of the hippocampus and parahippocampal gyri (arrows). The brain parenchyma is otherwise unremarkable (c)

3.6  Neurodegenerative Disorders

(de Leon et al. 1983; George et al. 1990; Holodny et al. 1998a). High resolution coronal images are ideal for visualization of the hippocampal formations (Fig. 3.6.3). When atrophy is only mild, it can still be detected on routine CT or MR imaging. Hippocampal atrophy with corresponding widening of the adjacent CSF spaces can be used to predict the development of AD (de Leon et al. 1983). Assessment of the dilatation of the peri hippocampal CSF spaces at two time intervals was shown to be 91% sensitive and 89% specific in predicting AD. Volumetric measurements from coronal MR imaging may be useful to differentiate normal aging brain from AD, but they overlap significantly. The greatest difference in the volumetric measurements of normal patients and those with AD is found in the hippocampus. Hippocampal and parahippocampal gyrus volumes correlate best with the mini-mental state examination. In one study, MR volumetric measurements of the temporal lobe structure were effective in differentiating AD patients from agematched controls (p < 0.001), with overlap in only 3 of the 42 patients in the two groups (Jack et al. 1992). MRI methods have been used to determine regional volumes of gray and white matter as well as CSF (Rusinek et al. 1991). One study which investigated 14 AD patients and 14 age-matched controls revealed that the percentage of gray matter in AD patients was significantly lower than that among the controls (44.9 +/–4.4% vs. 50.2 +/–3.2%). The highest gray matter loss (13.8%, p < 0.001) and the highest increase in adjacent CSF (p < 0.01) were observed in the temporal lobes of AD patients. Less extensive gray matter volume loss was also noted in the frontal and occipital lobes in AD patients. Regional white matter changes were not significant. The postmortem studies of the brains of patients with AD correlate with these findings (de la Monte 1989; Terry 1981). A more recent MRI volumetric study showed that patients with mild cases of AD had significant atrophy of the hippocampal formation and amygdale compared with normal age-matched controls. The sizes of the ventricles of the patients in the two groups did differ significantly (Lehericy et al. 1994). Initially it was suggested that the measurement of the intrauncal distance (IUD) might enable the differentiation of AD patients (Dahlbeck et al. 1991), but other reports have not confirmed this. A study using MR diffusion tensor imaging showed axonal degeneration in temporoparietal lobes that are commonly affected in AD patients (Rose et al. 2000). Therefore, diffusion imaging and tractography may help in understanding the neuropathological changes in AD. As outlined above, advances in neuroimaging can provide invaluable insight into AD and aid in its diagnosis. The role of neuroimaging is therefore no longer limited to excluding other causes of dementia such as multiple infarcts, subdural hematoma, metastatic disease, and hydrocephalus.

Fig. 3.6.3  Coronal T1-weighted image of a patient with Alzheimer’s disease. Note marked atrophy of the hippocampal formations and widening of the choroidal fissures and the temporal horns of the lateral ventricles

3.6.2.2  Pick’s Disease Pick’s disease is a form of frontotemporal dementia. It is a much less common cause of dementia than AD is, and up to 40% of the cases may be of familial origin. The affected patients often present with signs of frontal lobe damage. The patients usually have more personality and behavioral changes than patients with AD. Memory and visuospatial skills tend to be preserved until late in the disease. Patients with Pick’s disease slowly but steadily progress to a condition of marked dementia. The pathological hallmarks include atrophy of the brain, particularly involving frontal lobes and the anterior aspects of the temporal lobes, with a tendency to spare the parietal and occipital lobes (Fig. 3.6.4). The atrophy may affect the dominant hemisphere (usually the left) more severely (Dickson 2001). Histologically, there is neuronal loss and gliosis in the affected regions. Pick bodies are characteristic inclusions within the neurons seen histologically with Bielschowsky silver stain and antibody immunostaining. The so-called classic Pick’s disease is extremely rare, and its features overlap with other causes of frontotemporal dementia. Some investigators have attempted to classify Pick’s disease into different subgroups (Rose et al. 2000; Rossor 2001).

383

384

3  Brain, Head, and Neck

Fig. 3.6.4  Striking severe atrophy of the frontal and anterior aspect of the temporal lobes on gross photograph of brain of a patient with classic Pick’s disease

On imaging, patients with Pick’s disease demonstrate marked atrophy of the anterior temporal and frontal lobes with impressive enlargement of the corresponding CSF spaces (Fig. 3.6.5). The gyri of these lobes have been described as icicle-like because of severe atrophy. The posterior aspect of the superior frontal gyrus and the brain posterior to it are usually spared. 3.6.2.3  Multiple-Infarct Dementia, Leukoencephalopathy of Aging, and Binswanger’s Disease The second most common cause of dementia in the elderly population is multiple-infarct dementia (MID), which accounts for 10–20% of cases in patients over the age of 65. These patients often have evidence of arteriosclerosis such as strokes, coronary artery disease, and hypertension. They can present with a sudden onset of dementia with day-to-day fluctuation and spontaneous improvement early in the disease. On imaging, there are focally dilated fissures and multiple areas of cortical thinning and tissue loss (Fig. 3.6.6). These patients invariably demonstrate evidence of subcortical ischemia. It is thought that the cumulative effects of multiple infarcts leads to the observed dementia but multiple subcortical infarcts or a single infarct at a strategic location (i.e., parts of the thalamus, caudate, or the genu of the internal capsule) can lead to dementia. Periventricular white matter disease which has been shown not to be dementing in itself is not the same as MID (George et al. 1986). Periventricular white matter disease lesions appear hypodense on CT and hyperintense on long TR MRI and are referred to as UBOs (unidentified bright objects) or leukoaraiosis. These white matter lesions are thought to be due to hypertensive microvascular disease and resultant demyelination.

Fig. 3.6.5  Sagittal T1-weighted image (a) and axial T2-weighted image (b) of a patient with Pick’s disease. Marked atrophy of the gyri in anterior frontal and temporal lobes with dilatation of the adjacent sulci is characteristic of frontotemporal dementia

MRI is very sensitive in the detection of white matter abnormalities. Long TR images (T2, proton density, and fluid-attenuated inversion recovery [FLAIR]) are ideal for detecting white matter lesions (Fig 3.6.7). Among the elderly, white matter lesions are commonly patchy and involve the periventricular region. The brain stem is a less common location for white matter lesions in the elderly, but when present they are usually found in the pons. These lesions typically spare the arcuate fibers and the corpus callosum. They have been described with terms such as white matter hyperintensities of aging, leukoencephalopathy, leukoaraiosis, or deep white matter ischemic lesions. They are strongly related to age and

3.6  Neurodegenerative Disorders

Fig. 3.6.6  Axial FLAIR image of a patient with history of strokes and dementia. Multiple chronic infarcts are evident with extensive subcortical ischemic changes

their prevalence increases in individuals 60 years of age and older. The lesions are detected earlier and are more severe in patients with microvascular diseases, such as those with diabetes or hypertension (Fig. 3.6.8). Since the white matter receives its blood supply from thin, long blood vessels it is more prone than gray matter to chronic ischemic changes. It has been found that these lesions are related to gait disturbances and increased incidence of falls in the el-

derly population, fine motor coordination deficits, and an increased frequency of infarcts. The increased incidence of falls is particularly of concern in the elderly (Udaka et al. 1992). Autopsy work has shown hypertensive-type white matter changes are associated with hyalinosis of arterioles and white matter rarefaction. Mild gliosis and increased interstitial fluid “myelin pallor” can also be detected. Binswanger in 1894 described a slowly progressive condition characterized by memory loss, intellectual impairment, and recurring neurologic deficits of slow but progressive course. It seems that he was describing dementia with multiple infarcts and severe microvascular ischemic changes. The diagnosis of Binswanger’s disease or multi-infarct dementia should not be entertained in the setting of periventricular T2 prolongation unless evidence of infarcts and dementia is present. Dementia may be seen in the setting of multiple bilateral cortical or subcortical infarcts. Even in the cases where there are numerous subcortical infarcts patients may only have mild forms of dementia. The terms multi-infarct dementia and Binswanger’s disease should be avoided and not used to refer to periventricular T2 prolongation because these are commonly seen among the elderly (Pantoni and Garcia 1996). 3.6.2.4  CADASIL Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a hereditary disorder linked to mutations of the Notch 3 gene on chromosome 19. It accounts for only a small number of the cases of vascular dementia. The patients clinically have transient ischemic attacks, strokes, mood disturbances, progressive subcortical dementia, and migraines. A family history of dementia can often be elicited. The

Fig. 3.6.7  Axial FLAIR images (a,b) of a 60-year-old asymptomatic individual. Note mild periventricular white matter hyperintensity, without significant atrophy or ischemic changes

385

386

3  Brain, Head, and Neck

ischemic lesions are due to small vessel disease and are essentially confined to subcortical structures with sparing of the cortex (Fig. 3.6.9). Abnormalities of glucose

metabolism and cerebral blood flow in Notch 3 mutation carriers have been shown in PET and MR perfusion studies (Tatsch et al. 2003; van den Boom et al. 2003). 3.6.2.5  Creutzfeldt-Jakob Disease

Fig. 3.6.8  Axial FLAIR image of a 64-year-old patient with history of hypertension and hypercholesterolemia. Moderately severe patchy and more confluent areas of white matter hyperintensity are evident due to chronic microvascular ischemia

Creutzfeldt-Jakob disease (CJD) is a rare cause of dementia with rapid and progressive course. It has an incidence of approximately one case per million individuals each year. Most cases of CJD are sporadic but up to 10–15% are inherited and transmitted by autosomal dominant means. CJD is one of several neurodegenerative diseases which are caused by a non-viral, 30- to 35-kDa proteinaceous particle called a prion. Pathologically, there is neuronal loss and astrocytosis with formation of intracytoplasmic vacuoles within the neurons and glia, which is responsible for the spongiform appearance on light microscopy. The pathology in CJD involves the cerebral and cerebellar cortices and the basal ganglia. The patients often develop dementia rapidly and many develop myoclonus particularly late in the disease. The disease is uniformly fatal with survival of less than 1 year after the onset of symptoms. The CT and MRI scans early on may be normal. Symmetrical increased T2 signal intensity of the basal ganglia may be the earliest changes and may occasionally appear before clinical signs (Barboriak et al. 1994; Di Rocco et al. 1993; Milton et al. 1991; Tartaro et al. 1993). Diffusionweighted imaging (DWI) is very sensitive in detecting early stages even prior to development of FLAIR signal abnormality. In fact, DWI is more sensitive in identifying the abnormalities of the gray matter than routine MRI sequences (Parazzini et al. 2003). On DWI symmetrical

Fig. 3.6.9  Axial FLAIR images (a–c) of a 43-year-old female with family history of dementia. Extensive confluent subcortical, deep, and periventricular white matter hyperintensity with

multiple chronic lacunar infarcts are evident. Note lack of cortical tissue loss in this patient with CADASIL confirmed by skin biopsy

3.6  Neurodegenerative Disorders

Fig. 3.6.10  Symmetrically hyperintense striatum and asymmetrical cortical hyperintensity in the left frontal lobe and insula on the FLAIR image (a) of a patient with CJD. Corresponding decreased diffusion on the DWI (b) and ADC map (c)

decreased diffusion in the basal ganglia is characteristic of CJD (Fig. 3.6.10). Decrease in diffusion may also be detected in the thalami and cerebral cortex, but the abnormality in these regions tends to be less symmetrical (Fig. 3.6.11). Analyzing ADC measurements may increase sensitivity for detection of gray matter abnormality in patients with CJD (Henriette et al. 2003). Reduction in movement of water in the intracytoplasmic vacuoles of spongiform encephalitis is thought to be the cause of Brownian motion derangements that are detected on DWI (Bahn and Parchi 1999; Demaerel et al. 1997). Decrease in N-acetyl-aspartate and other metabolites in late CJD have been shown on MR spectroscopy (Graham et al. 1993). In late stages, the patients develop marked atrophy, and serial MRI scans reveal rapidly progressive atrophy. It is important to keep in mind that a normal initial MRI does not rule out CJD.

The classic pathological finding in HD patients is atrophy of the basal ganglia, particularly the caudate nuclei (Fig. 3.6.12). Interestingly, the atrophy progresses from medial-to-lateral and dorsal-to-ventral direction. The earliest atrophic changes can be detected in the head of

3.6.3  Disorders with Prominent Motor Disability 3.6.3.1  Huntington’s Disease Huntington’s disease (HD) is inherited in an autosomal dominant manner, with complete penetrance. The responsible gene has been localized to the short arm of chromosome 4. The patients often become symptomatic by the age 50 and usually present with affect and personality problems. As the disease progresses dementia becomes evident and motor abnormalities emerge. Choreoathetoid movements are typically described in these patients. A group of patients may have rigidity as the dominant motor abnormality, the Westphal variant.

Fig. 3.6.11  Diffusion-weighted image of a different patient with CJD than in Fig. 3.6.10 demonstrates asymmetrical decreased diffusion in the cerebral hemispheres in addition to more symmetrical abnormality in the basal ganglia

387

388

3  Brain, Head, and Neck

Fig. 3.6.12  Axial CT images (a,b) of a 45-year-old patient with Huntington’s disease. Marked atrophy of the caudate nuclei resulting in enlargement of the frontal horns of the lateral ventricles. Note that the lateral walls of the frontal horns are straight to slightly concave instead of having the normal convex mor-

phology. Axial T2-weighted image (c) of a different patient with Huntington’s disease demonstrates atrophic caudate nuclei with associated abnormal contour of the lateral walls of the frontal horns

the caudate. The frontal and temporal cortices may also be affected. On the molecular level, diminished neurotransmitter concentrations of acetylcholine and γ-aminobutyric acid occur in the affected regions. A number of investigators have reported that areas of abnormal T2 prolongation in the basal ganglia occur in HD patients. There are reports that all patients with the rigid form of HD have these changes, but they only seldom present in patients with the classic hyperkinetic form of the disease (Oliva 1993; Savoiardo et al. 1991). In severe cases the patients may have atrophy of other parts including the olives, pons, cerebellum, thalamus, white matter tracts, mesial temporal lobe, and the cortex.

gait, stooped posture, and what is commonly described as the “cog-wheel” type of rigidity. These patients may develop dementia up to 30% of the time, which may be due to the presence of concurrent AD. Narrowing of the pars compacta portion of the substantia nigra may be detected on MRI and is best appreciated on T2-weighted images. The changes are often bilateral; however, unilateral cases do exist.

3.6.3.2  Parkinson’s Disease Approximately 1% of individuals over 50 years of age are affected by Parkinson’s disease (PD), which is probably the most common cause of progressive motor dysfunction among the elderly. Histologically, it is characterized by loss of pigmented dopaminergic neurons of the pars compacta portion of the substantia nigra as well as loss of pigmented cells within the locus ceruleus and the dorsal motor nucleus of the vagus. Reactive astrocytosis and eosinophilic cytoplasmic inclusions (Lewy bodies) may be present as well. The characteristic clinical findings include tremors at rest, bradykinesis, masked facies, hypophonia, shuffling

3.6.3.3  Multiple-System Atrophy Motor dysfunction in multiple-system atrophy (MSA), which is also referred to clinically as Parkinson’s plus syndrome, results from degenerative processes that affect several subcortical anatomic structures. Patients with different forms of MSA can present with symptoms that are very similar to those seen in PD. Differentiating these patients from PD patients can be very difficult particularly during the early stages. Clinically when a patient with parkinsonian symptoms fails to respond to L-DOPA therapy, MSA should be suspected as the pathological cause of motor dysfunction. Pathological changes may affect the striatum, cerebellum, brainstem, and the spinal cord nuclei. In a pattern of MSA referred to as striatonigral degeneration (SND) (Fig. 3.6.13), progressive atrophy and neuronal loss of the putamen and caudate as well as cell loss within the pars compacta of the substantia nigra without Lewy bodies are characteristic histologically. The patients with SND may also demonstrate

3.6  Neurodegenerative Disorders

Fig. 3.6.13  Axial T2-weighted images (a,b) of a patient with SND demonstrating putaminal atrophy

thinning of the pars compacta on imaging similar to that seen in patients with PD. In contrast to PD, patients with SND may exhibit hypointensities in the putamen, which are thought to be due to increased iron deposition. The signal intensity of the putamen is dependent on the balance between hyperintensity from gliosis and hypointensity due to iron deposition. Olivopontocerebellar atrophy (OPCA) is another form of MSA in which degenerative changes involve the pontine nuclei and the transverse pontine fibers, middle cerebellar peduncles, inferior olives, and cerebellar cortex. Patients with OPCA have progressive ataxia and bulbar abnormalities. OPCA is typically seen in adults. There are familial as well as sporadic cases. The familial cases are commonly transmitted by autosomal dominant means. On MRI, atrophy of the transverse fibers of the pons, the cerebellum and the middle cerebellar peduncles may be detected (Fig. 3.6.14). A mild decrease in the width of the pars compacta may also be evident. Abnormal T2 hypointensity, seen in patients with SND, is not detected in patients with OPCA. In one form of MSA, there is neuronal loss involving the autonomic intermediolateral nuclei of the spinal cord. This condition is referred to as the Shy-Drager syndrome (SDS). In addition to symptoms of MSA, these patients also have symptoms of orthostatic hypotension, incontinence, and sexual dysfunction due to autonomic cell loss. On MRI, the findings in SDS reflect atrophic changes and abnormalities present in MSA. 3.6.3.4  Amyotrophic Lateral Sclerosis The most common form of motor system disease is amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s disease. In ALS the motor neurons within the brain, brain stem, and spinal cord undergo degeneration, affecting both the upper and lower motor neurons. The

anterior horn neurons as well as pyramidal Betz cells of the precentral gyrus are selectively affected by the degenerative process. The degenerative process eventually leads to secondary corticospinal tract fiber loss. Astrocytosis, lipofuscin deposition, and at times intracytoplasmic inclusion bodies may be seen on histology. The affected patients typically have upper motor neuron dysfunction such as spasticity and hyperreflexia as well as lower motor neuron dysfunction like weakness and atrophy. Symptoms progress from a distal distribution to a proximal fashion. Decreased T2 prolongation in the precentral gyrus is nonspecific and thought to be related to elevated iron deposition (Ishikawa et al. 1993; Oba et al. 1993). The corticospinal tract may demonstrate abnormal T2 prolongation at the level of centrum semiovale, corona radiata, and pons (Fig. 3.6.15) (Abe et al. 1993; Goodin et al. 1988; Iwasaki et al. 1991; Udaka et al. 1992; Yagashita et al. 1994). Abnormal T2 signal intensity may also be seen in the corticospinal tracts and within the spinal cord (Freidman and Tartaglino 1993). Care should be taken not to call subtle hyperintensity of the corticospinal tracts seen in normal individuals as a sign of ALS. Therefore, it is crucial to seek the clinical history and confirm the abnormal signal on the balanced images in equivocal cases. Changes in fractional anisotropy and ADC (Ellis et al. 1999) have been reported in patients with ALS. It has been shown that apparent diffusion coefficient correlates with disease duration and fractional anisotropy correlates with disease severity (Graham et al. 2004; Hong 2004). 3.6.4  Hydrocephalus Increase in the amount of intracranial CSF associated with ventricular dilatation is referred to as hydrocephalus. It can be caused by structural abnormalities (e.g., tumor, aqueductal stenosis, incisural block) or functional

389

390

3  Brain, Head, and Neck

Fig. 3.6.14  Axial balanced image (a) of a 48-year-old female with progressive spasticity, cerebellar dysarthria, and abnormal eye movement demonstrating “hot-cross-bun” sign due to atrophy of the transverse fibers of the pons. Atrophy of the cerebellum and pons are readily evident of the sagittal T1-weighted im-

age (b). Axial T2-weighted images reveal atrophy of the middle cerebellar peduncles (c) and olives (d). The pathological diagnosis after autopsy was in agreement with the clinical diagnosis of olivopontocerebellar degeneration

3.6  Neurodegenerative Disorders

Fig. 3.6.15  A series of Axial FLAIR images (a–f) show abnormal hyperintensity involving the corticospinal tracts extending from subcortical level to internal capsules in a patient with ALS

abnormalities (e.g., overproduction of CSF or resorption deficits). Traditionally hydrocephalus is classified as communicating or non-communicating. Since complete CSF obstruction is incompatible with life, the term non-communicating, or obstructive hydrocephalus, is a misnomer. A more accurate way of describing hydrocephalus is to consider it as obstructive intraventricular or obstructive extraventricular. Non-communicating hydrocephalus (obstructive intraventricular) is due to a lesion causing stenosis along the ventricular CSF pathways. This results in dilatation of the ventricular system proximal to the stenotic lesion. Downstream of the lesion the ventricular system is normal or small. This type of hydrocephalus can result from both malignant and benign lesions. Structural abnormalities such as colloid cysts (Fig. 3.6.16), aqueductal stenosis (Fig. 3.6.17), or tumors along the CSF pathways (Fig. 3.6.18), or mass effect from any cause on the CSF pathways (Fig. 3.6.19) can result in obstructive hydrocephalus.

When an area of stenosis or obstruction is not identified on MRI, then the diagnosis of communicating hydrocephalus can be made. Communicating hydrocephalus is usually due to inflammatory or neoplastic meningitis or can occur after surgery, trauma, or subarachnoid hemorrhage. The level of obstruction may be at the incisura of the tentorium. This is referred to as an incisural block. Structural and functional abnormalities of the pacchionian granulations may also lead to decreased CSF resorption and therefore communicating hydrocephalus (Fig. 3.6.20). In early stages of hydrocephalus, the only imaging findings may be dilatation of the temporal horns. Dilatation of the temporal horn from atrophy is invariably seen with enlargement of the perihippocampal fissures. In the setting of hydrocephalus, dilation of the temporal horn may compress the hippocampus medially and compress the perihippocampal fissures (Fig. 3.6.21) (Holodny et al. 1998a).

391

392

3  Brain, Head, and Neck

Fig. 3.6.17  Sagittal T1-weighted image. Chronic obstructive hydrocephalus due to aqueductal stenosis. The lateral and third ventricles are markedly dilated, but the fourth ventricle is normal in size

Fig. 3.6.16  Axial (a) and sagittal (b) T1-weighted images. A typical colloid cyst is seen as a hyperintense lesion in the anterior third ventricle. Note the presence of mild hydrocephalus

In more severe hydrocephalus the frontal horns are ballooned. The angle formed by the medial wall of the frontal horns known as the septal angle is acute in the setting of hydrocephalus but obtuse when enlargement

is due to atrophy. The third ventricle may also become enlarged and rounded. With atrophy, enlargement of the third ventricle maintains the parallel orientation of the walls. The roof of the third ventricle may become flat due to mass effect from dilated lateral ventricles. With more severe hydrocephalus, the corpus callosum is bowed and when hydrocephalus is long standing, the corpus callosum becomes thin. The height of the interpeduncular cistern may be decreased. The patients often have abnormalities of the gait with hydrocephalus. Dilatation of the forth ventricle often indicates communicating hydrocephalus. It should be noted that with a chronic obstructive type of hydrocephalus, the fourth ventricle may dilate as well. This may occur in the following manner. As the temporal horns dilate the temporal lobes get displaced into the incisural notch causing an incisural block. The fourth ventricle may not always dilate or may only minimally dilate with hydrocephalus. Sulcal effacement is helpful in the diagnosis of hydrocephalus particularly among elderly patients, who typically have some degree of atrophy. Hydrocephalus can occasionally be present with large sulci and fissures, particularly when there is high convexity pacchionian granulation obstruction. In this setting the CSF gets dammed, causing widening of the sulci and fissures. In such cases the sulci and fissures return to normal width after shunt placement. In acute and subacute hydrocephalus there is transependymal flow of CSF into the periventricular white matter. This is seen as periventricular T2 prolongation particularly adjacent to frontal horns and atria of the lateral

3.6  Neurodegenerative Disorders

Fig. 3.6.18  Axial FLAIR images (a,b). Obstructive hydrocephalus at the level of the aqueduct due to right thalamic expansile glioma (a). Mild periventricular hyperintensity due to transependymal CSF resorption in this young patient (b)

Fig. 3.6.19  Axial post-contrast image of a patient with a large posterior fossa meningioma. Note mass effect and distortion of the aqueduct and dilatation of the temporal horns due to obstructive hydrocephalus

Fig. 3.6.20  Coronal T1-weighted post-contrast image demonstrates enhancement within the internal auditory canals (arrows) from leptomeningeal carcinomatosis. Note mild hydrocephalus secondary to decreased CSF resorption by the pacchionian granulations

393

394

3  Brain, Head, and Neck

Fig. 3.6.22  Axial long TR image of the same patient as in Fig. 3.6.17. Note lack of any periventricular hyperintensity from transependymal CSF resorption in this young patient with longstanding hydrocephalus

periventricular edema is often absent on MRI or CT (Fig. 3.6.22). 3.6.4.1  Normal Pressure Hydrocephalus

Fig. 3.6.21  Coronal image of a patient with aqueductal stenosis and hydrocephalus (a). Note dilatation of the temporal horns and effacement of the perihippocampal fissures due to medial displacement and mild rotation of the hippocampal formations. The perihippocampal fissures are visualized (arrows) in a different patient without hydrocephalus but who has mild atrophy (b)

ventricles (Fig. 3.6.18b). At times it may be difficult if not impossible to differentiate between microangiopathic and hydrocephalus-induced periventricular T2 prolongation. It should be noted that microvascular changes may result from hydrocephalus as shown in animal models (Holodny et al. 1998b). In chronic hydrocephalus the

Normal pressure hydrocephalus NPH is one of the few causes of gait impairment, motor deficits, and dementia in the elderly which can potentially be treated. NPH is classically characterized by dementia, urinary incontinence, and a gait disturbance that is often described as magnetic or apraxic. These clinical features can be seen in approximately a third of the patients and are nonspecific. A dramatic improvement can be seen in some patients after ventricular shunting. Improvement in motor dysfunction can be dramatic, but only modest cognitive improvement is usually seen after shunting. Severe dementia is suggestive of irreversible damage or the presence of coexisting AD (Golomb et al. 1994). Preoperative memory deficits can improve in some patients after ventricular shunting. In patients who are markedly demented the surgical procedure may not be justified because the patient’s symptoms may not improve sufficiently after shunting. Therefore, it is crucial to identify other coexisting conditions preoperatively (Golomb et al. 1994).

3.6  Neurodegenerative Disorders

Fig. 3.6.23  Axial T2-weighted images (a,b) of a 70-year-old patient with gait disturbances demonstrate ventricular enlargement and white matter changes typical of chronic microvascular

ischemia. Note lack of significant enlargement of the sulci. The patient’s symptoms improved following ventricular shunting

The patients with NPH often have very large ventricles which are not proportional to the degree of atrophy (Fig. 3.6.23). They often have only mild cognitive deficits, and at times NPH may be suspected in normal volunteers. Motor deficits on the other hand are often severe, rendering the patients wheelchair bound. A number of investigators have tried to develop MRI criteria to predict response to ventricular shunting, with variable results. Quantitative CSF studies may be helpful in differentiating atrophy from hydrocephalus (Bradley et al. 1991). PET and SPECT studies have shown diffuse metabolic and cerebral blood flow abnormalities in patients with hydrocephalus. Elevated periventricular apparent diffusion coefficient correlated with poor post-surgical outcome in patients with NPH in one study. Quantitative ADC measurements may help in selecting patients who will respond favorably to surgical intervention (Corkill et al. 2003).

3.6.5  Mesial Temporal Sclerosis Effective medical treatment of partial complex seizures originating in the temporal lobe can be difficult, and surgical intervention may be the only suitable option for some patients. Preoperative MR imaging is imperative for localization of the lesion prior to surgery. Mesial temporal sclerosis (MTS) changes on histology include neuronal loss and gliosis of the hippocampal formation. MRI has been shown to detect MTS reliably by demonstrating atrophy and T2 prolongation of the hippocampal formation (Fig. 3.6.24). Thin-section MRI in the coronal plane works best in the detection of MTS. Quantified T2 relaxation times are more sensitive in the evaluation of MTS than visual inspection of the T2-weighted images. Some patients with MTS and T2 prolongation may have no appreciable atrophy.

395

396

3  Brain, Head, and Neck

Fig. 3.6.24  Images of a patient with MTS and temporal lobe epilepsy coronal FLAIR image (a) shows mild hyperintensity of the left hippocampal formation with mild associated atrophy best seen on the coronal T2-weighted image (b)

References 1.

2.

3.

4.

5.

6.

Abe K, Yorifuji S, Nishikawa Y (1993) Reduced isotope uptake restricted to the motor area in patients with amyotrophic lateral sclerosis. Neuroradiology 35:410–411 Bahn MM, Parchi P (1999) Abnormal diffusion-weighted magnetic resonance images in Creutzfeldt-Jakob disease. Arch Neurol 56:577–583 Ball MJ, Fisman M, Hachinski V, Blume W, Fox A, Kral VA, Kirshen AJ, Fox H, Merskey H (1985) A new definition of Alzheimer’s disease: a hippocampal dementia, Lancet 1:14–16 Barboriak DP, Provenzale JM, Boyko OB (1994) MR diagnosis of Creutzfeldt-Jakob disease: significance of high signal intensity in the basal ganglia. Am J Roentgenol 162:137–140 BradleyWG, Whittemore AR, Kortman KE, Watanabe AS et al (1991) Marked cerebrospinal fluid void: indicator of succesful shunt in patients with suspected normal-pressure hydrocephalus. Radiology 178:459–466 Corkill RG, Garnett MR, Blamire AM et al (2003) Multi-modal MRI in normal pressure hydrocephalus identifies pre-operative haemodynamic and diffusion coefficient changes in normal appearing white matter correlating with surgical outcome. Clin Neurol Neurosurg105:193–202

7.

8.

9.

10. 11.

12.

13.

14.

Dahlbeck JW, McCluney KW, Yeakley JW, Fenstermacher MJ, Bonmati C, Van Horn G III, Aldag J (1991) The ineruncal distance: a new MR measurement for the hippocampal atrophy of Alzheimer disease. AJNR Am J Neuroradiol 12:931–932 Demaerel P, Baert AL, Vanopdenbosch L, Robberecht W, Dom R (1997) Diffusion-weighted magnetic resonance imaging in Creutzfeldt-Jakob disease. Lancet 22:847–848 Di Rocco A, Molinari S, Stollman AL, Decker A, Yahr MD (1993) MRI abnormalities in Creutzfeldt-Jakob disease. Neuroradiology 35:584–585 Dickson DW (2001) Neuropathology of Pick’s disease. Neurology 56:S16–S20 Ellis CM, Simmons A, Jones DK, Bland J, Dawson JM, Horsfield MA, Williams SC, Leigh PN (1999) Diffusion tensor MRI assesses corticospinal tract damage in ALS. Neurology 53:1051–1058 Freidman DP, Tartaglino LM (1993) Amyotrophic lateral sclerosis: hyperintensity of the corticospinal tracts on MR images of the spinal cord. Am J Roentgenol 160:604–606 George AE, de Leon MJ, Kalnin A, Rosner L, Goodgold A, Chase N (1986) Leukoencephalopathy in normal and pathologic aging: 2. MRI of brain lucencies. AJNR Am J Neuroradiol 7:567–570 George, AE, de Leon MJ et al (1990) CT diagnostic features of Alzheimer disease: importance of the choroidal/hippocampal fissure complex. AJNR Am J Neuroradiol 11:101–107

3.6  Neurodegenerative Disorders 15. Golomb J, deLeon MJ, George AE, Kluger A, Convit A et al (1994) Hippocampal atrophy correlates with severe cognitive impairment in elderly patients with suspected normal pressure hydrocephalus. J Neurol Neurosurg Psychiatr 57:590–593 16. Goodin DS, Rowley HA, Olney RK (1988) Magnetic resonance imaging in amyotrophic lateral sclerosis. Ann Neurol 23:418–420 17. Graham GD, Petroff OA, Blamire AM, Rajkowska G, Goldman-Rakic P (1993) Proton magnetic resonance spectroscopy in Creutzfeldt-Jakob disease. Neurology 43:2065–2068 18. Graham JM, Papadakis N, Evans J, Widjaja E, Romanowski CA, Paley MN, Wallis LI, Wilkinson ID, Shaw PJ, Griffiths PD (2004) Diffusion tensor imaging for the assessment of upper motor neuron integrity in ALS. Neurology 63:2111–2119 19. Holodny AI, George AE, Golomb J, de Leon MJ, Kalnin AJ (1998a) The perihippocampal fissures: normal anatomy and disease states. Radiographics 18:653–665 20. Holodny AI, Waxman R, George AE, Rusinek H, Kalnin AJ, de Leon MJ (1998b) MR differential diagnosis of normal –pressure hydrocephalus and Alzheimer disease: significance of the perihippocampal fissures. AJNR Am J Neuroradiol 19:813–819 21. Hong YH, Lee KW, Sung JJ, Chang KH, Song IC (2004) Diffusion tensor MRI as a diagnostic tool of upper motor neuron involvement in amyotrophic lateral sclerosis. J Neurol Sci 227:73–78 22. Hyman BT, Van Hoesen GW, Damosia AR, Barnes CL (1984) Alzheimer’s disease: cell-specific pathology isolated the hippocampus. Science 225:1168–1170 23. Ishikawa K, Nagura H, Yokota T, Yamanouchi H (1993) Signal loss in the motor cortex on magnetic resonance images in amyotrophic lateral sclerosis. Ann Neurol 33:218–222 24. Iwasaki Y, Kinoshita M, Ikeda K, Takamiya K, Shiojima T (1991) MRI in patients with amyotrophic lateral sclerosis: correlation with clinical features. Int J Neurosci 59:253–258 25. Jack CR, Petersen RC, O’Brien PC, Tangalos EG (1992) MR-based hippocampal volumetry in the diagnosis of Alzheimer’s disease. Neurology 42:183–188 26. Joachim CL, Morris, JH Selkoe D (1988) Clinically diagnosed Alzheimer’s disease: autopsy neuropathological results in 150 cases. Ann Neurol 24:50–56 27. Kemper T (1984) Neuroanatomical and neuropathological changes in normal aging and in dementia. In: Albert ML (ed) Clinical neurology of aging. Oxford University Press, New York, pp 9–52 28. Kido DK, Caine ED, LeMay M, Ekholm S, Booth H, Panzer R (1989) Temporal lobe atrophy in patients with Alzheimer disease: a CT study. AJNR Am J Neuroradiol 10:551–555 29. Lehericy S, Baulac M, Chiras J et al (1994) Amygdalohippocampal MR volume measurements in the early stages of Alzheimer disease. AJNR Am J Neuroradiol 15:929–937

30. LeMay M, Stafford JL, Sandor T, Albert M, Haykol H, Zamani A (1986) Statistical assessment of perceptual CT scan ratings in patients with Alzheimer’s-type dementia. J Comput Assist Tomogr 10:802–809 31. Leon MJ de, George AE, Stylopoulos LA et al ( 1989) Early marker for Alzheimer disease: The atrophic hippocampus. Lancet 2:672–673 32. Leon MJ de, Golomb J, George AE et al (1983) The radiologic prediction of Alzheimer disease: the atrophic hippocampal formation. AJNR Am J Neuroradiol 14:897–906 33. McKhann G (1984) Clinical diagnosis of Alzheimer’s disease. Neurology 34: 939–944 34. Milton WJ, Atkas SW, Lavi E, Mollman JE (1991) Magnetic resonance imaging of Creutzfeldt-Jakob disease. Ann Neurol 29:438–440 35. Monte SM de la (1989) Quantification of cerebral atrophy in pre-clinical and end-stage Alzheimer’s disease. Ann Neurol 25:450–459 36. Morris JC, McKeel DW Jr, Fulling K, Torack RM, Berg L (1988) Validation of clinical diagnostic criteria for Alzheimer’s disease. Ann Neurol 24:17–22 37. Mu Q, Xie J, Wen Z, Weng Y, Shuyun Z (1999) A quantitative study of the hippocampal formation, the amygdala and the temporal horn of the lateral ventricle in healthy subjects 40 to 90 years of age. AJNR Am J Neuroradiol 20:207–211 38. Narkiewicz O, de Leon MJ, Convit A, George AE, Wegiel J, Morys J, Bobinski M, Golomb J, Miller DC, Wisniewski HM (1993) Dilatation of the lateral part of the transverse fissure of the brain in Alzheimer disease. Acta Neurobiol Exp 53:457–465 39. Oba H, Araki T, Ohtomo K, Monzawa S, Uchiyama G, Koizumi K et al (1993) Amyotrophic lateral sclerosis: T2 shortening in motor cortex at MR imaging. Radiology 189:843–846 40. Oliva D, Carella F, Savoiardo M, Strada L, Giovannini P, Tasta D et al (1993) Clinical and magnetic resonance features of the classical and akinetic-rigid variants of Huntington’s disease. Arch Neurol 50:17–19 41. Pantoni L, Garcia JH (1996) The significance of cerebral white matter abnormalities 110 years after Binswanger’s report. A review. Stroke 26:1293–1301 42. Parazzini C, Mammi S, Comola M, Scotti G (2003) Magnetic resonance diffusion-weighted images in CreutzfeldtJakob disease: case report. Neuroradiology 45:50–52 43. Prince JL, Davis PB, Morris LC, White DL (1991) The distribution of tangles, plaques and related immunohistochemical markers in healthy aging and Alzheimer’s disease, Neurobiol Aging 12:295–312 44. Rose SE, Chen F, Chalk JB et al (2000) Loss of connectivity in Alzheimer’s disease: an evaluation of white matter tract integrity with color coded MR diffusion tensor imaging. J Neurol Neurosurg Psychiatr 69:528–530 45. Rossor MN (2001) Pick’s disease: a clinical overview. Neurology 56:S3–S5

397

398

3  Brain, Head, and Neck 46. Rusinek H, deLeon MJ, George AE et al (1991) Alzheimer disease: measuring loss of cerebral gray matter with MR imaging. Radiology 178:109–114 47. Sandor T, Albert M, Stafford J, Harpley S (1988) Use of computerized CT analysis to discriminate between Alzeimer patients and normal control subjects. AJNR 9:1181–1187 48. Savoiardo M, Strada L, Oliva D, Girotti F, D’Incerti L (1991) Abnormal MRI signal in the rigid form of Huntington’s disease. J Neurol Neurosurg Psychiatr 54:888–891 49. Tartaro A, Fulgente T, Delli Pizzi C et al (1993) MRI alterations as an early finding in Creutzfeldt-Jakob disease. Eur J Radiol 17:155–158 50. Tatsch K, Koch W, Linke R, Poepperl G, Peters N, Holtmannspoetter M, Dichgans M (2003) Cortical hypometabolism and crossed cerebellar diaschisis suggest subcortically induced disconnection in CADASIL: an 18F-FDG PET study. J Nucl Med 44:862–869

51. Terry RD, Peck A, DeTheresa R, Schecter R, Horoupian DS (1981) Some morphometric aspects of the brain in senile dementia of the Alzheimer type. Ann Neurol 10:184–192 52. Tschampa HJ, Mürtz P, Flacke S, Paus S, Schild HH, Urbach H (2003) Thalamic involvement in sporadic CreutzfeldtJakob disease: a diffusion-weighted MR imaging study. AJNR Am J Neuroradiol 24:908–915 53. Udaka F, Sawada H, Seriu N, Shindou K, Nishitani N, Kameyama M (1992) MRI and SPECT findings in amyotrophic lateral sclerosis. Neuroradiology 34:389–393 54. van den Boom R, Lesnik Oberstein SA, Spilt A, Behloul F, Ferrari MD, Haan J, Westendorp RG, van Buchem MA (2003) Cerebral hemodynamics and white matter hyperintensities in CADASIL. J Cereb Blood Flow Metab 23:599–604 55. Yagashita A, Nakano I, Oda M, Hirano A (1994) Location of the corticospinal tract in the internal capsule at MR imaging. Radiology 191:455–460

3.7  Pituitary Gland and Parasellar Region

3.7  Pituitary Gland and Parasellar Region M. Kanagaki, N. Sato, and Y. Miki 3.7.1  Introduction A wide variety of diseases involve the sellar/parasellar region, mainly because numerous important anatomical structures are present in this region, such as the anterior and posterior pituitary, pituitary stalk, hypothalamus, cavernous sinus, circle of Willis, cranial nerves II–VI, bone, and the sphenoid sinus. The differential diagnoses of sellar/parasellar diseases are broad and may not always be straightforward, as many of these lesions, neoplastic or non-neoplastic, can display similar clinical, endocrinological, and radiological presentations. Overall knowledge of the anatomy, radiological findings, and clinical presentations of each sellar/parasellar disease is thus essential. In most instances, magnetic resonance (MR) imaging is the imaging modality of choice for the sellar/parasellar region, due to multiplanar capability and good soft tissue contrast. Computed tomography (CT) is also occasionally used to evaluate bony structures and calcifications. 3.7.2  Examination Techniques 3.7.2.1  Patient Positioning The prone position is used in most situations. Other positions are usually only used if the patient is incapable of assuming the prone position.

readout gradient is set to eliminate overlap between fatty marrow and the pituitary (Table 3.7.1. This technique is important, particularly when posterior pituitary high signal is to be evaluated or when the height of the pituitary gland is to be measured (Sato et al. 1991; Taketomi 2004). 3.7.2.4  Imaging Planes Sagittal T1-weighted images should be obtained in all sellar/parasellar examinations. This is essential for observing midline structures, as mid-sagittal T1-weighted imaging normally displays the anterior and posterior lobes and stalk in the same plane. Sagittal T1-weighted imaging is often very useful in evaluating lesions that involve the neurohypophyseal system, as posterior pituitary high signal is most clearly identified on these images. Coronal images most clearly demonstrate relationships between the pituitary, pituitary adenoma, cavernous sinuses, and optic chiasm. Coronal T2- and T1-weighted pre- and post-gadolinium (including dynamic scanning, Miki 1990) images are thus usually obtained in cases of pituitary adenoma. Axial T2-weighted imaging is often useful to evaluate relationships between lesions and the brain and to exclude the presence of brain lesions. Sagittal post-gadolinium imaging is useful to determine the extent of lesions involving the neurohypophyseal system (including dynamic scanning, Sato et al. 1993) and to exclude the presence of pineal lesions (important for germinoma cases). 3.7.2.5  Thickness of Slices

A head coil, used for routine head imaging, is usually used for imaging of the sellar/parasellar region. A multichannel phased-array head coil can be used for reducing acquisition time when the MR unit is equipped with parallel acquisition capability.

Obtaining high-resolution images is important, as various tiny structures are present in the sellar/parasellar region and lesions in this region are also often small. To obtain high spatial resolution, a thin slice (≤3 mm without gap), fine matrix (256 × 256 or more) and small field of view (≤20 cm) are needed (Elster 1993a; Kucharczyk et al. 1996).

3.7.2.3  Examination Sequences

3.7.2.6  Preferred Coverage

Sellar/parasellar imaging is vulnerable to magnetic susceptibility artifacts because of the presence of adjacent bone and air (Elster 1993b). Spin-echo sequences are thus preferable to gradient-echo sequences, although even spin-echo sequences cannot completely eliminate these artifacts (Sakurai et al. 1992). Usually, conventional spinecho sequences are used for pre- and post-gadolinium T1-weighted imaging, while fast spin-echo sequences are used for T2-weighted and dynamic contrast-enhanced imaging. It has been recommended that the direction of

In principle, if a parasellar lesion is suspected, the focus of examination should be on this lesion, and good quality images of the fine structures are needed. However, this can vary with lesions and disease. For example, in cases with pituitary adenoma, both the pituitary gland and the entire cavernous sinus should be imaged. In cases with lymphoma, wider coverage is needed as the lesion can extend diffusely along the dura. The whole brain sometimes needs to be imaged to exclude the presence of brain lesions.

3.7.2.2  Selection of Coils

399

400

3  Brain, Head, and Neck Table 3.7.1  Example of sellar/parasellar imaging protocols Sequence

Plane

TR

TE

ETL

Matrix

Thickness

a

FOV

NEX

Pituitary adenoma T1 (SE)

Sagittal

500

Min

NA

256 × 256

3

18

2

T1 (SE)

Coronal

500

Min

NA

256 × 256

3

18

2

T2 (FSE)

Coronal

2,500

90

16

256 × 256

3

18

1

Dynamic T1 (FSE)

Coronal

500

Min

8

256 × 256

3

18

1

T1 (SE) Gd

Coronal

500

Min

NA

256 × 256

3

18

1

T1 (SE) Gd (optional)

Sagittal

500

Min

NA

256 × 256

3

18

1

Sagittal

500

Min

NA

256 × 256

3

18

2

Coronal

500

Min

NA

256 × 256

3

18

2

90

16

256 × 256

3

18

1

Parasellar tumor T1 (SE) T1 (SE) T2 (FSE)

Axial

T1 (SE) Gd

Coronal

500

Min

NA

256 × 256

3

18

1

T1 (SE) Gd b

Axial

500

Min

NA

256 × 256

3

18

1

T1 (SE) Gd

Sagittal

500

Min

NA

256 × 256

3

18

1

b

2,500

Diabetes insipidus, pituitary dwarfism T1 (SE)

Sagittal

500

Min

NA

256 × 256

3

18

2

T1 (SE)

Coronal

500

Min

NA

256 × 256

3

18

2

T2 (FSE)

Coronal

2,500

90

16

256 × 256

3

18

1

Dynamic T1 (FSE) (optional)

Sagittal

500

Min

8

256 × 256

3

18

1

T1 (SE) Gd (optional)

Sagittal

500

Min

NA

256 × 256

3

18

1

T1 (SE) Gd (optional)

Axial

500

Min

NA

256 × 256

3

18

1

(Whole brain)

c

TR repetition time, TE echo time, ETL echo train length, FOV field of view, NEX number of excitations, SE spin-echo, FSE fast spinecho, Gd gadolinium, Min minimum, NA not applicable a Thickness for dynamic study may be increased for macroadenoma to cover the whole mass b Imaging of the whole brain is occasionally needed c Post-gadolinium images of the whole CNS should be additionally obtained to exclude tumor dissemination if germinoma is suspected

3.7.3  Normal Anatomy 3.7.3.1  Adenohypophysis and Neurohypophysis The pituitary lies within the bony sella turcica, which surrounds this glandular tissue inferiorly and laterally. Superiorly, the pituitary is covered by the diaphragma sella, a extension of the dura mater. Lateral to the sella are the

cavernous sinuses, while anteroinferior is the sphenoid sinus and anterosuperior is the optic chiasm. The pituitary comprises two anatomically and functionally distinct parts, the adenohypophysis, and the neurohypophysis. The adenohypophysis arises from Rathke’s pouch from the oral cavity, and the neurohypophysis arises from neural ectoderm at the floor of the forebrain. The adenohypophysis is composed of three parts: an-

3.7  Pituitary Gland and Parasellar Region

terior lobe; intermediate lobe (pars intermedia); and pars tuberalis, an extension of epithelium covering around the infundibulum. The adenohypophysis contains specialized cell types (somatotrophs, lactotrophs, mammosomatotrophs, and thyrotrophs).These are thought to maintain fluidity of transdifferentiation; for example, somatotrophs convert to mammosomatotrophs and lactotrophs during pregnancy and to thyrotrophs in hypothyroidism. Most of the anterior pituitary lacks any major direct arterial blood supply; instead, it is bathed in a dense capillary network of pituitary portal blood containing both hypothalamic hormones released in the median eminence and anterior pituitary hormones. The portal blood empties into the cavernous sinuses and then into the peripheral circulation via internal jugular veins. The intermediate lobe represents a vestigial remnant of Rathke’s pouch in humans and is not normally visualized on imaging studies. However, a small cyst can be seen at the location of the pars intermedia. The neurohypophysis is composed of the infundibulum, pituitary stalk, and pars nervosa. The neuro­ hypophysis stores and releases the hypothalamic hormones oxytocin and vasopressin, and contains axons and nerve terminals of larger neurons originating in the paraventricular and supraoptic nuclei of the hypothalamus. The neurohypophysis is supplied by the inferior hypophyseal artery and drains into the cavernous sinus via inferior–posterior pituitary veins, where vasopressin and oxytocin are released directly into the systemic circulation. The size of the normal pituitary greatly varies according to age, sex, and physiological status. In males, height is usually ≤8 mm. The pituitary in males is at its maximum size during adolescence, usually with a flat or concave superior surface. In females, the pituitary is ≤10 mm in height during adolescence and reaches up to 12 mm in late pregnancy or postpartum periods (Elster 1993a). At these times, the superior surface of the pituitary in females is usually convex. After delivery, the pituitary decreases in size over a period of around 8 months (Miki et al. 2005). The pituitary is significantly higher in preterm infants than in full-term infants (Kiortsis et al. 2004). In both males and females, pituitary height tends to decline with age, but increases again in females after menopause, due to a temporary increase in gonadotropin-releasing hormone (Tsunoda et al. 1997). An enlarged pituitary may be seen in patients with primary hypothyroidism, but will resolve with treatment of hypothyroidism (Shimono et al. 1999). Intensities on T1-weighted images of both anterior and posterior pituitaries also vary according to physiological and pathological status. The anterior pituitary is usually slightly hypointense to cerebral white matter (Fig. 3.7.1), but may be hyperintense on T1-weighted image in neonates and females during late pregnancy or postpartum periods (Cox and Elster 1991; Miki et al. 2005), representing increased hormonal activity. After delivery,

Fig. 3.7.1  Sagittal T1-weighted MR image. 1 anterior pituitary gland, 2 posterior pituitary gland, 3 infundibulum, 4 tuber cinereum, 5 mammillary body, 6 infundibular recess, 7 optic chiasm, 8 sphenoid sinus, 9 clivus, 10 pons

intensity of the anterior pituitary decreases over a postpartum period of around 8 months (Miki et al. 2005). Increased signal intensity in the anterior pituitary gland on T1-weighted imaging is also seen in patients receiving intravenous hyperalimentation or in patients with chronic hepatic failure, due to deposition of manganese. Conversely, the posterior pituitary is normally hyperintense on T1-weighted images, due to the effect of stored neurosecretory granules on T1 relaxation time (Fig. 3.7.1) (Fujisawa et al. 1989; Sato et al. 1995b). However, this hyperintensity may involve only a portion of the posterior pituitary (Miki et al. 1992). Such hyperintensity of the posterior pituitary is absent in patients with diabetes insipidus, and may be absent in patients with diabetes mellitus (Fujisawa et al. 1996) or those undergoing hemodialysis (Sato et al. 1995a), which will be discussed later in detail. 3.7.3.2  Hypothalamic–Pituitary Axis The hypothalamus is located at the base of the brain, below and beside the third ventricle and just above the optic chiasm and pituitary gland. The infundibulum appears as a thin stalk that tapers from the floor of the third ventricle to the pituitary gland. The diameter of the stalk normally does not exceed that of the basilar artery (about 3 mm). The optic chiasm is seen anterior to the infundibulum. The hypothalamus is considered the coordinating center of the endocrine system and is also involved in several

401

402

3  Brain, Head, and Neck

important non-endocrine functions such as temperature regulation, activity of the autonomic nervous system and appetite control. Hormone-releasing systems differ between the adenohypophysis and neurohypophysis. The adenohypophysis secretes growth hormone (GH), thyroid-stimulating hormone (TSH), gonadotropins (LH/FSH), and adrenocorticotropin (ACTH). These secretions are in turn regulated by growth hormone-releasing hormone (GRH), thyrotropin-releasing hormone (TRH), gonadotropin-releasing hormone (GRH), and corticotrophin-releasing hormone (CRH), respectively. Releasing hormones are produced in the hypothalamus, and terminals of the hypothalamic neurons are located in the median eminence, where the releasing hormones are secreted into a primary capillary plexus. These capillaries empty into the portal venous circulation, travel down the pituitary stalk, and bathe the anterior pituitary gland in a secondary capillary plexus. After receiving stimulation from the releasing hormones, adenohypophyseal hormonal secretion is regulated. Neurohypophyseal hormones, on the other hand, are directly transported to the posterior pituitary without a portal system. Two kinds of hormones, vasopressin and oxytocin, are produced in the hypothalamus. These pass through the neural component of the pituitary stalk and

are stored in nerve terminals in the posterior pituitary. As a result, no hormone-producing cells are present in the posterior pituitary, which instead represents a storage organ for the release of hormones into the systemic circulation. 3.7.3.3  Cavernous Sinus The cavernous sinuses are clinically important structures, as major vessels and many cranial nerves run through these paired structures, which are situated lateral to the pituitary gland. The cavernous sinus contains the venous plexus, internal carotid artery, periarterial sympathetic nerve fibers, fibrous tissue, and cranial nerves III (oculomotor nerve), IV (trochlear nerve), V1 (ophthalmic nerve), V2 (maxillary nerve) and VI (abducens nerve) (Fig. 3.7.2a). Cranial nerves III, IV, V1, and V2 line up vertically along the dura matter from top down. Only cranial nerve VI runs inside the cavernous sinus, inferior and lateral to the internal carotid artery. Cranial nerve VI frequently splits into anywhere from two to five rootlets. When multiple unilateral cranial nerve symptoms from cranial nerve III to VI occur, disease location in the cavernous sinus should be strongly suspected.

Fig. 3.7.2  a Structures of the cavernous sinus. 1 pituitary gland, 2 internal carotid artery, 3 periarterial sympathetic nerve fibers, 4 oculomotor nerve (III), 5 trochlear nerve (IV), 6 first division of the trigeminal nerve (ophthalmic nerve, V1), 7 second division of the trigeminal nerve (maxillary nerve, V2), 8 abducens nerve (VI), 9 inner dura mater (meningeal layer), 10 outer dura mater (periosteal layer). b Coronal contrastenhanced CISS image of the cavernous sinus. 1 optic chiasm, 2 sphenoid sinus, 3 pituitary gland, 4 internal carotid artery, 5 oculomotor nerve (III), 6 trochlear nerve (IV), 7 abducens nerve (VI), 8 first division of the trigeminal nerve (ophthalmic nerve, V1), 9 second division of the trigeminal nerve (maxillary nerve, V2)

3.7  Pituitary Gland and Parasellar Region

The venous plexus is a blood-filled, serpentine network passing through the connective tissues of the cavernous sinus. These cavernous venous channels communicate with both cavernous sinuses (anterior and posterior intercavernous venous sinuses), which are situated just inferior to the pituitary gland. In addition to the internal carotid artery and venous channels, branches of the internal carotid artery also pass through the sinus. The meningohypophyseal trunk and inferior lateral trunk represent two major branches from the intracavernous segment of the internal carotid artery, and the tentorial artery, dorsal meningeal artery, and inferior hypophyseal artery branch from the meningohypophyseal trunk. The cavernous sinus communicates with pituitary veins and surrounding vessels, such as the clival venous plexus, inferior and superior petrosal venous sinus, superior ophthalmic vein, and sphenoparietal sinus. When a carotid–cavernous fistula or a dural arteriovenous fistula of the cavernous sinus occurs, the pituitary gland becomes swollen and these vessels and cavernous venous channels are dilated (Sato et al. 1997). The cavernous sinus is covered by the inner dura matter (meningeal layer), and transits to the arachnoid matter in the suprasellar area without inverting along the pituitary capsule. The outer dura matter (periosteal layer) is separated at and located inferior to the cavernous sinus. No dural wall is present between the cavernous sinus and pituitary gland, facilitating the invasion of pituitary adenoma into the cavernous sinus (Dietemann et al. 1998). Cavernous sinus anatomy is seen well on coronal MR imaging, showing isointensity on T1-weighted images and high intensity on T2-weighted images with a covering hypointense line indicating the dura matter. After injection, the internal carotid artery appears as a signal void in the well-enhanced cavernous sinus, and cranial nerves III, V, and VI may be seen on conventional T1-weighted imaging (Korogi et al. 1991). Three-dimensional (3D) constructive interference in steady state (CISS) is a highspatial-resolution, refocused, gradient-echo sequence that shows increasing contrast with increasing concentration of gadolinium-based contrast agent. Contrast-enhanced 3D CISS is the best sequence to demonstrate fine structures in the cavernous sinus (Yagi et al. 2005). After the injection of gadolinium, with the well-enhanced venous plexus of the cavernous sinus playing a role similar to that of cerebrospinal fluid (CSF), cranial nerves stand out as dark spots or lines. Contrast-enhanced 3D CISS offers clear images of almost all of cranial nerves III, IV, V1, and V2 resembling an anatomic diagram (Fig. 3.7.2b), although, because of its small size and proximity to cranial nerve III, the detectability of cranial nerve IV is lower than that of other cranial nerves. Some non-enhanced areas are also observed in addition to cranial nerves and are considered to correspond to periarterial sympathetic fibers, fibrous tissues or branches of the internal carotid artery as described above.

3.7.3.4  Liliequist’s Membrane Liliequist’s membrane represents arachnoid trabeculae dividing the chiasmatic, interpeduncular, and prepontine cisterns. This membrane is composed of sellar, diencephalic, and mesencephalic segments (Fushimi et al. 2003, 2006). Sellar segments extend from the dorsum sellae, and divide posterosuperiorly into diencephalic segments that attach to the mamillary body and posteroinferiorly into the mesencephalic segments that spread over the basilar artery in front of the pons (Fig. 3.7.3). In most microneurosurgical operations, arachnoid membranes are incised and the subarachnoid cisterns are opened. Liliequist’s membrane represents an important anatomical landmark in neurosurgical approaches for parasellar and skull-base operations (Matsuno et al. 1988). This membrane is also important for endoscopic third ventriculostomy (Fushimi et al. 2003). Suprasellar arachnoid cyst, which may compress the third ventricle floor upward, is considered to arise from Liliequist’s membrane. The cyst wall may consist of the diencephalic segment or of both diencephalic and mesencephalic segments. CISS imaging may help visualize details of the cyst wall (see Fig. 3.7.16) (Fushimi et al. 2006). Liliequist’s membrane also plays an important role in dynamic movement of CSF. During neurosurgery, Liliequist’s membrane may prevent hematoma in the chi-

Fig. 3.7.3  Liliequist’s membrane. Sagittal CISS MR image shows the sellar (white arrow), diencephalic (white arrowhead), and mesencephalic segments (black arrowhead) of Liliequist’s membrane

403

404

3  Brain, Head, and Neck

asmatic cistern from spreading into the interpeduncular cistern. In perimesencephalic subarachnoid hemorrhage, Liliequist’s membrane also usually prevents hematoma in the prepontine cistern from spreading into the chiasmatic cistern, although hematoma rarely traverses the diencephalic segment of Liliequist’s membrane to reach the chiasmatic cistern (Matsuno et al. 1988). MR imaging using high-spatial-resolution MR cisternography, such as a CISS sequence, can depict the details and variations of this membrane (Fig. 3.7.3) (Fushimi et al. 2003, 2006).

that differentiates this condition from other cystic lesions in which the stalk is displaced or obliterated (Fig. 3.7.4). 3.7.4.2  Central Diabetes Insipidus

Empty sella results from arachnoid herniation through an incomplete diaphragma sellae. This condition is either primary, resulting from a congenital diaphragmatic defect that is thought to allow CSF pressure to enlarge the sella, or secondary, resulting from injury to the diaphragm by pituitary surgery, radiation, or infarction. Primary empty sella is found in more than 5% of adults, but is rare in children. Females are more frequently affected than males. Pituitary function is typically normal in primary empty sella, and no treatment is necessary. However, “empty sella syndrome” has been described in patients who suffer from a combination of headache, pituitary dysfunction, and visual disturbance (De Marinis et al. 2005). MR imaging shows the pituitary fossa filled with fluid isointense with CSF. Differential diagnoses include arachnoid cyst, Rathke’s cleft cyst, pars intermediate cyst of the pituitary, and epidermoid. In cases of empty sella, the pituitary stalk courses directly from the hypothalamus to the flattened gland without deviation, a finding

Central diabetes insipidus is a lack of vasopressin that causes excessive production of diluted urine. This condition can result from decreased production of vasopressin by the hypothalamus or a failure to release this hormone. Both primary (idiopathic) and secondary types exist. The latter can be caused by brain injury, surgery, neoplasms (germ cell tumor, craniopharyngioma, metastasis, etc.), sarcoidosis, tuberculosis, and Langerhans’ cell histiocytosis. A diagnosis of diabetes insipidus should be made following endocrinological examination, but MR imaging is useful in the diagnosis of this condition, as the bright signal of the posterior pituitary on T1-weighted image is absent in patients with diabetes insipidus (Fig. 3.7.5) (Fujisawa et al. 1987b). MR imaging is also helpful to exclude secondary causes of central diabetes insipidus. Psychogenic polydipsia is clinically characterized by polyuria and polydipsia similar to diabetes insipidus. MR imaging is again useful in differential diagnosis between these two conditions, because the bright posterior pituitary on T1-weighted imaging is intact in patients with psychogenic polydipsia (Moses et al. 1992). Conversely, attention is needed to the fact that some conditions display disappearance of the posterior pituitary bright signal without central diabetes insipidus. Since the bright signal depends on the amount of stored vasopressin granules, the signal is deceased or absent in hyperstimulated states, such as in patients undergoing hemodialysis (Sato et al. 1995a) or with nephrogenic diabetes insipidus (Sato et al. 1993). Imura et al. (1993) proposed the disease entity of “lymphocytic infundibuloneurohypophysitis,” which

Fig. 3.7.4  Empty sella. Cerebrospinal fluid (CSF) surrounds the pituitary infundibulum (arrows), but does not cause displacement on sagittal T1-weighted image (a), coronal T2-weighted

image (b), and coronal contrast-enhanced T1-weighted image (c). CSF collection is seen in the expanded sella turcica, which displaces the pituitary gland inferiorly (arrowheads)

3.7.4  Pathological Conditions 3.7.4.1  Empty Sella

3.7  Pituitary Gland and Parasellar Region

evation up to 100 ng/ml may be due to stalk compression by lesions other than prolactinoma (Freda et al. 1996). For the differential diagnosis of hyperprolactinemia, basal PRL levels >85 ng/ml, in the absence of renal failure and PRL-enhancing drugs, and a PRL increment of <30% after TRH accurately rule out functional hyperprolactinemia, and are typical of prolactinomas (Gsponer et al. 1999). 3.7.4.4  Microadenomas

Fig. 3.7.5  Diabetes insipidus. Sagittal T1-weighted image shows absence of posterior pituitary high signal (arrow)

may represent a cause of what had been considered as “idiopathic” central diabetes insipidus. This condition will be discussed in detail later in this section. 3.7.4.3  Pituitary Adenoma Pituitary adenomas are benign tumors arising from cells of the anterior pituitary gland. Most pituitary adenomas are seen in adults, but 2–10% occurs in pediatric populations. Pituitary adenomas are classified according to size (microadenoma vs. macroadenoma), types of hormone secreted and immunohistochemical findings. Presenting symptoms may reflect hypersecretion of anterior pituitary hormones such as amenorrhea/galactorrhea (prolactin [PRL]), Cushing’s syndrome (ACTH), acromegaly or gigantism (GH), and hyperthyroidism (TSH). Majorities of pituitary adenomas are hormonally positive on immunohistochemical analysis, but are not always “clinically” active. Non-functioning pituitary adenomas are those that are either non-secretory or are clinically silent and typically present with clinical symptoms related to effects on parasellar structures, including visual disturbance, headache and symptoms related to hypopituitarism. Diabetes insipidus, caused by mass effects on the hypothalamus or infundibulum, is rare in patients with pituitary adenoma. Patients with prolactinoma present with amenorrhea, galactorrhea, infertility, loss of libido, and impotence in males. Prolactinoma tends to be larger in males than in females, because diagnosis is often delayed in men. Serum prolactin levels in patients with prolactinoma are usually >100 ng/ml (normal, <20 ng/ml). Mild PRL el-

Functioning pituitary adenomas tend to be detected as microadenomas (adenomas ≤1cm in diameter), as the clinical symptoms related to hormonal activity usually occur in the early stages. MR imaging is preferable to CT for identification of microadenoma, although a normal MR examination cannot completely exclude the possibility of undetected pituitary microadenoma. Microadenomas will typically appear hypointense to isointense on T1-weighted images and iso- or hyperintense on T2-weighted images. However, GH-producing adenoma tends to be hypointense on T2-weighted images (Fig. 3.7.6) (Hagiwara 2003). Secondary signs, such as deviation of the infundibulum to the side opposite the tumor or upward convexity of the gland, are not always reliable. The pituitary gland, both anterior and posterior lobes, is outside the blood–brain barrier and hypervascular, and thus enhances intensely after intravenous injection of contrast medium. After administration of gadolinium, adenoma typically displays fewer enhancements compared with the normal pituitary gland. The sensitivity of precontrast MR imaging for detecting microadenomas is about 60–80%, and an additional conventional post-gadolinium study detects about 5–10% more lesions (Elster 1994). Dynamic contrast-enhanced imaging has been widely used to detect microadenomas (Bartynski and Lin 1997; Davis et al. 1994; Miki et al. 1990; Rand et al. 2002). On dynamic MR imaging, peak enhancement of adenoma is delayed compared with normal pituitary tissue, and good contrast between adenoma and normal pituitary can thus be achieved on early images (<2 min) after rapid (≥4 ml/s) injection of gadolinium (Fig. 3.7.7). The usefulness of spoiled gradient-recalled acquisition in the steady-state sequence for ACTH-producing microadenomas has also been reported (Batista et al. 2005). 3.7.4.5  Macroadenomas Pituitary macroadenomas are defined by a diameter >10 mm. Approximately 25% of adenomas are non-functioning and are usually diagnosed relatively late, at a time when the lesion is large enough to compress adjacent structures such as the optic chiasm, resulting in visual field deficits, or pituitary gland, causing hypopituitarism. Such tumors may extend upward to compress the chiasm

405

406

3  Brain, Head, and Neck

and infundibulum, laterally to displace or invade the cavernous sinus, or inferiorly to expand the sellar floor and invade into the sphenoid sinus (Kucharczyk et al. 1986). When the tumor extends superiorly into the suprasellar cistern, a characteristic “figure-of-eight” appearance may be seen due to the effect of the diaphragma sella. Pituitary macroadenoma, especially prolactinoma and GH-producing adenoma, can destroy bony structures and invade inferiorly and laterally, mimicking a very ex-

tensive lesion. Even in very extensive lesions that may at first be thought to be non-pituitary in origin, adenoma should always be excluded (Freda et al. 1996). Signal intensities of macroadenomas are variable. MR imaging typically shows uniform contrast enhancement after intravenous contrast medium administration. However, MR imaging shows a non-enhancing region within the tumor in some macroadenomas with cystic change, necrosis, or hemorrhage (Fig. 3.7.8).

Fig. 3.7.6  GH-producing pituitary microadenoma. a Coronal T1-weighted image shows no apparent abnormalities. b A small, lowsignal intensity lesion is apparent on coronal T2-weighted image. c,d After gadolinium enhancement, the tumor is well demarcated

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.7  Pituitary microadenoma (prolactinoma). Coronal T1-weighted (a) and T2-weighted (b) images do not clearly depict a pituitary microadenoma located in the right side of the pituitary gland in this case (white arrows). c Coronal dynamic

gadolinium-enhanced MR images depict a relatively less enhancing microadenoma in the right side of the anterior pituitary (arrow)

Fig. 3.7.8  Pituitary macroadenoma. a Sagittal T1-weighted image shows an ectopic posterior pituitary high signal at the dorsal part of the pituitary adenoma (arrow). b,c This mass is

heterogeneous on both T2- and contrast-enhanced T1-weighted images. The left cavernous sinus is invaded by the mass

407

408

3  Brain, Head, and Neck

Dynamic MR imaging is useful to show normal pituitary tissue that is compressed by macroadenoma (Miki et al. 1990). Defining the location of normal pituitary tissue preoperatively is important and will help to prevent postoperative pituitary dysfunction. Ectopic posterior pituitary high signal may be seen in cases with pituitary macroadenoma, due to blocked transportation of neurosecretory granules from the hypo‑thalamus to the posterior pituitary (Fig. 3.7.8a) (Bonneville et al. 2002; Miki et al. 1999; Saeki et al. 2002; Takahashi et al. 2005). This ectopic posterior pituitary high signal can be seen with large adenomas (>2 ml in volume), and is usually seen on the upper posterior margin of adenoma (Takahashi et al. 2005). The ectopic posterior pituitary high signal remains “ectopic” postoperatively, indicating that this is an irreversible change (Takahashi et al. 2005). 3.7.4.6  Cavernous Sinus Invasion by Pituitary Adenoma Both micro- and macroadenomas may invade the cavernous sinus. Preoperative evaluation of the cavernous sinus invasion is important, particularly in cases with functioning adenoma, as total removal of the tumor is desirable to relieve symptoms caused by excessive hormone secretion. MR imaging shows tumors that have invaded the cavernous sinus as a region of diminished enhancement compared with normal cavernous sinus, but the presence or absence of invasion is often difficult to evaluate as the medial wall of the cavernous sinus is extremely thin. Invasion is certain when encasement of the internal carotid artery by the tumor is ≥ 67%, and highly probably when the venous compartment in the carotid sulcus is not depicted (Cottier et al. 2000). Multidetectorrow CT is also useful to evaluate cavernous sinus invasion (Miki et al. 2007). 3.7.4.7  Postoperative Pituitary Adenoma As total surgical removal of all adenoma tissue is not usually possible, follow-up MR imaging is performed annually to detect tumor regrowth. The effect of surgery is best demonstrated on follow-up MR examination after 4 months (Rodriguez et al. 1996). Common postoperative sellar changes include formation of postoperative fibrosis, re-expansion of normal pituitary gland, thickening of the infundibulum, swelling of the optic chiasm, and resorption of implanted material. Early postoperative MR imaging (≤3 days after surgery) is useful in differentiating residual tumor from normal gland, implanted materials, or postoperative granulation tissue. Early postoperative dynamic MR imaging showing

the presence of nodular enhancements in the postoperative sella indicates residual tumor (Yoon et al. 2001). 3.7.4.8  Pituitary Adenocarcinoma Pituitary adenocarcinoma is extremely uncommon. Pituitary adenocarcinoma can only be diagnosed unequivocally on the basis of metastases, as rapidly growing adenomas may demonstrate histological changes similar to those seen in adenocarcinoma. 3.7.4.9  Pituitary Apoplexy Pituitary apoplexy refers to an intrasellar hemorrhage or infarction characterized by headache, visual loss, ophthalmoplegia, or altered mental status. This condition often occurs with pre-existing adenoma, and iatrogenic pituitary apoplexy may occur during pituitary-stimulating tests in patients with pituitary adenoma. A variety of other conditions can trigger pituitary apoplexy, including pregnancy, postpartum status, bleeding tendency, pituitary irradiation, general anesthesia, head trauma, and a wide variety of medications including bromocriptine. This could represent an emergency condition, as permanent visual loss may result if acute compression of the optic chiasm is not relieved rapidly. Urgent operation may thus be required. However, intrasellar hemorrhage has increasingly been recognized as an incidental MR finding with pituitary macroadenoma. Diagnosis is therefore made using a combination of history and appearance on neuroimaging. CT shows a high-density pituitary mass, representing acute hemorrhage. Occurrence of infarction does not necessarily show as a high-density area, even in the acute stage. MR signal intensities vary, depending on the hemorrhagic stage (Fig. 3.7.9). A fluid-fluid level may be seen on sagittal or axial MR images. Diffusion-weighted imaging may demonstrate a typical bright signal if infarction is present (Rogg et al. 2002). A pituitary apoplexy mass little or partially enhances under both hemorrhage and infarction. 3.7.4.10  Craniopharyngioma Craniopharyngiomas are solid or mixed solid-cystic benign tumors (WHO grade I) that arise from the embryonic remnants of Rathke’s pouch. They account for 1–3% of intracranial tumors. As the third most common intracranial tumor in children, after gliomas and medulloblastomas, craniopharyngiomas account for up to 50% of suprasellar tumors in children. A bimodal distribution is seen in peak incidence rates. The first peak occurs

3.7  Pituitary Gland and Parasellar Region

among infants and children aged 5–14 years, while the second peak occurs between the ages of 55 and 75 years (Bunin et al. 1998). Males are more commonly affected than females. Patients with craniopharyngioma commonly present with pituitary hypofunction, visual-field defects, and severe headaches. Central diabetes insipidus

Fig. 3.7.9  Pituitary apoplexy. a Sagittal T1-weighted image shows a hyperintense pituitary mass with suprasellar extension (white arrow). b Sagittal T2-weighted image shows heterogeneous signal with a low signal intensity rim due to the presence of intracellular methemoglobin/deoxyhemoglobin or hemosiderin (black arrow)

occurs in 10–20% of patients (Honegger et al. 1999). Craniopharyngiomas are divided into two histological types, adamantinomatous craniopharyngiomas, and squamouspapillary craniopharyngiomas. Adamantinomatous craniopharyngiomas generally occur in children and adolescents, and are most frequently suprasellar in location. A cystic component is present in 90%, and at least partial calcification will be apparent on CT in 90%. On MR imaging, these tumors typically appear as multilobular, multicystic suprasellar masses (Fig. 3.7.10). Increased signal intensity on T1-weighted images can be caused by cyst contents including proteinaceous fluid, cholesterol, keratin, hemorrhage, and necrotic debris. Both cystic and solid components tend to show high signal intensity on T2-weighted images. The solid portions enhance heterogeneously after contrast administration. The thin walls of the tumor nearly always enhance. Squamous-papillary craniopharyngiomas generally occur in adults, and are usually round and predominantly solid (Sartoretti-Schefer et al. 1997). Calcification and cyst formation are less frequent. Appearance is heterogeneous appearance and enhancement patterns are similar to the solid portion of adamantinomatous craniopharyngiomas (Fig. 3.7.11). Distinguishing non-neoplastic cystic lesions such as Rathke’s cleft cyst and arachnoid cyst from craniopharyngioma is important. These lesions can be differentiated from craniopharyngiomas by the absence of calcification and enhancing solid components. Edema-like changes along the optic tracts are seen in nearly half of craniopharyngiomas. It is speculated that regional inflammation or microscopic leakage of cystic contents may evoke adjacent edema (Nagahata et al. 1998). This finding may be seen in sellar tumors other than craniopharyngioma (Saeki et al. 2003). Astrocytomas of the optic chiasm and hypothalamus Astrocytomas of the optic chiasm and hypothalamus comprise 10–15% of supratentorial tumors in children. The most common symptom is diminished visual acuity, noted in almost 50% of patients. Endocrine dysfunction, most commonly short stature secondary to reduced growth hormone, is present in about 20% (Medlock et al. 1997). Astrocytomas arising in the hypothalamus often grow anteriorly or inferiorly to involve the optic chiasm. As the primary site of origin cannot be determined in many cases, tumors arising from these two locations are often discussed together. Clinical evidence or a positive family history of neurofibromatosis type 1 is seen in 20–50% of patients with astrocytomas of the optic chiasm and hypothalamus. Patients with neurofibromatosis type 1 may develop tumors primarily involving the optic nerves or optic chiasm/hypothalamus. Tumors originating within the optic nerve grow extremely slowly and most are histologically classi-

409

410

3  Brain, Head, and Neck

fied as pilocytic astrocytomas. These tumors behave in a more benign fashion than their counterparts in children without neurofibromatosis type 1 (Listerneck et al. 1999). However, tumors originating from the optic chiasm and hypothalamus may be invasive and frequently exhibit a higher histological grade. MR imaging is the procedure of choice for evaluating tumors of the optic nerves and chiasm, as the intracranial portions of the optic nerves and chiasm are better

seen on MR imaging than on CT. Pilocytic astrocytomas of the optic chiasm and hypothalamus are almost always hypointense on T1-weighted images and markedly hyperintense on T2-weighted images, due to microcystic components (Fig. 3.7.12). When large, these tumors are typically heterogeneous, with large cystic and solid components. The solid portions typically enhance markedly after contrast administration. Hemorrhage and calcification are unusual.

Fig. 3.7.10  Adamantinomatous craniophryngioma. Precontrast sagittal T1-weighted (a) and axial T2-weighted (b) images demonstrate a cystic portion of a craniopharyngioma. c Tumor rim

is well enhanced on sagittal gadolinium-enhanced T1-weighted image. d Axial precontrast CT reveals coarse calcifications within the solid portion of the tumor

3.7  Pituitary Gland and Parasellar Region

3.7.4.11  Pilomyxoid Astrocytoma Pilomyxoid astrocytoma is a recently described astrocytic tumor that usually occurs in the chiasmatic–hypothalamic region in young children. This tumor has been previously considered as pilocytic astrocytoma. Pilomyxoid astrocytoma appears to have a higher rate of recurrence and cerebral spinal fluid dissemination than does typical pilocytic astrocytoma. Definite histological differences exist

Fig. 3.7.11  Squamopapillary craniopharyngioma. a Sagittal precontrast T1-weighted MR image shows a suprasellar solid tumor with intact pituitary gland (arrow). b The tumor is inhomogeneously hyperintense on T2-weighted axial image. c, d Axial and

between these two tumors. Eosinophilic granular bodies, Rosenthal fibers, plump protoplasmic cells and a biphasic pattern are extremely rare in pilomyxoid astrocytoma, but are common in pilocytic astrocytoma. The most distinguishing radiological features suggestive of pilomyxoid astrocytoma are predominantly solid tumor, homogeneous enhancement, hydrocephalus, highly homogeneous T2 signal intensity extending into the deep white and gray matter, and CSF dissemination (Arslanoglu et al. 2003).

coronal post-gadolinium T1-weighted MR images show a wellmarginated tumor, with strong and slightly inhomogeneous enhancement

411

412

3  Brain, Head, and Neck

Fig. 3.7.12  Chiasmatic-hypothalamic astrocytoma. The high water content of this tumor causes low signal intensity on T1-weighted image (a) and markedly high signal intensity on

T2-weighted image (b). c,d Heterogeneous enhancement is demonstrated following contrast administration

3.7.4.12  Meningioma

tumor contacts the dura. This sign is useful for diagnosis of meningioma, but is not specific (Guermazi et al. 2005). Meningiomas involving the pituitary fossa can usually be seen distinct from the pituitary gland. Meningiomas involving the cavernous sinus may show narrowing of the internal carotid artery, due to common encasement of the carotid artery (Fig. 3.7.14) (Grossman and Yousem 2003). MR imaging can also demonstrate thickening of the low signal intensity representing cortical bone, due to the presence of hyperostosis in some cases of intra- and parasellar meningioma. CT shows sclerotic cortical bone thickening (hyperostosis) and blistering more effectively than MR imaging.

Meningiomas are benign extra-axial tumors that arise from meningoepithelial cells of the arachnoid. Comprising approximately 15% of all intracranial tumors, these tumors display a peak incidence at around 45 years and a 2:1 female predominance. They can arise from the tuberculum sella, planum sphenoidale, or diaphragm sella. Meningiomas are typically iso- or hypointense to gray matter on T1-weighted images and iso- or hyperintense on T2-weighted images and markedly enhance (Fig. 3.7.13). MR imaging may also demonstrate a linear duralbased enhancement, called a “dural-tail sign,” where the

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.13  Tuberculum sellae meningioma. a,b Sagittal T1- and coronal T2-weighted images demonstrate a lobulated mass. c,d The tumor enhances homogeneously, and extends anteriorly along the planum and posteriorly along the clivus.

3.7.4.13  Rathke’s Cleft Cyst Rathke’s cleft cysts are thought to arise from the remnants of Rathke’s pouch, and are a relatively common incidental finding on autopsy or MR imaging. These lesions typically remain intrasellar, but can enlarge, extend into the suprasellar region and cause visual impairment, hypothalamic–pituitary dysfunction, and headaches in a small population. Diabetes insipidus can be present in those cases. MR imaging shows a round to ovoid, sharply defined mass that most often lies either between the anterior and posterior pituitary lobes or anterior to the pituitary stalk.

The cyst shows variable signal intensity, ranging from iso- to hyperintense compared to CSF on T1-weighted images and from iso- to hypointense compared to CSF on T2-weighted images, as the composition of cyst fluid varies (Figs. 3.7.15, 3.7.16) (Kucharczyk et al. 1987). The cyst wall generally does not enhance after administration of paramagnetic contrast, but a compressed and elongated pituitary gland may mimic cyst wall enhancement (Hua et al. 1992). An intracystic nodule displaying high signal intensity on T1-weighted images and low signal intensity on T2-weighted images that represents a mucin clump is seen in 77% of Rathke’s cleft cysts (Byun et al. 2000).

413

414

3  Brain, Head, and Neck

Fig. 3.7.14  Parasellar meningioma. Axial T1-weighted (a) and T2-weighted images (b) show a left cavernous sinus mass (arrows). Axial (c) and coronal (d) gadolinium-enhanced T1-

weighted images demonstrate a well-enhanced cavernous sinus tumor extending into the left orbital apex

3.7.4.14  Arachnoid Cyst

as smooth-contoured masses, and are isointense to CSF on all MR sequences because contents are very similar to CSF (Fig. 3.7.17).

Arachnoid cyst is a fluid-filled enclosed space between arachnoid tissue. Such cysts can be congenital in origin, or can develop from adhesions within the subarachnoid space. Suprasellar arachnoid cysts are considered to develop from an imperforate Liliequist’s membrane (Fushimi et al. 2003). Sagittal CISS MR imaging best shows a thin-walled cyst filled with CSF, herniating upward to displace the floor of the third ventricle (Fig. 3.7.17). Cysts can appear

3.7.4.15  Metastasis Symptomatic metastasis to the pituitary is rare, as patients usually die of the malignancy well before pituitary function is disordered. The most common tumors that metastasize to the pituitary are breast and lung carcinoma

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.15  Rathke’s cleft cyst. Sagittal T1-weighted (a) and coronal T2-weighted (b) images of the sella turcica show a proteinaceous intrasellar cyst (arrows)

Fig. 3.7.16  Rathke’s cleft cyst. Sagittal T1-weighted (a) and coronal T2-weighted (b) images show a cystic lesion located in the pars intermedia filled with serous fluid showing similar signal intensity to CSF (arrows). Note anterior displacement of the infundibulum by the cyst and the compressed anterior lobe (arrowheads)

(McCormick et al. 1989). Patients with intrasellar metastases may present with diabetes insipidus, endocrine abnormalities, or visual disturbance. Imaging studies are not specific for metastases, but metastases appear to show a propensity for the posterior pituitary, resulting in diabetes insipidus (McCormick et al. 1989; (Schubiger and Haller 1992). MR imaging

may show an enhancing lesion in the pituitary gland, frequently involving the infundibulum and hypothalamus. The combination of age >50 years, cranial nerve abnormalities, diabetes insipidus and a rapidly progressive pituitary mass on MR imaging should heighten the suspicion of metastasis (Branch, Jr., and Laws, Jr. 1987).

415

416

3  Brain, Head, and Neck

Fig. 3.7.17  Suprasellar arachnoid cyst. Sagittal T1-weighted image (a) and CISS image (b) show a large cyst compressing the third ventricle and displacing the brainstem posteriorly. c Axial T2-weighted image demonstrates resultant hydrocephalus. Note that the cyst wall is clearly demarcated on CISS image (arrowheads)

3.7.4.16  Germ Cell Tumors Non-gonadal germ cell tumors are embryologically divided into germinomas, which originate from primordial germ cells, and non-germinomatous germ cell tumors, originate from cells with embryonal differentiation. Germinomas are potentially malignant, but display a relatively benign overall prognosis due to a high susceptibility to irradiation and chemotherapy. Non-germinomatous germ-cell tumors are generally highly malignant and have poor prognosis. Non-germinomatous germ cell tumors are classified into embryonal carcinoma, choriocarcinoma, yolk sac tumors, teratomas, and any mixture of these tumor types (mixed germ cell tumor).

Germ cell tumors are diseases of children and adolescence. Approximately 90% occur in patients less than 20 years old. In central nervous system (CNS) germ cell tumors, the pineal gland is the most common site of origin (50%), followed by the suprasellar region (20–30%). Other sites include the basal ganglia, thalamus, brainstem, and spinal cord. Germinomas are the most common germ cell tumors originating in the CNS, while nongerminomatous germ cell tumors are usually seen in the pineal region. Males are more commonly affected than females, with an estimated ratio of 4:1 in CNS germ cell tumors. Germ cell tumors of the pineal gland or basal ganglia are seen predominantly in males, whereas males and females are equally affected by suprasellar germ cell

3.7  Pituitary Gland and Parasellar Region

tumors. Suprasellar germinoma is more precisely called neurohypophyseal germinoma, since germinomas of this region commonly involve the hypothalamoneurohypophyseal axis and cause diabetes insipidus (Fujisawa et al. 1991). CNS germinomas sometimes involve the pineal and suprasellar compartments simultaneously or sequentially. This finding is nearly specific for germinomas, particularly in young patients. Klinefelter’s syndrome is known to be associated with various malignancies, including CNS germ cell tumor (Kaido et al. 2003). Elevated CSF or blood tumor markers are most specific in the diagnosis of non-germinomatous germ cell tumors. Predominantly associated with yolk sac tumors, α-fetoprotein (AFP) can also be expressed by embryonal cell carcinomas and immature teratomas. While β-human chorionic gonadotropin (HCG) is markedly elevated in association with choriocarcinomas, 10–50% of germinomas also secrete β-HCG at low levels due to the presence of syncytiotrophoblastic giant cells. Typically, the solid portion of the tumor is iso- or slightly hypointense on T1-weighted images and iso- or hyperintense on T2-weighted images. Marked contrast enhancement is commonly seen after gadolinium injection. Cyst formation is a common finding, but calcification is usually not seen. MRI findings are rather non-specific, although infundibular thickening and an absence of normal signal hyperintensity in the posterior pituitary on T1-weighted MR images and relative hyperdensity without calcification on unenhanced CT represent common imaging features for neurohypophyseal germinomas (Fig. 3.7.18) (Kanagaki et al. 2004b). Anterior compression of the normally enhanced pituitary gland by a mass may be seen on sagittal gadolinium-enhanced T1-weighted images. At the time that patients present with diabetes insipidus, germinoma may be small or not yet be visible on MR imaging. In such patients, repeated imaging should be performed in 3–6 months. 3.7.4.17  Primary Sellar Lymphoma Primary CNS lymphomas are extranodal malignant lymphomas arising in the CNS in the absence of lymphoma outside the CNS at the time of diagnosis. The incidence of primary CNS lymphomas has recently increased markedly worldwide, particularly in immunocompromised patients. Primary CNS lymphoma sometimes involves the pituitary and hypothalamus. Pituitary involvement may cause neurological symptoms including headaches, visual and oculomotor impairment, hypopituitarism, and diabetes insipidus. Compared to gray matter, lymphoma may appear iso- or hypointense on T1-weighted images and iso- or hypointense on T2-weighted images (Kaufman et al. 2002). Such tumors usually enhance homogeneously after contrast administration. On CT, lesions appear iso- to

hyperdense. Calcification, necrosis, cystic appearance, and ring enhancement are uncommon. If the infundibulum or hypothalamus is affected, the posterior pituitary high signal will disappear (Fig. 3.7.19). 3.7.4.18  Granular Cell Tumor of the Sellar and Suprasellar Regions Granular cell tumor of the neurohypophysis, also called granular cell myoblastoma, granular cell neuroma, choristoma, and pituicytoma are extremely rare benign tumors arising from the neurohypophysis or infundibulum (Rosenblum et al. 2000). Less than 100 symptomatic cases have been reported. These tumors consist of cells with granular, eosinophilic cytoplasm due to abundant intracytoplasmic lysosomes. Such tumors grow very slowly and rarely reach sufficient size to compress the pituitary gland, optic chiasm, and hypothalamus, and are thus typically diagnosed during adulthood. Females are affected twice as often as males are. Symptoms include visual disturbance, headache, hormonal abnormality with amenorrhea, galactorrhea, decreased libido, infertility, and diabetes insipidus. MR imaging shows a well-circumscribed isointense intrasellar and/or suprasellar mass with marked contrast enhancement. Diameter is usually <3cm, and infiltration of surrounding structures is rare. Sagittal MR imaging best demonstrates tumor location, usually in the posterior pituitary. Although rare, calcification, necrosis, and cystic degeneration can be seen in some cases. 3.7.4.19  Hypothalamic Hamartoma Hypothalamic hamartomas are not true neoplasms, but are congenital malformations composed of disorganized neural tissue, and are typically located between the infundibulum and mamillary bodies (tuber cinereum). The most frequent presenting complaint is precocious puberty, usually seen before the age of 2 years. Other symptoms include gelastic seizures, developmental delay, and hyperactivity. Hypothalamic hamartomas consist of two distinct types. The pedunculated type, attached to the tuber cinereum or mammillary bodies with a thin stalk, is characterized by precocious puberty. Conversely, the sessile type presents as a round mass involving the tuber cinereum, and is characterized by gelastic seizures of the partial complex type, intellectual impairment, or psychiatric disturbances (Arita et al. 1999). MR appearance is typically that of a well-circumscribed round mass lying within or suspended from the floor of the third ventricle. Sometimes the tumor will be located in the lateral wall of the hypothalamus. Hypothalamic hamartomas are isointense with gray matter

417

418

3  Brain, Head, and Neck

on T1-weighted images and iso- to slightly hyperintense on T2-weighted images (Fig. 3.7.20). No enhancement is seen after gadolinium injection (Boyko et al. 1991). Calcification is absent.

Fig. 3.7.18  Neurohypophyseal germinoma. Intra- and suprasellar tumor components are homogeneous and isointense to cerebral cortex on both T1-weighted (a) and T2-weighted (b) MR images. a Sagittal T1-weighted MR image shows absence of posterior pituitary high signal (white arrow). c Sagittal postgadolinium T1-weighted MR image shows a heterogeneously

3.7.4.20  Langerhans Cell Histiocytosis Langerhans cell histiocytosis (LCH) represents non-malignant proliferation of histiocytic granulomas. This dis-

enhancing tumor involving the intrasellar region and infundibulum (arrow). The anterior pituitary is compressed anteroinferiorly (arrowhead). The tumor is less enhancing than the anterior pituitary is. d Non-contrast CT reveals a homogeneously hyperdense tumor with no calcification

3.7  Pituitary Gland and Parasellar Region

ease is rarely seen in adults, but is common in children. CNS symptoms include hypothalamic/pituitary dysfunction such as diabetes insipidus, visual disturbances, and ataxia. Diabetes insipidus is present in 5% of patients with LCH at the time of diagnosis, but 10–50% of patients are affected on follow-up examinations. LCH is the most common cause of childhood diabetes insipidus in Western countries. Involvement of the CNS is most common in patients with this multisystemic disease. In a large series of LCH cases, 42 of 44 patients with diabetes insipidus displayed involvement of other organ systems, including bone, skin, lung, and lymph nodes (Howarth et al. 1999). When the pituitary stalk and hypothalamus are involved by LCH, MR imaging shows infundibular thickening or a focal hypothalamic mass, isointense on both T1- and T2-weighted images. LCH usually enhances markedly and homogeneously after administration of paramagnetic contrast. After the onset of diabetes insipidus, the posterior pituitary high signal on T1-weighted images becomes absent (Fig. 3.7.21) (Maghnie et al. 2000). Lytic calvarial lesions and enhancing parenchymal and meningeal lesions of the brain may occur. MR imaging may also demonstrate white matter lesions predominantly located in the pons, cerebellar peduncles and cerebellar white matter. Children with isolated diabetes insipidus should be carefully observed for onset of other symptoms characteristic of LCH. In a study in Italy, 15% of patients with

isolated diabetes insipidus were found to have LCH (Maghnie et al. 2000). 3.7.4.21  Sarcoidosis Sarcoidosis is a systemic granulomatous disease of undetermined etiology. CNS abnormalities are noted in approximately 5% of patients. Sarcoidosis may involve the leptomeninges of the brain, producing granulomas on the pituitary stalk, optic chiasm, cranial nerves, or cerebral parenchyma, and causing hypopituitarism, diabetes insipidus or cranial neuropathy. Granulomas demonstrate marked contrast enhancement after gadolinium injection (Fig. 3.7.22). Meningeal enhancement is a common finding of CNS sarcoidosis. Hypothalamic–pituitary sarcoidosis is usually accompanied by other evidence of neurosarcoidosis or systemic sarcoidosis. Typical MR appearance of hypothalamic–pituitary sarcoidosis is an enhancing thickened infundibulum with meningeal enhancement, usually accompanied by other evidence of neurosarcoidosis or systemic sarcoidosis (Dumas et al. 2000; Seltzer et al. 1991). Lymphoma, metastatic carcinoma, LCH, and germ cell tumors may show similar findings. Other granulomatous diseases such as Wegener’s granulomatosis, ChurgStrauss syndrome, or Rosai-Dorfman disease can also involve the hypothalamic neurohypophyseal axis, and result in similar imaging and clinical findings.

Fig. 3.7.19  Malignant lymphoma. Sagittal T1-weighted images before (a) and after (b) gadolinium injection demonstrate thickening of the infundibulum (arrow). a Sagittal T1-weighted image shows absence of the posterior pituitary high signal (arrowhead)

419

420

3  Brain, Head, and Neck

3.7.4.22  Lymphocytic Hypophysitis Lymphocytic hypophysitis is an autoimmune inflammatory infiltration of the anterior and/or posterior pituitary gland by lymphocytes and plasma cells that leads to destruction of the gland. Patients typically present with hypo‑thalamic-pituitary dysfunction, headache, or impaired vision. Lymphocytic infiltration and destruction of the pituitary can present with an enlarging intrasellar and/or suprasellar mass and varying degrees of pituitary

Fig. 3.7.20  Hypothalamic hamartoma. MR images (a sagittal T1-weighted image, b axial T2-weighted image, c coronal T1-weighted image) show a well-demarcated mass isointense to

insufficiency. The presence of associated autoimmune diseases can offer a clue to the diagnosis. Lymphocytic adenohypophysitis involves adenohypophysis and lymphocytic infundibuloneurohypophysitis affect the neurohypophysis. If both lobes are involved, then the lesion is properly called lymphocytic hypophysitis. Lymphocytic adenohypophysitis causes pan- or partial hypopituitarism by inflammatory infiltration. This disease exhibits a female predilection and frequently affects young women during pregnancy or in the postpar-

gray matter, arising from the region of the tuber cinereum. d No enhancement is seen after gadolinium injection

3.7  Pituitary Gland and Parasellar Region

tum period. However, the condition is now known to occur at all ages and in both genders. Various concomitant autoimmune diseases can be present and anti-pituitary antibody is detected in some patients. MR imaging is the best modality to detect the enlarged, well-enhanced pituitary mass with or without thickened stalk (Fig. 3.7.23). These findings are non-specific, but adjacent dural thickening or cavernous sinus involvement with heterogenous hypointensity on T2-weighted image is suggestive of this disease. Sometimes the internal carotid artery of the af-

fected cavernous portion is encased. This pituitary mass regresses naturally or with steroid treatment and neither operation nor biopsy is necessary. After regression, morphological abnormality can sometimes not be detected, but dynamic studies clearly demonstrate destruction of the pituitary vasculature, representing evidence of past inflammatory disease (Sato et al. 1998). Lymphocytic infundibuloneurohypophysitis is a pituitary disorder first described in 1993 by Imura, in which lymphocytic inflammation confined to the hypotha-

Fig. 3.7.21  Langerhans cell histiocytosis. a T1-weighted sagittal image of the pituitary gland shows abnormal thickening of the infundibulum (arrow). Note the absence of the posterior pi-

tuitary high signal in this patient, who presented with diabetes insipidus (arrowhead). b CT demonstrates lytic calvarial lesions

Fig. 3.7.22  Neurosarcoidosis. a Coronal T1-weighted image after gadolinium injection demonstrates abnormal hypothalamic enhancement. b Sagittal T1-weighted image shows loss of pos-

terior pituitary high signal consistent with the presence of diabetes insipidus

421

422

3  Brain, Head, and Neck

Fig. 3.7.23  Lymphocytic hypophysitis. Sagittal T1-weighted images before (a) and after (b) gadolinium injection demonstrate diffuse enlargement of the pituitary gland, thickening of the infundibulum (white arrow), absence of posterior pituitary bright spot (white arrowhead) and heterogeneous enhancement

lamic–neurohypophyseal system causes diabetes insipidus (Imura et al. 1993). Mild anterior pituitary hormonal deficiency may be observed. Complication with various autoimmune diseases is again commonplace, particularly with diseases such as Hashimoto disease, collagen disease, or rheumatoid arthritis. Lymphocytic infundibuloneurohypophysitis is thought to be a common cause of what was previously considered to be idiopathic diabetes insipidus over a wide range of ages. MR imaging shows thickening of the pituitary stalk or enlargement of the neurohypophysis, disappearing during followup. Whether these two diseases actually represent the same entity remains controversial, because common and differing findings exist between lymphocytic adenohypophysitis and lymphocytic infundibuloneurohypophysitis. 3.7.4.23  Abscess of the Pituitary Gland Pituitary abscess is a rare but potentially life-threatening disease. This disease can be caused by direct extension of infection from the sphenoid sinus, cavernous sinus or CSF, or secondary to bacteremia. Headache, endocrine abnormalities, and visual changes are the most common clinical presentations. Most cases are caused by suppurative organisms accompanying other sellar lesions, such as a pituitary adenoma, Rathke’s cleft cyst, or craniopharyngioma. However, peripheral leukocytosis and meningismus are present in less than a third of the patients. Given the ambiguous clinical features and imaging findings, most abscesses are difficult to diagnose before treatment. Pituitary abscess may be indistinguishable from pituitary adenoma on T1- and T2-weighted images alone, but typical MR appearance is a fluid-filled intrasellar mass with thick peripheral enhancement (Wolansky et al. 1997). Pituitary stalk thickening or meningeal enhance-

ment may be seen in some cases. Diffusion-weighted imaging may be useful, as pus is usually very hyperintense on such images. 3.7.4.24  Hemochromatosis Hemochromatosis is a disease caused by excess iron deposition in parenchymal cells, leading to cellular damage and organ dysfunction. Primary hemochromatosis is an inherited autosomal recessive form with elevated intestinal iron absorption. Secondary hemochromatosis usually results from iron overload due to increased intake following repeated blood transfusions. The excess iron is deposited in the liver, spleen, skin, heart, and bone marrow. Iron deposition can also occur in endocrine organs, including the pancreas, thyroid, and anterior pituitary gland. Excess iron deposition in the anterior pituitary may lead to secondary hypogonadism in advanced cases of transfusion-induced hemochromatosis (Sparacia et al. 2000). The presence of excess iron deposition can be demonstrated on MR imaging in the form of reduced signal intensity in the anterior lobe, due to decreased T2 relaxation time and magnetic field inhomogeneities caused by excess intracellular iron (Fig. 3.7.24) (Fujisawa et al. 1998). Posterior pituitary hormonal function and the posterior pituitary bright spot on T1-weighted images are usually preserved, as iron deposition does not generally occur in the posterior pituitary. 3.7.4.25  Pituitary Dwarfism Pituitary dwarfism is a disease in which growth hormone secretion is insufficient, with or without concomitant other anterior pituitary hormone deficiency, and etiology

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.24  Hemochromatosis. a Sagittal T1-weighted image shows nearly normal signal intensity of the anterior pituitary. b However, sagittal T2-weighted image shows markedly low signal intensity of the anterior pituitary due to iron deposition

varies. Some cases show morphological alterations of the hypothalamic-pituitary region or mutations in growth hormone-related genes (Osorio et al. 2002), while others still appear idiopathic. Pituitary stalk interruption with or without an ectopic posterior lobe represents a morphological abnormality causing dwarfism, and has been explained by traumatic episode or congenital anomaly. Transection of the pituitary stalk may result from acceleration or deceleration mechanisms. Panhypopituitarism can result from dissociation of the anterior pituitary from hypothalamic neurohormones. An ectopic posterior pituitary can be seen in the proximal stump of the stalk. Diabetes insipidus occurs only if the hypothalamus is involved. A high incidence of breech delivery has been reported in patients with dwarfism (about 32%), and traumatic transection of the pituitary stalk during delivery can be one pathogenesis of this condition. Preservation of posterior lobe functions in these patients can be explained by the presence of the high signal seen in the median eminence, the so-called ectopic posterior lobe (Fujisawa et al. 1987a). MR imaging also shows absence or hypoplasia of the pituitary stalk and hypoplasia of the anterior pituitary (Fig. 3.7.25). For other patients with normal delivery without any evidence of trauma or cesarean delivery, dysgenesis or abnormal embryonic development of both adeno- and neurohypophysis can be related to the pathogenesis. A report of familial congenital hypopituitarism with diabe-

tes insipidus in which the pituitary stalks are missing also supports this theory (Yagi et al. 1994).

Fig. 3.7.25  Ectopic posterior pituitary in pituitary dwarfism. The posterior pituitary high signal is not found within the sella turcica and a small ectopic high signal spot is seen at the median eminence (arrows). The sella turcica is small and the infundibular stalk is absent (arrowheads)

423

424

3  Brain, Head, and Neck

3.7.4.26  Cavernous Sinus Diseases 3.7.4.26.1  Aneurysm Cavernous sinus aneurysms exert mass effects on the intracavernous cranial nerves. When rupture occurs, a carotid–cavernous fistula is created, as opposed to intradural aneurysms resulting in subarachnoid hemorrhage. Cavernous sinus aneurysms are typically confined to the parasellar region, but may rarely extend to the sella turcica and mimic a pituitary mass (Fig. 3.7.26). Intrasellar aneurysms may produce sellar enlargement and mass effects. MR imaging of non-thrombosed aneurysm typically shows a “flow void” or lack of signal within a portion of the aneurysm, and may show ghost artifacts related to pulsatile flow of blood within the aneurysm. In cases with thrombus within an aneurysm, MR imaging shows multiple laminar signals representing blood products in various stages of degradation. 3.7.4.26.2  Carotid–Cavernous Fistula Carotid–cavernous fistula (CCF) is a fistulous connection caused by rupture of the internal carotid artery or its branches into the cavernous sinus. Patients can present with pulsatile exophthalmos, chemosis, diplopia, blurred vision, or headache. CCF can be classified as direct and indirect. Direct CCF is more common and is caused by direct rupture of the internal carotid artery into the cavernous sinus. Such a fistula is usually high flow and found in patients with either traumatic episodes or rupture of a carotid artery aneurysm. Indirect CCF is caused by rupture of small intracavernous branches of the internal carotid

Fig. 3.7.26  Cavernous sinus aneurysm. Sagittal (a) and coronal (b) T1-weighted images show a hyperintense intrasellar mass mimicking hemorrhagic pituitary adenoma or Rathke’s cleft

artery or external carotid artery into the cavernous sinus. Indirect CCF is usually low flow and is more commonly seen in postmenopausal females. Symptoms are greatly influenced by shunt size, but are also affected by the pattern of venous drainage. Anterior drainage through the superior and inferior ophthalmic veins is the most common pattern. Posterior drainage into the inferior and superior petrosal sinuses is the other important drainage route, and in this situation orbital symptoms may be less pronounced. MR imaging findings depend on the size and type of CCF and the pattern of venous drainage. Direct CCF typically produces marked distension of the ipsilateral cavernous sinus and ophthalmic veins. MR imaging may identify enlarged orbital veins or cavernous sinus (Fig. 3.7.27). Signal voids can be seen in the affected cavernous sinus. The contralateral cavernous sinus and ophthalmic veins will occasionally be enlarged, since the cavernous sinuses are normally connected through intercavernous sinus. MR findings may be more subtle if the fistulous communication is small or posterior drainage is predominant. Angiography is essential to confirm location of the fistula, arterial supply, and patterns of venous drainage. Direct CCF rarely resolves spontaneously and usually requires therapeutic intervention. Indirect CCF often resolves spontaneously and can be conservatively observed if symptoms are mild. The pituitary gland may be enlarged due to venous congestion in CCF (Sato et al. 1997). 3.7.4.26.3  Tolosa-Hunt Syndrome Tolosa-Hunt syndrome is characterized by painful ophthalmoplegia with associated paresis of one or more of cranial nerves III, IV, and VI due to non-infectious gran-

cyst (arrows). c Contrast-enhanced CT demonstrates that this mass is a partially thrombosed aneurysm (arrowheads)

3.7  Pituitary Gland and Parasellar Region

ulomatous inflammation involving the cavernous sinus, superior orbital fissure, and/or orbital apex. Corticosteroid therapy dramatically resolves both clinical and radiological findings of this syndrome. Other causative lesions must be excluded on neuroimaging, especially for the region of the cavernous sinus and orbit. MR imaging can sometimes visualize inflammatory tissue in Tolosa-Hunt syndrome, demonstrating asymmetrical enlargement of the clinically affected cavernous sinus. However, positive MR imaging findings compatible with inflammatory tissue neither exclude nor confirm Tolosa-Hunt syndrome and remain suspect until malignant tumor and inflammation other than TolosaHunt syndrome can be excluded. Clinical and radiological follow-up examinations must be performed, even in patients with negative findings on MR imaging at onset. Conventional MR imaging sometimes fails to depict abnormal findings. In such cases, dynamic MR imaging may show gradually enhancing abnormal soft tissue lesions in the affected cavernous sinus, facilitating the diagnosis of Tolosa-Hunt syndrome (Fig. 3.7.28) (Haque et al. 2004). 3.7.4.26.4  Cavernous Sinus Invasion of Nasopharyngeal Carcinoma Head and neck tumors may demonstrate perineural extension to the cavernous sinus through the skull base foramen. The second and third divisions of the trigeminal nerve are common sites for perineural spread of head and neck tumors such as adenoid cystic carcinoma, squamous cell carcinoma, basal cell carcinoma, lymphoma, mucoepidermoid carcinoma, melanoma, and schwannoma. Infections such as actinomycosis can also show perineural involvement. MR findings of perineural spread

Fig. 3.7.27  Carotid–cavernous fistula. Axial T2-weighted (a) and coronal T1-weighted (b) images show enlarged left cavernous sinus with multiple abnormal flow voids (arrowheads).

include asymmetrical enlargement of neural foramina, enlarged nerves with irregular or nodular enhancement and replacement of normal fat with infiltrating soft tissue (Fig. 3.7.28). 3.7.4.26.5  Neurinoma Extending into the Cavernous Sinus Neurinomas develop from the Schwann cells of sensory nerves, most commonly along the superior vestibular branch of cranial nerve VIII (acoustic nerve). The trigeminal nerve is the third most commonly involved nerve after the acoustic and facial nerves. Neurinoma may arise from any portion of the trigeminal nerve, but most often arises from the Gasserian ganglion portion of the nerve within Meckel’s cave, and occasionally extends into the cavernous sinus. Rarely, neurinomas may arise in the cavernous sinus. On MR imaging, neurinomas are usually iso- to hypo­intense on T1-weighted images, hyperintense on T2-weighted images and enhance homogeneously (Fig. 3.7.30). Necrosis and cyst formation are commonly seen. Calcification, dural tail, bony reaction, and narrowing of the internal carotid artery are more commonly associated with meningioma than with neurinoma, helping to distinguish between these two tumors. MR imaging is also sensitive for detection of other possible neurinomas in patients with neurofibromatosis type 2. 3.7.4.26.6  Cavernous Hemangioma of the Cavernous Sinus Cavernous hemangiomas are hamartomas of blood vessels and are predominantly found in middle-aged women.

c Axial T2-weighted image of the left orbit demonstrates enlarged superior ophthalmic vein (arrow)

425

426

3  Brain, Head, and Neck

Extra-axial cavernous hemangiomas are uncommon and rarely arise in the cavernous sinus. The symptoms of cavernous sinus cavernous hemangioma are caused by growth, and include headaches, retro-orbital pain, and dysfunction of the cranial nerves passing through or near the cavernous sinus. Extra-axial cavernous hemangiomas display a different natural history and radiological features to the intra-axial type (Meyer et al. 1990). Cavernous hemangiomas in the cavernous sinus show no evidence of previous hemorrhage and blood flow is homogeneously maintained. MR imaging reflects this history well. The mass is well-demar-

cated, homogenously hypo- or isointense on T1-weighted images, and markedly hyperintense on T2-weighted images, just like CSF (Sohn et al. 2003). Dynamic studies demonstrate a characteristic enhancement pattern. After the injection, peripheral stains gradually increase and cover the entire portion of the mass with long-standing persistence of contrast materials (Fig. 3.7.31) (Suzuki et al. 2005). These MR signals and enhancement pattern are exactly as seen in the liver or retrobulbar space. MR imaging findings of typical hyperintensity on T2-weighted images and gradual enhancement on dynamic studies lead to the correct diagnosis without difficulty.

Fig. 3.7.28  Tolosa-Hunt syndrome. Coronal T1-weighted pre- (a) and post-contrast (b) images show enlarged left cavernous sinus with outer bulging of the lateral wall (arrowheads). Signal intensity and contrast enhancement of the left affected cavernous sinus are similar to those of the right unaffected cavernous sinus. c Dynamic images depict slowly enhancing lesions in the lateral (arrowhead), inferior-lateral (short arrow) and superior-medial (long arrow) parts in the left enhanced cavernous sinus, suggestive of granulation tissues. (Reprinted from Haque et al. (2004) Eur J Radiol 51:209–217, with permission)

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.29  Cavernous sinus and sellar invasion of nasopharyngeal carcinoma. Axial T1-weighted (a) and T2-weighted (b) images show irregular tumor invading the sella and right cavernous sinus (white arrowheads). c Gadolinium-enhanced coro-

nal T1-weighted image demonstrates replacement of normal fat with infiltrating soft tissue showing decreased signal intensity (black arrow)

Fig. 3.7.30  Trigeminal neurinoma. a Coronal T1-weighted image shows low signal intensity mass in the left Meckel’s cave. Coronal (b) and axial (c) gadolinium-enhanced T1-weighted images show marked enhancement of the mass. d Axial T2-weighted image demonstrates dilatation of Meckel’s cave well

427

428

3  Brain, Head, and Neck

Fig. 3.7.31  Cavernous hemangioma in the cavernous sinus. a Coronal T1-weighted image shows the right cavernous sinus occupied by an isointense mass with the pituitary gland shifted to the opposite side. b The mass is markedly and homogenously

high intensity on T2-weighted image. c Coronal dynamic gadolinium-enhanced MR images demonstrate gradual enhancement from the lateral edge. d Enhancement remains apparent in the late phase

3.7.4.27  Chordomas

Chondroid chordomas contain both chordomatous and chondromatous components. This variant accounts for 5–15% of all chordomas and up to 33% of cranial chordomas. Compared to conventional chordomas, chondroid tumors occur in a slightly younger age group and appear less aggressive. MR imaging is superior to CT in delineating extent of the tumor and relationships to adjacent structures. The normal pituitary gland may be separately visible on MR imaging, distinguishing chordomas from pituitary adenomas. Intermediate to low signal intensity is apparent on T1-weighted images and very high signal intensity on T2-weighted images likely reflects high fluid content of vacuolated cellular components (Fig. 3.7.32). Chondroid chordomas may not be as bright as conventional chordomas on T2-weighted images due to the presence of cartilaginous components (Sze et al. 1988). These tumors usually produce bony destruction. CT typically shows bone destruction and calcification with invasion of the sella. Calcification occurs in 30–70% of cases.

Chordomas are rare, slowly growing, locally aggressive neoplasms of bone that arise from embryonic remnants of the primitive notochord. In a recent review of 400 patients with chordoma, tumors were relatively evenly distributed between the cranial (32%), spinal (32.8%) and sacral (29.2%) regions (McMaster et al. 2001). Intracranial chordomas most often originate from the sphenooccipital synchondrosis of the clivus. Chordomas account for 1% of intracranial tumors and 4% of all primary bone tumors. A 2:1 male predilection is seen, with peak prevalence in the fourth decade of life. Chordomas of the sella or parasellar regions present with visual disturbances, pituitary insufficiency, or cavernous sinus syndrome. Chordomas have been divided into two histopathological subtypes, conventional chordomas and chondroid chordomas. Conventional chordomas are the more common, and are characterized by the absence of cartilaginous or additional mesenchymal components.

3.7  Pituitary Gland and Parasellar Region

Fig. 3.7.32  Parasellar chordoma. a Axial T2-weighted MR image demonstrates a hyperintense mass with lateral extension to the middle cranial fossa and cavernous sinus. The mass also exhibits areas of hypointensity (arrowheads) corresponding to

punctate calcifications seen on CT (not shown). Axial (b) and coronal (c) contrast-enhanced T1-weighted MR images show a large lobulated parasellar mass with variable enhancement (honeycomb appearance)

References

9.

1.

2.

3.

4.

5.

6.

7.

8.

Arita K, Ikawa F, Kurisu K, Sumida M, Harada K, Uozumi T, Monden S, Yoshida J, Nishi Y (1999) The relationship between magnetic resonance imaging findings and clinical manifestations of hypothalamic hamartoma. J Neurosurg 91:212–220 Arslanoglu A, Cirak B, Horska A, Okoh J, Tihan T, Aronson L, Avellino AM, Burger PC, Yousem DM (2003) MR imaging characteristics of pilomyxoid astrocytomas. AJNR Am J Neuroradiol 24:1906–1908 Bartynski WS, Lin L (1997) Dynamic and conventional spin-echo MR of pituitary microlesions. AJNR Am J Neuroradiol 18:965–972 Batista D, Courkoutsakis NA, Oldfield EH, Griffin KJ, Keil M, Patronas NJ, Stratakis CA (2005) Detection of adrenocorticotropin-secreting pituitary adenomas by magnetic resonance imaging in children and adolescents with cushing disease. J Clin Endocrinol Metab 90:5134–5140 Bonneville F, Narboux Y, Cattin F, Rodiere E, Jacquet G, Bonneville JF (2002) Preoperative location of the pituitary bright spot in patients with pituitary macroadenomas. AJNR Am J Neuroradiol 23:528–532 Boyko OB, Curnes JT, Oakes WJ, Burger PC (1991) Hamartomas of the tuber cinereum: CT, MR, and pathologic findings. AJNR Am J Neuroradiol 12:309–314 Branch CL Jr, Laws ER Jr (1987) Metastatic tumors of the sella turcica masquerading as primary pituitary tumors. J Clin Endocrinol Metab 65:469–474 Bunin GR, Surawicz TS, Witman PA, Preston-Martin S, Davis F, Bruner JM (1998) The descriptive epidemiology of craniopharyngioma. J Neurosurg 89:547–551

10.

11.

12.

13.

14.

15.

16. 17. 18.

Byun WM, Kim OL, Kim DS (2000) MR imaging findings of Rathke’s cleft cysts: significance of intracystic nodules. AJNR Am J Neuroradiol 21:485–488 Cottier JP, Destrieux C, Brunereau L, Bertrand P, Moreau L, Jan M, Herbreteau D (2000) Cavernous sinus invasion by pituitary adenoma: MR imaging. Radiology 215:463–469 Cox TD, Elster AD (1991) Normal pituitary gland: changes in shape, size, and signal intensity during the 1st year of life at MR imaging. Radiology 179:721–724 Davis WL, Lee JN, King BD, Harnsberger HR (1994) Dynamic contrast-enhanced MR imaging of the pituitary gland with fast spin-echo technique. J Magn Reson Imaging 4:509–511 De Marinis L, Bonadonna S, Bianchi A, Maira G, Giustina A (2005) Primary empty sella. J Clin Endocrinol Metab 90:5471–5477 Dietemann JL, Kehrli P, Maillot C, Diniz R, Reis M, Jr., Neugroschl C, Vinclair L (1998) Is there a dural wall between the cavernous sinus and the pituitary fossa? Anatomical and MRI findings. Neuroradiology 40:627–630 Dumas JL, Valeyre D, Chapelon-Abric C, Belin C, Piette JC, Tandjaoui-Lambiotte H, Brauner M, Goldlust D (2000) Central nervous system sarcoidosis: follow-up at MR imaging during steroid therapy. Radiology 21:4411–420 Elster AD (1993a) Modern imaging of the pituitary. Radiology 1871–14 Elster AD (1993b) Sellar susceptibility artifacts: theory and implications. AJNR Am J Neuroradiol 14:129–136 Elster AD (1994) High-resolution, dynamic pituitary MR imaging: standard of care or academic pastime? [Comment]. AJR Am J Roentgenol 16:3680–682

429

430

3  Brain, Head, and Neck 19. Freda PU, Wardlaw SL, Post KD (1996) Unusual causes of sellar/parasellar masses in a large transsphenoidal surgical series. J Clin Endocrinol Metab 81:3455–3459 20. Fujisawa I, Kikuchi K, Nishimura K, Togashi K, Itoh K, Noma S, Minami S, Sagoh T, Hiraoka T, Momoi T et al (1987a) Transection of the pituitary stalk: development of an ectopic posterior lobe assessed with MR imaging. Radiology 165:487–489 21. Fujisawa I, Nishimura K, Asato R, Togashi K, Itoh K, Noma S, Kawamura Y, Sago T, Minami S, Nakano Y et al (1987b) Posterior lobe of the pituitary in diabetes insipidus: MR findings. J Comput Assist Tomogr 11:221–225 22. Fujisawa I, Morikawa M, Nakano Y, Konishi J (1988) Hemochromatosis of the pituitary gland: MR imaging. Radiology 168:213–214 23. Fujisawa I, Asato R, Kawata M, Sano Y, Nakao K, Yamada T, Imura H, Naito Y, Hoshino K, Noma S et al (1989) Hyperintense signal of the posterior pituitary on T1-weighted MR images: an experimental study. J Comput Assist Tomogr 13:371–377 24. Fujisawa I, Asato R, Okumura R, Nakano Y, Shibata T, Hamanaka D, Hashimoto T, Konishi J (1991) Magnetic resonance imaging of neurohypophyseal germinomas. Cancer 68:1009–1014 25. Fujisawa I, Murakami N, Furuto-Kato S, Araki N, Konishi J (1996) Plasma and neurohypophyseal content of vasopressin in diabetes mellitus. J Clin Endocrinol Metab 81:2805–2809 26. Fushimi Y, Miki Y, Ueba T, Kanagaki M, Takahashi T, Yamamoto A, Haque TL, Konishi J, Takahashi JA, Hashimoto N, Konishi J (2003) Liliequist membrane: three-dimensional constructive interference in steady state MR imaging. Radiology 229:360–365; discussion, 365 27. Fushimi Y, Miki Y, Takahashi JA, Kikuta K, Hashimoto N, Hanakawa T, Fukuyama H, Togashi K (2006) MR Imaging of Liliequist’s membrane. Radiat Med 24:85–90 28. Grossman R, Yousem D (2003) CNS anatomy: sella and central skull base. In: Thrall JH (ed) The requisites, 2nd edn. Mosby, St. Louis, pp 517–564 29. Gsponer J, De Tribolet N, Deruaz JP, Janzer R, Uske A, Mirimanoff RO, Reymond MJ, Rey F, Temler E, Gaillard RC, Gomez F (1999) Diagnosis, treatment, and outcome of pituitary tumors and other abnormal intrasellar masses. Retrospective analysis of 353 patients. Medicine (Balt) 78:236–269 30. Guermazi A, Lafitte F, Miaux Y, Adem C, Bonneville JF, Chiras J (2005) The dural tail sign—beyond meningioma. Clin Radiol 60:171–188 31. Hagiwara A, Inoue Y, Wakasa K, Haba T, Tashiro T, Miyamoto T (2003) Comparison of growth hormone-producing and non-growth hormone-producing pituitary adenomas: imaging characteristics and pathologic correlation. Radiology 228:533–538 32. Haque TL, Miki Y, Kashii S, Yamamoto A, Kanagaki M, Takahashi T, Fushimi Y, Asato R, Murase N, Shibasaki H, Konishi J (2004) Dynamic MR imaging in Tolosa-Hunt syndrome. Eur J Radiol 51:209–217

33. Honegger J, Buchfelder M, Fahlbusch R (1999) Surgical treatment of craniopharyngiomas: endocrinological results. J Neurosurg 90:251–257 34. Howarth DM, Gilchrist GS, Mullan BP, Wiseman GA, Edmonson JH, Schomberg PJ (1999) Langerhans cell histiocytosis: diagnosis, natural history, management, and outcome. Cancer 85:2278–2290 35. Hua F, Asato R, Miki Y, Okumura R, Hashimoto N, Kikuchi H, Konishi J (1992) Differentiation of suprasellar nonneoplastic cysts from cystic neoplasms by Gd-DTPA MRI. J Comput Assist Tomogr 16:744–749 36. Imura H, Nakao K, Shimatsu A, Ogawa Y, Sando T, Fujisawa I, Yamabe H (1993) Lymphocytic infundibuloneurohypophysitis as a cause of central diabetes insipidus. N Engl J Med 329:683–689 37. Kaido T, Sasaoka Y, Hashimoto H, Taira K (2003) De novo germinoma in the brain in association with Klinefelter’s syndrome: case report and review of the literature. Surg Neurol 60:553–558; discussion, 559 38. Kanagaki M, Miki Y, Takahashi JA, Shibamoto Y, Takahashi T, Ueba T, Hashimoto N, Konishi J (2004b) MRI and CT findings of neurohypophyseal germinoma. Eur J Radiol 49:204–211 39. Kaufmann TJ, Lopes MB, Laws ER, Jr., Lipper MH (2002) Primary sellar lymphoma: radiologic and pathologic findings in two patients. AJNR Am J Neuroradiol 23:364–367 40. Kiortsis D, Xydis V, Drougia AG, Argyropoulou PI, Andronikou S, Efremidis SC, Argyropoulou MI (2004) The height of the pituitary in preterm infants during the first 2 years of life: an MRI study. Neuroradiology 46:224–226 41. Korogi Y, Takahashi M, Sakamoto Y, Shinzato J (1991) Cavernous sinus: correlation between anatomic and dynamic gadolinium-enhanced MR imaging findings. Radiology 180:235–237 42. Kucharczyk W, Davis DO, Kelly WM, Sze G, Norman D, Newton TH (1986) Pituitary adenomas: high-resolution MR imaging at 1.5 T. Radiology 161:761–765 43. Kucharczyk W, Peck WW, Kelly WM, Norman D, Newton TH (1987) Rathke cleft cysts: CT, MR imaging, and pathologic features. Radiology 165:491–495 44. Kucharczyk W, Bishop JE, Plewes DB, Keller MA, George S (1994) Detection of pituitary microadenomas: comparison of dynamic keyhole fast spin-echo, unenhanced, and conventional contrast-enhanced MR imaging. AJR Am J Roentgenol 163:671–679 45. Kucharczyk W, Montanera WJ, Becker LE (1996) The sella turcica and parasellar region. In: Atlas SW(ed) Magnetic resonance imaging of the brain and spine, 2nd edn. Lippincott-Raven, Philadelphia, pp 871–930 46. Listernick R, Charrow J, Gutmann DH (1999) Intracranial gliomas in neurofibromatosis type 1. Am J Med Genet 89:38–44 47. Maghnie M, Cosi G, Genovese E, Manca-Bitti ML, Cohen A, Zecca S, Tinelli C, Gallucci M, Bernasconi S, Boscherini B, Severi F, Arico M (2000) Central diabetes insipidus in children and young adults. N Engl J Med 343:998–1007

3.7  Pituitary Gland and Parasellar Region 48. Matsuno H, Rhoton AL, Jr., Peace D (1988) Microsurgical anatomy of the posterior fossa cisterns. Neurosurgery 23:58–80 49. McCormick PC, Post KD, Kandji AD, Hays AP (1989) Metastatic carcinoma to the pituitary gland. Br J Neurosurg 3:71–79 50. McMaster ML, Goldstein AM, Bromley CM, Ishibe N, Parry DM (2001) Chordoma: incidence and survival patterns in the United States, 1973–1995. Cancer Causes Control 12:1–11 51. Medlock MD, Madsen JR, Barnes PD, Anthony DS, Cohen LE, Scott RM (1997) Optic chiasm astrocytomas of childhood. 1. Long-term follow-up. Pediatr Neurosurg 27:121–128 52. Meyer FB, Lombardi D, Scheithauer B, Nichols DA (1990) Extra-axial cavernous hemangiomas involving the dural sinuses. J Neurosurg 73:187–192 53. Miki Y, Matsuo M, Nishizawa S, Kuroda Y, Keyaki A, Makita Y, Kawamura J (1990) Pituitary adenomas and normal pituitary tissue: enhancement patterns on gadopentetate-enhanced MR imaging. Radiology 177:35–38 54. Miki Y, Asato R, Okumura R, Hua F, Konishi J (1992) Contrast enhanced area of posterior pituitary gland in early dynamic MRI exceeds hyperintense area on T1-weighted images. J Comput Assist Tomogr 16:845–848 55. Miki Y, Asato R, Hashimoto N, Konishi J (1999) Ectopic posterior pituitary in macroadenomas: demonstration by dynamic MR imaging. Book of Abstracts, 7th Annual Meeting of Internal Society for Magnetic Resonance in Medicine, May 22nd-28th, 1999, Philadelphia, p. 913 56. Miki Y, Kanagaki M, Takahashi JA, Ishizu K, Nakagawa M, Yamamoto A, Fushimi Y, Okada T, Mikuni N, Kikuta K, Hashimoto N, Togashi K (2007) Evaluation of pituitary macroadenomas with multidetector-row CT (MDCT): comparison with MR imaging. Neuroradiology 49:327–333 57. Miki Y, Kataoka ML, Shibata T, Haque TL, Kanagaki M, Shimono T, Okada T, Hiraga A, Nishizawa S, Ueda H, Rahman M, Konishi J (2005) The pituitary gland: changes on MR images during the 1st year after delivery. Radiology 235:999–1004 58. Moses AM, Clayton B, Hochhauser L (1992) Use of T1weighted MR imaging to differentiate between primary polydipsia and central diabetes insipidus. AJNR Am J Neuroradiol 13:1273–1277 59. Nagahata M, Hosoya T, Kayama T, Yamaguchi K (1998) Edema along the optic tract: a useful MR finding for the diagnosis of craniopharyngiomas. AJNR Am J Neuroradiol 19:1753–1757 60. Osorio MG, Marui S, Jorge AA, Latronico AC, Lo LS, Leite CC, Estefan V, Mendonca BB, Arnhold IJ (2002) Pituitary magnetic resonance imaging and function in patients with growth hormone deficiency with and without mutations in GHRH-R, GH-1, or PROP-1 genes. J Clin Endocrinol Metab 87:5076–5084 61. Rand T, Lippitz P, Kink E, Huber H, Schneider B, Imhof H, Trattnig S (2002) Evaluation of pituitary microadenomas with dynamic MR imaging. Eur J Radiol 41:131–135

62. Rodriguez O, Mateos B, de la Pedraja R, Villoria R, Hernando JI, Pastor A, Pomposo I, Aurrecoechea J (1996) Postoperative follow-up of pituitary adenomas after transsphenoidal resection: MRI and clinical correlation. Neuroradiology 38:747–754 63. Rogg JM, Tung GA, Anderson G, Cortez S (2002) Pituitary apoplexy: early detection with diffusion-weighted MR imaging. AJNR Am J Neuroradiol 23:1240–1245 64. Rosenblum MK, Matsutani M, Van Meir EG (2000) CNS germ cell tumours. In: Kleihues P, Cavenee W (eds) Pathology and genetics tumours of the nervous system. IARC, Lyon, pp 208–214 65. Saeki N, Hoshi S, Sunada S, Sunami K, Murai H, Kubota M, Tatsuno I, Iuchi T, Yamaura A (2002) Correlation of high signal intensity of the pituitary stalk in macroadenoma and postoperative diabetes insipidus. AJNR Am J Neuroradiol 23:822–827 66. Saeki N, Uchino Y, Murai H, Kubota M, Isobe K, Uno T, Sunami K, Yamaura A (2003) MR imaging study of edemalike change along the optic tract in patients with pituitary region tumors. AJNR Am J Neuroradiol 24:336–342 67. Sakurai K, Fujita N, Harada K, Kim SW, Nakanishi K, Kozuka T (1992) Magnetic susceptibility artifact in spin-echo MR imaging of the pituitary gland. AJNR Am J Neuroradiol 13:1301–1308 68. Sartoretti-Schefer S, Wichmann W, Aguzzi A, Valavanis A (1997) MR differentiation of adamantinous and squamous-papillary craniopharyngiomas. AJNR Am J Neuroradiol 18:77–87 69. Sato N, Ishizaka H, Matsumoto M, Matsubara K, Tsushima Y, Tomioka K (1991) MR detectability of posterior pituitary high signal and direction of frequency encoding gradient. J Comput Assist Tomogr 15:355–358 70. Sato N, Ishizaka H, Yagi H, Matsumoto M, Endo K (1993) Posterior lobe of the pituitary in diabetes insipidus: dynamic MR imaging. Radiology 186:357–360 71. Sato N, Endo K, Kawai H, Shimada A, Hayashi M, Inoue T (1995a) Hemodialysis: relationship between signal intensity of the posterior pituitary gland at MR imaging and level of plasma antidiuretic hormone. Radiology 194:277–280 72. Sato N, Tanaka S, Tateno M, Ohya N, Takata K, Endo K (1995b) Origin of posterior pituitary high intensity on T1-weighted magnetic resonance imaging. Immunohistochemical, electron microscopic, and magnetic resonance studies of posterior pituitary lobe of dehydrated rabbits. Invest Radiol 30:567–571 73. Sato N, Putman CM, Chaloupka JC, Glenn BJ, Vinuela F, Sze G (1997) Pituitary gland enlargement secondary to dural arteriovenous fistula in the cavernous sinus: appearance at MR imaging. Radiology 203:263–267 74. Sato N, Sze G, Endo K (1998) Hypophysitis: endocrinologic and dynamic MR findings. AJNR Am J Neuroradiol 19:439–444 75. Schubiger O, Haller D (1992) Metastases to the pituitary—hypothalamic axis. An MR study of 7 symptomatic patients. Neuroradiology 34:131–134

431

432

3  Brain, Head, and Neck 76. Seltzer S, Mark AS, Atlas SW (1991) CNS sarcoidosis: evaluation with contrast-enhanced MR imaging. AJNR Am J Neuroradiol 12:1227–1233 77. Shimono T, Hatabu H, Kasagi K, Miki Y, Nishizawa S, Misaki T, Hiraga A, Konishi J (1999) Rapid progression of pituitary hyperplasia in humans with primary hypothyroidism: demonstration with MR imaging. Radiology 213:383–388 78. Sohn CH, Kim SP, Kim IM, Lee JH, Lee HK (2003) Characteristic MR imaging findings of cavernous hemangiomas in the cavernous sinus. AJNR Am J Neuroradiol 24:1148–1151 79. Sparacia G, Iaia A, Banco A, D’Angelo P, Lagalla R (2000) Transfusional hemochromatosis: quantitative relation of MR imaging pituitary signal intensity reduction to hypogonadotropic hypogonadism. Radiology 215:818–823 80. Suzuki M, Matsui O, Ueda F, Matsushita T, Fujinaga Y, Kobayashi K, Horichi Y, Hayashi Y, Tachibana O, Yamashita J (2005) Dynamic MR imaging for diagnosis of lesions adjacent to pituitary gland. Eur J Radiol 53:159–167 81. Sze G, Uichanco LS III, Brant-Zawadzki MN, Davis RL, Gutin PH, Wilson CB, Norman D, Newton TH (1988) Chordomas: MR imaging. Radiology 166:187–191 82. Takahashi T, Miki Y, Takahashi JA, Kanagaki M, Yamamoto A, Fushimi Y, Okada T, Haque TL, Hashimoto N, Konishi J, Togashi K (2005) Ectopic posterior pituitary high signal in preoperative and postoperative macroadenomas: dynamic MR imaging. Eur J Radiol 55:84–91

83. Taketomi A, Sato N, Aoki J, Endo K (2004) The effects of frequency-encoding gradient upon detectability of the margins and height measurements of normal adult pituitary glands. Neuroradiology 46:60–64 84. Tsunoda A, Okuda O, Sato K (1997) MR height of the pituitary gland as a function of age and sex: especially physiological hypertrophy in adolescence and in climacterium. AJNR Am J Neuroradiol 18:551–554 85. Wolansky LJ, Gallagher JD, Heary RF, Malantic GP, Dasmahapatra A, Shaderowfsky PD, Budhwani N (1997) MRI of pituitary abscess: two cases and review of the literature. Neuroradiology 39:499–503 86. Yagi H, Nagashima K, Miyake H, Tamai S, Onigata K, Yutani S, Kuroume T (1994) Familial congenital hypopituitarism with central diabetes insipidus. J Clin Endocrinol Metab 78:884–889 87. Yagi A, Sato N, Taketomi A, Nakajima T, Morita H, Koyama Y, Aoki J, Endo K (2005) Normal cranial nerves in the cavernous sinuses: contrast-enhanced three-dimensional constructive interference in the steady state MR imaging. AJNR Am J Neuroradiol 26:946–950 88. Yoon PH, Kim DI, Jeon P, Lee SI, Lee SK, Kim SH (2001) Pituitary adenomas: early postoperative MR imaging after transsphenoidal resection. AJNR Am J Neuroradiol 22:1097–1104

3.8  The Orbits

3.8  The Orbits N. Hosten, C. Zwicker, and M. Langer 3.8.1  Introduction MRI and ultrasound have successfully been introduced as first-line imaging modalities of orbital and ocular space-occupying lesions; the role of CT is confined to those disease entities for which detection of calcifications is of importance (retinoblastoma, perioptic meningioma, osteoma). Primary and secondary orbital tumors are rare compared with brain tumors. Primary lesions of the orbit constitute 80–90% of all cases; optic nerve glioma, meningioma, lymphoma, tumors of the lacrimal gland and metastases are the extraocular lesions seen most often. Malignomas of the paranasal sinuses may infiltrate the orbits secondarily and the same holds true for infiltration by systemic disease of the hematopoietic system. Benign lesions, such as osteoma, may also be found (Table 3.8.1). As for tumors of the eye, melanoma is seen most frequently in adults and retinoblastoma in young children. Malignant melanoma is the most frequent ocular tumor of adulthood. It nearly always is unilateral; 85% of these tumors are located in the choroid, 9% in the ciliary body, and 6% in the iris. Retinoblastoma is a tumor of early childhood; two thirds of all tumors are diagnosed before children are 3 years old. Benign lesions most often found in the orbit are Graves’ ophthalmopathy, orbital pseudotumor, hamartoma, and cavernous hemangioma. Neurofibromas are somewhat less often seen, depending on the population imaged.

Table 3.8.1  Tumors of the eye and orbit Tumor

Relative frequency (%)

Peak incidence (years)

Malignant melanoma

1–4

50–70

Retinoblastoma

<1

Up to 3

Optic glioma

2–3

10–60

Lymphoma

7–14

50–60

Lacrimal gland tumors

5

30–50

Metastases

2–9

50–70

Hemangioma, cavernous

8–16

50–70

Pseudotumor

8

40–70

Others

49

Mesenchymal tumors like liposarcoma and lipofibroma, chondroma, and osteoma constitute only a small portion of orbital tumors. 3.8.2  Examination Techniques 3.8.2.1  Patient Preparation and Selection of Coils Space-occupying lesions of the eye and tumors of the anterior and medial third of the orbit should be imaged using surface coils (Lemke 2004). If symptoms are bilateral or if a comparison to the not affected eye is necessary, a larger surface coil (>10 cm diameter) may be used to image both eyes simultaneously. Surface coils are usually available in different diameters, with those suited for examination of the eye starting at a diameter of 4 cm. A smaller diameter reduces the noise, thus improving on signal to noise ratio and making possible better spatial resolution, and a smaller slice thickness. Larger diameter surface coils (diameter 10 to 14 cm) and head coils may be employed for good results when imaging the orbits’ apex. Higher field strength tomographs (3 T) generally improve signal-to-noise and contrast-to-noise ratios. As higher field strength examinations are more sensitive to motion artifacts, anesthesia of the retrobulbar space is necessary to suppress eye movements (Hosten et al. 1997; Lemke et al. 2006a). 3.8.2.2  Imaging Planes The transverse plane is standard for MR imaging of the eye and orbit. Coronal images add information when imaging the lacrimal gland, Graves’ ophthalmopathy, and diseases of the eye muscles (Table 3.8.2). Sagittal images are indicated when lesions are expected in the superior or inferior part of either eye or orbit. To minimize partial volume effects slice thickness should not exceed 3 mm.

Table 3.8.2  Imaging planes and preferred use Image plane

Region imaged

Transverse

All

Frontal

Eye muscles, orbital apex

Sagittal

Apical and caudal orbit

Paraxial

Optic nerve, superior, and inferior rectus muscle

433

434

3  Brain, Head, and Neck

3.8.2.3  Pulse Sequences Delineation of orbital anatomy is best achieved by T1-weighted SE images with a TR (at 1.5 T) of 300–800 ms and a TE of 10–20 ms (Table 3.8.3). Fast SE sequences should be employed for acquisition of T2-weighted images. IR sequences, especially STIR sequences with short inversion times (TI between 100 and 150 ms), help to null signal from fat; they may be helpful in highlighting tumors located in retrobulbar fat. GRE sequences have certain drawbacks in imaging the orbit: due to differences in magnetic susceptibility caused by air-filled sinuses or by the bones of the skull base, artificially reduced signal may cover pathologic processes. Motion artifacts play an additional role. For dynamic contrast-enhanced examinations, which may be of use in differentiation between schwannoma and hemangioma, T1-weighted sequences with reduced TR and 1 average (Tanaka et al. 2004) offer the advantage of shorter imaging times, with better temporal resolution of the dynamic contrast. 3.8.2.4  Use of Contrast Media Regarding the role of contrast media in orbital and ocular imaging, only a few published reports exist. For uveal melanoma contrast enhancement is caused by hypervascularization of the tumor. Differentiation between melanoma and hemorrhage is thus facilitated. Perioptic meningioma also enhances after intravenous paramagnetic contrast application; the enhancement may serve for differentiation from other pathologies in this area. Analogous to the blood–brain barrier, the intactness or breakdown of the blood–ocular barrier could be demonstrated in animal experiments by absence or presence of contrast (Barkovich et al. 1997) extravasations into the anterior chamber and/or the vitreous. 3.8.3  Normal Anatomy Orbital anatomy is best delineated when surface coils are used with 3-mm-thick T1-weighted spin-echo sequences. Orbital anatomy should always be imaged in multiple

Table 3.8.3  Pulse sequences (1.5 T) Sequence (TR/TE [ms])

Slice thickness (mm)

NEX

T1-weighted

SE (400/20)

3

3

T2-weighted

TurboSE (3,000/70)

3

1

planes. Lens and vitreous may be separated from surrounding tissue using this approach (Fig. 3.8.1). On T1-weighted images the vitreous and fluid in the anterior chamber have low signal and may be separated from the capsule of the lens. T2-weighted images have lower spatial resolution; lens and vitreous have high signal while the lens’ signal is low. Eye movements may cause spatial misregistration; in the direction of the ghosting multiple images of the globe are often seen on T2-weighted images and—though to a much lesser degree—on T1-weighted images. Improved spatial resolution may be achieved through use of small surface coils. They allow differentiating retina and choroid, which are contrast enhancing, from the sclera, which is not. The orbital septum is seen on high-resolution T1-weighted images as a thin band of low signal intensity (Hoffmann et al. 1998). When higher field strengths are used for imaging the orbit, chemical shift induces artifacts (Kim et al. 2006), which run along the globe or the optic nerve, depending on the direction of frequency- and space-encoding gradients. They should not be mistaken for capsules or other anatomical structures. Retrobulbar fat serves as a natural contrast medium on T1-weighted images. The use of fat suppressed images should therefore be restricted to those cases where contrast medium is applied. Precontrast images should additionally be acquired without fat saturation in different orientations. 3.8.4  Pathological Lesions 3.8.4.1  Intrabulbar Space-Occupying Lesions 3.8.4.1.1  Uveal Malignant Melanoma Malignant uveal melanoma is the most frequent malignant ocular tumor seen in adults. It is mostly encountered one-sided and has a peak in the sixth decade of life. Melanotic melanomas are characterized by high signal on plain T1-weighted images. On T2-weighted images the tumor’s signal is low (Fig. 3.8.2). This phenomenon is caused by T1 and T2 relaxation times, which are shorter than those seen in other tumors. Melanin has a paramagnetic effect and causes this shortening of relaxation times (Table 3.8.4; Lemke et al. 1999). Spin-echo images may differentiate malignant melanotic melanoma from subacute hemorrhage (5 –7 days). Subacute hemorrhage has high signal on both T1- and T2 weighted images (Fig. 3.8.3). The higher signal from subacute hemorrhage is caused by the well-known conversion of hemoglobin to methemoglobin, which is the reason for shorter T1 and T2 relaxation times (Table 3.8.4). Amelanotic melanoma, which is rather rare (5% of all melanomas), differs in its signal from melanotic melanoma due to the absence of melanin’s paramagnetic effect. Amelanotic melanoma may therefore not be differentiated from other patholo-

3.8  The Orbits

Fig. 3.8.1  Normal MR anatomy delineated with a larger surface coil (a–c) and a small surface coil (d,e). Use of surface coil with a diameter of 11 cm allows simultaneous imaging of both orbits. Pathological lesions are easy to recognize. Transverse images are well suited for imaging the medial and lateral rectus muscles (a,b). On frontal images the inferior, superior rectus muscles and the superior oblique muscle is seen. Note that with use of a surface coil signal in regions at a distance from the coil

decreases. This effect is seen with both coils; it is more pronounced with the smaller coil (d,e). On the other hand, spatial resolution is even better with the smaller surface coil. 1 lateral rectus muscle, 2 medial rectus muscle, 3 optic nerve, 4 superior ophthalmic vein, 5 inferior ophthalmic vein, 6 lacrimal gland, 7 superior oblique muscle, 8 inferior rectus muscle, 9 superior rectus muscle, 10 levator palpebrae muscle

435

436

3  Brain, Head, and Neck

Fig. 3.8.2  Melanotic uveal melanoma. Appearance on different pulse sequences. A characteristic finding in melanotic melanoma is the high signal on T1-weighted images (a). Paramagnetic melanin causes this phenomenon. At the posterior pole of the globe there is a small subretinal effusion. After intravenous application of contrast medium, signal of the tumor increases. The band consisting of uvea and retina also shows some enhancement, whereas the subretinal effusion does not (b). The melanoma has low signal on the T2-weighted image (c). Subretinal effusion is somewhat brighter

Fig. 3.8.3  Vitreous hemorrhage. Vitreous has high signal on the T2-weighted image due to its high water content. This is seen on the right eye, which is somewhat deformed (a). In the left eye there is a complete retinal detachment, appearing as a linear structure between the posterior surface of the lens and the pos-

terior pole of the eye. Hemorrhage appears as a circular signal change in this eye. On the plain T1-weighted image (b), there is incomplete retinal detachment in the right eye, appearing as a V-shaped linear structure. Hemorrhage in the left vitreous has high signal on the T1-weighted image

3.8  The Orbits

gies by its signal alone. Intravenous application of paramagnetic contrast medium is therefore recommended for the differentiation of hemorrhage on the one hand, amelanotic melanoma on the other. Paramagnetic contrast media lead to an increase in signal on T1-weighted images in cases of malignant melanoma (Fig. 3.8.4). It is probably caused by vascularization and a wide interstitium. Choroidal metastases, however, may not easily be differentiated from small melanoma. MRI is often useful to exclude extraocular growth of malignant melanoma; it is superior to ultrasound in this regard (Hosten et al. 1999). Due to the high signal of choroidal hemangioma on T2-weighted images MRI can differentiate this benign condition from melanotic melanoma (Hosten et al. 1998).

Table 3.8.4  Signal intensities of selected pathologies on T1- and T2-weighted images

Fig. 3.8.4  Contrast enhancement in a melanotic melanoma (a,b) and appearance of retinal detachment and sediment at 3T (c,d; different patient) (a,b 1.5 T; c,d 3 T). On the plain T1weighted image (a) the melanoma has broken through Bruch’s membrane, as is apparent from its mushroom shape. Melanin content is high, as is the signal intensity on the T1-weighted im-

age. Despite the high signal on a, some increase in signal intensity is visible after intravenous contrast (b). At 3 T (c,d) the hemorrhagic component of subretinal fluid accompanying the tumor decreases signal intensity. This is marked in the sediment (a,b) and appearance of retinal detachment and sediment at 3 T (c,d; different patient)

Lesion

T1-weighted

T2-weighted

Malignant melanoma, melanotic

+

–

Hemorrhage, subacute

+

+

Retinoblastoma

–

–

Hemangioma

–

+

Lymphoma

–

(+)

Orbital pseudotumor

(–)

(+)

437

438

3  Brain, Head, and Neck

3.8.4.1.2  Retinoblastoma

3.8.4.2  Retrobulbar Space-Occupying Lesions

Retinoblastoma (Fig. 3.8.5) is the most frequent ocular tumor of early childhood; 80–90% of all cases are observed in children younger than 3 years. Due to the different age of manifestation, a differential diagnosis between melanoma and retinoblastoma is not necessary (Table 3.8.4). Retinoblastoma is treated without a definite histological diagnosis from clinical findings and CT detection of intratumoral calcification alone. Low-signal-intensity spots inside the tumor may on MR images suggest calcifications; however, calcifications cannot be diagnosed with the same certainty as on CT. Ultimately they cannot be differentiated from air or hemorrhage. Retinoblastoma enhances after IV contrast application; choroidal invasion is difficult to diagnose. The same holds true for invasion of the optic nerve by the tumor (Schueler et al. 2003; Lemke et al. 2007).

3.8.4.2.1  Optic Glioma

Fig. 3.8.5  Retinoblastoma. The contrast-enhancing tumor is visible on the enhanced T1-weighted image (a). There is retinal detachment. Small dots with very low or no signal correspond to calcification. Calcification (b) is usually better seen on CT, as in this image from another patient. The macroscopic specimen (c)

Optic nerve glioma is classified as a pilocytic astrocytoma. Seventy-five percent of these tumors occur in patients under the age of 10 years and 90% occur in patients under the age of 20 years. Gliomas of the optic nerve may be highly malignant when they occur in older patients (Astrup, 2003; Miller et al. 2004). Signal intensity of optic nerve glioma does not differ significantly from that of gray matter on both T1- and T2-weighted images. The optic nerve is surrounded by a small, fluid-filled space, which communicates with the intracranial external liquor spaces. The high signal intensity of cerebrospinal fluid on T2-weighted- images and its low signal on T1-weighted highlights both the optic nerve and optic nerve tumors on paraxial images.

demonstrates white tumor masses besides the detached retina. An important differential diagnosis is Coats’ disease (d). Coats’ disease becomes clinically apparent at about the same patient age, but no space-occupying mass is demonstrated by MRI

3.8  The Orbits

As noted above, IV application of contrast medium helps differentiate optic nerve glioma from perioptic meningioma. In the case of optic nerve glioma the optic nerve shows fusiform enlargement. Enhancement completely covers the enlarged optic nerve. In perioptic meningioma, enhancing tissue surrounds the otherwise normal optic nerve. MRI may not necessarily differentiate neuritis of the optic nerve from glioma (Hesselink et al. 1991). 3.8.4.2.2  Meningioma Orbital meningioma may appear in the form of perioptic meningioma or as intraorbital sphenoid wing meningioma. Alternatively, intraorbital meningioma may very rarely stem from ectopic arachnoid cells (Bilaniuk et al. 1985; Moster et al. 2005; Thiagalingam et al. 2004). Meningioma accounts for 3–7% of all orbital tumors. Eighty percent of meningiomas are seen in women. The incidence peaks between the third and fifth decade. Signal intensity is close to that of extraocular eye muscles on T1-weighted images, on T2-weighted images the tumors are isointense or slightly hypointense to retrobulbar fat. Tumor calcification may be suspected (“tram-track sign”), but CT is superior to MRI in this respect. Intracranial extension, however, is best demonstrated by contrastenhanced MRI when an enhancing mass is in continuity with the intracranial meninges (Moster et al. 2005). 3.8.4.2.3  Hemangioma Hemangioma is the most frequent benign retrobulbar space-occupying lesion. In adults, cavernous hemangiomas are seen, while capillary hemangiomas are seen in children. Cavernous hemangiomas have variable appearance on MR images (Abe et al. 2000): on T1-weighted images hemangiomas usually have low signal, isointense to extraocular eye muscles (Fig. 3.8.6; Table 3.8.4). In typical cases, the tumors show small dots of high signal on T1-weighted images. These probably correspond to small thrombi. On T2 w images cavernous hemangiomas are hyperintense to extraocular eye muscles and to the optic nerve. On heavily T2-weighted images (very long TR) the tumor is nearly always demarcated from its surroundings by its high signal. Differentiation from schwannoma which resembles cavernous hemangioma in size, localization, and in its smooth margins is possible on contrast-enhanced dynamic series. Schwannoma shows homogeneous contrast enhancement on dynamic imaging. In cavernous hemangioma, contrast enhancement usually starts in a circumscript point of the lesion. It spreads from there through the entire tumor during later phases (Yan et al. 2004). Enhancement in these tumors starts at the entry point of the feeding vessel.

Fig. 3.8.6  Cavernous hemangioma. The tumor, which is the most frequent retrobulbar tumor, almost always is localized inside the muscle cone. On the T2-weighted image the signal is usually increased compared with muscle (a). Signal increase is spotty in the case shown. Note displacement of the optic nerve with compression of the subarachnoid space

439

440

3  Brain, Head, and Neck

Cavernous hemangioma is nearly always located in the intraconal space (inside the funnel built by the four straight extraocular eye muscles). Capillary hemangiomas are located extraconally and in the anterior parts of the bony orbit. They are characterized by a mass of tubular structures of low signal intensity. Contrast enhancement is dramatic in capillary hemangioma. 3.8.4.2.4  Lymphangioma Lymphangioma is a tumor of early childhood. A polycystic tumor with fluid-fluid levels in multiple cells, exhibiting low signal in the lower part of the cells is virtually pathognomonic (Lemke et al. 2004). 3.8.4.2.5  Lymphoma of the Orbit Signal intensity of orbital lymphoma does not differ from that of other retrobulbar lesions. It is hypointense to fat on T1-weighted images and shows a slight increase on T2-weighted images. The tumor is located anteriorly rather than in the posterior parts of the orbit. The tendineous parts of the extraocular eye muscles are involved, while inflammation and thyroid ophthalmopathy involve the eye muscles themselves (Hosten et al. 1991; Akansel et al. 2004).

Fig. 3.8.7  Lacrimal gland carcinoma. The lacrimal gland is located in the upper outer quadrant of the orbit. Caudally, a small rest of the gland is preserved. The gland is otherwise nearly completely destroyed by a space-occupying lesion with relatively smooth margins. This image was obtained after intravenous contrast. The tumor is strongly enhanced. The case is not characteristic for a lymphoma, which would enlarge the gland but would preserve its almond shape

3.8.4.2.6  Orbital Pseudotumor Orbital pseudotumor is caused by a diffuse or circumscript inflammation of retrobulbar tissue. Pathogenesis is not completely understood. The tumor is isointense to muscle on T1-weighted images and hypointense to orbital fat. It may replace parts of the retrobulbar fat. Extraocular eye muscles may also be involved. On T2-weighted images orbital pseudotumor is hardly discernible from orbital fat but it may be slightly hyperintense in some cases. After intravenous contrast application, enhancement is absent or minimal. In rare cases an involved muscle has prolonged T2 in comparison to non-involved eye muscles (Hosten et al. 1991). If follow-up MRI is performed during anti-inflammatory therapy, a normalization of MR findings may be observed.

On T1-weighted MRI pleomorphic adenoma is more or less homogenous and isointense to gray matter (Fig. 3.8.7). On T2-weighted images the signal of pleomorphic adenomas is more inhomogeneous; solid components are more or less isointense to gray matter, while liquid areas have prolonged T2 relaxation times. Lymphoma of the lacrimal gland and pseudotumors have low signal intensity on T2-weighted images and are practically isointense to fat. While lymphoma and pseudotumor do enlarge the lacrimal gland but do not alter its contour, carcinomas and pleomorphic adenomas are often visible as a round, space-occupying lesion inside the otherwise normally configured lacrimal gland. Contrast enhancement (Calle et al. 2006) is inhomogeneous in pleomorphic adenoma but homogeneous and often intense in lymphoma.

3.8.4.2.7  Lacrimal Gland Tumors

3.8.4.2.8  Orbital Metastases

From 50 to 60% of all lacrimal gland tumors are pleomorphic adenomas; 5–10% of these may show carcinoma inside the adenoma. Adenoid-cystic carcinomas account for another 20–30% of cases, other carcinomas for the rest.

Two to 9% of orbital space-occupying lesions are metastases. Carcinomas of the breasts and lungs are the most frequent primary diseases (Albert et al. 1967). The signal is low on T1-weighted images, intermediate on T2weighted images.

3.8  The Orbits

3.8.4.2.9  Dermoids Dermoids may be diagnosed from their vicinity to the fronto-zygomatic suture in the lateral orbital wall. On close inspection, the suture is often seen incompletely fused. Ectodermal tissue components which were supposed to migrate from inside the orbit into a subcutaneous position remain intra-orbitally. Dermoids are often fat-containing and therefore characterized by high signal on plain T1-weighted images (Smirniotopoulos et al. 1995). 3.8.4.3  Non-Tumorous Retrobulbar Lesions 3.8.4.3.1  Graves’ Ophthalmopathy, or Thyroid Ophthalmopathy Graves’ ophthalmopathy is an autoimmune disorder. Mucopolysaccharides accumulate in the eye muscles and in the retrobulbar fat. Painful exophthalmos is the result. The optic nerve may be compressed and vision compromised or lost. Graves’ ophthalmopathy most often affects the inferior rectus muscle, followed by the medial rectus muscle. Muscles enlarged by Graves’ ophthalmopathy are affected in their middle and in the parts closer to the optic foramen (Fig. 3.8.8). In the acute stage of Graves’ ophthalmopathy affected muscles have prolonged relaxation times compared with normal extraocular eye muscles. If patients respond to anti-inflammatory treatment or during the natural course of the disease, eye muscle edema resolves, T2 relaxation times return to normal. Fibrosis or fatty degeneration is the end stages of Graves’ ophthalmopathy. The former is characterized by slightly decreased T2 relaxation times of eye muscle, while the latter is often clearly visible on T1-weighted images. If T2 relaxation times of edematous eye muscles are calculated from spin-echo sequences with multiple TEs, a quantification of the changes is possible (Hosten et al. 1989; Hosten et al. 1992; Hosten et al. 1993; Higer et al. 1991).

Fig. 3.8.8  Graves’ ophthalmopathy in a 24-year-old patient. On the transverse T1-weighted image (a) the typical enlargement of eye muscles is visible. Note that the anterior tendineous part is not involved. The frontal T2 parameter image (b) demonstrates increased signal in the enlarged medial and inferior rectus muscles corresponding to edema, seen in acute eye muscle inflammation

close to insertion at the globe. This finding may help differentiate myositis from Graves’ ophthalmopathy.

3.8.4.3.2  Myositis

3.8.4.3.3  Vascular Lesions

Myositis is occasionally seen as a complication of inflammatory disease of the paranasal sinuses, of autoimmune disease (lupus erythematosus) or idiopathically. Contractility of affected eye muscles is restricted due to pain and diplopia may result. Eye muscle enlargement due to myositis tends to be more expressed than that caused by Graves’ ophthalmopathy. Eye muscles have low signal on T1-weighted images (Fig. 3.8.9) and do not differ from normal eye muscles. Myositis tends to enlarge the anterior and tendineous parts of extraocular eye muscles,

Capillary hemangiomas have been described above. Tubular structures corresponding to vessels are characteristic findings. Their signal is low if flow is present, but flow phenomena with different signal intensities may be observed. Fistulae between the internal carotid artery and the cavernous sinus often result in an increased diameter of the superior and/or inferior ophthalmic vein (Fig. 3.8.10). Both veins usually have no signal due to flow. Carotid– cavernous fistulae may drain through other structures as

441

442

3  Brain, Head, and Neck

melanin. Choroidal hemangioma has very high signal on T2-weighted images and low signal on T1-weighted images. Other bulbar lesions may not be differentiated by a different signal but choroidal metastases tend to grow in a plaque-like manner. In the extraocular space myositis of an extraocular eye muscle may be differentiated from eye muscle enlargement due to Graves’ ophthalmopathy by its location (Grannemann et al. 1988). Capillary and cavernous hemangiomas have distinct forms (capillary hemangioma has a tubular structure) and contrast behavior (fill-in in the case of cavernous hemangioma). Intravenous administration of paramagnetic contrast material may help in diagnosing other hypervascularized tumors as well (Dwyer et al. 1986; Fig. 3.8.9). In retinoblastoma CT is important due to its ability to detect calcifications, which support the diagnosis. MRI does not detect calcifications with the same certainty. The role of MRI would be enhanced if it could detect invasion of the optic nerve and of the choroid with more certainty. Differentiation between perioptic meningioma and optic glioma is improved by application of paramagnetic contrast material. In glioma the enlarged optic nerve enhances as a whole, while in perioptic meningioma the meninges surrounding the optic nerve show enhancement (tram-track sign). Intracranial involvement is suspected if the meninges close to the optic foramen enhance. The normal lacrimal gland has the shape of an almond. This shape is preserved in the case of lymphoma and inflammatory pseudotumor while carcinoma and pleomorphic adenoma are round, space-occupying lesions inside the lacrimal gland. 3.8.6  Diagnostic Procedures Fig. 3.8.9  Myositis. The contrast enhanced coronal image (a) demonstrates massive enlargement of the superior rectus muscle. Signal is higher than that of other extraocular eye muscles. On the sagittal image (b) involvement of the anterior tendineous part of the muscle is visible. This is a typical finding in myositis, while Graves’ ophthalmopathy, which also enlarges eye muscles, always spares the tendon

well. If drainage is via the petrous sinus, no dilatation of the ophthalmic veins is seen. 3.8.5  Differential Diagnosis Signal intensity on T1-weighted and T2-weighted images may help differentiation of melanotic intraocular melanoma from subacute hemorrhage or choroidal hemangioma (Grannemann et al. 1988). Amelanotic melanoma, which is rare, has signal characteristics not different from those of other tumors, due to the lack of paramagnetic

If orbital pathology is suspected, MRI should be the method of choice. Exceptions are the trauma patient and patients with suspected foreign bodies; in both cases CT is performed first. Work-up in all others should start with T1-weighted spin-echo images. High-resolution images with a slice thickness of 2 or 3 mm should be obtained. No fat suppression should be used. Transverse and coronal images depicting both eyes for comparison should then be studied for the detection of space-occupying lesions. The rest of the examination should be tailored according to the results of these two image sets. • If no lesion is seen, T2-weighted coronal images should be obtained to exclude discrete inflammatory changes. • If eye muscles are enlarged, sagittal and coronal T2weighted images and sagittal T1-weighted images should be obtained to differentiate myositis and Graves’ ophthalmopathy and detect acute inflammation. • If a round lesion inside the muscle cone is seen, transverse T2-weighted images should be obtained as well as a dynamic study with a paramagnetic contrast medium.

3.8  The Orbits

Ocular pathology will only be imaged in selected cases when ophthalmologists treat tumors in sufficient numbers. Examination should include two or three T1-weighted

spin-echo sequences and a single T2-weighted and contrast-enhanced image set. Orientation should be chosen in such a way that the tumor can be visualized in a perpendicularly oriented plane. Ocular and orbital MRI can be extremely worthwhile if clinical partners are found. While ophthalmologists treat tumors of the eye, the treatment of retrobulbar tumors often involves the interdisciplinary collaboration of neurosurgeons, head and neck specialists and ophthalmologists. Given the right patient selection, the appropriate MRI technique should be implemented for maximum patient benefit.

Fig. 3.8.10  Cavernous sinus fistula. The transverse T1-weighted image (a) demonstrates a linear structure where the right superior ophthalmic vein is suspected. Compare the left orbit, where the vessel is visible as a small dot. Massive enlargement of the vein is due to increased flow after arterialization. The coronal image (b) demonstrates the increase in caliber of the vein, which is visible right below the superior rectus muscle

Fig. 3.8.11  Rhabdomyosarcoma. The plain (a) and enhanced (b) T1-weighted images demonstrate a flat space-occupying lesion in the medial superior quadrant of the orbit. The signal intensity of this lesion is lower than that of retrobulbar fat on the plain image (a). There is some enhancement after contrast application (b). In children with findings like the one shown here, MRI of the head should also be performed to exclude intracranial extension

• An enlarged optic nerve should be studied with angulated T2-weighted and plain and contrast-enhanced T1-weighted images with fat suppression. • Lesions in contact with the orbital roof should be imaged with coronal images and possibly a head scan to exclude intracranial extension (Fig. 3.8.11). • Capillary hemangioma, dermoid, and pseudotumor may be excluded on the initial images.

443

444

3  Brain, Head, and Neck

References 1.

2.

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Abe T, Kawamura N, Homma H, Sasaki K, Izumiyama H, Matsumoto K (2000) MRI of orbital schwannomas. Neuroradiology 42:466–468 Akansel G, Hendrix L, Erickson BA, Demirci A, Papke A, Arslan A, Ciftci E (2005) MRI patterns in orbital malignant lymphoma and atypical lymphocytic infiltrates. Eur J Radiol 53:175–181 Albert DM, Rubenstein RA, Scheie HG (1967) Tumour metastases to the eye: 1. Incidence in 213 adult patients with generalized malignancy. Am J Ophthalmol 63:723–726 Astrup J (2003) Natural history and clinical management of optic pathway glioma. Br J Neurosurg 17:327–335 Barkovich AJ, Latal-Hajnal B, Partridge JC, Sola A, Ferriero DM (1997) MR contrast enhancement of the normal neonatal brain. Am J Neuroradiol 18:1713–1717 Born C, Rademaker J, Hosten N Felix R (2001) Hemorrhagic cavernoma or ruptured dermoid of the orbit: diagnosis with MRI. Orbit 20: 291–295 Calle CA, Castillo IG, Eagle RC, Daza MT (2006) Oncocytoma of the lacrimal gland: case report and review of the literature. Orbit 25:243–247 Grannemann D, Zwicker C, Langer M (1988) (Diagnosis of orbital space-occupying lesions-a comparison of magnetic resonance imaging and computerized tomography). Fortschr Ophthalmol 192: 327–329 Hesselink JR, Szumoswski J Tien RD (1991) MR fat suppression combined with Gd-DTPA enhancement in optic neuritis and perineuritis. J Comput Assist Tomogr 15:223–227 Higer HP, Just M, Kahaly G et al. (1991) Graves ophthalmopathy: role of MR imaging in radiation therapy. Radiology 179:187–190 Hoffmann KT, Hosten N, Lemke AJ, Sander B, Zwicker C, Felix R (1998) Septum orbitale: high-resolution MR in orbital anatomy. Am J Neuroradiol 19:91–94 Hosten N, Beckrakis NE, Lietz A, Noske W (1993) Endokrine Orbitopathie. Korrelation magnetresonanztomographischer und histopathologischer Befunde. Fortschr Röntgenstr 159:304–306 Hosten N, Bornfeld N, Fellix R, Lemke AJ, Sander B, Waßmuth R (1997) MR of the eye with retrobulbar anesthesia. Am J Neurorad 18:1788–1790 Hosten N, Bornfeld N, Felix R, Foerster P, Hoffmann KT, Lemke AJ, Schüler A, Strosczcynski C, Wiegel T (1998) Choroidal haemangioma: MR findings and differentiation from uveal melanoma. Am J Neurorad 19:1441–1447 Hosten N, Cordes M, Sander B et al (1989) Graves’ ophthalmopathy: MR imaging of the orbits. Radiology 172:759–762

16. Hosten N, Lietz A, Zwicker C et al (1991) Lymphozytäre Infiltrationen der Orbita in MRT und CT: Lymphom, Pseudolymphom und entzündlicher Pseudotumor. Fortschr Röntgenstr 155/5:445–451 17. Hosten N, Lietz A, Schörner W, Wenzel KW (1992) Der Krankheitsverlauf bei der endokrinen Orbitopathie: Magnetresonanztomographische Dokumentation. Fortschr Röntgenstr 157:210–214 18. Kim JH, Hwang JM (2006) Imaging of the superior rectus in superior rectus overaction after retrobulbar anesthesia. Ophthalmology 113: 1681–1684 19. Lemke AJ, Hosten N, Bechrakis NE, Bornfeld N, Felix R, Gurvit Ö, Richer M, Schüler A, Stroszczynski C (1999) Histopathological-radiological correlation of choroidal melanoma using high resolution MRI with a surface coil. Radiology 210: 775–783 20. Lemke AJ, Kazi I, Landeck LM, Zaspel U, Hosten N, Felix R (2004) Differenzialdiagnostik intrakonaler orbitaler Raumforderungen unter Verwendung der hochauflösenden MRT mit Oberflächenspulen anhand von 78 Patienten. Fortschr Röntgenstr 176:1436–1446 21. Lemke AJ, Alai-Omid M, Hengst Sa, Kazi I, Felix R (2006a) Eye imaging with a 3.0-T MRI using a surface coil—a study on volunteers and initial patients with uveal melanoma. Eur Radiol 16:1048–1049 22. Lemke AJ, Kazi I, Mergner U et al (2007) Retinoblastoma—MR appearance using a surface coil in comparison with histopathological results. Eur Radiol 17:49–60 23. Miller NR (2004) Primary tumours of the optic nerve and its sheath. Eye 18:1026–1037 24. Moster ML. Detection and treatment of optic nerve sheath meningioma (2005) Curr Neurol Neurosci Rep 5:367–75 25. Schueler AO, Hosten N, Bechrakis NE, Lemke AJ, Foerster P, Felix R, Foerster MH, Bornfeld N (2003) High-resolution magnetic resonance imaging of retinoblastoma. Br J Ophthalmol 87:330–335 26. Smirniotopolous JG, Chiechi MV (1995) Teratomas, dermoids, and epidermoids of the head and neck. Radiographics 15:1437–1455 27. Tanaka A, Mihara F, Yoshiura T, Togao O, Kuwabara Y, Natori Y, Sasaki T, Honda H (2004) Differentiation of cavernous haemangioma from schwannoma of the orbit: a dynamic MRI study. Am J Roentgenol 183:1799–1804 28. Thiagalingam S, Flaherty M, Billson F, North K (2004) Neurofibromatosis type 1 and optic pathway gliomas: follow-up of 54 patients. Ophthalmology 111:568–577 29. Yan J, Wu Z (2004) Cavernous haemangioma of the orbit: analysis of 214 cases. Orbit 23:33–40

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

3.9  Magnetic Resonance of the Skull Base and Petrous Bone R. Maroldi, D. Farina, A. Borghesi, E. Botturi, and C. Ambrosi 3.9.1  Introduction Both the skull base and the temporal bone are composed of bone covered by the dura mater on its intracranial surface. Bone thickness ranges from less than 1 mm—at the very thin cribriform plate—up to 20–25 mm of the petrous bone or the clivus. While the inner covering of the skull base faces the CSF from all points, its outer surfaces contact very different structures. These include mucosa-lined air cavities or complex compartments (as the orbit), which face the outer surfaces of the anterior and central skull base, and deep spaces of the face containing muscles, vessels, and nerves separated by fat (as in the nasopharynx or in the parapharyngeal space), which face the external surfaces of the central and posterior skull base. Foramina, fissures, and channels discontinue and perforate the bony framework, allowing the passage of the brain stem, nerves, and vessels. The size, shape, and direction of the passages vary considerably according to the different structures traveling along their path. In addition, located within the petrous bone are very small

cavities of complex shape (cochlea and labyrinth) filled with fluid. Therefore, MR imaging of skull base and temporal bone is technically demanding, as it requires combining both high-spatial and high-contrast resolution. Moreover, numerous lesions have to be considered. In fact, pathological conditions not only arise directly from the structures which form the skull base and temporal bone, but they may extend into or through these structures either from below or from above. 3.9.1.1  Examination Techniques 3.9.1.1.1  Techniques for Imaging the Skull Base • Patient is imaged in the supine position. Careful adjustment of localizer sequences is recommended to obtain the best symmetry of skull base structures. No special patient preparation is necessary. • Head coils are generally sufficient to cover the skull base, the brain, the sinonasal tract, the nasopharynx, or the deep spaces adjacent to the skull base. • T2-weighted sequences are preferably obtained at first (Tables 3.9.1, 3.9.2). In fact, on T2-weighted images most lesions show intermediate signal intensity (SI) compared with muscles and can be easily distinguished from bone structures, retained secretions

Table 3.9.1  Parameters of the main sequences used for skull base and temporal imaging Sequence

Slice thickness (mm)

FOV

Matrix

Average

Acquisition time

TSE T2-weighted

3

180-210

256 × 512

3 or 2

2′06′′

SE T1-weighted

3

180-210

256 × 512

2

2′50′′

FS TSE T2-weighted

3

180-210

256 × 512

3

2′28′′

SE T1-weighted Gd

3 or 2

180-210

256 × 512

2

2′50′′

VIBE

0.5

230

320 × 448

1

3′40′′

5

180-230

346 × 512

2

3′10′′

1

256

256 × 256

1

9′14′′

3DFT-CISS a

0.7

180

256 × 256

1

3′ to 5′

3D FSE

0.5

160-180

320 × 320

4

12′ to 15′

CE-MRA

1

300

230 × 512

1

22′′

1

220

256 × 256

1

6′53′′

0.8

230

256 × 256

1

19′′ × 7 measurements (2′09′′)

4

230

128 × 128

4

1′30′′

a

FLAIR MPRAGE

a

TOF DCE-MRI (VIBE) EPI-DWI

a

Isotropic sequences allowing MPR (multiplanar reconstruction)

a

445

446

3  Brain, Head, and Neck Table 3.9.2  Skull base imaging protocol Contrast agent

Sequence

Plane

when

Pre-

TSE T2-weighted

Axial, coronal, and sag

Always

SE T1-weighted

Axial or coronal

Always

Fat-suppressed TSE T2-weighted

Variable

↑Contrast resolution of tumor vs. normal tissues

3DFT-CISS or 3D FSE

Variable

To demonstrate fluid-content lesions

EPI-DWI

Axial

To enhance tissue characterization

TOF or PC

Variable

To provide angiographic information

DCE-MR

Axial or coronal

Helps to separate recurrences vs. post-treatment changes

CE MRA

Coronal

To provide angiographic information

SE T1-weighted

Axial or coronal

Always

VIBE (MPRAGE)

Axial and/or coronal (sagittal)

Always

During

Post-

within sinusal cavities, nasopharyngeal muscular walls, and fat spaces adjacent to the skull base. Turbo (or fast) spin-echo T2-weighted sequences (TSE) are acquired more quickly than are conventional SE T2or T1-weighted sequences, thus maximizing the feasibility of acquiring sufficient data in uncooperative patients. A single—rarely more—non-enhanced SE (or TSE) T1-weighted sequence is usually the second step in the examination protocol. Besides enabling a more detailed anatomy than T2-weighted images do, this sequence provides a reference for assessing the degree of enhancement of lesions, and supplies valuable information about the signal intensity of both the cortical and the spongy/diploic bone of the skull base. Fat-saturated TSE T2-weighted sequences (Lenz 2000) have been advocated to enhance the conspicuity of the abnormal signal of both tumors and inflammatory lesions. However, suppression of fat signal may hamper the detectability of the boundaries of normal structures. After Gd-based contrast-agent administration, enhancement of both normal structures and lesions may be demonstrated either by SE (or TSE) T1-weighted sequences or gradient-echo sequences as the VIBE (volume-interpolated GE) or MPRAGE (ultrafast GE 3D). Because enhanced SE T1-weighted sequences benefit from the overall increased signal due to contrast agent presence within tissues, either the number of acquisitions (averages/measurements) or the slice thickness can be reduced without significant deterioration of image quality.

According to the first choice, the overall acquisition time can be noticeably reduced, whereas with the second solution an increased spatial resolution is obtained by employing thinner slices. On the contrary, fat suppression requires additional acquisition time in SE T1-weighted sequences, which may prove to be critical in the follow up of treated, oftenuncooperative, patients. VIBE sequences overcome this limitation, enabling faster acquisition of post-contrast fat-saturated high resolution T1-weighted images than fat-sat SE T1-weighted sequences (Fig. 3.9.1). In addition, VIBE sequences may provide 3D isotropic imaging, suitable for MPR and MIP reconstructions, with voxel size up to 0.5 mm (Maroldi 2005). Like VIBE, the MPRAGE sequence obtains a volume that can be examined in different planes, though images show greater T1 dominance. However, the practical use of the MPRAGE sequence is hampered by its longer acquisition time. MR angiographic techniques have to be used when it is necessary to image the arterial or venous vessels which approach the skull base from below, run across it or lie along its intracranial surface. Three different methods are available. Though time of flight (TOF) provides high-resolution angiograms, the volume coverage is limited, and the stationary tissue never totally suppressed. Nevertheless, 3D TOF details the circle of Willis or venous anatomy, and helps in identifying most large tumor feeders in highly vascularized neoplasms, like paragangliomas (van den Berg 2004).

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Fig. 3.9.1  VIBE sequence post contrast-agent administration. Axial slices of 0.5 mm thickness. a The plane cuts the basisphenoid level (BS). The high spatial resolution permits a detailed demonstration of the mandibular branch of the trigeminal nerve (V3) running through the foramen ovale. Posterior to the nerve is the bright signal of the middle meningeal artery (MMA). The internal carotid artery (ICA) is seen entering the external ca-

rotid canal. TS transverse sinus b A nasopharyngeal carcinoma (T) is demonstrated, confined within the pharyngobasilar fascia. The sequence precisely displays progressive branching of V3, the extracranial ICA in front of the jugular fossa filled by the internal jugular vein (IJV). The hypoglossal nerve (arrowheads) is shown running through the canal, surrounded by a venous plexus. SS sigmoid sinus

Phase contrast (PC) techniques enable excellent tissue suppression. In addition, quantification of flow velocity and selective demonstration of high or low-flow vessels are feasible. However, phase contrast angiography is technically demanding and requires a long acquisition time. Contrast-enhanced MR angiography (CE MRA) is a very rapid technique that offers high spatial and contrast resolution. Its optimal use requires accurate timing of the contrast bolus to achieve maximum arterial or venous concentration in the period corresponding to the k-space acquisition of high-contrast data. CE MRA reduces or eliminates most of the artifacts of TOF or PC angiography (Yang 2005), thus it is the method of choice whenever a long vessel segment has to be imaged. Demonstration of neoplastic involvement (displacement, encasement, wall invasion, occlusion) of the internal carotid artery, the dural sinuses or the jugular vein requires differentiating the signal intensities of the vessel wall from both tumor and blood flow. This can be done by using SE sequences with pre-saturation pulses which cause strong intravascular signal loss due to flow effects. These sequences use the flow-void effect as blood passes rapidly through the selected slice. An alternative solution is offered by the high resolution VIBE sequence, which shows the vessel wall as a low SI structure between the enhancing tumor and the hyperintense blood. When it is necessary to demonstrate fluid-content lesions extending beyond skull-base boundaries (like cephaloceles or meningoceles) high resolution 3D T2weighted fast SE sequences (FSE or FASE) or the 3D GE

T2-weighted Fourier transformation constructive interference in steady state (3DFT-CISS) sequence are indicated. Both methods provide isotropic sub-millimetric slices. Compared with 3DFT-CISS, 3D FSE and/or FASE are free from susceptibility artifacts but have the disadvantage of a very long acquisition time. Dynamic contrast-enhanced (DCE-MR) imaging is a still evolving and promising technique which has proved to be helpful in the differentiation of recurrent neoplasms from post-treatment scar. The short acquisition time of GE T1-weighted sequences (like VIBE) allows both sufficient time and spatial resolution for the analysis of neoplastic lesions at the skull base. As the enhancement pattern of tissues reflects capillary blood flow, permeability, and the relative volume of extra-vascular extra-cellular space, it provides semi-quantitative parameters related to tumor angiogenesis (Fig. 3.9.2). A significant technical improvement is expected from the introduction of blood-pool contrast agents, like gadofosveset trisodium. These agents have a mainly intravascular distribution in normal vessels, due to an inability to pass through the endothelial pores. This property will probably enhance the role of DCE-MR because evaluation of tissue perfusion will be more accurate than with low-molecular conventional contrast agents which rapidly diffuse into the extravascular, extracellular space. Diffusion-weighted imaging (DWI) of skull base lesions is not yet routinely performed because of difficulties with severe susceptibility artifacts caused by inhomogeneous signal from substances such as bone, air, fat, and

447

448

3  Brain, Head, and Neck

Fig. 3.9.2  TSE T2 sequence vs. DCE MRI. Suspected recurrence of undifferentiated nasopharyngeal carcinoma 7 years after RT. a The T2-weighted images shows heterogeneous high SI of the basisphenoid, associated with irregular outline (arrowheads).

b The parametric image obtained at the maximum enhancing slope shows remarkable signal from the recurrent neoplasm. BA basilar artery

soft-tissue (Fig. 3.9.3). Nonetheless, new techniques with sensitivity encoding (SENSE) have been shown to reduce blurring and off-resonance artifacts, thereby dramatically improving image quality of EPI-DWI images (Parmar 2005). • The skull base should be imaged in all three planes. Axial and coronal sections are usually acquired first, as both sequences permit a comparison between one side of the skull base and the opposite. The planes are oriented parallel and perpendicular to the hard palate, respectively. The sagittal plane is usually acquired later. • The choice of slice thickness depends on several aspects. Because it is necessary to achieve optimal space resolution (512 matrix), TSE T2-weighted images are usually acquired with 3-mm slice thickness. On a 1.5-T scanner, this combination requires an examination time of about 2 min for each TSE T2-weighted sequence to acquire up to 16 slices separated by an 80% inter-slice gap. Parallel imaging saves time; however, it takes a toll on signal intensity and homogeneity. A plain SE (TSE) T1-weighted sequence can be obtained with the same 3-mm slice thickness. If the examination time appears to be critical because a patient’s cooperation is poor, spatial resolution has to be reduced (down to 448, 384 or 256 matrix), while slice thickness should be increased up to 3.5–4 mm, and averages should be reduced to one. As a result, examination time decreases to approximately 1 min.

Conversely, on post-contrast SE T1-weighted sequences, the slice thickness may be reduced (from 3 to 2 mm) in fully cooperating patients, without deteriorating the image quality. In these patients, post-contrast isotropic VIBE may be acquired with 0.5-mm slice thickness, 448 matrix, and acquisition times less than 4 min. If necessary because of time constraints, the same anatomic coverage can be acquired in less than 2 min with a 0.7-mm isotropic resolution. Standard MPRAGE slice thickness is 1 mm, matrix 256. CE MRA is obtained with approximately 1-mm slice thickness. Isotropic sub-millimetric slice thickness is provided both by 3DFT-CISS (0.7 mm) and 3D FASE (up to 0.5 mm), the latter requiring 12–15 min of acquisition time. As MR imaging may be indicated for lesions primarily arising within or involving the skull base from above or from below, the whole intracranial content should be imaged with at least T2-weighted and contrast-enhanced T1-weighted sequences. • Very few lesions do not require contrast agent administration. If MR has to provide information about extra- and/or intra-cranial vessels alternative to digital angiography, small-molecular-weight contrast agents containing gadolinium as their active element are injected. Timing is dictated by the need to achieve the maximum

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Fig. 3.9.3  VIBE sequence and DWI. a Follow-up of an undifferentiated nasopharyngeal carcinoma 3 years after RT. The VIBE sequence shows persistent enhancement in front of the basiocciput (arrows). DWI (b = 1,000) indicates absence of recurrence

concentration during high-contrast k-space acquisition. Care-bolus or bolus tracking methods can be applied. When tumors are imaged, a delay of 2–3 min after contrast agent administration improves the differentiation of the enhanced neoplasm (usually washing-out at that time) from surrounding inflammatory reactive changes (very often showing brighter signal). 3.9.1.1.2  Techniques for Imaging the Temporal Bone The patient is imaged in the supine position by means of a head coil; a temporomandibular joint coil can be used in some cases. The head coil has the advantage of permitting

the examination of the whole head, the cerebellopontine angle (CPA), and both ears simultaneously. After a localizer sequence, the whole brain is imaged by a fluid attenuation recovery sequence (FLAIR) or a T2-weighted SE to detect or exclude lesions that could cause hearing impairment or disturbances similar to inner ear or CPA lesions (Table 3.9.3). Then, a high resolution 3D T2-weighted FSE or GE 3DFT CISS sequence is acquired in the axial plane. The 0.7-mm isotropic partitions provided by CISS are subsequently processed to obtain MPR reconstructions in the coronal plane and the parasagittal plane, the latter perpendicular to the internal auditory canal (IAC). Before and after Gd-based contrast agent administration, a 2D SE or 3D GE T1-weighted sequence (VIBE) are acquired in the axial and coronal planes. The VIBE se-

Table 3.9.3  Temporal bone imaging protocol Contrast agent

Sequence

Plane

When

Pre-

SE T2-weighted or FLAIR

Axial, coronal, and sag (whole brain)

Always

3DFT-CISS or 3D FSE

Axial (MPR should be obtained)

Always

TOF

Axial

To demonstrate neurovascular conflicts

During

CE MRA

Coronal

To demonstrate neurovascular conflicts

Post-

SE T1-weighted

Axial, and/or coronal, and/or sag

Always

VIBE or MPRAGE

Axial or coronal (MPR should be obtained)

Always

449

450

3  Brain, Head, and Neck

quence permits isotropic partitions up to 0.5 mm; 2D SE T1-weighted images are obtained with 2-mm thickness. 3.9.2  Normal Anatomy Viewed from above, the cranial floor appears separated into three naturally contoured regions: the anterior, middle, and posterior cranial fossae which approximately correspond to the anterior, central, and posterior skull base. 3.9.2.1  Key Anatomy of the Anterior Skull Base The anterior skull base is a narrow space limited anteriorly by the frontal bone (posterior frontal sinus wall), and posteriorly by the sphenoid bone (lesser wings and central planum sphenoidale). The central part is the deepest portion. It is formed by the thin cribriform plates of the ethmoid bone (roof of the ethmoid sinus) which are bordered laterally by the thickest fovea ethmoidalis (the medial portion of the orbital plates of the frontal bone). A thick triangular process, the crista galli, projects upwards from the midline of the cribriform plate and serves as a point of attachment of the falx cerebri. Laterally to the crista galli are numerous small foramina transmitting the olfactory nerves from the nasal mucosa to the olfactory bulb. Posteriorly, the lesser sphenoid wings form the roof of the optic canal. The crista galli is a key reference for the MR anat-

Fig. 3.9.4  Key references for anatomy of the anterior skull base. a The crista galli (CG) appears as a leaf-like midline structure. The vertical lamella of the cribriform plate (double white arrowheads) and the fovea ethmoidalis (black arrowhead) border the

omy of the anterior skull base. When imaged on a TSE T2-weighted coronal plane, the crista galli stands as a leaf-like midline bony structure (Fig. 3.9.4). Its short stem rises at a right angle from the horizontal part of the cribriform plate. On each side of the stem is the olfactory fossa, filled by CSF. Laterally to the fossa, the fovea ethmoidalis is detected as a vertical or sloping hypointense thick bony lamina. Within the fossa, the olfactory bulbs are imaged as a pair of low SI structures with a round shape, as they are cut perpendicularly to the long axis. As the fossa continues posteriorly into the planum sphenoidale (flat to slightly convex), the round shape of the olfactory bulb is replaced by the flat olfactory tract. Cranially to the olfactory bulb and tract is the olfactory sulcus, bordered by the gyrus rectus (medially) and medial orbital gyrus (laterally). The planum sphenoidale ends posteriorly at the level of the lesser wings, which mark the boundary with the central skull base. Their hypointense cortical outlines separate the oval hypointense optic nerve (medially located) from the hyperintense fat signal of the superior orbital fissure (lateral and inferior). 3.9.2.2  Key Anatomy of the Central Skull Base The body of the sphenoid bone and its greater wings form the central and lateral aspects of the central skull base, which includes the anterior part of the clivus (the basisphenoid). The petrous and squamous parts of the temporal bone are included by some authors in the central

olfactory fossa. b The round shape of the olfactory bulb (OB) is located at the floor of the olfactory fossa. OS olfactory sulcus, FC falx cerebri c The olfactory tract (OT) is a flat structure below the gyrus rectus (GR)

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

skull base. They will be discussed separately, as lesions arising at the petrous apex form a unique group of clinicopathological lesions. The key structure for the MR anatomy of the central skull base is the body of the sphenoid bone. On coronal MR, it is centrally located and shows a grossly quadrangular shape with cortical hypointense rounded outlines, spongy internal signal, and air content where the sphenoid sinus extends (Fig. 3.9.5). But the most important structures are located laterally to the sphenoid body (cavernous sinus, Meckel’s cave) and on its top (hypophysis, optic chiasm). On TSE T2-weighted and plain T1weighted images, the intermediate SI of the cavernous sinus borders the body of the sphenoid bone. Within the cavernous sinus, the carotid siphon appears as a flow-void structure. Lateral and inferior to the external straight profile of the cavernous sinus is the fluid signal of Meckel’s cave. It is characterized by an oval shape on both axial and coronal planes. On high-resolution TSE T2-weighted coronal sequences, Meckel’s cave is filled with the bright CSF signal and crossed by several small hypointense dots which correspond to the fibers of the trigeminal nerve within the cistern. On the floor of Meckel’s cave, the Gasser ganglion appears as a small convex defect within the hyperintense CSF. This defect continues downwards into

the hypointense mandibular nerve which exits the skull base via the foramen ovale. Anterior to the cavernous sinus level, the maxillary nerve—cut perpendicular to its long axis—is imaged as a hypointense tiny round structure on the top of the pterygoid process. Whether the nerve runs along a groove or a complete bony canal (foramen rotundum), it is surrounded by the high SI of a tiny venous plexus. Within the pterygoid process runs the smaller vidian canal, medial and inferior to the foramen rotundum. Larger at its anterior opening into the pterygopalatine fossa, the vidian canal narrows while leading posteriorly towards the foramen lacerum. On coronal images it appears as a small bony canal with a thin, round, cortical rim. Although the whole canal is usually detectable on thin axial and sagittal images, side-by-side comparison on the coronal plane helps to identify even subtle asymmetrical enhancement/ thickening of the vidian nerve (greater petrosal plus deep petrosal nerves) in case of early perineural spread. When examining the central skull base on MR, the pterygoid process of the sphenoid bone is a useful landmark. In fact, the pterygoid process is at a crossroads with several structures such as the pterygopalatine fossa (a common target for perineural spread from palate or maxillary sinus), the inferior and superior orbital fissure,

Fig. 3.9.5  Key references for anatomy of the central skull base. A large portion of the sphenoid (body, wings, and pterygoid processes) is imaged in the four coronal planes. a Lateral to the sphenoid sinus (SS) walls are the superior orbital fissures (SOF), the maxillary nerve (V2), and lesser wing of the sphenoid (LWS); close to the roof is the optic nerve (ON). b The vidian nerve (VN) is seen on the floor of the sphenoid sinus. c The cavernous sinus (CS) is bordered by a thick dural layer. Along the lateral aspect are the III, IV, VI nerves and V1 and V2 branches. Optic nerves reach the chiasm (OC). The pterygoid muscles converge toward the pterygoid laminae (MPM, LPM). Hypophysis (H). d Within the cavernous sinus a long portion of the internal carotid artery (ICA) is seen. Lateral to the cavernous sinus are the Meckel’s caves. Several low SI dots within the CSF filling the cistern (MC) indicate trigeminal nerve components. The Gasser ganglion is on the cistern floor (GG) emitting the mandibular branch (V3) imaged through the foramen ovale (arrowheads) and in the masticator space

451

452

3  Brain, Head, and Neck

the nasopharynx, and the sphenoid sinus. Careful evaluation of the adjacent structures is recommended in case the normal MR signal of its spongy bone content appears abnormal, or the cortical outline is eroded. 3.9.2.3  Key Anatomy of the Posterior Skull Base The posterior skull base has a deep inclining anterior wall—the clivus—which is formed by the posterior part of the sphenoid body and the basilar part of the occipital bone (the basiocciput). The anterior wall continues laterally into the temporal bones. The occipital bone accounts also for the central and posterior portions. Key structures for MR anatomy are the clivus, the hypo­glossal canal, the jugular fossa and the jugular vein. On the sagittal plane, the clivus appears as a triangular bone sloping down and posteriorly, towards the foramen magnum (Fig. 3.9.6). Its base corresponds to the posterior wall of the sphenoid sinus, while its apex lines the foramen magnum. Its signal intensity is homogeneous and reflects the gross composition of the bone marrow, mainly replaced by fat in the adult. The basisphenoid and the basiocciput meet at the spheno-occipital synchondrosis, a linear hypointense cleft seen on MR in children (in most subjects it fuses at the age of 8). On axial MR images obtained at the floor of the pos-

Fig. 3.9.6  Key reference for anatomy of the posterior skull base: sagittal plane. Contrast enhanced VIBE in the sagittal plane. The clivus (C) is the key structure between the nasopharynx (N), sphenoid sinus (SS), and the cisterns separating the brain stem from the bone framework. (H hypophysis, A anterior arch of the atlas, D dens of the axis). Arrowheads indicate the foramen magnum

terior cranial fossa, the basiocciput appears as a triangular bony structure which borders the large opening of the foramen magnum (Fig. 3.9.1). On both sides of the triangle there are the occipital condyles, traversed by the hypoglossal canal, which is obliquely directed. Within the canal, the hypoglossal nerve is detected as a central hypointense linear structure surrounded by venous signal. Lateral to the canal, is the jugular fossa where the internal jugular vein is reached by the sigmoid sinus, which borders the internal surface of the mastoid. The coronal plane is critical for demonstrating the foramen lacerum, the narrow space between the petrous apex and the lateral surface of the sphenoid body. This space is a potential route for the intracranial spread of nasopharyngeal neoplasms. Posterior to this cutting plane, is a second landmark: the anterior occipital condyle (Fig. 3.9.7). The hypoglossal canal, running through the condyle, gives it a “comma/inverted comma” shape, on the right and left sides, respectively. The condyle articulates with the lateral mass of the atlas. 3.9.2.4  Key Anatomy of the Temporal Bone The temporal bone is mostly made of osseous structures. Greatly compact around the membranous labyrinth, its bony structure can be spongy within the petrous apex

Fig. 3.9.7  Key reference for anatomy of the posterior skull base: coronal plane, enhanced VIBE sequence. In the center is the complex joint formed by axis (dens, D), atlas (A) and occipital condyles (OC). The hypoglossal nerve (XII), surrounded by enhancing veins, exits the canal, and joins mixed nerves (black arrowhead) in front of the internal jugular vein. On the left side the jugular bulb (JB) is depicted. IAC (white arrowheads) and the cisternal portion of the trigeminal nerve (V) are demonstrated

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

and variably hollowed out into air cavities and pneumatized cells, such as in the mastoid portion. Five parts form the temporal bone: squamous, mastoid, tympanic, styloid, and petrous portions. The squamous portion acts as the lateral floor of the central skull base, its external surface being the roof of the temporal fossa. The mastoid fuses anteriorly with the squamous portion above. Below, it enters into the formation of the tympanic cavity and the external acoustic canal, with the tympanic portion. The styloid process projects downwards from the inferior surface just anterior to the stylomastoid foramen, where the facial nerve exits the mastoid. The petrous portion is shaped like a pyramid. The internal orifice of the carotid canal opens at its vertex. The base fuses with the mastoid and the squamous portions of the temporal bone. The side faces consist of two intracranial surfaces (anterior, posterior) divided by a ridge, and an inferior (extracranial) surface which helps in forming the carotid canal and jugular foramen. The IAC, the vestibular and cochlear aqueducts open on the posterior intracranial surface. The vestibular aqueduct runs mainly parallel to the petrous ridge, whereas the cochlear aqueduct is parallel to and directly below the IAC. The petrous portion contains the inner ear, i.e. the membranous labyrinth, set within the bony labyrinth (otic capsule), and separated from it by the perilymphatic space. The bony labyrinth consists of the cochlea, vestibule, and semicircular canals. Within these bony spaces are the structures forming the membranous labyrinth.

These consist of the cochlear duct, the vestibule contents (utricle and saccule), the membranous semicircular canals, and the endolymphatic duct (Fig. 3.9.8). The cochlea is a spiral osseous canal consisting of two and three quarter turns encircling a central bony axis (modiolus). Within the cochlea are three parallel fluidfilled canals. In the center is the cochlear duct (scala media), a spiral tube containing endolymph, incompletely surrounded by the perilymphatic canals of the scalae tympani and vestibuli. High resolution T2-weighted images allow the scala tympani to be distinguished from the scala vestibuli at the level of the basal turn of the cochlea, but do not allow identification of the smaller triangular cochlear duct (Fig. 3.9.8a). The perilymphatic cavity of the cochlea continues posteriorly with the vestibule, the largest part of the labyrinth. The membranous labyrinth structures contained within the vestibule are the utricle and saccule which are the sensory organs of static balance. Three bony semicircular canals communicate with the vestibule. Each makes about two thirds of a circle. Two of them (the superior and posterior canals) are oriented vertically at an approximately right angle to one another. The third—the lateral semicircular canal—makes an angle of approximately 30° with the horizontal plane and gives rise to a ridge in the medial wall of the tympanic cavity. The canals contain the corresponding semicircular ducts, which are part of the membranous labyrinth and are filled with endolymph. The ducts occupy one quarter of the diameter of the bony canal. The endolymphatic duct collects the endolymph from

Fig. 3.9.8  Membranous labyrinth. a The single axial partition of 3DFT-CISS enables separation of the scala timpani (st) from the scala vestibuli (sv). b The thin MIP reconstruction is useful to display the whole complex of the membranous vestibule and semicircular ducts. lsd lateral semicircular ducts, psd posterior semicircular ducts

453

454

3  Brain, Head, and Neck

the utricle and saccule, enters the vestibular aqueduct, and terminates in a dilated blind extremity, the endolymphatic sac, on the posterior surface of the petrous portion of the temporal bone, beneath the dura mater. The perilymphatic spaces of each semicircular canal are in communication with themselves and with the scala tympani and vestibuli. Perilymphatic pressure is probably regulated by means of the cochlear aqueduct. High resolution T2-weighted sequences (3DFT-CISS, FSE, TSE) are excellent to demonstrate the nerves and vessels within the IAC (Casselman 1997). On axial images obtained close to the IAC floor, the cochlear and inferior vestibular branches of the vestibulocochlear nerve (VCN) can be seen as linear hypointense structures that bifurcate prior to reaching the cochlea and vestibule, respectively (Fig. 3.9.9). In the upper part of the IAC, the facial nerve (anteriorly) and the superior branch of the VCN (posteriorly) are usually seen as parallel structures running close to the roof. The facial nerve is the only one that reaches the IAC fundus. Parasagittal MPR perpendicular to the IAC shows the four nerves running along the canal. As most key structures of the tympanic cavity consist of tiny bones (the ossicular chain) which are surrounded by air, proper imaging of the middle ear requires high resolution CT, rather than MR. Nevertheless, indirect reference structures for the middle ear on MR T2-weighted images are the lateral semicircular canal fluid signal, which marks the medial wall of the tympanic cavity, and the thin hypointense bony plate of the tegmen tympani, which marks the roof of the mastoid antrum. It should be noted that in the normal temporal bone, only the fluid signal from the membranous labyrinth, the perilymphatic spaces, and the CSF filling the IAC should be detected by T2-weighted images.

Fig. 3.9.9  Nerves in the IAC. 3DFT-CISS in the axial plane. a The plane cuts the inferior portion of the canal where the vestibulocochlear nerve divides into the cochlear (c) and inferior vestibular (ivn) branches. b In the superior portion of the canal the superior vestibular (svn) branch of VIII nerve and the facial nerve (VII) are shown. C cochlea, U utricle, lsd lateral semicircular canal

3.9.3  Lesions of the Skull Base Lesions involving the skull base may be differentiated as intrinsic, arising from its bone structures, and extrinsic, arising either from below, in the extracranial head and neck, or from above, in the intracranial compartment. 3.9.3.1  Congenital Lesions Cephalocele is the most frequent congenital lesion of the skull base. It is defined as the herniation of intracranial content through a congenital or acquired defect in the skull and dura. According to the presence or absence of brain within the herniated meninges, cephaloceles can be subdivided into meningoencephaloceles and meningoceles, respectively (Naidich 1992). An exception is the atretic cephalocele, which contains meningeal structures with arachnoid, glial, or central nervous system rests. It occurs only in the parietal and occipital regions. Congenital cephaloceles involving the skull base are classified, according to the site of the cranial defect, as either sincipital (15%), basal (10%), or occipital (75%) (Kennard 1990). Sincipital cephaloceles are located in the most anterior portion of the anterior skull base. They are visible externally. Basal cephaloceles exit through the central and/or the anterior skull base. They may be occult to clinical view, unless large enough to protrude from the nostrils or the mouth. Occipital cephaloceles refer to defects located between the foramen magnum and the lambda. Sincipital cephaloceles include the interfrontal and the frontoethmoidal types (more common). Frontoethmoidal cephaloceles develop through two defects present in the embryonic anterior skull base. In the embryo, a small fontanel (fonticulus nasofrontalis) temporarily separates the frontal from the nasal bones. In the same period, the nasal bones are still separated from the nasal capsule (from which the upper nasal cartilages and the ethmoid will develop) by the transitory prenasal space. This space extends upwards to the base of the brain. Normally, midline diverticula of the dura mater pro­ ject into both the fonticulus nasofrontalis and the prenasal space, to regress later on. The fonticulus nasofrontalis is then closed to make the nasofrontal suture. The prenasal space is closed by the frontal and the ethmoid bones, leaving a small depression in front of the crista galli, filled with fibrous tissue (the foramen cecum). Three main conditions result from faulty regression of the embryologic dural diverticula: • If the diverticula contain meninges, CSF, and neural tissue, then they form frontonasal and nasoethmoidal meningoencephaloceles. • If the developing meningoencephalocele is severely constricted by the dura and bone, and it becomes iso-

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

lated from the intracranial compartment, then it results in heterotopic foci of meninges and neural tissue (extranasal and intranasal gliomas). • If the diverticula become adherent to the ectoderm, then the ectoderm may be pulled with the diverticula when they retreat, thereby forming an ectodermal tract (nasal dermal sinuses and dermoid cysts). Frontonasal meningoencephaloceles is identified by persistent herniation of dura, CSF, and brain through both the foramen cecum and the fonticulus frontalis into the glabellar region. Nasoethmoidal meningoencephaloceles are characterized by herniation via the foramen cecum into the prenasal space and nasal cavity. MR imaging is indicated for the initial assessment of encephaloceles because it can demonstrate their extent, size, and content. Additionally, associated intracranial malformations may be detected. On MR, findings of frontoethmoidal meningoencephaloceles encompass a soft-tissue mass connected to the subarachnoid space through an enlarged foramen cecum, extending to the glabella or into the nasal cavity (Fig. 3.9.10). Brain tissue within the herniated dural protrusion usually shows SI equal to grey matter, but gliosis may account for abnormal hyperintensity on T2-weighted images. CT is indicated for demonstrating indirect signs of intracranial connection as a bifid or deformed crista galli, a widened foramen cecum, or a defect of the cribriform plate. Though, only few nasal gliomas are reported to contain neurons, the presence of heterotopic brain tissue may even show the gyral structure of grey matter on T1- and T2-weighted images. MR is essential in ruling out the communication between the intracranial CSF and any fluid space surrounding the glioma, thus excluding the diagnosis of meningoencephalocele (Fig. 3.9.11). Nonetheless, meningoencephaloceles may be difficult to

Fig. 3.9.10  Nasoethmoidal meningocele. a On a coronal enhanced T1-weighted image a fluid content nasoethmoidal poly­ poid lesion (asterisk) abuts the anterior skull base floor. It is located between the middle turbinate and the nasal septum. b A

differentiate from nasal gliomas. MR cisternography with heavily T2-weighted thin sections (3DFT-CISS) has been advocated (Lowe 2000). Among basal cephaloceles, the sphenopharyngeal type is the most frequent. The skull base defect lies within the ethmoid and/or the sphenoid bones. These cephaloceles are subcategorized as trans-ethmoidal, spheno-ethmoidal, or trans-sphenoidal according to the specific site of the ostium. Sphenopharyngeal cephaloceles are associated with hypertelorism, midline facial clefting, ocular clefting and coloboma, optic nerve dysplasia, and midline cerebral defects. They may escape detection for years, and first present themselves in childhood or even adulthood, sometimes being the unexpected cause of an “idiopathic” CSF rhinorrea (Fig. 3.9.12). In this clinical setting, an occult spontaneous defect of the tegmen tympani (form of basal cephaloceles of the temporal bone) has to be ruled out as well. Occipital cephaloceles are the most common. The osseous defect(s) may be limited to the occipital bone (superior or inferior to the external occipital protuberance) or may involve the posterior arch(es) of adjacent cervical vertebrae. Both supra- and infratentorial structures (occipital lobes, cerebellum and fourth ventricle) may protrude through the defect, together with the tentorium and major venous sinuses. Ultrasounds usually provide a feasible prenatal diagnosis of occipital cephaloceles. Thus, the goals of MR imaging, which may be antenatal as well, are partly to confirm the lesion, but primarily to detail which structures are involved, particularly the dural sinuses, and to detect possible associated anomalies of the nervous system. Because occipital cephaloceles are midline congenital anomalies, the midline sagittal T1-weighted plane is very helpful. MR angiography may be necessary to demonstrate the location and the course of dural venous sinuses. CT is better for the evaluation of the osseous defect(s).

sagittal T2-weighted image shows the meningeal stalk connecting the lesion to the subarachnoid space via a persistent defect at the embryonic prenasal space area (opposite arrows)

455

456

3  Brain, Head, and Neck

Fig. 3.9.12  Basal meningocele. The meningocele (asterisk) protrudes through a defect at the roof of a pneumatized pterygoid process (arrowheads)

3.9.3.2  Inflammatory Lesions Infectious and inflammatory lesions may involve the skull base from below: the anterior skull base from sinonasal tract processes; the central and posterior skull bases from lesions arising in the nasopharynx, parapharyngeal space, and temporal bone. Mucocele, sinonasal polyposis, fungal infection, Wegener’s granulomatosis, and sarcoidosis are among the most frequent lesions arising from the sinonasal tract. 3.9.3.2.1  Mucocele

Fig. 3.9.11  Nasal glioma. a On coronal T2-weighted image the lesion splays the septum and middle turbinate (black arrowheads). The skull base floor is continuous (white arrows). b No connection of the glioma (asterisk) with the intracranial compartment is demonstrated

Mucocele may be defined as an accumulation of products of secretion, desquamation, and inflammation within a paranasal sinus with expansion of its bony walls. The lesion is limited by a wall made by respiratory mucosa. About 90% of the lesions are located in the frontoethmoidal area, while maxillary, ethmoidal, and sphenoidal mucoceles are rarely observed. Mucocele develops as the result of sinus ostium blockage. The accumulation of mucus creates a positive pressure inside the cavity, which might explain the reabsorption of the surrounding bony walls. Sinus expansion with displacement and thinning of the bony walls is the most typical presentation of a mucocele. CT and MR findings largely depend on the composition of the entrapped material. At MR, the basic expected signal pattern of fluids may be greatly modified by progressive dehydration and a consequent increase of viscosity and protein concentra-

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

tion (Fig. 3.9.13). In detail, protein concentration within the range 20–25% will result in hyperintensity on both T2-weighted and T1-weighted images. Further increase of concentration changes the signal pattern, thus turning into hypointensity at 30% (T2-weighted) and 40% (T1-weighted) (Som 1989). Similarly to the various signal patterns observed on MR, CT density ranges from

fluid-like (cystic appearance) to progressively higher values as more of the entrapped material desiccates. Hypointense signal and high density on plain T1-weighted images can be observed also in the case of eosinophilic fungal rhinosinusitis. After contrast administration, enhancement is observed exclusively at the periphery, along the thin mucosal layer, both at CT and MR.

Fig. 3.9.13  Frontoethmoidal mucocele secondary to nasal polyposis. a,b TSE T2-weighted axial plane. The mucocele (M) remodels and displaces the frontal sinus and orbital walls. Its SI is intermediate: lower than fluid in supraorbital ethmoidal cell (asterisk) and in the adjacent frontal sinus (double asterisk). Polyps with heterogeneous SI extend into the left frontal sinus

(black arrows). The crista galli (CG).c,d TSE T2-weighted sagittal plane. The polyps fill the nasoethmoidal cavity and extend through the frontal sinus ostium (black arrows). Focal reabsorption of posterior frontal sinus wall and planum sphenoidale is shown (white arrows)

457

458

3  Brain, Head, and Neck

3.9.3.2.2  Sinonasal Polyposis Sinonasal polyposis is quite a common finding in chronic rhinosinusitis. Nasal polyps appear as edematous formations consisting of respiratory epithelium lining an edematous stroma infiltrated by inflammatory cells. Eosinophils are found in 80% of cases. Bilateral involvement of nasoethmoidal cavities is always observed. Though diffuse sinonasal polyposis may cause bone remodeling of the skull base, bone destruction with eventual intracranial extent is usually observed in complications by fungal infections (see section 3.2.3) or mucocele formation, which are more frequent at the level of the frontal and sphenoid sinuses. MR findings of non-complicated inflammatory polyposis include hyperintensity on T2-weighted images, and a combination of intense enhancement of the mucosal lining with the hypointensity of the edematous stroma on contrast-enhanced T1-weighted sequences.

Fig. 3.9.14  Sphenoid sinus fungus ball. The sphenoid sinus is filled by a non-homogeneous low SI lesion (F) which is separated from the walls by the thickened and enhancing mucosa (arrows)

3.9.3.2.3  Fungal Infections Fungal rhinosinusitis can be defined as an infection of paranasal sinuses in which fungi play the role of primary pathogens or cause an inflammation due to their presence. According to the presence or lack of sinonasal mucosa invasion, fungal rhinosinusitis is classified into: • Non-invasive form: fungus ball and eosinophilic fungal rhinosinusitis • Invasive form: acute fulminant rhinosinusitis, chronic invasive fungal rhinosinusitis, and granulomatous invasive fungal rhinosinusitis In non-invasive fungal rhinosinusitis, skull base involvement is mainly due to the mass-like growth pattern of fungal debris and mucus within a sinus (fungus ball, usually an isolated lesion) or to the mechanical pressure exerted by diffuse accumulation of mucine (esinophilic fungal rhinosinusitis, frequently associated with sinonasal polyposis). CT and MR findings in non-invasive forms depend on the high content of calcium, iron and manganese within fungal hyphae. On CT, spontaneous hyperdensity and scattered calcifications may be observed. Both iron and manganese cause shortening of T1 and T2. Therefore, on MR, both fungus ball and eosinophilic fungal rhinosinusitis will appear as hypointense/signal-void lesions filling the nasosinusal cavity (Fig. 3.9.14). They are bordered by the thickened non-invaded mucosa, which has high SI on T2-weighted images and enhances on post-contrast T1-weighted images. Expansion of sinusal walls and bone thinning are more commonly observed in eosinophilic fungal rhinosinusitis, whereas sclerosis of sinusal walls is more typical in a fungus ball. A more aggressive pattern of tissue involvement characterizes the invasive form. The capability of entering the

vascular bed is typical of the acute fulminant rhinosinusitis. The infection generally occurs in immunocompromised patients. Early extent into the orbit and invasion of the skull base are frequently observed with rapid and fatal course. On MR, due to the vascular invasion, the necrotic mucosa does not enhance. It may show variable signal intensity on T2-weighted images, probably reflecting different stages of tissue ischemia. Frequently, the infected, devascularized tissue extends beyond the sinusal walls with involvement of the dura, dural sinuses, and the brain. In chronic invasive fungal rhinosinusitis, tissues are progressively and slowly infiltrated without vessel invasion. Patients may or may not be immunocompromised. MR features do not significantly differ from the acute fulminant form: intracranial and intraorbital infiltration is almost always observed, usually with inhomogeneous enhancing tissue invading the orbital apex or the cavernous sinus (Howells 2001). Granulomatous invasive fungal rhinosinusitis, which occurs in immunocompetent patients with a geographical predominance in Sudan, is a slowly progressive chronic infection that extends beyond the boundaries of the sinus involved. 3.9.3.2.4  Osteomyelitis Osteomyelitis of the central skull base is an aggressive infectious process related either to bacterial or fungal infections. It may mimic a diffuse neoplastic process, like a non-Hodgkin’s lymphoma. It often presents with subtle nonspecific symptoms, such as persistent headaches, and it eventually leads to cranial nerves deficits. Underlying

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

conditions causing immune system depression such as diabetes mellitus, corticosteroid use, HIV infection, or chronic inflammatory sphenoid sinus disease are predispositions to the infection. In most patients, the initial lesion is an otitis externa, which rapidly spreads with extensive involvement of the bone marrow of the mastoid and petrous parts of the temporal bone and of the adjacent soft tissues of the parapharyngeal and masticator spaces. Alternative infection sources are the sphenoid sinus and, less frequently, the nasopharynx. Whether due to Pseudomonas, Aspergillus or other infectious agents, the assessment of osteomyelitis is best accomplished with MR, as it is more sensitive than CT in demonstrating “early” findings of osteomyelitis. These include diffuse decrease of SI of bone marrow and fat-tissues on T1-weighted and high SI on T2-weighted images (Fig. 3.9.15). Particularly, on T1-weighted images the diffuse hypointensity of the abnormal fat (within the bone marrow or in the parapharyngeal and masticator spaces) makes it difficult to clearly separate muscles from bones and fat-spaces. This diffuse inflammatory soft tissue greatly enhances after contrast agent administration. Post-contrast changes are highlighted by the use of fat saturated SE or GE T1-weighted sequences. Intracranial extent on the infection leads to cavernous sinus invasion, encasement of the cavernous or petrous internal carotid artery, and Meckel’s cave involvement. Vessel narrowing can be demonstrated precisely by MR angiography. Diffusion-weighted MR has been shown to correctly identify abscesses associated with skull base osteomyelitis.

Wegener’s granulomatosis is a chronic, granulomatous necrotizing vasculitis affecting the upper and lower respiratory tract and the kidneys.

At imaging, sinonasal mucosal changes in the early stage of the disease are non-specific and very similar to chronic inflammatory changes. Only in the late stage of the disease, signal intensity of mucosa and submucosa switches to hypointensity on both T2-weighted and T1-weighted sequences, with variable degrees of contrast enhancement. This is mostly due to submucosal granuloma formation. In advanced stages of the disease, inflammatory infiltrate and granulomatous lesions within small vessel walls lead to obliteration of the lumen and to avascular necrobiosis. This is the pathologic basis of bone destruction, often involving midline structures like the nasal septum. A similar pattern of bone destruction can be observed in advanced cocaine abusers. Unfortunately, Wegener’s granulomatosis and cocaine abuse share overlapping histopathological features and ANCA testing may give positive results also in cocaine induced midline destructive lesions. Imaging may be useful for the differential diagnosis. In cocaine abusers, not only the septum but also the adjacent turbinates may be destroyed, in a sort of centri­ fugal pattern. In addition, unlike Wegener’s granulomatosis, cocaine abuse may lead to the destruction of hard and soft palate, usually in late stages (Trimarchi 2001). Central skull base involvement by Wegener’s granulomatosis may be due to spread from contiguous deep spaces of the face (Fig. 3.9.16). MR is decisive in identifying the cause of nerve impairment, as it can: • Show direct granuloma extension into fissures or foramina of the skull base, such as the pterygopalatine fossa, the orbital fissure, or the vidian canal. On MR, the granulomatous lesions show hypointense signal on both T2-weighted and plain T1-weighted sequences. Contrast enhancement is usually observed and ranges from mild and inhomogeneous to hyperintense.

Fig. 3.9.15  Skull base osteomyelitis. Previous external otitis on left side. a The pre-contrast T1-weighted axial image demonstrates a mass-like lesion (asterisk) extending into the retrostyloid compartment and encroaching the pharyngobasilar fascia with nasopharyngeal wall involvement. Bone marrow in the

basiocciput shows decreased SI, sclerotic changes (arrows) and erosion. b Both the inflammatory soft tissue and the abnormal bone marrow content show inhomogeneous enhancement. Destructive changes in the left mastoid (arrows) and involvement of left glenoid fossa (double asterisk) are present

3.9.3.2.5  Wegener’s Granulomatosis

459

460

3  Brain, Head, and Neck

• Show perineural granulomatous spread along trigeminal/parasympathetic nerve branches. MR findings include asymmetrical nerve thickening, enlargement, and late destruction of the related foramina and fissures. • Aid in central nervous system localization of the disease. However, involvement of deep spaces of the face and/or perineural spread is indistinguishable from malignant neoplasms, like non-Hodgkin lymphoma. Otologic manifestations of Wegener’s granulomatosis are related to the involvement of the Eustachian tube or of the middle ear. Active granulomatous tissue within the middle ear exhibits enhancement, unlike the retained fluid observed in serous otitis media (Maroldi 2001). Remote granulomatous lesions may rarely affect the brain parenchyma or dura. Plaque thickening and marked contrast enhancement of the dura (without pial involvement) are reported. Similar findings are detected in a list of different conditions including sarcoidosis, primary and secondary dural tumors, infectious diseases, and hypertrophic pachymeningitis. 3.9.3.3  Benign Neoplasms Most frequent benign neoplasms involving the skull base can be subdivided into three groups according to their origin: • Neoplasms arising from sinonasal structures (i.e., below the skull base) or from structures crossing the skull base (i.e., nerves). • Neoplasms arising from components of the skull base itself (i.e., intrinsic lesions developing from bone, cartilage, fibrous tissue, residual embryonic tissue). • Neoplasms developing from tissues covering the intra­ cranial surface of the skull base (i.e., meninges).

Fig. 3.9.16  Wegener’s granulomatosis. a Hypointense soft tissue lesions border the basiocciput, bilaterally (asterisks) on TSE T2-weighted image. Effacement of fat planes is present. b The lesions enhance remarkably on a T1-weighted sequence. Submu-

3.9.3.3.1  Benign Neoplasms Arising from below or from Structures Crossing the Skull Base As a large variety of tissues constitute the sinonasal tract, various histotypes may be detected among the neoplasms arising from this anatomic district. The anterior or central skull bases are most frequently accessed from the sinonasal tract by the inverted papilloma, the juvenile angiofibroma, and the ossifying fibroma. Foramina and fissures of the skull base are conduits through which schwannomas (or neurofibromas) may extend into (or exit) the intracranial compartment. Inverted Papilloma Inverted papilloma is an epithelial benign neoplasm originating from the schneiderian membrane, the ectodermally derived mucosa that lines the nasal cavity and paranasal sinuses. It is characterized by the alternation of quite regular parallel folds made of a highly cellular metaplastic epithelium and of an underlying less cellular stroma. The imaging features of the inverted papilloma are based on the site of origin, the changes of the lateral nasal wall, and—respectively—the lobulated surface contour on CT and the striated inner pattern on MR. While CT can only identify a lobulated contour (i.e., air surrounding the fan-shaped folds), on MR the inner macroscopic arrangement is demonstrated as a striated pattern or a convoluted cerebriform pattern both on T2-weighted and contrast-enhanced T1-weighted images (Ojiri 2000) (Fig. 3.9.17). On T2-weighted images the parallel folds appear as thick striations of high SI alternated with thinner ones. The thinner striations have been correlated with the metaplastic epithelium, while the thicker ones have been considered to correspond to the less cellular edematous stroma. On contrast-enhanced T1-weighted images, the stroma enhances strongly while

cosal involvement of the prestyloid space medially to tensor veli palatini is seen (arrows). The granulomas encircle the internal carotid artery (arrowheads)

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

the thinner epithelium enhances less (Maroldi 2004). High spatial resolution techniques improve detection of the pattern. When assessing the extent of inverted papilloma, two main tasks must be performed: • Distinguishing the neoplasm from intrasinusal-retained secretions. • Detailing the relationship with both the orbit and the skull base. MR can easily separate tumor signal (see above) from retained secretion (SI varies depending on the protein concentration within the fluid). If the inverted papilloma accesses the ethmoidal labyrinth, CT is superior in demonstrating the destruction of the thin cells. Involvement of the skull base is usually late, being observed after several failed surgical excisions. The sphenoid and the frontal sinus are critical sites where the lesion may extend. Extensive erosion of the lamina papyracea or of the cribriform plate is more often associated with a coexistent squamous cell carcinoma.

A second pattern of growth is the encroachment through the inferior and the superior orbital fissures. Erosion of the anterior skull base is rare. Very rare observations of juvenile angiofibromas trespassing the dura have been documented (Danesi 2000). CT and MR features include: • The site of origin of the lesion. In small angiofibromas this area is limited to a tiny triangle encompassing the sphenopalatine foramen, the pterygopalatine fossa, and the vidian canal. • Hypervascular appearance after contrast enhancement. MR sequences may show large vessels within the mass (flow-void, serpiginous structures) or enlargement of arterial feeders (the internal maxillary artery). Strong enhancement is observed post–contrast agent administration (Fig. 3.9.18). • Pattern of growth. Juvenile angiofibroma tends to grow along the paths of least resistance, displacing adjacent soft tissues rather than invading them. Adherence can be present, particularly when the lesion contacts the dura, but dural or brain infiltration is rare.

Juvenile Angiofibroma Juvenile angiofibroma is a lesion composed of vascular and fibrous elements, which typically occurs in adolescent males. The point of origin is identified in the area of the sphenopalatine foramen, or in the pterygopalatine fossa, at the aperture of the vidian canal. Juvenile angiofibroma has the peculiar tendency to grow submucosally into the nasopharynx, along the vidian canal into the basisphenoid, and laterally towards the pterygopalatine and infratemporal fossa. Early invasion of the spongy bone of the pterygoid root is frequently present. From there, the angiofibroma can subsequently extend laterally to involve the greater wing of the sphenoid bone. Intracranial extension mainly occurs into the central skull base along the maxillary nerve into the parasellar region and the cavernous sinus.

On imaging, a dual pattern of bone involvement (remodeling and destruction) can be observed in juvenile angiofibromas. When the angiofibroma contacts and displaces the cortical rim of an adjacent bone structure, remodeling and thinning usually result. Though extensive demineralization eventually occurs, the neoplasm may still be confined until the periosteal covering is preserved. Further growth of the angiofibroma may lead to bone destruction. This final step requires the angiofibroma to break through the periosteal covering, which acts as a protection against invasion. An alternative pattern of bone involvement consists of the direct growth of juvenile angiofibroma along perforating arteries into the cancellous sphenoid root. From this starting point, further spread into the medullary content of the adjacent greater sphenoid wing or floor of the sphenoid sinus may occur.

Fig. 3.9.17  Inverted papilloma. a The neoplasm arises from the anterior ethmoid. The floor of the anterior skull base is shifted upwards (arrows). b On the sagittal plane both the “convoluted cerebriform” and the “columnar” patterns are shown by tumor

bulk and posterior projections, respectively (arrows). c Tumor extent into the frontal sinus is key information for treatment planning

461

462

3  Brain, Head, and Neck

On CT, strong enhancement within the diploe suggests invasion by juvenile angiofibroma. On MR, the normal fatty signal of the cancellous bone is replaced

Fig. 3.9.18  Juvenile angiofibroma. a Sagittal contrast-enhanced T1-weighted image. The lesion presents remarkable enhancement. The angiofibroma fills the sphenoid sinus, limited by the roof. The mass projects into the nasal cavity and the nasopharynx. b On the coronal enhanced VIBE the angiofibroma extends into pterygopalatine fossa and superior orbital fissure (arrows)

by low SI on plain T1-weighted images and by high SI on post-contrast T1-weighted sequences (Fig. 3.9.19). Fat-suppressed post-contrast T1 sequences greatly improve the differentiation of the enhanced juvenile angiofibroma from the suppressed signal of the bone marrow. Although spread from the sphenoid sinus may result in extensive medullary invasion of the clivus, a break through both the periosteal intracranial lining and the dura is uncommon. Intracranial invasion occurs via tumor extent along a canal (the foramen rotundum to reach the cavernous sinus) and/or via spread through bone destruction. Generally, the second pattern occurs when huge lesions break through the inner table of the greater wing or the lateral sphenoid sinus walls. Nerve Sheath Benign Neoplasms Schwannomas arise from the nerve sheath. Among the nerves crossing the skull base, the trigeminal nerve is the most frequently involved. Rarely observed are schwannomas of cranial nerves within the cavernous sinus (III, IV, and VI) or jugular foramen (IX, X, XI) or arising from the hypoglossal nerve. The olfactory and optic nerves do not have a Schwann cell layer. Trigeminal nerve schwannomas may arise in Meckel’s cave. When large, they acquire a dumbbell appearance with an anterior component in the middle cranial fossa (Meckel’s cave and cavernous sinus), and a posterior component protruding into the posterior cranial fossa through the porus trigeminalis, which causes a waist in the tumor shape—and extending into the pre-pontine cistern (Fig. 3.9.20). From Meckel’s cave the lesion may extend along the trigeminal branches into the superior orbital fissure, the foramen rotundum, and the foramen ovale. On MR, schwannomas show high SI on T2-weighted images, low SI on T1-weighted images, and heterogeneous enhancement. A cystic component may be detected. Schwannomas of the jugular foramen are typically centered or based on an enlarged jugular foramen with sharply rounded bone borders (scalloping) having a sclerotic rim. Extra-cranial vagus nerve schwannoma presents as a fusiform mass anteriorly displacing the internal carotid artery. Its long axis parallels the nerve course. Extension into the jugular foramen is possible. Hypoglossal schwannoma usually impinges the hypoglossal canal with a dumbbell appearance and bone remodeling (Fig. 3.9.21). Neurofibromas consist of Schwann cells, perineural and fibroblast cells. Un-encapsulated, they may have a nodular or infiltrative pattern of growth (plexiform neurofibroma). In patients with neurofibromatosis I, bony orbital deformities frequently occur, and always with an optic nerve glioma or orbital plexiform neurofibroma.

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Fig. 3.9.19  Recurrent juvenile angiofibroma. a The epicenter of the lesion is at the right pterygoid process, replaced by the low SI of the angiofibroma. The mass extends into the superior orbital fissure (SOF), the parapharyngeal and masticator spaces, and shifts up the sphenoid sinus floor. b Marked enhancement

is obtained on the T1-weighted image. Several serpiginous flowvoid structures indicate enlarged tumoral vessels. The extensive intradiploic invasion of the pterygoid process is associated with thickening of the overlying dura (arrows). V2 left maxillary nerve, VN left vidian nerve and canal

3.9.3.3.2  Benign Neoplasms Arising from Intrinsic Skull Base Structures

tertrabecular spaces may produce scattered hyperintense foci on both T1- and T2-weighted sequences.

Intrinsic benign neoplasms of the skull base develop either from cartilage (chondroma), bone (osteoma, giant cell tumor), fibrous tissue (ossifying fibroma), and embryonic rests (chordoma, craniopharyngioma, epidermoid and dermoid cysts). Chondromatous neoplasms (chondroma and chondrosarcoma) will be discussed together in the section of malignant tumors. Fibrous dysplasia, which is not a tumor, is included in this section as it may present clinically with signs and symptoms suggesting an expansile lesion.

Fibrous Dysplasia Fibrous dysplasia is a non-neoplastic, slowly progressing disorder, characterized by reabsorption of normal bone, replaced by fibrous tissue and immature woven bone. Cortex is unaffected by this process. Skull base may be involved either focally or diffusely. Imaging appearance depends on the quantity of fibrous tissue in the immature woven bone. Therefore, CT density may range from radiolucent to ground glass (equal proportions of fibrous and osseous tissue), or even sclerotic (predominance of dense osseous tissue). On MR T2-weighted sequences, the signal intensity is rather variable, ranging from overall hypointense signal with cystic/necrotic hyperintense areas, to hyper- or isointense signal as compared to subcutaneous fat tissue. The lesion is hypointense on T1-weighted images; nonhomogeneous enhancement after contrast administration is observed (Fig. 3.9.22). As the lesion progressively enlarges, remodeling of the affected bone can occur. This may result in encroachment of the optic canal and skull base foramina and fissures, entailing the risk of potentially severe neurologic complications.

Osteoma Osteoma is a benign, slow-growing osteoblastic tumor. Though it is the most frequent benign tumor of the nose and paranasal sinuses, involvement of the anterior cranial skull base is rare. Intracranial and orbital extension of these tumors is mainly associated with frontal sinus localization. Better demonstrated on CT, osteoma presents as a well defined mass with bone density confined to the sinus of origin where the lesion conforms to the available space. Large lesions may extend into the anterior skull base. Osteomas may exhibit variable degrees of density, according to the amount of mineralized bone within the lesion. No enhancement is to be expected. MR is intrinsically inferior to CT in imaging osteomas, as both the low water content and the lack of mobile protons within osteomas result in very low signal intensity in all sequences. Small amounts of fat tissue within the in-

Ossifying Fibroma Ossifying fibroma is a slow-growing benign neoplasm composed of fibroblastic and osseous components, which

463

464

3  Brain, Head, and Neck

is considered to arise from mesenchymal blast cells giving origin to the periodontal ligament. It appears as an expansile, well-defined lesion displacing, remodeling, and eventually destroying the adjacent sinusal or orbital walls, or the anterior skull base floor. A multiloculated lesion, bordered by a peripheral eggshell-like dense rim is the CT imaging pattern. At MR, the ossifying fibroma shows high SI on T2-weighted sequences. More heterogeneous, the signal pattern on plain T1-weighted images is characterized by intermediate-tohigh SI in the central part, combined with hypointensity of the outer shell. Strong enhancement after paramagnetic contrast administration is usual (Fig. 3.9.23).

Fig. 3.9.20a,b  Trigeminal nerve schwannoma. On TSE T2weighted image a large trigeminal schwannoma appears as a dumbbell lesion, larger at the central skull base. A waist is seen at the porus trigeminalis (arrows) on both the axial and sagittal planes

Chordoma Chordoma takes origin from embryonic remnants of the primitive notochord. The base of the skull is the second most common location, after the sacrococcygeal region. Chordomas develop from the bone, so they initially grow extradurally with bone destruction and secondary extension into the adjacent soft tissues. They present some of the typical features of a malignant tumor, such as local invasiveness, tendency to recur, and a potential for developing distant metastases in up to 40% of patients. The site of origin of most skull base chordomas is the basiocciput-basisphenoid, where the terminal portion of the notochord ends, reaching the sphenoid bone just inferior to the sella turcica and dorsum sellae. Nasopharyngeal and intracranial locations are rarely observed. Their origin is explained by the extra-osseous path of the notochord, which may have short segments running outside the bone, either within nasopharyngeal soft tissues or within the posterior cranial fossa. On CT, chordoma appears as a midline clival soft tissue mass with bone destruction. No sclerotic changes are detected at the boundary with the invaded bone. Cystic components are frequently observed. Enhancement is present in at least some parts of the soft tissue component. Intracranial extent frequently leads to posterior displacement of the basilar artery and mass effect on the brain stem. In up to 80% of lesions, MR shows heterogeneous high SI on T2-weighted images (Meyers 1992) with low SI areas, indicating mucoid or old hemorrhage, respectively. The low SI of chordoma on sagittal plain T1-weighted sequences helps to demonstrate tumor replacement of the hyperintense clival bone marrow. Cystic areas may show high SI on plain T1-weighted sequences. A lobulated honeycomb pattern of contrast enhancement has been described as a typical feature. Large intra-tumoral calcifications account for low to void SI areas in all sequences. Chondromatous tumors and metastases are the other most frequent midline destructive lesions in the clival area to be included in the differential diagnosis.

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Craniopharyngioma Craniopharyngioma is thought to arise in the adenohypophysis from residual ectodermal clusters of the primitive craniopharyngeal duct or from metaplasia of residual squamous epithelium in the adenohypophysis. According to Giangaspero et al. (1984), the adamantinous craniopharyngioma (childhood) arises from embryonic remnants, whereas the squamous papillary (adulthood) develops from metaplastic foci of adenohypophysis cells. Since craniopharyngioma arises from the pituitary

stalk axis, it is usually located on the midline, more frequently within an area extending from the infundibular recess of the third ventricle to the sphenoid body. It is an epithelial neoplasm with solid and cystic components and calcifications. It shows a tendency to infiltrate and to produce a glial reaction which causes strong adherences. The most typical features on CT include a cyst with nodular or rim calcifications (90% of cases) and some solid enhancing components. Though variable, the den-

Fig. 3.9.21  Hypoglossal nerve schwannoma. a The schwannoma is characterized by non-homogeneous high SI on a TSE T2-weighted image. Anterior displacement of the internal carotid artery (arrow) and lateral displacement of the styloid muscles (arrowheads) indicate a post-styloid neoplasm. Sharp and regular margins suggest a benign tumor. b The schwannoma (T)

enhances on the post-contrast VIBE. The lesion extends into the jugular fossa where it compresses the internal jugular vein (arrowheads). Medially, there is erosion of the right condyle with replacement of the bone (black arrows). Retrograde spread into the hypoglossal canal is demonstrated (white arrows). XII left hypoglossal nerve, SS sigmoid sinus

Fig. 3.9.22  Fibrous dysplasia. a The lesion involves the body and right greater wing of the sphenoid, appearing deformed and enlarged (arrows). On the plain T1-weighted image, homogeneous low SI is detected within the abnormal bone. b Minimal to mild heterogeneous enhancement is observed after contrast agent administration

465

466

3  Brain, Head, and Neck

Fig. 3.9.23  Ossifying fibroma. Post-contrast VIBE sequences. a The neoplasm appears as a large nodular and enhancing naso­ ethmoidal mass. Though markedly displaced, the hypointense lining of the medial orbital wall separates tumor from intra­

orbital content (black arrows). Left anterior skull base floor is displaced upwards (white arrows). Retained secretion within maxillary sinus (asterisk). b Part of the ossifying fibroma pro­ jects into the frontal sinus (arrows)

sity of the cystic part is generally greater than the density of CSF. On MR, the adamantinous craniopharyngioma presents as a heterogeneous suprasellar-sellar mass containing a well-defined cystic component, with internal homogeneous high SI on both T1- and T2-weigheted sequences. However, a variety of different MR patterns may be present, including solid, calcified, CSF-like, hematin-like, and protein-like signals. A solid component is also invariably present, often partially calcified. Very frequently, more than one pattern may be found in the same lesion. Craniopharyngioma often elicits a relevant inflammatory gliosis reaction from adjacent tissues. Infrasellar extent is rare. Papillary craniopharyngioma is more frequently located within the floor of third ventricle. It appears more solid. Calcifications are usually absent. Cystic components are less relevant than in the adamantinous type, and do not show high SI on T1-weighted images. Postcontrast studies demonstrate a partially cystic mass that enhances peripherally. The mass is usually encapsulated and readily separable from adjacent structures. The imaging features useful to differentiate the two tumor types are: • Encasement of vessels, lobulated shape, and presence of hyperintense cysts in the adamantinous type. • Rounder shape, presence of hypointense cysts, and predominantly solid appearance in the squamouspapillary type (Sartoretti-Schefer 1997).

cellular debris are denser on CT and show heterogenous signal on MR. Unlike craniopharyngioma, Rathke’s cleft cyst does not enhance except for a thin, peripheral enhancement of its wall.

The differential diagnosis includes the Rathke’s pouch, or cleft cyst, which shares the same common embryologic origin of craniopharyngioma, but is lined by a single layer of epithelial cells, filled with mucoid fluid or cellular debris. Mucoid content accounts for the hyperintensity on both T1 and T2 sequences, making the distinction with craniopharyngioma difficult. The cysts containing

3.9.3.3.3  Benign Neoplasms Arising from Above Meningiomas are the most frequent benign neoplasms arising from the intracranial compartment and extending into or through the skull base. Pituitary (macro)adenomas may also encroach the central skull base. Pituitary adenomas are discussed in Sect. 3.7 Meningioma takes origin from arachnoidal cap cells of the meninges. Only a third of intracranial meningiomas are found along the skull base. Recent improvements in the surgical treatment of skull base meningiomas are parallel to the increasing detailed assessment provided by imaging techniques. Actually, imaging is essential in detailing the size, site of origin, and extent, the involvement of the carotid artery (and its branches) or of the basilar artery, the relationship with the brain stem and the cranial nerves. The selection of a particular approach or combination is based on a careful consideration of these factors, as well as on the neurological status of the patient. According to the site of origin, meningiomas are classified as: • Anterior skull base meningiomas. Most of these arise in the olfactory groove. • Central skull base meningiomas. These can be found in almost any portion of the sphenoid, including the greater wing, the planum sphenoidale, the tuberculum sella, the anterior clinoid, and the cavernous sinus walls. Imaging enables stratification of the risk related to surgical treatment. In fact, if the sphenoid wing is divided in thirds, meningiomas arising from the medial third tend to involve cranial nerves early and en-

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

tail a higher surgical risk compared to those developing from the external third. • Petroclival meningiomas. These are among the most challenging for surgical treatment. They can be separated according to their site of origin from the upper, middle, or lower third of the clivus. Each subgroup requires a specific surgical approach to maximize tumor resection and minimize the extent of brain retraction. • Petrosal meningiomas. These encompass posterior fossa neoplasms arising lateral to the trigeminal nerve. They are classified into anterior and posterior petrosal meningiomas (see temporal bone). Most skull base meningiomas belong to the en plaque subtype, which is characterized by hyperostosis disproportionate to the volume of the intracranial tumor (Fig. 3.9.24). Thus, the tumor appears as a thin layer of tissue investing an abnormal inner table of the skull. Skull base meningiomas can potentially extend towards the sinonasal tract and the nasopharynx through the bone or growing along the arachnoidal cells that invest the cranial nerves (Fig. 3.9.25). CT findings consist of a plaque-like enhancing mass along with hyperostosis of the adjacent bone. On MR T1-weighted images, meningiomas are commonly isointense to grey matter. On T2-weighted imaging they are more frequently heterogeneous, due to the presence of cysts and calcifications. Relevant enhancement is invariably obtained after contrast agent administration. The enhancement is related to the absence of blood barrier in meningioma capillaries. Frequently, thickening and enhancement of the dura at the margins of the tumor may be observed (dural tail sign). This finding can be due either to direct dural infiltration with accompanying dural congestion or to non-specific inflammatory changes (Kawahara 2001). Meningioma may involve the paranasal sinuses either by means of skull base encroachment or, rarely, as an ectopic lesion. When the meningioma arises from the planum sphenoidale, the bone may show a characteristic blistering, consisting of upward expansion. An associated pneumosinus dilatans has been described. 3.9.3.4  Malignant Neoplasms The first step in the diagnostic work-up of malignant sinonasal and nasopharyngeal neoplasms consists of fiber optic examination. Endoscopy allows adequate demonstration of the superficial spread of the lesion and may guide a biopsy. The discrimination between benign and malignant tumors and the detailed characterization of the lesion are, in most cases, far beyond the capabilities of CT or MR. Therefore, the main goal of imaging is to provide a

Fig. 3.9.24  Meningioma of the greater wing of the sphenoid bone. a The meningioma grows through the bone causing marked hyperostosis and enlargement of the left greater wing. b A very thin layer of enhancing lesion invests the intracranial surface of the sphenoid wing (arrows)

precise map of the deep tumor extension in all the areas blinded at fiber optic examination, especially the skull base, the orbit, and the pterygopalatine fossa. In this setting, MR is the technique of choice because it clearly differentiates tumor from retained secretions, and it allows early detection of perivascular/perineural spread. On the other hand, the advantages of CT include superior definition of bone structures even in the case of subtle erosions, faster and easier examination, greater accessibility, and lower cost. Despite the relevant improvements provided by multislice technology (i.e., fast coverage of the volume of interest, thin collimation, and acquisition of nearly isotropic voxels) CT indications are nowadays restricted to patients not previously examined by the otolaryngologist (to rule out non-neoplastic lesions) or to patients with contraindications to MR. The most frequent malignant neoplasms involving the skull base can be subdivided in three groups according to their origins: • Neoplasms arising from below the skull base with direct encroachment of the bony-periosteal barrier (sinonasal, nasopharyngeal neoplasms), or with intracranial extent via perineural spread.

467

468

3  Brain, Head, and Neck

Fig. 3.9.25  Recurrent meningioma of the greater wing of the sphenoid bone. Patient operated on 5 years before. TSE T2weighted sequence in the coronal plane. a The meningioma invades the left Meckel’s cave and abuts the lateral wall of the sphenoid sinus (black arrows). Hyperostosis of the pterygoid process (white asterisk) is seen. Part of the extracranial component displaces downwards the external pterygoid muscle (white arrows).

Post-surgical changes are present in the left temporal lobe (black asterisk). b The tumor (T) extends into the masticator space via enlargement and hyperostosis of the foramen ovale (white opposite arrows). Extensive invasion of Meckel’s cave appears as replacement of the CSF SI by the low SI of meningioma (black arrows). Hyperostosis of the left basisphenoid (arrowheads)

• Neoplasms arising from structures of the skull base itself (i.e., primary lesions developing from bone, cartilage, fibrous tissue, residual embryonic tissue) or metastases. • Neoplasms developing from tissues investing the intracranial surface of the skull base (i.e., meninges).

wood and shoe-leather workers—prevail in the ethmoid and nasal cavity; squamous cell carcinomas in the maxillary sinus; adenoid cystic carcinomas in the hard and soft palate; undifferentiated carcinomas and squamous cell carcinomas in the nasopharynx. Non-Hodgkin’s lymphoma may involve the sinonasal tract, the nasopharynx, or the deep spaces of the face. Skull base invasion is due to neoplastic growth along cranial nerves and through permeative or destructive bone invasion. Olfactory neuroblastoma is a neuroectodermal malignant tumor arising from olfactory epithelium covering the upper portion of the ethmoidal labyrinth. MR signal features are non specific. Among the factors suggesting the diagnosis are the site of origin and the tendency to early invade the anterior skull base or the orbit. Although the diagnostic imaging of sinonasal diseases has evolved with the improvement of imaging technology, the characterization of neoplasms remains a challenge for imaging techniques. On T2-weighted images, MR usually succeeds in differentiating retained secretions—which have a high component of water and thus high SI—from malignant neoplasms, which have a predominant cellular component and consequently a solid pattern with low SI. Epithelial neoplasms containing a relevant component of intra/extra-cellular mucin, such as intestinal type adenocarcinomas, tend to have intermediate to bright SI on T2-weighted images, (Fig. 3.9.26).

3.9.3.4.1  Skull Base Invasion by Extra-Cranial Malignant Neoplasms Intracranial invasion by extracranial neoplasms is most frequently related to sinonasal tract or nasopharyngeal malignant neoplasms. The main routes of spread are a direct infiltration of the periosteum-bone-dura composite layer, and perineural and perivascular spread along structures traveling through foramina and fissures of the skull base. While most ethmoidal neoplasms invade the anterior skull base via bone erosion and (eventual) dura infiltration, maxillary sinus neoplasms and tumors of the palate more commonly reach the central skull base by means of perineural spread. Both bone invasion (foramen lacerum, clivus) and perineural spread (mandibular branch of the trigeminal nerve) are shown by nasopharyngeal neoplasms. The frequency with which histotypes occur differs according to the site of tumor origin. For example, adenocarcinomas—including the intestinal type found in hard-

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Fig. 3.9.26  Nasoethmoidal adenocarcinoma, intestinal type (ITAC). A large tumor with high SI on TSE T2-weighted coronal plane fills the nasoethmoidal cavities. The neoplasm abuts the cribriform plate, which is intact (arrows). The mass displaces laterally the right maxillary sinus wall (arrowheads). Mucous retention within maxillary and frontal sinuses (asterisks) is shown

Fig. 3.9.27  Recurrent nasoethmoidal adenocarcinoma. a TSE T2-weighted coronal plane. Tumor remodels the left sphenoid sinus roof where focal thinning of the bone is detected (arrows). Irregular blurring of the lamina left papyracea suggest focal invasion (arrowheads). b After contrast agent administration,

Precise assessment of the deep extension of sinonasal tumors towards the dural layer covering the anterior, middle, and posterior floor of the skull base significantly influences the treatment planning. It is important to note that extra-dural invasion of the anterior cranial fossa floor may be successfully managed by surgery (by means of craniofacial resection), whereas malignant neoplastic spread through the clivus is a challenging factor for radiation therapy. On MR, assessment of the relationship between tumor and the adjacent skull base requires detection and analysis of specific changes in the composite appearance of the bone signal, of the investing dura and of the overlying subarachnoid space. On enhanced sagittal and coronal T1-weighted or 3D GE fat-suppressed T1-weighted sequences the three layers (skull base floor with its double periosteal covering, dura mater, subarachnoid space) give rise to a “sandwich” of different signals (bone–periosteal complex, dura mater, CSF). If a malignant sinonasal neoplasm impinges the anterior/central skull base, three different conditions may occur: • The neoplasm contacts an uninterrupted, low SI cribriform plate or fovea ethmoidalis (Fig. 3.9.26). • The neoplasm erases the low SI of the cribriform plate, extends through the anterior skull base, and displaces an uninterrupted, enhanced, and thickened dura (Fig. 3.9.27).

tumor enhances. The focal erosion on the TSE T2-image corresponds to a focal bulging of the sphenoid roof. A residual low SI layer (arrows) separates the tumor (below) from the thickened dura (above)

469

470

3  Brain, Head, and Neck

• The neoplasm encroaches the enhanced dura, without erasing the hypointense signal of CSF (Fig. 3.9.28). • The neoplasm extends beyond the dura; it encroaches on the low SI of CSF and invades brain tissue. This last sign is easier to detect when the signal intensity of the neoplasm is lower than the enhanced dura surrounding the invaded bone segment (Fig. 3.9.29). Resectability of tumors invading the brain does not stand only upon the assessment by imaging techniques of the depth of tumor extent into the brain or on the detection of bilateral brain invasion. It requires a thorough evaluation of several other issues, the most important being the histotype and the patient’s performance status. Contraindications to surgery, other than brain invasion, are considered the involvement of the internal carotid artery or of the cavernous sinus. Though CT is superior in demonstrating bone erosion of the clivus and in planning radiation therapy, MR is considered the technique of choice in the assessment of the local extent of nasopharyngeal carcinomas because of a more precise evaluation of the parapharyngeal space, retrolateropharyngeal nodes, and perineural spread. In ad-

Fig. 3.9.28  Nasoethmoidal squamous cell carcinoma. The tumor (T) invades the planum sphenoidale through complete erosion of bone (arrows). The thickened dura has a higher SI than the enhanced tumor. The dura is displaced upwards but it still separates the tumor from the subarachnoid space. Retained secretion within frontal sinus (asterisk)

dition, intracranial invasion, which occurs without bone destruction in about 65% of cases—via foramen ovale and lacerum, is better demonstrated by MR (Figs. 3.9.30, 3.9.31). On T2-weighted and plain T1-weighted sequences in the coronal planes, particular attention has to be paid to detecting even focal changes of the low SI signal of the cortical rim lining the lateral boundaries of the clivus. Besides cortical erosion, bone invasion may be demonstrated by the replacement of bone marrow fat signal by tumor, which usually shows intermediate-to-high SI on T2-weighted, low SI on T1-weighted, and remarkable enhancement on fat-sat T1-weighted sequences. Intracranial Perineural Spread Perineural spread is defined as neoplastic extension along the peripheral nerve sheath via endoneurium, perineurium, or perineural lymphatics. Though most frequently centripetal, i.e., towards the skull base foramina, perineural spread may run in the opposite direction (centri­ fugal). Though adenoid cystic carcinoma is frequently associated with perineural spread, other malignant neoplasms of the head and neck may show this pattern of growth.

Fig. 3.9.29  Nasoethmoidal undifferentiated carcinoma. Permeative destruction (black arrows) of the anterior skull base floor by the nasoethmoidal mass (T) is shown. Intracranial invasion is associated with focal effacement (black arrowheads) of the thickened dura (white arrows) and brain edema (asterisk), which suggests brain invasion. Contemporary extent into the frontal sinus (white arrowheads) is present

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

Fig. 3.9.30  Nasopharyngeal adenoid cystic carcinoma. Both the TSE T2-weighted and the post-contrast T1-weighted coronal planes demonstrate the invasion of the left pterygoid process (arrows) by the nasopharyngeal neoplasms (T). The cortical rim of the sphenoid bone is not invaded (arrowheads)

Among them are squamous cell carcinomas arising from either the skin or the mucosal epithelium, desmoplastic melanoma of the skin, and virtually any salivary gland carcinoma, and lymphomas. A key issue to improve perineural spread detection with imaging consists in selecting technical parameters that maximize both spatial and contrast resolution. On MR, improved contrast resolution is strongly recommended, particularly by means of fat-saturated T1-weighted sequences after contrast agent administration. In our experience, 3D VIBE sequences provide an excellent solution by obtaining high resolution fat-saturated images in an acceptable study time (Fig. 3.9.32). On this sequence, the normal nerve is hypointense, clearly detectable where it is surrounded by an enhanced venous plexus, for example along bony grooves and canals—like the maxillary and mandibular nerves through the respective foramina, or the hypoglossus nerve at the condylar canal (Maroldi 2005). The purpose of high resolution MR imaging is to demonstrate even subtle signal changes of the nerve itself and/or to detect asymmetrical thickening of the enhanced signal surrounding the nerve. Perineural spread is characterized by progressive accumulation of neoplastic cells around a nerve with a segmental increased diameter. Destruction of the blood–nerve barrier accounts for extravasation of contrast agent, causing asymmetrical nerve enhancement. As the nerve enlarges, foramina/fissures are remodeled, widened, and, eventually, eroded. CT findings of perineural spread include widening/erosion of foramina/ canals, and asymmetrical enhancement within the same bony conduits. Extension of perineural spread into the central skull base leads to the replacement of the fluid signal of Meck-

el’s cave by solid and enhancing tissue, and causes an increased convexity of the lateral boundary of the cavernous sinus (Fig. 3.9.33). Atrophy of masticator, tongue, or oral floor muscles should be considered as an indirect sign: in such cases, T1-weighted images show fatty degeneration (high SI) of denervated muscles. The thickness of the neoplastic layer “coating” the nerve may rapidly switch from millimeters to micro­ meters where the nerve enters a foramen or runs along a bony canal. This transition from a macroscopic scale nerve abnormality to a microscopic scale accounts for false-negative MR results (skip lesions). At the exit of

Fig. 3.9.31  Maxillary sinus adenoid cystic carcinoma invading the clivus. Extensive permeative bone marrow invasion presents as diffuse and homogeneous low SI of the posterior nasal septum (black arrowheads), and of the diploic bone within the clivus. The tumor erodes the posterior cortical rim of the clivus (short white arrows), and invades the meningeal layer on its intracranial surface (white arrowhead). The normal signal of the meninges is seen at the basisphenoid area (long white arrow)

471

472

3  Brain, Head, and Neck

Fig. 3.9.32  Hard palate adenoid cystic carcinoma invading the clivus. On post-contrast axial VIBE images, multiple paths of perineural spread are demonstrated in this hard palate tumor. a Right greater palatine nerve invasion (black arrow), normal nerve for comparison on the opposite side (white arrow); b perineural spread along into the pterygopalatine fossa (asterisk) with invasion of the vidian nerve (white arrows); c invasion

of right V2 (white arrows); d Retrograde perineural spread into the right orbit via the V1 (white arrows). Increased convexity of the course of V2 (white arrowheads). The normal right VI nerve is seen (black arrowhead); e, f intraorbital spread is associated with optic nerve compression (black arrowheads). Normal nerve III at the upper portion of the cavernous sinus (black arrows)

Fig. 3.9.33  Recurrent undifferentiated nasopharyngeal carcinoma. Post-contrast VIBE sequences. a The submucosal recurrent tumor (long arrow) reaches the left anterior condyle (short arrows) and invades the basiocciput and the hypoglossal canal. b The coronal plane shows the destruction on the left condyle (curved arrows), tumor tissue within the left hypoglossal canal (white arrowheads), and perineural spread along the cisternal course of left trigeminal nerve (white arrows). The complex of

mixed nerves is detected on both sides, thanks to the low SI contrasting the high SI of the adjacent internal jugular vein (black arrowheads). c The recurrent tumor invades the right Meckel’s cave (arrows) and the cavernous sinus with encasement of the internal carotid artery (black arrowheads). Tumor shows retrograde spread along the cisternal tract (black arrows). From cavernous sinus, the tumor invades the lateral sphenoid sinus wall (white arrowhead)

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

the canal, the neoplastic coating of perineural spread re-gains a macroscopic size, becoming detectable by MR immediately distal to a foramen or a canal (resurfacing phenomenon). False positive MR findings may occur when nerve enhancement is detected. Actually, the blood–nerve barrier may be disrupted in several conditions different from tumor spread, such as inflammation, demyelination, axonal degeneration, ischemia, and trauma.

are larger (probably representing residual bone) in chordoma, and appear smaller (denoting organic matrix production) in chondrosarcoma. Nonetheless, the site of origin (midline for chordoma, paramedian for chondrosarcoma) is probably the most reliable criterion. Primary malignant bone tumors other than chondrosarcomas include osteosarcoma, which is infrequent. Benign bone lesions, such as osteoblastoma, giant cell tumors, or aneurysmal bone cysts, may show a very aggressive pattern of growth and mimic malignant tumors.

3.9.3.4.2  Malignant Neoplasms Arising from Skull Base Structures

Hemangiopericytoma Hemangiopericytomas are rare vascular neoplasms. They may commonly reach a large size and present as multilobulated masses. The anterior skull base can be invaded from either an inferior (intracranial) or superior (sinonasal origin) extension of the neoplasm. Hemangiopericytomas show moderate to intense enhancement on CT. The neoplasm demonstrates low to intermediate SI on T1-weighted and intermediate to high SI on T2-weighted images. Remarkable enhancement is obtained after contrast agent administration.

Chondrosarcoma Chondrosarcoma of the skull base is a rare primary intraosseous neoplasm. It can be isolated or multiple, being part of one of the enchondromatosis syndromes (Ollier disease, Maffucci syndrome, metachondromatosis). Chondrosarcoma may arise from primitive mesenchymal cells or from embryonic rests of the cranium’s cartilaginous matrix. Conventional chondrosarcomas are grouped into three grades, according to the degree of their cellularity, cytologic atypia, and mitotic activity. Grade 1 is the least aggressive neoplasm with features of a benign tumor, while grade 3 is the most aggressive type. Other histological types are clear cell, myxoid, mesenchymal, and dedifferentiated chondrosarcoma. Conventional chondrosarcoma grows with an infiltrative pattern. The clivus, the sphenoid bone—particularly in the parasellar area and the petrous apex—are the most frequent locations of skull base chondrosarcomas. Chondrosarcomas have more commonly a paramedian location and they involve the sphenoethmoidal area in a third of cases. On CT, the chondroid matrix within the lesion appears hypodense and shows variable degrees of enhancement. Intra-tumoral calcifications are common. They range from small and scattered, often arranged as a peripheral rim, to large, dense, and diffuse. They tend to have an inter­rupted ring-like shape. Destruction of bone is a rather constant finding. The non-mineralized portion of the lesion shows moderately high-to–very high SI on T2-weighted images. Intratumoral low SI areas correspond to coarse chondroid mineralized areas. Small calcifications commonly go undetected by MR. Chondrosarcomas show low SI on T1-weighted images. Marked nonhomogeneous contrast enhancement may be observed. Based on pathological findings, it is often difficult to discriminate low-grade chondrosarcoma from its benign counterpart. Imaging may play a relevant role, providing information about cortical bone destruction. The differential diagnosis in the skull base includes chordoma. Some clues are offered by calcifications, which

Metastasis Metastases to the central skull base arise more commonly from breast carcinoma, bronchogenic carcinoma, and, less frequently, from tumors of the gastrointestinal tract, prostate, or melanoma. They are less common than calvaria metastases (Figs. 3.9.34, 3.9.35).

Fig. 3.9.34  Metastasis. Metastases to the calvaria (arrowheads) and to the cavernous sinus (arrows) from adenoid cystic carcinoma of the right submandibular gland. Metastases occurred after 8 years from primary surgical treatment and after extensive perineural spread

473

474

3  Brain, Head, and Neck

Fig. 3.9.35  Metastasis. Basiocciput and left condyle metastasis from recurrent adenocarcinoma ex-pleomorphic adenoma of right parotid gland. Patient has been previously treated by extensive surgery, RT. a CT shows destruction of left condyle with erosion of the cortical rim (arrows). b on MR the lesion has a large necrotic portion with peripheral enhancing rim (arrows)

Fig. 3.9.36  Dysplasia of the membranous labyrinth. a On TSE T2-weighted axial plane, the left vestibule is larger than normal; the lateral semicircular duct is not recognizable (arrow). b Aplasia of the lateral semicircular duct and abnormal shape of the vestibule (arrow) are confirmed by the thin MIP reconstruction obtained by means of a 3DTF-CISS

3.9.4  Lesions of the Temporal Bone

Detection and assessment of inner ear malformations is an essential step in planning cochlear implant surgery. Complete aplasia of the vestibulocochlear nerve is a contraindication for cochlear implant procedures. MR clearly shows abnormal enlargement of the endolymphatic duct (its diameter being larger than that of the posterior semicircular canal), which is the corresponding “membranous” finding of the large vestibular aqueduct syndrome (LVAS). Even the absence of the bony wall between the cochlea and the IAC wall (Gusher’s syndrome) can be identified by high resolution studies, though the bone abnormality is easier to detect on CT. This finding is critical in surgical planning. Finally, bone abnormalities of the otic capsule can be imaged by MR, mostly on T2-weighted sequences where a sort of “band-like” high SI can be shown around the cochlea, as in otosclerosis or osteogenesis imperfecta. Variable enhancement can be appreciated after contrast agent administration. Findings are nonspecific and require correlation to the clinical history for the diagnosis.

High-resolution CT (HRCT) is the first choice imaging technique for most lesions of the temporal bone. The inherently high contrast resolution between air, bone, and soft tissues allows excellent evaluation of the middle ear and mastoid air spaces. Similarly, the contrast between dense bone and fluid enables fine assessment of the otic capsule. Nonetheless, due to its overall superior contrast resolution, MR of the temporal bone is indicated in a varied list of congenital and infectious conditions and in all neoplastic lesions, both to allow characterization and to precisely define extension of the disease. 3.9.4.1  Congenital Lesions When a congenital abnormality of the inner ear is suspected, the whole brain has to be surveyed, by means of FLAIR or a T2-weighted sequence, to rule out lesions mimicking an inner ear or a CPA abnormality. Malformations of semicircular canals, vestibule, and cochlea may range from complete absence of the labyrinthine structures to severe/moderate dysplasia of the bone and/or membranous inner ear. Congenital anomalies of nerves range from hypoplasia to aplasia. Though the identification of the branches of the vestibulocochlear nerve can be obtained directly on axial planes, a precise assessment of size and course requires multiplanar reformatted planes from 3DFT-CISS or 3D T2-weighted volumes (Fig. 3.9.36).

3.9.4.1.1  Vascular Anomalies At clinical examination, three major conditions may suggest a retro-tympanic vascular mass, namely aberrant internal carotid artery (ICA), persistent stapedial artery, and jugular bulb anomalies (high riding bulb, dehiscent bulb). Aberrant ICA is due to agenesis of the first embryonic segment of the ICA, resulting in a collateral pathway cre-

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

ated by the anastomosis between the inferior tympanic branch of the ascending pharyngeal artery and the caroticotympanic artery of the ICA. As the former runs through the Jacobsen canal at the skull base, a characteristic narrowing of the aberrant ICA is appreciated at imaging. At the skull base, turbulent flow within the vessel increases spin dephasing, thus lessening signal intensity on MRA sequences (Roll 2003). MRA diagnosis is based also on anomalous location of the carotid genu in the middle ear (lateral and superior on anteroposterior view, posterior and superior on lateral reconstructions) (Roll 2003, Davis 1991). During embryogenesis, the stapedial artery connects the ICA to branches of the external carotid artery (ECA). Although rare, failure of this vessel to regress in the postembryonic life results in a condition named persistent stapedial artery (PSA). Seldom is this anomaly associated with an aberrant ICA (Roll 2003, Silbergleit 2000). Several imaging signs indicate a persistent stapedial artery, including absence of foramen spinosum and meningeal artery (supplied by PSA), and demonstration of a soft tissue/vascular mass along the promontory, which enlarges or duplicates the facial canal (Thiers 2000). This diagnosis is basically obtained by HRCT, and, in uncertain cases, it may be confirmed by MRA. Asymmetrically large jugular bulb, high-riding jugular bulb (extending above the level of floor of the internal auditory canal), and jugular bulb diverticulum are also generally demonstrated with HRCT. On MR, anomalous signals due to flow related phenomena may simulate a disease. MR venography proves helpful in suggesting the correct diagnosis (Maroldi 2001). 3.9.4.2  Traumatic Lesions Traumas of the inner ear may cause intra-labyrinthine hemorrhage. Plain SE T1-weighted sequences are expected to increase the sensitivity for the detection of blood. 3.9.4.3  Inflammatory Lesions Inflammatory processes can be related to viral or bacterial infection. Inflammation of membranous labyrinth structures or nerves may be demonstrated earlier by contrast-enhanced fat-suppressed SE or GE T1-weighted sequences than by CT. Inflammation or infection may damage the membranous labyrinth, where the fluid signal or endolymph/perilymph is replaced by fibrous tissue. Later on, calcification of the abnormal tissue leads to positive CT demonstration (labyrinthitis ossificans). Aggressive autoimmune lesions as histiocytosis × are characterized by destructive changes (Fig. 3.9.37).

Fig. 3.9.37  Histiocytosis X. TSE T2-weighted and post-contrast VIBE on coronal plane. a On the left side a recurrent lesion postsurgery appears as a heterogeneous hyperintense signal within the residual middle ear and roof of the left IAC. The low SI of the complex bone-periosteum-dura separates the disease from the subarachnoid space (black arrows). On right side, high SI in the epitympanum is seen (white arrow). b On left side, both the tissue within the residual bone and the cochlea enhance (double white parallel arrows), the latter suggesting inflammatory involvement. Mild enhancement of the dura overlying the tegmen tympani is present (black arrows). On right side, enhancement is observed in the epitympanum (single arrow), associated with thickening of soft tissues lining the external auditory canal roof (short arrows).

3.9.4.3.1  Cholesteatoma Cholesteatoma is a soft tissue mass composed of a sac filled with desquamative debris and lined by keratinizing epithelium, expanding in the middle ear or in any other pneumatized area of the petrous bone. The lesion does not exhibit cellular growth, thus it is not a true neoplasm, as its name would imply. Primary acquired choleasteatomas develop behind an intact tympanic membrane in patients with no history or with chronic otitis media (primary); secondary acquired lesions are the result of Eustachian tube dysfunction causing a complex chain of events (retraction of the tympanic membrane entrapping the epithelium, perforation, and accumulation of keratinizing epithelium within the middle ear). Congenital cholesteatomas are thought to arise from aberrant embryonic ectodermal remnants entrapped within the temporal bone.

475

476

3  Brain, Head, and Neck

Regardless of their classification, all these lesions are characterized by a locally aggressive behavior that may manifest with erosion of the ossicular chain, scutum, mastoid, and Körner’s septum destruction (Maroldi 2001). CT is the imaging technique of choice because of the inherently high contrast between bone, air and soft tissues, high spatial resolution and the multiplanar capabilities provided by multidetector technology. On MR, cholesteatoma exhibits high SI on T2-weighted images, low SI on T1-weighted images, and no contrast enhancement (Fig. 3.9.38). The role of this technique is basically restricted to the assessment of potential complications, such as epidural extension or meningo(encephalo)celes— both secondary to tegmen tympani erosion—or of more severe conditions, such as meningitis and dural sinus thrombosis. In patients complaining of facial nerve impairment, vertigo, or sensorineural hearing loss, MR may adequately demonstrate subtle changes, such as thickening and contrast uptake of the nerve or enhancement of the membranous labyrinth (Maroldi 2001). The role of MR during follow up has been recently enhanced by both diffusion-weighted sequences (Aikele 2003) and delayed contrast enhancement. In the post-operative middle ear, the detection of a mass on any of the conventional pulse sequences combined with a hyperintense area on diffusion-weighted images has high positive predictive value for cholesteatoma. However, due to the lack of spatial resolution, lesions smaller than 5 mm can be missed on diffusion-weighted sequences. Sensitivity and positive predictive value up to 100% have been reported with delayed post-contrast sequences acquired 30–45 min after administration. The rationale of this strategy exploits the difference between residual cholesteatoma (avascular, i.e., never enhancing) and scar tissue (hypovascular, thus showing very late enhancement) (Ayache 2005). 3.9.4.3.2  Cholesterol Granuloma Cholesterol granuloma may occur in any obstructed air cavity, and it is the most frequent primary lesion of the petrous apex. When occurring in that site in a patient with no history of chronic otitis media, it is also named giant cholesterol cyst. In the petrous apex, cholesterol granuloma may be a rather destructive lesion, involving the

Fig. 3.9.38  Cholesteatoma of right middle ear. On the 3DFT CISS, the lesion has an intermediate signal (arrows). Discontinuity of the bone limit is possible

carotid canal, cavernous sinus, or cerebellopontine angle. MR signal is bright on all pulse sequences. It allows an accurate diagnosis in most cases and it differentiates cholesterol granuloma from cholesteatoma (long T1-weighted and T2-weighted relaxation times). Peripheral hypointense rim (both on T2-weighted and T1-weighted images) can be observed, due to expanded cortical bone and hemosiderin deposits (Bonneville 2001). 3.9.4.3.3  Petrous Apicitis It is generally secondary to middle ear inflammation spread into the petrous apex, via either longitudinally oriented air cells and/or intraosseous venous channels (Lee Y. H. 2005). Immunosuppression (due to diabetes, HIV infection, chronic steroid treatment) predisposes to this complication. The classic triad of symptoms first described by Gradenigo (otorrhea, pain in the territory of trigeminal nerve, abducens nerve paralysis) occurs infrequently nowadays, probably because prompt antibiotic treatment has made this disease rarer and less severe. MR plays a prominent role in this condition. Early bone marrow changes can be easily detected on plain T1weighted sequences. T2-weighted and contrast enhanced T1-weighted images precisely depict soft tissue lesions that may be associated (particularly fluid collections). Vascular complications—such as cavernous or dural sinus thrombosis—can be ruled out by MR venography sequences (TOF/CE MRA). 3.9.4.4  Benign Neoplasms 3.9.4.4.1  Glomus Tumors Glomus tumor (or paraganglioma) is a generally benign neoplasm arising from chemoreceptor cells located in several subsites of the head neck, such as the middle ear, jugular bulb, carotid bulb, ganglion nodosum (vagal nerve) and larynx. Glomus tympanicum is the most common middle ear neoplasm arising from the paraganglionic tissue scattered along both Jacobson and Arnold nerves (branches of cranial nerves IX and X, respectively) at the level of cochlear promontory. The lesion is generally confined within the middle ear, with infrequent extension into the mastoid and—via the Eustachian tube—into the nasopharynx. The ossicular chain may be entrapped by large lesions, although bone erosion is not typical (Rao 1999). Glomus jugulare originates from paraganglia located along both the anterior and posterior aspects of the jugular foramen. Enlargement and destruction of the foramen (with a peculiar moth-eaten pattern) are common. The lesion may spread inferiorly to invade the jugular vein

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

and the retrostyloid compartment of the parapharyngeal space. Posterior fossa invasion may occur, either through petrous bone destruction or through the intrameatal pathway of spread. When the lesion spreads superiorly through the middle ear floor into the hypo- and mesotympanum it is referred to as glomus jugulotympanicum (Rao 1999) (Fig. 3.9.39). Imaging plays a pivotal role in discriminating between glomus tympanicum and jugulotympanicum. Even subtle changes of bone at the level of the middle ear floor and jugular fossa should be carefully scrutinized because failure in differentiating the two entities may result in improper surgical approach and, therefore, in an increased risk of complications. At MR, paraganglioma exhibits high signal on T2weighted and low signal on T1-weighted images. Multiple punctuate and serpiginous flow-voids may be detected

when lesions are greater than 1 cm in diameter, thus representing large caliber intralesional vessels. This results in a peculiar salt and pepper appearance, more pronounced after contrast application due to the high degree of enhancement of the lesion (Farina 1999). On T1-weighted imaging, focal areas of spontaneous hyperintensity may be observed, indicating hemorrhage (Bonneville 2001). The role of MR angiography is debatable, even though some authors advocate the use of MR venography to discriminate between glomus jugulare and non-neoplastic vascular anomalies (see below) (Vogl 1994). 3.9.4.4.2  Temporal Bone Schwannomas Schwannoma is a benign tumor arising from Schwann cells of the sheath of cranial, spinal, and peripheral

Fig. 3.9.39  Glomus jugulotympanicum. a On TSE T2-weighted, the right jugular fossa is occupied by a low SI mass. b The paraganglioma has bright enhancement, a small component projects into the middle ear. Erosion of the inferior aspect of the internal auditory is detected. c The lesion invades the right jugular foramen (arrows)

477

478

3  Brain, Head, and Neck

nerves. For unknown reasons, the acoustic (i.e. vestibulocochlear nerve) is the commonest nerve of origin. In the temporal bone schwannomas may develop at the jugular foramen, or along the facial nerve course. The acoustic nerve consists of two anatomically and functionally distinct nerves, namely the cochlear and vestibular nerve. Most acoustic neuromas arise from the intra­ canalicular tract of either the superior or inferior vestibular nerves, at the level of the glial-Schwann cell junction, thus at the fundus of the canal. Conversely, intralabyrinthine schwannomas are thought to arise more frequently from the cochlear nerve (Mafee 1990). Large acoustic schwannomas may extend towards the cerebellopontine angle and, less frequently, reach the jugular foramen. Schwannomas of the jugular foramen more frequently arise from the ninth cranial nerve. Enlargement of the foramen itself and intraosseous spread (cortical remodeling, medullary replacement) are frequently observed, as well as variable degree of posterior fossa invasion. The neurologic findings often do not allow the discrimination between vestibulocochlear and jugular foramen schwannomas extending into the CPA. Careful scrutiny of the IAC on MR images is required, as this is rarely left intact by a large acoustic neuroma). Facial nerve schwannomas may occur along the entire course of the nerve, though they are more often reported at the level of the geniculate ganglion. From this site, they may extend either along the labyrinthine or the tympanic portion of the nerve. Intracranial extension through the roof of the temporal bone is uncommon (Ginsberg 1999). MR appearance of schwannomas basically consists of slightly high T2-weighted signal and marked enhancement after contrast administration (Fig. 3.9.40). The latter is due to extravascular leakage of contrast material and poor venous drainage rather than to hypervascularization (Farina 1999). This fact is confirmed by the avascular appearance of this lesion on digital subtraction angiography (Caldemeyer 1997). Partial or complete cystic degeneration of the lesion may result in high SI on T2-weighted images and decreased or absent enhancement on post-contrast images. Focal inhomogeneities may also be related to fatty degeneration, intralesional hemorrhage, or calcification (Bonneville 2001, Ginsberg 1999, Caldemeyer 1997). The differential diagnosis of acoustic neuromas basically includes CPA meningioma and glomus tumors. The former shows similar enhancement, but generally a lower SI on T2-weighted images. Additional findings suggesting a meningioma include calcification and hyperostosis (see below). Tumor growth through the jugular foramen is uncommon. Glomus tumors are characterized by serpiginous flow voids. Moreover, the vascularization of these lesions is much higher than that of schwannomas, particularly when the dynamic enhancement pattern is analyzed.

Fig. 3.9.40  Schwannoma of the cochlear nerve. a Although limited in size, the schwannoma extends from the right cochlea (arrows) into the IAC fundus (black arrowhead), without blocking the CSF flow. b Remarkable enhancement of both the intra-cochlear (arrow) and intrameatal portion (arrowheads) is obtained

The use of high resolution thin slices from 3DFT-CISS MRI is advantageous in the detection of acoustic neuromas. Even small lesions are superbly detected as focal, hypointense, filling defects within the hyperintense signal of fluid contained within the internal acoustic canal or the inner ear (Held 1997) (Figs. 3.9.40, 3.9.41). 3.9.4.4.3  Endolymphatic Sac Tumors These neoplasms arise from the endolymphatic sac, which is located in the distal portion of the vestibular aqueduct. Though rare, endolymphatic sac tumors are observed in up to 7–11% of patients affected by von Hippel-Lindau disease, a syndrome manifesting with multiple hemangioblastomas of the retina and central nervous system, pancreatic and renal cysts, renal carcinoma, pheochromocytoma and epididymal cystadenomas (Mukherji 1997). Endolymphatic sac tumors display a locally aggressive pattern of growth with destruction of the retrolabyrinthine portion of the petrous bone, resulting in invasion of the cerebellopontine cistern and cerebellum. Large lesions may erode the labyrinth, causing sensorineural hearing loss. Mastoid destruction and encasement of the facial nerve may be observed. MR signal is heterogeneous on all sequences. Focal areas of spontaneous high SI on T1-weighted images, scattered within the tumor, may be due to different

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

breakdown products of a subacute hemorrhage, including methemoglobin, cholesterol clefts, blood filled cysts, proteinaceous cysts (Mukherji 1997, Kilickesmez 2006). Contrast enhancement is marked but heterogeneous. Multiple intralesional flow voids may be detected in lesions larger than 2 cm. Focal hypointense spots within

the lesion likely correspond to bone fragments rather than new bone formation (Bonneville 2001). As endolymphatic sac tumors and several other papillary neoplasms share common histopathological features, MR signs may play a major role in the differential diagnosis of small retrolabyrinthine lesions.

Fig. 3.9.41  Schwannoma of the vestibular nerve. a The TSE T2-weighted axial plane allows the detection of a small hypointense lesion within left vestibule (arrows). b Marked enhancement is obtained on the VIBE sequence. c On the series of coronal MPR obtained by the 3DFT CISS, it is possible to appreciate the abnormal signal of the fluid filling the vestibule and semicircular canals (arrows) on the side of the vestibular tumor

479

480

3  Brain, Head, and Neck

3.9.4.4.4  Meningioma Approximately 45% of posterior fossa meningiomas arise from the petrous bone (Wu 2005), where they account for the second most common lesion (10–15% of lesions); seldom are they described at the level of jugular foramen. In the petrous bone, meningiomas are classified as anterior or posterior to the internal acoustic canal, mirroring different surgical approaches and prognosis. Only sporadically are they centered at the IAC level. MR appearance consists of high SI on T2-weighted and iso intense SI on T1-weighted images (as compared to grey matter). Bright enhancement after contrast application (Laudadio 2004) is typical. Meningiomas are generally well demarcated but not encapsulated. They may induce extensive bone changes, more frequently characterized by hyperostosis (Fig. 3.9.42). Lysis is rare. The presence of the dural tail sign may help in the diagnosis, but it should not be considered pathognomonic (Laudadio 2004). The differential diagnosis includes acoustic neuroma. As both lesions brightly enhance after contrast agent administration, the lesion site and its morphology are the main differential criteria. Meningioma is only rarely centered in the IAC, and it exhibits a mushroom-like shape, which forms obtuse angles with the

petrous bone outline. Acoustic schwannomas generally have their epicenter in the IAC. When they extend into the CPA, they display the typical ice cream-cone shape. 3.9.4.5  Malignant Neoplasms Although rare on the whole, malignant tumors of the temporal bone include a varied list of histotypes. In the pediatric population, rhabdomyosarcoma is probably the most common malignancy of the temporal bone. Several other sarcomas are described in literature, mostly observed in children or young adults: reports can be found on cases of Ewing’s sarcoma, angiosarcoma, kaposiform hemangioendothelioma, and chondrosarcoma. All share a rather aggressive pattern of growth with bone invasion and destruction. Bone permeation, rather than erosion, is exhibited by lymphoma and Langerhans cell histiocytosis, the latter being an idiopathic proliferation of a specialized subpopulation of histiocytes, characterized by high SI on T2-weighted images and bright contrast enhancement (Caldemeyer 1997). Other hematological disorders potentially involving the temporal bone include plasmacytoma (neoplastic

Fig. 3.9.42  Giant cell tumor. a The tumor arises from the lateral greater wing of the sphenoid bone, which appears remarkably hypointense (arrows) on the TSE T2-weighted image. Fluid within the tympanic cavity and mastoid cells show high SI. b Intermediate and heterogeneous enhancement is observed after contrast agent administration (arrows)

3.9  Magnetic Resonance of the Skull Base and Petrous Bone

proliferation of a single clone of plasma cells) and granulocytic sarcoma (chloroma) (Lee B. 2002). Additionally, the temporal bone (particularly the petrous apex) may be secondarily involved by neoplasms arising from adjacent spaces/structures, such as chordoma, chondrosarcoma, and nasopharyngeal carcinoma. Most temporal bone tumors lack MR features allowing a precise characterization based on signal intensity. Generally, the presence of destructive bone changes along with soft tissue invasion is suggestive of a malignant tumor. Nonetheless, it must be emphasized that locally aggressive behavior may be exhibited also by benign neoplasms such as giant cell tumor (Fig. 3.9.42). Thus, the major role of MR in patients with a temporal bone aggressive mass is restricted to an accurate assessment of tumor extension, particularly when surgical treatment has to be planned.

7.

3.9.4.5.1  Metastasis

13.

Temporal bone metastases are infrequent. They are reported as hematogenous spread from a varied list of primary sites (breast, lung, kidney, prostate, thyroid, gastrointestinal tract) (De Vos 2005, Nagai 2005). MR findings vary according to the histotype of the primary tumor: hypervascular metastases (kidney, thyroid) may mimic paragangliomas. Bone changes may range from destruction to sclerosis.

8.

9.

10.

11.

12.

14.

15.

16.

References 1.

2.

3.

4.

5.

6.

Aikele P, Kittner T, Offergeld C, Kaftan H, Huttenbrink KB, Laniado M (2003) Diffusion-weighted MR imaging of cholesteatoma in pediatric and adult patients who have undergone middle ear surgery. AJR Am J Roentgenol 181:261–265 Ayache D, Williams MT, Lejeune D, Corre A (2005) Usefulness of delayed postcontrast magnetic resonance imaging in the detection of residual cholesteatoma after canal wall-up tympanoplasty. Laryngoscope 115:607–610 Berg R van den, Schepers A, de Bruine FT, Liauw L, Mertens BJ, van der Mey AG, van Buchem MA (2004) The value of MR angiography techniques in the detection of head and neck paragangliomas. Eur J Radiol 52:240–245 Bonneville F, Sarrazin JL, Marsot-Dupuch K, Iffenecker C, Cordoliani YS, Doyon D, Bonneville JF (2001) Unusual lesions of the cerebellopontine angle: a segmental approach. Radiographics 21:419–438 Caldemeyer KS, Mathews VP, Azzarelli B, Smith RR (1997) The jugular foramen: a review of anatomy, masses, and imaging characteristics. Radiographics 17:1123–1139 Casselman JW, Offeciers FE, Govaerts PJ, Kuhweide R, Geldof H, Somers T, D’Hont G (1997) Aplasia and hypoplasia of the vestibulocochlear nerve: diagnosis with MR imaging. Radiology 202:773–781

17.

18.

19.

20.

21. 22.

23.

24.

Danesi G, Panizza B, Mazzoni A, Calabrese V (2000) Anterior approaches in juvenile nasopharyngeal angiofibromas with intracranial extension. Otolaryngol Head Neck Surg 122:277–283 Davis WL, Harnsberger HR (1991) MR angiography of an aberrant internal carotid artery. AJNR Am J Neuroradiol 12:1225 De Vos C, Gerard JM, Decat M, Gersdorff M (2005) Metastatic renal cell carcinoma to the temporal bone: case report. B-ENT 1:43–46 Farina D, Hermans R, Lemmerling M, Op Beeck K (1999) Imaging of the parapharyngeal space. JBR-BTR 82:234–239 Giangaspero F, Burger PC, Osborne DR, Stein RB (1984) Suprasellar papillary squamous epithelioma (“papillary craniopharyngioma”). Am J Surg Pathol 8:57–64 Ginsberg LE, DeMonte F (1999) Diagnosis please. Case 16: facial nerve schwannoma with middle cranial fossa involvement. Radiology 213:364–368 Held P, Fellner C, Fellner F, Seitz J, Graf S, Hilbert M, Strutz J (1997) MRI of inner ear and facial nerve pathology using 3D MP-RAGE and 3D CISS sequences. Br J Radiol 70:558–566 Howells RC, Ramadan HH (2001) Usefulness of computed tomography and magnetic resonance in fulminant invasive fungal rhinosinusitis. Am J Rhinol 15:255–261 Kawahara Y, Niiro M, Yokoyama S, Kuratsu J (2001) Dural congestion accompanying meningioma invasion into vessels: the dural tail sign. Neuroradiology 43:462–465 Kennard CD, Rasmussen JE (1990) Congenital midline nasal masses: diagnosis and management. J Dermatol Surg Oncol 16:1025–1036 Kilickesmez O (2006) Endolymphatic sac tumor in a patient with von Hippel-Lindau disease: MR imaging findings. Diagn Interv Radiol 12:14–16 Laudadio P, Canani FB, Cunsolo E (2004) Meningioma of the internal auditory canal. Acta Otolaryngol 124:1231–1234 Lee B, Fatterpekar GM, Kim W, Som PM (2002) Granulocytic sarcoma of the temporal bone. AJNR Am J Neuroradiol 23:1497–1499 Lee YH, Lee NJ, Kim JH, Song JJ (2005) CT, MRI and gallium SPECT in the diagnosis and treatment of petrous apicitis presenting as multiple cranial neuropathies. Br J Radiol 78:948–951 Lenz M, Greess H, Dobritz M, Kersting-Sommerhoff B (2000) Methods: MRT. Eur J Radiol 33:178–184 Lowe LH, Booth TN, Joglar JM, Rollins NK. Midface anomalies in children. Radiographics 20:907–922; quiz 1106–1107, 1112 Mafee MF, Lachenauer CS, Kumar A, Arnold PM, Buckingham RA, Valvassori GE (1990) CT and MR imaging of intralabyrinthine schwannoma: report of two cases and review of the literature. Radiology 174:395–400 Maroldi R, Nicolai P (2005) Imaging in treatment planning for sinonasal diseases. Springer, Berlin Heidelberg, New York

481

482

3  Brain, Head, and Neck 25. Maroldi R, Farina D, Palvarini L, Marconi A, Gadola E, Menni K, Battaglia G (2001) Computed tomography and magnetic resonance imaging of pathologic conditions of the middle ear. Eur J Radiol 40:78–93 26. Maroldi R, Farina D, Palvarini L, Lombardi D, Tomenzoli D, Nicolai P (2004) Magnetic resonance imaging findings of inverted papilloma: differential diagnosis with malignant sinonasal tumors. Am J Rhinol 18:305–310 27. Meyers SP, Hirsch WL, Jr., Curtin HD, Barnes L, Sekhar LN, Sen C (1992) Chordomas of the skull base: MR features. AJNR Am J Neuroradiol 13:1627–1636 28. Mukherji SK, Albernaz VS, Lo WW, Gaffey MJ, Megerian CA, Feghali JG, Brook A, Lewin JS, Lanzieri CF, Talbot JM, Meyer JR, Carmody RF, Weissman JL, Smirniotopoulos JG, Rao VM, Jinkins JR, Castillo M (1997) Papillary endolymphatic sac tumors: CT, MR imaging, and angiographic findings in 20 patients. Radiology 202:801–808 29. Nagai M, Yamada H, Kitamoto M, Ikeda J, Mori Y, Monzen Y, Fukuhara T (2005) Facial nerve palsy due to temporal bone metastasis from hepatocellular carcinoma. J Gastroenterol Hepatol 20:1131–1132 30. Naidich TP, Altman NR, Braffman BH, McLone DG, Zimmerman RA (1992) Cephaloceles and related malformations. AJNR Am J Neuroradiol 13:655–690 31. Ojiri H, Ujita M, Tada S, Fukuda K (2000) Potentially distinctive features of sinonasal inverted papilloma on MR imaging. AJR Am J Roentgenol 175:465–468 32. Parmar HA, Sitoh YY (2005) Diffusion-weighted imaging findings in central skull base osteomyelitis with pharyngeal abscess formation. AJR Am J Roentgenol 184:1363–1364 33. Rao AB, Koeller KK, Adair CF (1999) From the archives of the AFIP. Paragangliomas of the head and neck: radiologicpathologic correlation. Armed Forces Institute of Pathology. Radiographics 19:1605–1632

34. Roll JD, Urban MA, Larson TC III, Gailloud P, Jacob P, Harnsberger HR (2003) Bilateral aberrant internal carotid arteries with bilateral persistent stapedial arteries and bilateral duplicated internal carotid arteries. AJNR Am J Neuroradiol 24:762–765 35. Sartoretti-Schefer S, Wichmann W, Aguzzi A, Valavanis A (1997) MR differentiation of adamantinous and squamous-papillary craniopharyngiomas. AJNR Am J Neuroradiol 18:77–87 36. Silbergleit R, Quint DJ, Mehta BA, Patel SC, Metes JJ, Noujaim SE (2000) The persistent stapedial artery. AJNR Am J Neuroradiol 21:572–527 37. Som PM, Dillon WP, Fullerton GD, Zimmerman RA, Rajagopalan B, Marom Z (1989) Chronically obstructed sinonasal secretions: observations on T1 and T2 shortening. Radiology 72:515–520 38. Thiers FA, Sakai O, Poe DS, Curtin HD (2000) Persistent stapedial artery: CT findings. AJNR Am J Neuroradiol 21:1551–1554 39. Trimarchi M, Gregorini G, Facchetti F, Morassi ML, Manfredini C, Maroldi R, Nicolai P, Russell KA, McDonald TJ, Specks U (2001) Cocaine-induced midline destructive lesions: clinical, radiographic, histopathologic, and serologic features and their differentiation from Wegener granulomatosis. Medicine (Balt) 80:391–404 40. Vogl TJ, Juergens M, Balzer JO, Mack MG, Bergman C, Grevers G, Lissner J, Felix R (1994) Glomus tumors of the skull base: combined use of MR angiography and spinecho imaging. Radiology 192:103–110 41. Wu ZB, Yu CJ, Guan SS (2005) Posterior petrous meningiomas: 82 cases. J Neurosurg 102:284–289 42. Yang CW, Carr JC, Futterer SF, Morasch MD, Yang BP, Shors SM, Finn JP (2005) Contrast-enhanced MR angiography of the carotid and vertebrobasilar circulations. AJNR Am J Neuroradiol 26:2095–2101

3.10  Head and Neck

3.10  Head and Neck H.E. Stambuk and N.J. Fischbein 3.10.1  Introduction The head and neck region is a complex collection of anatomic subsites, and a spectrum of diseases occurs at each anatomic subsite with unique implications in terms of diagnosis and treatment. For most diseases, radiologic imaging is an essential modality in diagnosis, work up, treatment planning and post-treatment surveillance. In addition, certain benign conditions of the head and neck can be easily diagnosed using appropriate radiologic imaging without the need for histopathological confirmation. As a general rule, CT is the more frequently used initial study, but MRI is extremely useful in answering specific questions. We begin with a general discussion of some technical aspects of MRI as they pertain to the head and neck. Classically, evaluation of head and neck lesions is best separated into mucosal and non-mucosal diseases because of distinct biologic behavior. Mucosal lesions tend to directly invade adjacent structures and violate fascial planes so that they require a distinct diagnostic approach, compared with non-mucosal pathology that tends to respect the fascial planes and allows a spatial approach to diagnosis. Owing to the complexity of anatomy and the variety in histopathology, the more common disease processes are discussed under separate anatomic subsites and the reader will be directed to appropriate sources for more detailed information where indicated. 3.10.1.1  Examination Technique MRI imaging of the head and neck requires at least two different types of sequences in no less than two planes for

adequate characterization of lesions (Table 3.10.1). Certain modifications in technique may be indicated for certain anatomic subsites, and these will be presented when appropriate. The precontrast T1-weighted image sequence is of particular importance in the head and neck because the presence of high-signal fat provides excellent innate contrast against which most pathology can be easily delineated. Infiltrative processes that violate tissue planes or extend into the bone marrow fat (in adults) are especially well imaged on the T1-weighted image sequence, and may in fact be missed on the contrast-enhanced T1-weighted image sequence (Fig. 3.10.1). Lesions that contain blood or proteinaceous material can cause T1 shortening and are therefore bright on the T1-weighted image sequence. Fluid appears very bright on the T2-weighted image sequence; normal muscle is dark while most pathology is conspicuously brighter. Artifacts secondary to motion and magnetic susceptibility in a conventional spin-echo T2-weighted image sequence can be overcome to some extent by using the fast spin-echo technique. However, fat saturation should be applied since fat appears bright on the fast spin-echo image. Most pathology enhances with gadolinium contrast administration and the enhancement characteristics of the lesion can be useful in differential diagnosis. For example, benign cystic lesions, edema, and calcification do not enhance while certain tumors such as paragangliomas enhance densely. In addition to delineating the borders of a lesion, contrast imaging is particularly helpful in detecting perineural spread of tumor, cavernous sinus invasion, and meningeal involvement (Fig. 3.10.2). As described above, fat saturation on post-contrast T1-weighted image and fast spin-echo T2-weighted image sequences is imperative in head and neck imaging to make enhancement of a lesion conspicuous relative to surrounding fat. Fat saturation is often not available on low-field scanners which are

Table 3.10.1  Standard technique for head and neck MRI Sequence

Imaging plane

Imaging characteristics

T1-weighted images

Sagittal, axial and coronal

Fat is bright Fluid (cerebrospinal fluid, mucus) is dark Muscle is intermediate in signal intensity Most pathology is isointense to muscle

Fast spin-echo T2-weighted images with fat saturation

Axial

Fluid (cerebrospinal fluid, mucus) is bright Fat is suppressed Most pathology is intermediate but some are high signal

Post-gadolinium T1-weighted images with fat saturation

Axial and coronal

Most malignant neoplasms enhance Benign lesions enhance variably depending on histology

Slice thickness

≤5 mm

Gadolinium-based contrast agent

0.1 mmol/kg body weight

483

484

3  Brain, Head, and Neck

Fig. 3.10.1  a Precontrast axial T1-weighted image of the mandible shows tumor replacing normal fatty marrow (arrow) of the right hemimandible with extraosseous extension of tumor. b Post-contrast axial T1-weighted image at the same level shows enhancing tumor in right hemimandible that is similar to and

blends in with fatty marrow on the left. c Axial post-contrast fat saturated T1-weighted image suppresses normal marrow in the left hemimandible so that the enhancing tumor in the right hemimandible is more apparent

therefore not ideal for head and neck imaging. Additionally, low-field scanners do not render the detail necessary for evaluation of head and neck anatomy. Therefore, highfield (1.5-T) MRI scanners are preferred for head and neck imaging. Most head and neck imaging is performed using a neurovascular array coil except for the orbits and the parotid where a head coil is preferred. Patients who are claustrophobic may need adequate sedation. Apart from the obvious contraindications to MR imaging such as cardiac pacemakers, cochlear implants, and ferromagnetic intracranial aneurysm clips, other relative contraindications include patients who are not able to cooperate or are too ill. MR imaging of each head and neck subsite takes about 30 min to accomplish and requires the patient to lie still. Although MR is ideal for assessment of tumors of the oral cavity and oropharynx, the exam can be non-diagnostic secondary to excessive swallowing artifact (Fig. 3.10.3), which is not uncommon if the patient has pooling of saliva and a large tumor. MRI techniques such as magnetic resonance spectroscopy, diffusion-weighted imaging, and functional MRI are under investigation and are not yet in standard use for head and neck imaging. Fig. 3.10.2  Post-contrast coronal T1-weighted image fat saturation image in a patient with SCC of the right gingivobuccal sulcus and new onset numbness of the chin. Note the abnormal enhancement and enlargement of the right inferior alveolar nerve in the mandibular canal (white arrow), representing perineural spread of tumor. The normal left inferior alveolar nerve (white arrowhead) is included for comparison. The intense enhancement of the nasal cavity mucosa is normal and is a cue that this is a post-contrast sequence

3.10.2  Mucosal Diseases of the Head and Neck The overwhelming majority of mucosal lesions in the head and neck are squamous cell carcinoma (SCC). Direct clinical examination of the mucosa is superior to imaging in assessing the surface extent of the lesion. Imaging, on the other hand, provides better evaluation of the depth of the lesion, its submucosal extent, perineural spread,

3.10  Head and Neck

bone invasion, and neurovascular encasement. These features must be addressed by the radiologist in evaluating patients with mucosal lesions of the head and neck. In terms of therapeutic decision-making and treatment planning, a thorough understanding of the pathways of tumor spread is essential. This information is beyond the scope of this chapter and is available elsewhere for the interested reader. Due to the complex anatomy and peculiar behavior of tumors at various locations, mucosal diseases of the head and neck are most conveniently discussed under distinct anatomic subsites as follows. 3.10.2.1  Paranasal Sinuses and Nasal Cavity 3.10.2.1.1  Normal Anatomy The paired maxillary sinuses are situated on either side of the nasal cavity while the frontal and ethmoid sinuses lie superiorly. The maxillary sinuses are related to the orbit superiorly, the hard palate inferiorly, and the pterygopalatine fossa posteriorly. The ethmoid sinuses consist of anterior, middle, and posterior air cells that lie in close proximity to the floor of the anterior cranial fossa superiorly and the orbits on either side. The roof of the ethmoid sinuses (fovea ethmoidalis) constitutes the floor of the anterior cranial fossa lateral and superior to the cribriform plate which forms the roof of the nasal cavity medially. The frontal sinus is located above the orbits and anterior to the contents of the anterior cranial fossa. The sphenoid sinus is contained within the body of the sphenoid bone and drains into the superior meatus via the sphenoeth-

Fig. 3.10.3  Sagittal T1-weighted image showing marked motion degraded artifact from excessive swallowing

moidal recess. The nasal cavity (NC) is divided by a relatively featureless midline septum while the superior, middle, and inferior turbinates project into the cavity from its lateral wall. The nasal cavity communicates with the pterygopalatine fossa (PPF) via the sphenopalatine foramen which allows spread of infection and tumor to several deep facial and intracranial spaces (Fig. 3.10.4). The bony anatomy of the paranasal sinuses (PNS) and nasal cavity is best delineated by thin section CT (Fig. 3.10.5) while MR is superior at defining mucosal pathology. CT is also the modality of choice in the evaluation of inflammatory sinus disease, which will not be discussed and the reader is referred to specialized texts for detailed description of the bony anatomy, anatomic variations, drainage flow patterns, and inflammatory sinus disease. In addition to the standard sequences listed in Table 3.10.1, coronal fast spin-echo T2-weighted images with fat saturation easily distinguish neoplastic masses from polyps, thickened mucosa, and retained secretions (Fig. 3.10.6). 3.10.2.1.2  Pathological Findings A wide spectrum of disease processes can affect the PNS and NC which are not within the scope of this book. Inflammatory sinus disease is best evaluated using CT. However, its intracranial complications such as subdural empyema, meningitis, brain abscess, and cavernous sinus thrombosis are better evaluated by MR. Since MR imaging is primarily used for evaluating neoplastic processes, only a limited discussion of the relevance of MR in some of the commoner congenital and neoplastic conditions follows. Congenital and developmental defects of this region include choanal atresia, pyriform aperture stenosis, nasal gliomas, dermoids, sinus tracts, and cephaloceles. Of these only cephalocele will be mentioned briefly, since recognition prior to surgery is crucial to avoid inadvertent entry into the central nervous system. As with many other situations in this region, CT and MR are complementary since CT demonstrates a bony skull base defect while MR reveals the precise nature of its contents. For example, a bony defect at the skull base on CT may suggest cephalocele but MR can delineate exactly which tissues have herniated through the defect (Fig. 3.10.7). Benign neoplasms are generally slow growing and tend to regressively remodel adjoining bone. Bone erosion and destruction, perineural spread and regional lymphadenopathy are reliable features of malignant pathology. CT is the modality of choice for evaluating bone while MR is useful for differentiating tumor from post-obstructive sinus disease, and assessment of other features such as orbital or intracranial extension and perineural spread. The most common benign neoplasm in this region is the inverting papilloma that usually arises from the lateral wall of the NC and middle meatus. CT generally shows a lobulated tumor that may have calcification and extends

485

486

3  Brain, Head, and Neck

Fig. 3.10.5  The bony anatomy of the paranasal sinuses and nasal cavity is best delineated by thin section coronal CT images in bone window. The following structures are demonstrated: frontal (fr), ethmoid (eth) and maxillary (max) sinuses, maxillary alveolus (alv) and hard palate (HP), and the infraorbital foramen (IOF) through which the infraorbital nerve (branch of V2) exits

Fig. 3.10.4  Thin section axial CT images in bone window nicely demonstrate the pterygopalatine fossa and its interconnections at the skull base. a The following structures are delineated: pterygopalatine fossa (PPF), pterygomaxillary fissure (horizontal white arrow), sphenopalatine foramen (white arrowhead), vidian canal (double, short white arrows), foramen ovale (O), foramen spinosum (oblique white arrow), clivus (Cl), jugular foramen (J) and carotid canal (C). b Slightly more cephalad, the PPF communicates with the foramen rotundum (R), which allows access to the middle cranial fossa (MCF) at the level of the inferior orbital fissure (arrowheads). The petrous bone (P) and the petroclival fissure (white arrow) are indicated for reference

Fig. 3.10.6  Coronal FSE T2-weighted image with fat saturation easily distinguishes the intermediate signal intensity sinonasal tumor (mass) from the very high signal intensity post-obstructive secretions and inflamed mucosa (M) in the left maxillary and ethmoid sinuses. Note the normal high signal intensity of the right inferior turbinate (IT)

3.10  Head and Neck

Fig. 3.10.7  a Coronal CT of the PNS in bone window obtained in a young woman who presented with chronic right nasal obstruction shows a large mass in the right nasal cavity with absence of the right fovea ethmoidalis and post-obstructive secretions in the right maxillary sinus (max). Note the normal appearance of the left fovea ethmoidalis (arrowheads) and cribriform plate (arrow). MR must be obtained to evaluate if the mass originated from the nasal cavity or anterior cranial fossa prior to biopsy or surgical manipulation. b Coronal FSE T2-weighted image with FS shows a CSF filled meningocele (M) with herniated brain tissue (B) through the ethmoid roof defect consistent with meningoencephalocele which required surgical repair

into the maxillary antrum from the middle meatus via a widened maxillary ostium (Fig. 3.10.8). The MR features are nonspecific but a convoluted, cerebriform appearance on T2-weighted image or post-contrast images has been reported to be suggestive of inverting papilloma. MR is also helpful in delineating the precise extent of disease at the skull base since it can differentiate between tumor and inflammatory mucosal change. Juvenile nasal angiofibroma (JNA) is classically seen in young males. These multilobulated benign tumors arise from the posterolateral nasal wall at the level of the sphenopalatine foramen through which they can extend into the PPF even when they are small. Extension into the nasopharynx, sphenoid and ethmoid sinuses, infratemporal fossa, and the middle cranial fossa can occur with increasing size. These tumors are highly vascular and enhance intensely on CT. Regressive remodeling of bone including anterior bowing of the posterior wall of the maxillary sinus is common, but bone invasion can occur. JNA have characteristic flow voids and are heterogenous on MR due to cyst formation and hemorrhagic areas (Fig. 3.10.9). Preoperative embolization via catheter angiography is helpful in minimizing intraoperative blood loss and increasing the likelihood of complete surgical resection. SCC of the maxillary sinus is the most common malignant neoplasm of the PNS and tends to present in advanced stages since symptomatology is nonspecific. Early tumors may be diagnosed incidentally on imaging obtained for other reasons such as inflammatory sinus disease. MR is especially useful if CT done for investigation of sinus disease is suspicious for a mass since T2-weighted image can easily distinguish tumor from inflammatory disease. Tumors appear as a bone-destructive mass with irregular invasion into adjacent soft tissue (Fig. 3.10.10). Extension through the posterior wall of the maxillary sinus allows the tumor access to the orbit and cranial cavity via the PPF. Esthesioneuroblastoma (ENB) is a neuroendocrine tumor of neural crest origin that arises from the olfactory epithelium in the high nasal vault and upper nasal septum. Because of their anatomic proximity to the cribriform plate, these tumors are particularly prone to intracranial extension. Although imaging characteristics are nonspecific, the presence of peripheral cysts within the intracranial component of tumor is highly suggestive of ENB (Fig. 3.10.11). Other rare malignant tumors include mucosal melanoma and lymphoma, which also are largely nonspecific on imaging. Mucosal melanoma can be focally or diffusely bright on T1-weighted image because of melanin or previous hemorrhage (Fig. 3.10.12). Extent of disease work-up is important to evaluate not only perineural spread and regional nodal disease, but also distant metastases. Non-Hodgkin’s lymphoma has a variable appearance but involvement of multiple locations without a dominant mass, infiltration of adjacent fat (premaxillary

487

488

3  Brain, Head, and Neck

Fig. 3.10.8  a Coronal CT image of the PNS in bone window shows a lobulated mass from the lateral nasal wall extending through a widened left maxillary ostium and also into the right nasal cavity (arrowhead) through the nasal septum. b Post-gadolinium coronal T1-weighted image with FS shows a heterogeneously enhancing mass involving the left inferior turbinate (IT). The lesion was found to be an inverting papilloma at surgery

3.10  Head and Neck 9 Fig. 3.10.9  a Axial T1-weighted image shows a large JNA within the large nasal cavity with widening of, and extension into, the right PPF (white arrowheads). Note the normal fat (F) in the left PPF. The posterior wall of the right maxillary remodeled anteriorly while the tumor has eroded the right pterygoid process. The normal fatty marrow signal in the left pterygoid process is shown for comparison (small arrow). b The tumor is densely heterogeneously enhancing with focal areas of cystic de-

generation (C) on axial postcontrast T1-weighted image. c Axial post-contrast T1-weighted image shows a smaller tumor in the classic location in another patient. The tumor arises in the nasal cavity adjacent to the sphenopalatine foramen (arrow). A common feature is the presence of intralesional vascular flow voids (arrowhead) (Images a and b courtesy of Sofia S. Cooperman, MD)

Fig. 3.10.10  a Axial T1-weighted image shows a soft tissue mass (white arrowheads) in the left maxillary sinus with posterior extension through the posterior wall of the sinus into the retroantral fat (black arrows) and pterygopalatine fossa (PPF). The contralateral retroantral fat (RAF) and PPF are normal. The pterygoid body and plates (P) are sclerotic. Slightly more

heterogeneous appearing post-obstructive secretions are noted anterior to the mass in the maxillary sinus. b The tumor is enhancing on post-contrast axial T1-weighted image with FS while the more anterior post-obstructive secretions (asterisk) do not. Invasive SCC was confirmed on histopathological examination of the surgical specimen

and PPF) and permeative rather than grossly destructive bone involvement should be considered indicative. The differential diagnosis of lesions that cause necrosis and midline destruction includes lymphoma, cocaine abuse, sarcoidosis, Wegener’s granulomatosis, syphilis, tuberculosis, leprosy, and fungal infection.

Eustachian tube orifice, and the lateral pharyngeal recess or the fossa of Rosenmuller (Fig. 3.10.14). The nasopharynx also contains lymphoid tissue (adenoids), minor salivary glands, and is lined by the pharyngobasilar fascia, and pharyngeal constrictor muscles. The pharyngobasilar fascia, which is the aponeurosis of the superior constrictor muscle, attaches it to the skull base and has a gap in its upper margin (the sinus of Morgagni) through which the distal Eustachian tube and levator palatini muscle pass on either side. The pharyngobasilar fascia is a natural barrier to lateral spread of nasopharyngeal tumors and serves to funnel tumors superiorly toward the skull base. The sinus of Morgagni however provides a path of least resistance for tumors to spread to the lateral skull base.

3.10.2.2  Nasopharynx 3.10.2.2.1  Normal Anatomy The boundaries of the nasopharynx are illustrated in Fig. 3.10.13. The nasopharynx serves as a conduit from the nasal cavity to the oropharynx and is therefore bounded anteriorly by the posterior choana, posterosuperiorly by the lower clivus, upper cervical spine, and prevertebral muscles, and inferiorly by a horizontal line drawn along the hard and soft palates. Landmarks of interest on the lateral wall of the nasopharynx include the torus tubarius,

3.10.4.2.2  Pathological Findings Some common lesions of the nasopharynx are listed in Table 3.10.2. Under normal circumstances, the lateral

489

490

3  Brain, Head, and Neck

Fig. 3.10.11  Coronal post-contrast T1-weighted image of the brain shows a large sinonasal tumor with gross intracranial extension. Note the peripheral cysts (white arrows) along the intra­ cranial component of the tumor, which is consistent with the diagnosis of esthesioneuroblastoma

Fig. 3.10.12  A large mucosal melanoma of the left nasal cavity is seen invading bilateral ethmoid air cells and the left maxillary sinus with left intraorbital extension (black arrows), and extension into the anterior cranial fossa through the skull base on this coronal T1-weighted image. Areas of hemorrhage or melanin appear as focal areas of high signal intensity (arrowheads)

Fig. 3.10.13  The anatomic soft tissue and bony relations of the nasopharynx (NP) are seen on this midline sagittal T1-weighted image. The inferior boundary of the NP is marked by the horizontal white line. Also shown are the sphenoid sinus (SS), pituitary gland in the sella turcica (white arrowhead), clivus (C), C2 vertebral body (C2), adenoids (A), soft palate (SP), and hard palate (HP)

Fig. 3.10.14  Important landmarks in the nasopharynx (NP) are shown on this axial T1-weighted image: Eustachian tube orifice (white arrow), torus tubarius (T), the lateral pharyngeal recess, or fossa of Rosenmuller (white arrowhead and the longus colli muscle (LC)

3.10  Head and Neck Table 3.10.2  Common lesions of the nasopharynx Benign

Adenoidal hypertrophy Post-inflammatory retention cyst Thornwaldt cyst Benign minor salivary gland tumors

Malignant

Nasopharyngeal carcinoma Non-Hodgkin’s lymphoma Malignant minor salivary gland tumors Rhabdomyosarcoma (pediatric age group)

pharyngeal recesses may appear asymmetrical due to mucosal coaptation giving the appearance of a pseudomass (Fig. 3.10.15). The most common benign condition is adenoidal hypertrophy (Fig. 3.10.16). It is commonly seen in children, young adults, chronic smokers, as well as in HIV (human immunodeficiency virus) positive patients. The incidental finding of asymmetric nasopharyngeal soft tissue in HIV positive patients or patients of southern Chinese descent should raise suspicion for neoplasm: lymphoma in the setting of HIV and nasopharyngeal carcinoma (NPC) in Chinese individuals. NPC is

Fig. 3.10.15  Axial T1-weighted image shows a normal right fossa of Rosenmuller (white arrow), which is well defined due to natural contrast provided by air. The left fossa is not well delineated and may give an impression of a mass, since the mucosal surfaces have collapsed against each other

the most frequent malignant lesion of the nasopharynx, and Fig. 3.10.17 illustrates its most common features. Standard treatment for NPC consists of chemoradiation therapy, and imaging plays a crucial role in planning radiation therapy. The finding of unilateral middle ear or mastoid fluid in an adult should alert the radiologist to assess the nasopharynx for obstruction of the Eustachian tube orifice from a nasopharyngeal tumor. Patients with NPC may present with asymptomatic but bulky cervical lymphadenopathy. The primary NPC may be missed if the nasopharynx is not carefully examined on imaging. In assessing NPC, the radiologist should look for asymmetric mass, skull base invasion (using sagittal T1weighted image to assess invasion of the clivus), widening of the petroclival fissure, parapharyngeal space invasion, lateral retropharyngeal lymphadenopathy, pterygopalatine fossa invasion, masticator space invasion, cavernous sinus or cranial nerve involvement (using post-contrast T1-weighted image with particular attention to V2 and V3), and cervical lymphadenopathy. MRI is the imaging modality of choice for NPC. It is superior to CT in the assessment of invasion of the clivus, perineural spread, and cavernous sinus invasion. Marrow invasion of the clivus is detected much earlier on MRI than on CT. Complications of treatment of NPC include radiation necrosis of the temporal lobes, osteoradionecrosis of the skull base, and cranial neuritis. These condi-

Fig. 3.10.16  Adenoidal hypertrophy (A) and prominent bilateral retropharyngeal lymph nodes (N) are commonly seen in young children and adolescents as seen in this axial T2-weighted image

491

492

3  Brain, Head, and Neck

Fig. 3.10.17  a Axial post-contrast T1-weighted image with fat saturation shows a small nasopharyngeal carcinoma (white arrowheads) within the right fossa of Rosenmuller, distinctive from the normal adjacent linear enhancing mucosa and venous plexus. b With larger NPC, there is effacement of adjacent fat planes, erosion of the left petrous bone (the normal fatty marrow of the right petrous bone is marked (P) and replacement of the normal fatty marrow signal within the clivus (C). c A more

cephalad image from the patient in b shows tumor infiltrating the left cavernous sinus (arrowheads) with encasement and mild narrowing of the cavernous portion of the left internal carotid artery. d Coronal post-contrast T1-weighted image with FS in a third patient with NPC who had numbness in the distribution of V3 shows obvious perineural spread of tumor into the middle cranial fossa (arrowheads). The right foramen ovale (FO) is widened compared with the normal opposite side (white arrow)

3.10  Head and Neck

tions need to be differentiated from recurrent tumor on imaging studies (Fig. 3.10.18). Post-treatment surveillance of NPC can be difficult using MRI due to treatment-related changes. FDG-PET scan can be especially helpful in this situation if there are borderline suspicious structural abnormalities on MRI and/or physical examination. FDG-PET scan is also useful for assessing distant metastases which may be particularly relevant in patients with high-volume cervical nodal disease. Recurrence of NPC in a lateral retropharyngeal node may be evident on post-treatment surveillance imaging, and tissue diagnosis is often required prior to re-treatment. These nodes are difficult to access surgically, especially in the post-chemoradiated neck, and CT-guided biopsy can easily provide tissue for histologic diagnosis (Fig. 3.10.19). 3.10.2.3  Oropharynx 3.10.2.3.1  Normal Anatomy The division between the oral cavity and the oropharynx is artificial, with the oropharynx bounded anteriorly by the circumvallate papillae of the tongue, the soft palate, and the anterior tonsillar pillars (Fig. 3.10.20). It extends inferiorly to the plane of the hyoid bone and is separated from the larynx by the epiglottis and glossoepiglottic fold, and from the hypopharynx by the pharyngoepiglottic folds. The superior and middle constrictor muscles constitute the musculature of the oropharynx and its contents also include the palatine and lingual tonsils, and minor salivary glands. 3.10.2.3.2  Pathological Findings Table 3.10.3 lists some common lesions of the oropharynx. Asymmetry in the palatine tonsils may be simply due to prior unilateral tonsillectomy causing the remaining contralateral tonsil to look like a mass. Lymphoid hyperplasia of the palatine or lingual tonsils is not uncommon and it can mimic SCC if it is asymmetrical. However, unlike SCC there is no invasive or infiltrative component (Fig. 3.10.21). Tonsillar hypertrophy is commonly seen in children as are tonsillar inflammatory processes such as peritonsillar abscess. These conditions are most often infectious in nature, but the radiologist should be aware of predisposing factors such as an unsuspected foreign body. It is equally important to evaluate for sequelae such as septic thrombophlebitis of the jugular vein. Infectious and inflammatory processes are best evaluated by CT since it more easily identifies calcification, foreign bodies, and gas within soft tissues. Although a rare condition, it is important to recognize a lingual thyroid on imaging since it may be the only

Fig. 3.10.18  a Coronal post-contrast T1-weighted image with FS shows irregular enhancement (white arrows) in bilateral inferior temporal lobes, which is the typical appearance of radiation necrosis following radiation therapy for NPC. b The same patient also has linear enhancement within the left optic nerve (arrowheads) from radiation-induced optic neuritis. The patient did not develop recurrent carcinoma on follow-up

Table 3.10.3  Common lesions of the oropharynx Benign

Lingual or palatine tonsillar hypertrophy Tonsillar/peritonsillar abscess Lingual thyroid Post-inflammatory retention cyst Dystrophic calcification (“tonsillolith”) Benign minor salivary gland tumors

Malignant

Squamous cell carcinoma Non-Hodgkin’s lymphoma Malignant minor salivary gland tumors

493

494

3  Brain, Head, and Neck

Fig. 3.10.19  CT guided FNA is a minimally invasive option for tissue diagnosis of an abnormal lateral retropharyngeal lymph node (LRPLN). a Axial T1-weighted image shows an abnormally enlarged right LRPLN (arrow) in this patient who had previously undergone right neck dissection followed by chemoradiation therapy for metastatic SCC of the neck from an “occult primary.” Metastatic papillary carcinoma of the thyroid was also found in the neck dissection specimen. b Contrast-enhanced sinus CT performed for biopsy planning shows coarse calcification in the abnormally enlarged right LRPLN. Calcification in lymph nodes is commonly seen in metastatic papillary thy-

roid carcinoma, but could represent post-treatment change in this patient who had received chemoradiation therapy. Tissue diagnosis of the abnormal right LRLN was necessary in order to differentiate between the two entities. The right internal carotid artery (arrowhead) is located immediately laterally. c CTguided FNA was performed via a right buccal space approach. This image shows the needle trajectory avoiding the right internal carotid artery and the tip of the needle within the parenchyma of the abnormal right LRPLN. Cytologic examination revealed metastatic papillary thyroid carcinoma

functioning thyroid tissue in the body and surgical resection could render the patient hypothyroid. These lesions are hyperdense on non-contrast CT due to normal iodine content, densely enhance, and appear as a well-defined midline soft tissue mass at the foramen cecum. The remainder of the neck should be closely examined for thyroid tissue along the normal course of descent of the thyroid gland, and a radioactive 123I scan can be diagnostic. SCC is the most common malignant neoplasm of the oropharynx. Patients may be asymptomatic at presentation and the radiologic findings may be very subtle. The patient may present with obvious metastatic cervical lymphadenopathy and an “unknown” or occult primary which may not be evident even after careful clinical examination (Fig. 3.10.22). The tonsil and the base of tongue are common sites for occult primary lesions especially in patients who have level II cervical lymphadenopathy. While CT or MR may demonstrate a primary lesion in these patients, FDG PET is also useful in identifying clinically occult primary lesions. More commonly, SCC of the tonsil and base of tongue presents as a bulky exophytic and/or infiltrative lesion (Fig. 3.10.23) that is often poorly marginated with infiltration of adjacent fat planes. On MR imaging, these lesions tend to be of intermediate signal intensity on T1-weighted image and T2-weighted image sequences with moderate enhancement. The most common pathway for metastasis from these tumors is to the regional lymph nodes. While level II lymph nodes are

the most frequently involved, the lateral retropharyngeal nodes are also at risk especially from tonsil SCC. Locally advanced SCC of the tonsil can invade the masticator space so that V3 is at risk for perineural invasion. SCC of the base of tongue can invade the sublingual space, where it can gain access to the lingual, glossopharyngeal, and/or hypoglossal nerves. Tumors of the soft palate can spread along the lesser and greater palatine nerves to the pterygopalatine fossa, from where they can extend intracranially along the vidian canal and foramen rotundum. Perineural spread of tumor is a particular characteristic of minor salivary gland tumors, especially adenoid cystic carcinoma. Minor salivary gland tumors generally present as a well-circumscribed submucosal mass, and the base of tongue is the most common site in the oropharynx. MR imaging is particularly good at demonstrating the interface between normal soft tissue and tumor. This feature is helpful in delineating the extent of base of tongue cancer, but the quality of imaging may be significantly degraded if the patient has a bulky tumor causing pooling of secretions and therefore marked swallowing artifact. In this scenario, contrast-enhanced CT is an excellent alternative. MRI is superior in the evaluation of perineural spread as discussed above, and should be used not only to investigate clinical symptoms but also to detect subclinical perineural spread for tumors at certain locations such as the soft palate.

3.10  Head and Neck

Fig. 3.10.21  Axial FSE T2-weighted image with FS shows asymmetrical soft tissue at the right base of tongue (arrowheads). Unlike lymphoma or SCC there is no obvious invasive component to this soft tissue, and histopathological examination of the surgical resection specimen did not show malignancy confirming the diagnosis of lymphoid hypertrophy

Fig. 3.10.20  Normal anatomy of the oropharynx. a The superior and inferior limits of the oropharynx are shown on this sagittal T1-weighted image. The vallecula (V) and the base of tongue (BOT) are labeled. b The normal palatine tonsils (T) are intermediate signal on axial FSE T2-weighted image with FS as is other lymphoid tissue, compared to muscle (Mp medial pterygoid muscle)

3.10.2.4  Oral Cavity 3.10.2.4.1  Normal Anatomy The anatomic boundaries of the oral cavity include the hard palate, maxillary alveolar ridge, and maxillary teeth superiorly, the cheek laterally and the circumvallate pa-

Fig. 3.10.22  This patient presented with metastatic cervical lymphadenopathy (N) from SCC but the primary tumor was not evident on clinical examination. An axial post-contrast T1-weighted image with FS shows a subtle, infiltrative lesion at the right base of tongue (arrowheads), which was found to be SCC of the base of tongue

495

496

3  Brain, Head, and Neck

pillae and anterior tonsillar pillars posteriorly. At its posterior border the oral cavity blends into the oropharynx. The mucosal surfaces of the oral cavity are designated buccal (cheek), labial (lip), gingival (gum), lingual, sublingual, and palatal. However, for anatomic purposes, the oral cavity is divided into specific subsites that are de-

Fig. 3.10.23  Axial post-contrast T1-weighted image with FS shows a bulky lesion of the left tonsillar fossa (white arrowheads) that was consistent with SCC on biopsy. The tumor abuts but does not invade the adjacent base of tongue

picted in Fig. 3.10.24. Squamous epithelium lines the oral cavity mucosa; however, minor salivary gland rests are scattered throughout the oral cavity and are most concentrated in the hard palate. The mucosal surfaces are easily examined by direct clinical inspection and bimanual palpation. Deep-seated pathology within the musculature of the tongue can be better delineated on radiologic imaging (Fig. 3.10.25). Two additional spaces in relation to the oral cavity include the sublingual space and the submandibular space with the mylohyoid muscle, the muscular supportive floor of the oral cavity, separating these two spaces. The sublingual space is a bowl-shaped potential space without fascial boundaries that is located above the mylohyoid muscle. It is artificially divided into two halves by the genioglossus–geniohyoid complex and is bounded anteriorly by the mandible. This potential space freely communicates over the posterior free margin of the mylohyoid muscle with the submandibular space as well as the inferior parapharyngeal space. The contents of the sublingual space include the paired sublingual salivary glands and their ducts (the ducts of Rivinus), the deep portion of the paired submandibular glands and their ducts (Wharton’s ducts), the lingual artery and nerve and the distal glossopharyngeal and hypoglossal nerves. The submandibular space is located inferolateral to the mylohyoid muscle and superior to the hyoid bone. Although it is defined in part by the superficial layer of deep cervical fascia, it also freely communicates with the sublingual and inferior parapharyngeal spaces. The contents of the submandibular space include the paired submandibular glands, facial artery and vein, level I lymph nodes, and

Fig. 3.10.24  Anatomic subsites of the oral cavity: 1 oral tongue, 2 floor of the mouth, 3 lower alveolus, 4 retromolar trigone, 5 upper alveolus, 6 hard palate, 7 buccal mucosa

3.10  Head and Neck

Fig. 3.10.25  Imaging anatomy pertinent to the oral cavity on sagittal T1-weighted image: G genu of the mandible, H hyoid bone, extrinsic muscles of the tongue—genioglossus (GG) and geniohyoid (GH), and intrinsic muscles—longitudinal (long) and transverse (trv) fibers, and the soft palate (SP)

the hypoglossal nerve. Similar pathologic processes are seen in both the sublingual and submandibular spaces. However, the surgical approach for resection or drainage is different and therefore accurate location of the lesion is imperative. 3.10.2.4.2  Pathological Findings The most common lesions of the oral cavity and its associated spaces are listed in Table 3.10.4. Infectious processes commonly extend into the oral cavity from odontogenic sources. It may be difficult to distinguish cellulitis and phlegmon from abscess. In general an abscess is characterized by a well-defined enhancing rim, a non-enhancing pus-filled center, and by mass effect on local tissues while cellulitis tends to infiltrate along fascial planes and obscures them. Other benign lesions include dermoid/ epidermoid and venolymphatic malformations. Injury or a lesion that affects cranial nerve XII causing acute and subacute denervation can be a source of confusion since the acutely denervated hemitongue appears swollen and bright on T2-weighted image and enhances with gadolinium contrast (Fig. 3.10.26). Unlike most tongue

Fig. 3.10.26  a Axial FSE T2-weighted image with FS for investigation of a right hypoglossal nerve palsy shows marked hyperintensity of the right hemitongue (arrows) limited by the lingual septum, which is deviated toward the side of the abnormality. In contrast, a true mass of the tongue would push the lingual septum to the opposite side. b Axial post-contrast T1-weighted image with FS shows homogeneous enhancement of the right hemitongue that is prolapsed into the oropharynx. The tonsil (T) is labeled. These are the classic imaging findings of acute/ subacute denervation atrophy. The etiology in this patient was right skull base metastasis involving the hypoglossal canal (not shown)

497

498

3  Brain, Head, and Neck

neoplasms, denervation atrophy is restricted to the hemitongue and the process does not transgress the midline lingual septum. These changes are also associated with other signs of XII nerve paralysis such as deviation of the septum toward the pseudomass and ptosis of the atrophic hemitongue into the oropharynx. Benign tumors of the minor salivary glands are infrequently encountered, and the hard palate is the most common site involved. Clinical and radiologic differential diagnosis from malignant tumors of minor salivary gland origin can be particularly difficult, and precise diagnosis requires histopathological confirmation. A similar range of pathology affects the sublingual and submandibular spaces. Calculus disease, stenosis, or obstruction by tumor can cause dilation of the submandibular salivary duct (Fig. 3.10.27). A ranula is usually restricted to the sublingual space but extension into the submandibular space can occur to produce a “diving” ranula (Fig. 3.10.28). Malignant soft tissue tumors of the substance of the tongue are rare and can involve the sublingual space. However, the sublingual space is more commonly involved by deep extension of SCC from the mucosa of the floor of mouth or tongue (Fig. 3.10.29) or from deep anterior extension of a tongue base tumor. This finding

is important to elicit since it can have important staging and treatment implications. In addition to these lesions, lymphadenopathy constitutes the most commonly encountered pathology in the submandibular space. Lymph nodes in this space are at particular risk for metastasis from primary SCC of the oral cavity. Pleomorphic adenoma is the most common benign primary tumor of the submandibular salivary gland but its radiologic characteristics may not allow reliable distinction from malignant lesions. Evaluation for perineural spread becomes an important consideration for malignant tumors of the submandibular gland, especially adenoid cystic carcinoma. Squamous cell carcinoma is the most common malignant tumor of the oral cavity. Mucosal lesions of the oral cavity are easily amenable to clinical examination but imaging studies are useful in assessing their deep extent, submucosal spread, bone invasion, perineural spread and neurovascular bundle encasement (Fig. 3.10.30). It may be difficult to delineate a tumor of the buccal mucosa from adjacent alveolar ridge and teeth on radiologic imaging, and a useful maneuver is to have the patient puff out their cheeks while the scan is being acquired (Fig. 3.10.31). Tumors of the

Table 3.10.4  Common lesions of the oral cavity

Fig. 3.10.27  Axial contrast-enhanced CT shows a left anterior floor of mouth SCC (arrow) causing obstruction of left Wharton’s duct as evidenced by ductal dilatation (arrowhead)

Congenital/ developmental

Hemangioma Venolymphatic malformation Dermoid/epidermoid Second branchial cleft cyst (submandibular space)

Infectious/ inflammatory

Cellulitis/Ludwig’s angina Abscess Dilated submandibular (Wharton’s) duct Ranula, simple or diving

Neoplastic, benign

Pleomorphic adenoma Other benign lesions of minor salivary origin Lipoma

Neoplastic, malignant

Squamous cell carcinoma Malignant neoplasm of minor salivary origin SCC nodal metastases (submandibular space) Lymphoma   (submandibular space)

3.10  Head and Neck

Fig. 3.10.28  a Coronal T1-weighted image shows a welldefined cystic lesion (C) in the left submandibular space, medial to the left submandibular gland. b Axial T2-weighted image shows a homogenously bright lesion in the left submandibular space with a collapsed “tail” (arrowhead) in the sublingual space consistent with a “diving” ranula (Images courtesy of Michelle S. Bradbury, MD, PhD)

retromolar trigone mucosa should be evaluated for invasion of the underlying ascending ramus of the mandible and the pterygomandibular raphe (Fig. 3.10.32). In the evaluation of tumors of the tongue, features such as the relationship of the tumor to the neurovascular bundles (Fig. 3.10.33), extension across the midline septum or into the base of tongue have the potential for influencing therapy. A tumor of the floor of the mouth or ventral tongue can produce obstruction of Wharton’s duct result-

ing in an enlarged submandibular salivary gland which may mimic nodal metastasis on physical examination. Radiologic imaging can easily differentiate between these two situations but more importantly, can also identify direct extension of tumor through the mylohyoid muscle into the submandibular space (Fig. 3.10.34). MR is generally preferred as the study of choice in imaging neoplasms of the oral cavity as it provides better soft tissue contrast (Fig. 3.10.35) and is less susceptible

Fig. 3.10.29  SCC of the anterior FOM (T) with invasion of the right sublingual space fat is easily identified on pre-contrast T1-weighted image. The normal fat in the left sublingual space (SLS) and the mandible (M) are labeled

Fig. 3.10.30  Perineural spread from SCC of the right gingivobuccal sulcus along the right inferior alveolar nerve is seen as abnormal thickened soft tissue within a widened alveolar canal (arrows) on axial T1-weighted image. The contralateral normal alveolar canal (arrowhead) is labeled for comparison

499

500

3  Brain, Head, and Neck

to dental artifact compared to CT. CT does however have the advantage of better delineation of cortical bone invasion from adjacent mucosal tumor and also is the preferred modality for assessing calculus disease and inflammatory pathology (Fig. 3.10.36). 3.10.2.5  Hypopharynx 3.10.2.5.1  Normal Anatomy The hypopharynx is the inferior continuation of the pharyngeal mucosal space from the oropharynx. Anatomically it begins at the level of the hyoid bone and ends at the lower border of the cricoid cartilage. The pyriform sinuses, the postcricoid area, and the posterior pharyngeal wall are the three major subsites of the hypopharynx (Fig. 3.10.37). The pyriform sinuses are bilateral, inverted pyramid-shaped spaces located on either side of the larynx with an open posterior wall. Superiorly, its medial wall is bordered by the aryepiglottic fold which is conventionally classified as part of the supraglottic larynx. Precise classification of tumors located in this “marginal zone” is often difficult in terms of their origin from the supraglottic larynx versus the hypopharynx, but the distinction is important to make. Although the most common malignant tumor at both sites is SCC, hypopharyngeal SCC has a significantly worse prognosis than SCC of the supraglottic larynx. The medial and lateral walls of the pyriform sinus converge onto each other anteriorly,

Fig. 3.10.31  “Puffed-cheek” maneuver on axial post-contrast T1-weighted image with FS makes it easier to delineate a tumor of the right buccal mucosa (T) because of the contrast provided by air between the tumor and the teeth

Fig. 3.10.32  a Anatomic depiction of the pterygomandibular raphe (R) which extends from the medial pterygoid plate to the medial aspect of the mandible. The buccinator (B) and superior constrictor (S) muscles interdigitate along the raphe, which lies in close proximity to the mucosa of the retromolar trigone. b Axial contrast-enhanced CT of the oral cavity shows a right retromolar trigone SCC (arrow) spreading anteriorly along the pterygomandibular raphe into the right buccal space and invading the posterior aspect of the buccinator muscle. Medially, the tumor has also involved the fibers of the right superior constrictor and medial pterygoid muscles

3.10  Head and Neck

and its apex is located at the level of the cricoarytenoid joint. The ala of the thyroid cartilage lines the lateral wall of the pyriform sinus, and its mucosa blends into that of the lateral pharyngeal wall which in turn blends with the posterior pharyngeal wall without any distinguishing clinical or radiologic landmarks. The lamina of the cricoid cartilage underlies the mucosa of the medial wall and the postcricoid region. The lateral and posterior pharyngeal walls of the hypopharynx merge imperceptibly with the oropharynx and the plane of the vallecula is generally

Fig. 3.10.33  This large tumor of the tongue clearly invades the right neurovascular bundle (arrowhead) in the sublingual space, while the contralateral neurovascular bundle is uninvolved (arrow) but in close proximity to the tumor. Surgical resection of this tumor with adequate margins would require sacrifice of both neurovascular bundles, and therefore total glossectomy

Fig. 3.10.34  Coronal post-contrast T1-weighted image shows a right oral tongue tumor that extends through the right mylohyoid muscle into the submandibular space (arrow), with direct invasion of the submandibular gland. The normal contralateral mylohyoid muscle (arrowhead) is labeled for comparison

Fig. 3.10.35  Beam hardening artifact from dental amalgam on this axial contrast-enhanced CT (a) decreases conspicuity of a squamous cell carcinoma in the left lateral oral tongue while axial FSE T2-weighted image b more clearly delineates the tumor (arrowheads) relative to tongue musculature

501

502

3  Brain, Head, and Neck

used as the demarcating line. Inferiorly, the hypopharynx continues into the cervical esophagus which is guarded by the cricopharyngeus muscle. The hypopharynx is lined by squamous mucosa on the inside and the inferior constrictor muscles on the outside, and contains minor salivary glands in its submucosa.

3.10.2.5.2  Pathological Findings Although other pathology such as retention cysts, minor salivary gland tumors, and extranodal lymphoma can occur, the most common tumors are squamous cell carcinoma. Most SCC arise from the pyriform sinus and patients commonly present with locally advanced lesions and nodal metastases because symptoms occur late with this disease (Fig. 3.10.38). Conversely, a small primary tumor of the pyriform sinus may evade clinical detection and present with nodal metastases from an “unknown primary.” Good clinical examination should be able to identify such a lesion, but it is imperative that the radiolo­gist surveys the pyriform sinus mucosa for early asymmetry (Fig. 3.10.39). The mucosa of the pyriform sinus is often coapted and may not be easily demonstrated unless the patient is instructed to perform the Valsalva maneuver. The relationship of a pyriform sinus tumor to the cricoid cartilage is important to define since involvement of the cartilage has important treatment implications. The cricoid is the only complete ring in the upper airway, and violation of the integrity of this structure precludes the possibility of conservation laryngeal surgical resection. The risk of cricoid invasion is obviously high with large tumors of the pyriform sinus and postcricoid mucosa, but even smaller tumors that arise from the apex of the pyriform sinus can invade the upper border of the cricoid by virtue of anatomic proximity (Fig. 3.10.40). Radiologic delineation of lesions of the posterior pharyngeal wall may not be easy, especially if they involve both sides of the midline symmetrically. Involvement of the prevertebral fascia and musculature by tumor is important to recognize for staging and therapeutic decision-making. In addition to cervical lymph nodes at levels II–IV, the retropharyngeal nodes are at risk for metastasis from hypo­pharyngeal carcinoma (Fig. 3.10.41). 3.10.2.6  Larynx 3.10.2.6.1  Normal Anatomy

Fig. 3.10.36  a Axial T1-weighted image obtained for investigation of tenderness in the right submandibular region shows a subtle ovoid low signal intensity structure in the right anterior floor of mouth (white arrow). b A calculus (arrow) obstructing and causing dilatation of the right submandibular duct (arrowheads) is easily identified on axial contrast-enhanced CT scan. There are inflammatory changes in the adjoining soft tissues including thickening of the platysma (P) and stranding of the subcutaneous fat (SQ fat) consistent with cellulitis

The larynx plays a crucial role in voice production, swallowing, and breathing but its fundamental function is protection of the airway. The integrity of the larynx and hence that of the airway is maintained by its supporting framework of cartilage, especially the cricoid cartilage. Its remaining components such as the ligaments, muscles, and mucosa work in a coordinated fashion to enable the larynx to perform its other functions. The larynx is anatomically divided into the supraglottis, glottis, and subglottis for descriptive purposes (Fig. 3.10.42). The supraglottis extends from the tip of the epiglottis to the level of the laryngeal ventricles and includes the laryngeal surface of the epiglottis, aryepiglottic folds, arytenoid and interarytenoid mucosa, laryngeal ventricles and the false

3.10  Head and Neck

Fig. 3.10.37  Radiologic anatomy of the normal hypopharynx. a The hypopharynx (white lines) extends from the level of the vallecula (V) and hyoid bone (H) superiorly to the inferior border of the cricoid cartilage on sagittal T1-weighted image. b The right aryepiglottic fold (arrow) and pyriform sinuses (P) are shown on

axial T1-weighted image. c Coronal T1-weighted image demonstrates the relationship of the pyriform sinus (P) to the epiglottis (E), aryepiglottic fold (white arrow), and thyroid cartilage (black arrow). The submandibular salivary gland (smg), trachea (Tr), tonsil (T), and soft palate (S) are labeled for orientation

vocal folds. The preepiglottic space is located anterior to the epiglottis and is normally filled with fat. It is defined superiorly by the hyoepiglottic ligament, which separates it from the vallecula. On its lateral aspects, it is continuous with and communicates on either side with the paraglottic spaces that lie laterally to the false vocal folds and the true cords. The glottis includes the true vocal cords, and the anterior and posterior commissures. The true

Fig. 3.10.38  Axial post-contrast T1-weighted image with FS in a patient who presented with bulky metastatic lymphadenopathy in the left neck (nodal mass) from a small primary lesion in the left pyriform sinus (arrowheads). Note the appearance of the right pyriform sinus (P)

Fig. 3.10.39  Axial contrast-enhanced CT scan shows abnormal thickening and enhancement of the left pyriform sinus, the socalled reverse-C configuration (arrow) consistent with earlystage left pyriform sinus carcinoma. Note that the normal mucosa of the right pyriform sinus enhances linearly and is pencil-thin

503

504

3  Brain, Head, and Neck

vocal cord is composed of mucosa, vocalis muscle, and the thyroarytenoid ligament and muscle. On its undersurface, the mucosa of the true cord continues into that of the subglottis which ends at the level of the inferior border of the cricoid cartilage. The mucosa of this part of the larynx has a high density of minor salivary glands. The laryngeal framework consists of three paired cartilages (the arytenoids, cuneiform, and corniculate) and three unpaired cartilages (the thyroid, cricoid, and epiglottic cartilages). The epiglottis contains elastic cartilage while the thyroid and cricoid are composed of hyaline cartilage that begins to ossify in early adulthood. The cricoid cartilage is the only complete ring in the framework of the airway. Its superior border slopes anteriorly because of a narrow arch anteriorly and a wider lamina posteriorly. The paired arytenoid cartilages form the cricoarytenoid joint on the superior aspect of the posterior cricoid lamina and their vocal processes are attached to the posterior margins of the vocal cords. A more detailed description of laryngeal anatomy in regards to its cartilage framework, ligaments and membranes is beyond the

scope of this section, but is well described in surgical anatomical texts. The supraglottic and subglottic larynx have a rich lymphatic network. The supraglottic lymphatics exit the larynx bilaterally through the thyrohyoid membrane and drain into the levels II and III lymph nodes. Lymphatics from the subglottis exit through the cricothyroid membrane and drain to the levels III and IV lymph nodes via the paratracheal and tracheoesophageal nodes. In contrast, the true vocal cords have a very sparse lymphatic supply and therefore the risk of nodal metastases from early glottic tumors is usually low. The posterior cricoarytenoid muscle is the only true abductor of the true vocal cords, and is supplied by the recurrent laryngeal nerve. All the remaining intrinsic laryngeal musculature adduct or tense the true cords, and are supplied by the internal branch of the superior laryngeal nerve. 3.10.2.6.2  Pathological Findings Some common lesions of the larynx are listed in Table 3.10.5. Vocal cord paralysis may be suspected radiologically if the true cord is paramedian in location and there is accompanying ipsilateral medial rotation of the arytenoid cartilage, dilation of the pyriform sinus, the laryngeal ventricle or vallecula. Laryngeal imaging is best performed during quiet respiration so that the true cords are relaxed and abducted. Assessment for paralysis or mechanical fixation can be done with the patient holding their breath which would adduct the normal cord. However, it is important to realize that vocal cord paralysis is easily diagnosed on clinical examination so that radiologic studies are generally ordered to investigate its cause

Table 3.10.5  Common lesions of the larynx Congenital/ developmental

Hemangioma Venolymphatic malformation

Trauma

Hematoma Fracture of laryngeal cartilage

Functional

Laryngocele

Neoplastic Mucosal Fig. 3.10.40  The superior border of the cricoid cartilage (C) lies in close proximity to the pyriform apex, so that the cricoid cartilage is at risk for involvement from a tumor (T) involving the pyriform apex as seen in this coronal T1-weighted image



Cartilaginous

Squamous cell carcinoma Malignant minor salivary gland tumors Chondroma Chondrosarcoma Metastatic disease

3.10  Head and Neck

Fig. 3.10.41  a Axial FSE T2-weighted image with FS shows a right pyriform sinus SCC with bilateral metastatic lymphadenopathy (N). The vallecula (V), epiglottis (E), and submandibular gland (smg) are labeled for orientation. b A pathologic right

lateral retropharyngeal lymph node (R) is present on a more cephalad slice. The normal left lateral retropharyngeal lymph node (white arrow) is labeled for comparison

and not for its detection. When imaging is indicated in the investigation of hoarseness, the clinical picture is useful in determining the modality used. Clinical evidence of multiple lower cranial nerve dysfunctions is indicative of a lesion at the skull base or in the carotid sheath, and MR is the preferred modality (Fig. 3.10.43). Isolated hoarseness, on the other hand, points to involvement of the vagus or the recurrent laryngeal nerve. In these instances, it is important to trace the entire course of the tenth nerve from its nucleus in the brain stem, down into the mediastinum and neck. It is important to remember that the left recurrent laryngeal nerve loops around the aortic arch and is susceptible to injury by tumor or surgery in the aorticopulmonary window. The right recurrent nerve hooks around the right subclavian artery and may occasionally be non-recurrent, so that it is given off directly from the vagus nerve. An indirect indicator of a non-recurrent right recurrent laryngeal nerve is anomalous origin of the right subclavian artery. Laryngeal trauma is not uncommon and is best evaluated clinically and endoscopically. Imaging may however be useful when there is clinical suspicion for cartilage fracture or soft tissue injury. CT is generally the modality of choice. The cricoid is prone to fracture in two or more places since it is a complete ring, but the thyroid cartilage generally fractures along its anterior margin. Laryngocele may be internal or external—internal laryngocele is a dilatation of the laryngeal ventricle into the

paraglottic space, whereas an external laryngocele has an additional external component that penetrates the thyrohyoid membrane and can present as a neck mass. The contents of the laryngocele can include air or fluid, and it may become superinfected. The etiology of laryngocele is thought to be increased intraglottic pressure with or without anatomic obstruction of the laryngeal ventricle due to trauma, post-inflammatory scar, or neoplasm. It is important to rule out an obstructing neoplastic lesion in the larynx especially in an adult patient with a unilateral laryngocele (Fig. 3.10.44). Cartilaginous lesions of the larynx including chondroma and chondrosarcoma arise most commonly from the cricoid cartilage. On clinical examination, they can be confused for minor salivary gland tumors of the subglottis which are also smooth, diffuse, and submucosal in location. Radiologic imaging is very characteristic since cartilaginous lesions have extremely high signal intensity on T2-weighted image. Calcifications may be noted on CT scan and cinch the diagnosis. Malignant involvement of cartilage from laryngeal cancer is thought to occur more frequently in ossified areas. These areas of the laryngeal framework are also thought to be at increased risk of involvement from blood-borne metastases from distant sites and hematologic malignancy such as lymphoma, leukemia, and multiple myeloma. The most common primary malignant tumor of the larynx is squamous cell carcinoma. The mucosal surface

505

506

3  Brain, Head, and Neck

Fig. 3.10.42  Normal radiologic anatomy of the larynx. a Slightly oblique sagittal T1-weighted image shows the vocal ligaments (white arrows) that extend from the vocal process of the arytenoid (A) to the anterior thyroid cartilage, the cricoid cartilage (C), epiglottis (E), pre-epiglottic space (pes), hyoid bone (H), and vallecula (V). b Coronal T1-weighted image shows the fatty paraglottic space (F) lateral to the right false vocal fold, the air-filled right laryngeal ventricle (arrow) and the thyroarytenoid muscle (tam) of the true vocal cord. c Contrast-enhanced

axial CT scan at the level of the supraglottis shows the lower epiglottis (E), aryepiglottic folds (white arrows), pyriform sinus (P), hyoid bone (H), and pre-epiglottic space (pes). d Contrastenhanced axial CT scan at the level of the glottic larynx shows the laryngeal ventricle (V) and true vocal cord (white arrow). In a young patient such as this, the thyroid lamina (white arrowheads), cricoid cartilage (C), and arytenoid cartilage (A) appear subtle since they have not yet ossified

3.10  Head and Neck 9 Fig. 3.10.43  a Axial contrast-enhanced CT scan of the neck at the level of the true vocal cord in a patient who presented with a left vocal cord paralysis. The classic imaging features of vocal cord paralysis include medial rotation of the arytenoid cartilage (black arrow), adducted and fatty true vocal cord/thyroarytenoid muscle (asterisk), and dilatation of the ipsilateral laryngeal ventricle (V). In this patient, the left sternocleidomastoid (SCM) and trapezius (Tr) muscles were relatively atrophic suggestive of XI nerve dysfunction, and therefore suspicious for a skull base lesion. (b) Axial FSE T2-weighted image with FS shows a well defined soft tissue mass (white arrows) occupying the left jugular foramen which was a schwannoma. The normal right jugular bulb (JB) and medulla (Me) are labeled. c The lesion (white arrow) heterogeneously but densely enhances on coronal postcontrast T1-weighted image with FS. The right normal jugular bulb (white arrowheads) can be confused with a lesion

Fig. 3.10.44  a Axial contrast-enhanced CT scan shows an airfilled sac at the level of the left supraglottic larynx consistent with an internal laryngocele (lar) with posterior displacement of the left pyriform sinus (arrow). The hyoid bone (H) is labeled for orientation. b More inferiorly, tumor is seen involving both true vocal cords (V) with erosion (arrowheads) of the thyroid cartilage (T). The cricoid cartilage (C) is labeled

507

508

3  Brain, Head, and Neck

extent of most laryngeal tumors is more accurately assessed on endoscopic exam, but imaging is indicated for assessment of involvement of the laryngeal spaces, cartilage framework and regional lymph nodes since these features have the potential for altering treatment. Early stage glottic carcinomas are easily evaluated clinically and infrequently metastasize to cervical nodes. Therefore, there is no cost-effective role for imaging in the diagnosis or staging of early glottic carcinoma. Imaging is indicated in the staging and workup of all other laryngeal carcinoma. The preepiglottic and paraglottic spaces are at risk for invasion from supraglottic and locally advanced glottic carcinoma. These fat-filled spaces are contiguous, and there are no normal barriers to tumor spread from one space to the other. Restricted mobility or paralysis of the true vocal cord is a clinical sign of paraglottic invasion of tumor, but there are no reliable clinical indicators of preepiglottic space invasion. Subglottic extension of tumor can be detected endoscopically, but accurate clinical evaluation may be difficult in the presence of a bulky, transglottic tumor that has compromised the airway. The normal subglottic mucosa is not discernible on imaging since it is closely applied to the cricoid cartilage. The presence of radiologically evident soft tissue in this location should be suspicious for tumor extension. Imaging provides a noninvasive and reliable means for assessing the extent of subglottic disease in such instances. Invasion of the laryngeal cartilages also has staging and treatment implications. The only reliable imaging indicator of cartilage invasion is the presence of tumor on either side of the cartilage. Early invasion of the thyroid or cricoid cartilages is not easy to identify on imaging, and while MR is generally more sensitive than CT both modalities can misinterpret reactive changes as tumor infiltration. Thin section contrast-enhanced CT is generally the modality of choice in most patients because MR is susceptible to degradation from swallowing and motion artifact. Benign conditions of the larynx such as trauma and laryngocele are best imaged by CT, while MR is particularly sensitive for cartilage invasion by tumor and can be a useful adjunct to CT to clarify other specific issues. For instance, sagittal T1-weighted image can easily show infiltration of the preepiglottic fat by tumor which may not be easily distinguishable on CT. 3.10.2.7  Lymph Nodes 3.10.2.7.1  Normal Anatomy A detailed description of the lymphatic anatomy of the head and neck is not within the scope of this discussion, and is available elsewhere. The superficial and deep lymphatics of the head and neck have a well recognized drain-

age pattern which allows categorization of named lymphatic groups into a level based classification system for tumor metastasis (Fig. 3.10.45). The imaging correlates of this classification system are depicted in Table 3.10.6. Other nodal groups such as the facial, parotid, pre-, and post-auricular and suboccipital nodes are not included in this classification scheme since they do not represent the usual sites of lymphatic drainage from SCC of the upper aerodigestive tract (UADT). The lateral retropharyngeal nodes which are also not included in the level based classification scheme are important to examine. These nodes are at particular risk for nodal metastases from SCC at certain sites such as the tonsil and nasopharynx, but also from thyroid cancer. 3.10.2.7.2  Pathological Findings The importance of assessment of the cervical lymph nodes in the presence of head and neck cancer is obvious, but these nodes can also be affected by benign conditions (infections, inflammation, or granulomatous disease), hematogenous malignancy (lymphoma), or metastasis from an infraclavicular malignant primary tumor. Normal lymph nodes are ovoid in shape and of homo­ genous density with a short axis diameter between 5 and 10 mm. They have intermediate signal on T1-weighted image, are hyperintense to muscle on T2-weighted image and show mild to moderate homogeneous enhancement after gadolinium administration (Fig. 3.10.46). An eccentric fatty hilum may be identifiable on MR, but is more readily and consistently seen on CT. Abnormal lymph nodes are generally diagnosed by certain size criteria and physical characteristics such as round shape, “cystic change” from necrosis or extracapsular extension. Necrotic lymph nodes may be seen in infectious conditions such as tonsillitis, mycobacterial infection, and cat-scratch disease, but it is crucial to remember that malignant lymphadenopathy from SCC can also be necrotic (Fig. 3.10.47). Necrosis may also be seen as a treatment effect, particularly in lymphoma. Calcification in lymph nodes is best demonstrated on CT and is generally associated with old granulomatous disease, metastatic thyroid carcinoma, or irradiated nodal metastases. Extension of the disease process beyond the confines of the capsule of the lymph node is demonstrated earlier on CT compared to clinical exam and MRI. Necrotic cystic-appearing nodes may be associated with infection, but they are not uncommonly seen with certain cancers such as SCC of the tonsil or base of tongue, as well as thyroid carcinoma. As expected, necrotic foci in lymph nodes are low signal on T1-weighted image, high signal intensity on T2-weighted image and show peripheral enhancement. A solitary cystic mass at level II in the adult should be considered suspicious for nodal metastases and is commonly mistaken for a

3.10  Head and Neck

Fig. 3.10.45  Anatomic depiction of lymph node levels in the neck. See Table 3.10.6 for detailed description

branchial cleft cyst. A cystic metastatic lymph node typically has a thick, irregular, enhancing wall as opposed to a branchial cleft cyst, which should not enhance unless it is or has been infected. These features are by no means diagnostic and these two entities may be difficult to distinguish radiologically. The diagnosis is established by histology and it is important for the radiologist to be aware that the diagnosis of branchial cleft cyst in an adult is relatively rare, and the clinician should be alerted to the possibility of cystic nodal metastasis. Other less common variations in the appearance of metastatic lymph nodes include high-signal foci within a lymph node on T1-weighted image from vascular metastases such as thyroid cancer or renal cell carcinoma and low signal foci on T2-weighted image due to keratin pools in SCC. The presence of necrosis is an important indicator of metastatic involvement of a lymph node but is by no means reliable. Size criteria have therefore been developed in an effort to predict the risk for metastatic involvement, based on the principle that larger lymph nodes are more likely to be metastatic compared to smaller ones. The smallest axial diameter of the lymph node is used as the measurement for these criteria, and the most widely

accepted cut-off for metastatic disease is >10 mm. Sizebased classification schemes are a compromise between sensitivity and specificity, and depending on the cut-off chosen, there may be a significant incidence of false-positive and false-negative nodes. Current imaging techniques have a low accuracy in estimating risk of metastasis in small or normal sized lymph nodes. Newer modalities such as PET scan, iron oxide–enhanced MR lymphography, MR spectroscopy, and dynamic contrast-enhanced (DCE) MRI are under investigation and may improve our ability to diagnose and stage cervical nodal metastases. 3.10.3  Non-Mucosal Diseases of the Head and Neck The previous section described imaging for pathology of the conventional anatomic divisions of the mucosal surfaces of the head and neck which is designed for staging and evaluating squamous cell carcinoma. Pathologic processes of the remainder of the head and neck require a different approach in diagnosis and staging since they have a distinct biologic behavior.

509

510

3  Brain, Head, and Neck Table 3.10.6  Image-based classification of cervical lymph nodes Level/ subclassification

  Boundaries

Previous terminology

I

Above hyoid bone, below mylohyoid muscle, anterior to a transverse line drawn through the posterior edge of the SMG

Submental and submandibular nodes

IA

Between medial margins of anterior bellies of digastric muscles

Submental nodes

IB

Posterior and lateral to medial edge of anterior belly of digastric muscle, anterior to posterior edge of SMG

Submandibular nodes

II

From skull base to the lower body of the hyoid bone, anterior to the posterior edge of the SCM, and posterior to the posterior edge of the SMG a

Upper internal jugular and spinal accessory nodes

IIA

Lie anteriorly, laterally, or medially to the IJV, or lie posteriorly to the IJV and are inseparable from it

Upper internal jugular nodes

IIB

Lie posteriorly to the IJV and have a fat plane separating the nodes and the vein

Upper spinal accessory nodes

III

Between the lower body of the hyoid bone and the lower margin of the cricoid arch, anterior to the posterior edge of the SCM and lateral to the medial margin of the CCA or ICA

Mid jugular nodes

IV

Between the lower margin of the cricoid cartilage arch and the level of the clavicle, anterior and medial to an oblique line drawn between the posterior edge of the SCM and the posterolateral edge of the ­anterior scalene muscle, lateral to the medial margin of the CCA

Low jugular nodes

V

From the skull base at the posterior border of the attachment of the SCM to the clavicle, anterior to the anterior edge of the trapezius muscle and posterior to the posterior edge of the SCM (skull base to bottom of cricoid) or posterior and lateral to an oblique line through the posterior edge of the SCM and the posterolateral edge of the anterior scalene muscle (bottom of cricoid to clavicle)

Posterior cervical

VA

From skull base superiorly to lower margin of cricoid cartilage inferiorly

Upper posterior cervical

VB

From lower margin of cricoid cartilage to level of clavicle

Lower posterior cervical

VI

Inferior to body of hyoid, superior to top of manubrium, and between the medial margins of the ICAs or CCAs

Visceral nodes

VII

Caudal to the top of the manubrium in the superior mediastinum, between the medial margins of the left and right common carotid arteries and superior to the innominate vein

Superior mediastinal

SMG submandibular gland, SCM sternocleidomastoid muscle, IJV internal jugular vein, CCA common carotid artery, ICA internal carotid artery A node located within 2 cm of the skull base and medial to the internal carotid arteries is classified as a retropharyngeal node. Within 2 cm of the skull base but anterior, lateral, or posterior to the ICAs, it is classified as a level II node. More than 2 cm below the skull base, level II nodes can lie in any position relative to the internal jugular vein (Modified from Som PM, Curtin HD, Mancuso AA (2000) Imaging-based nodal classification for evaluation of neck metastatic adenopathy. AJR Am J Roentgenol 174:837. PMID: 10701636)

3.10  Head and Neck 9 Fig. 3.10.46  a A normal lymph node (arrows) is isointense to muscle on T1-weighted image. The internal carotid artery (C) and internal jugular vein (J) are labeled. b The node (arrows) is homogeneously high signal on axial FSE T2-weighted image with FS. c There is mild and homogeneous enhancement of the lymph node (arrows) post-contrast administration

Fig. 3.10.47  a Axial post-contrast T1-weighted image with FS shows a heterogeneously enhancing abnormally enlarged right level II lymph node (arrow). The low-signal region is very bright on T2-weighted image b, and this is consistent with necrotic nodal metastasis from SCC

511

512

3  Brain, Head, and Neck

3.10.3.1  Spatial Approach Unlike squamous cell carcinoma, other pathology in the neck tends to remain confined to defined spaces limited by the fascia of the neck. The neck can therefore be visualized as a collection of spaces and this permits a logical differential diagnosis for pathology in each of these fascial-bound spaces. This approach to cross-sectional imaging analysis is termed the spatial approach. Since the fascial attachments of the hyoid bone can be viewed to functionally separate the neck into two segments, the neck is conveniently divided into the suprahyoid and infrahyoid compartments for analytic purposes. 3.10.3.1.1  Suprahyoid Neck The superficial (investing), middle (buccopharyngeal), and deep (prevertebral) layers of the deep cervical fascia divide the suprahyoid neck into the spaces depicted in (Fig. 3.10.48). Pharyngeal Mucosal Space (PMS) The PMS is the lining of the nasopharynx and oropharynx on the luminal aspect of the middle layer of the deep cervical fascia and is therefore not entirely enclosed by the three layers of the deep cervical fascia. Squamous mucosa lines this space which also contains minor salivary

glands and lymphoid tissue of the Waldeyer’s ring, and the constrictor muscles form the outer sheath. Squamous cell carcinoma is the most common pathology of the PMS and has been described above. Other tumors such as lymphoma and sarcoma of the PMS are less common and should be considered in the differential diagnosis. Parapharyngeal Space (PPS) The PPS is an inverted pyramid-shaped fat-filled space in the suprahyoid neck that extends from the skull base to the level of the hyoid bone (Fig. 3.10.49). The PPS is surrounded by the PMS, parotid space (PS), masticator space (MS), and carotid space (CS) (Fig. 3.10.50). The PPS is contiguous with the submandibular and sublingual spaces at its inferior extent, which allows disease processes from one space to extend into the other. Relatively few pathologic processes are intrinsic to the PPS because of the limited contents of this space. However, its strategic location relative to the surrounding spaces helps discern the space of origin of lesions by recognizing displacement patterns of the fat-filled PPS (Fig. 3.10.51) and aids differential diagnosis of suprahyoid neck masses. In addition to fat, the PPS contains arteries, veins and minor salivary gland rests, and although rare, primary pathology of this space is generally limited to lipoma and minor salivary gland tumors. More frequently, lesions from adjoining spaces compress the PPS and appear to arise from it. These lesions can generally be distinguished

Fig. 3.10.48  Schematic diagram of crosssectional anatomy of the suprahyoid neck at the level of the nasopharynx to show the various spaces and their fascial boundaries. The deep layer of the deep cervical fascia (prevertebral fascia) is shown as a dashed line, the middle layer of the deep cervical fascia as a dotted line, and the superficial layer of the deep cervical fascia (investing fascia) as a thick solid line. The pharyngobasilar fascia outlines the PMS and is shown by the heavy solid line. PPS parapharyngeal space, MS masticator space, PS parotid space, CS carotid space, BS buccal space, RPS retropharyngeal space, PVS perivertebral space (Modified and used with permission from Harnsberger HR [1987] CT and MRI of masses of the deep face. Curr Probl Diagn Radiol 16:141. PMID: 3297507)

3.10  Head and Neck

from primary PPS lesions by displaced eccentric PPS fat in contrast to a circumference of fat that is generally identifiable around a primary PPS mass. Notable exceptions to this rule include large lesions, infectious processes such as phlegmon or abscess, or malignant lesions such as SCC or lymphoma (Fig. 3.10.52) that are locally invasive and do not respect fascial boundaries. Parotid Space (PS) The PS is enclosed within the split superficial layers of the deep cervical fascia and lies posterolateral to the PPS. It contains the parotid gland, facial nerve, and blood vessels. The parotid gland is the only major salivary gland that is encapsulated late in development, and therefore contains normal lymph nodes within its parenchyma. The facial nerve divides the parotid gland into a superficial and deep lobe. The majority of parotid tumors arise from its superficial lobe and distinction from deep lobe tumors is important for adequate surgical planning. Cross-sectional imaging typically cannot identify the normal facial nerve but the plane of the nerve after its exit from the stylomas-

toid foramen can be estimated because it lies laterally to the retromandibular vein which is readily identifiable on imaging. Stenson’s duct arises from the anterior border of the gland, crosses the MS lateral to the masseter muscle, goes through the buccinator muscle, and drains into the oral cavity on the buccal mucosa at the level of the second maxillary molar. Accessory parotid tissue is not uncommonly located along the duct in the MS and can give rise to a similar spectrum of pathology as the parotid gland (Table 3.10.7). Multiple unilateral or bilateral PS lesions should prompt consideration of reactive or metastatic lymphadenopathy, Warthin’s tumors, lymphoepithelial lesions, and recurrent pleomorphic adenoma. Hemangioma Hemangioma is a proliferative vascular mass that is generally seen early in life and can grow large enough to replace the entire parotid gland before it begins slowly involuting with the increasing age of the child. It appears as a multilobulated mass that may involve the entire gland and classically appears isointense to muscle on

Fig. 3.10.49  The PPS is an inverted pyramid-shaped fat-filled space in the suprahyoid neck that extends from the skull base to the level of the hyoid bone

513

514

3  Brain, Head, and Neck

T1-weighted image, hyperintense on T2-weighted image, with prominent flow voids, and intense homogeneous enhancement post-gadolinium (Fig. 3.10.53). Calcification is not easily identified on MR but is apparent on CT.

The external carotid artery and its branches may be enlarged. First Branchial Cleft Cyst Less than 10% of branchial complex anomalies arise from the first branchial apparatus, and include cysts and sinuses with or without fistula. The radiologic appearance is of a cystic mass within or adjacent to the parotid gland (Fig. 3.10.54) and a tract connecting the cyst to the external auditory canal may be identifiable in some cases. Active infection of the cyst can produce surrounding soft tissue changes, while past infection may have produced thickening of the cyst wall. Lymphoepithelial Lesions Benign lymphoepithelial lesions of the parotid gland are most commonly seen with HIV but can be associated with connective tissue disorders such as Sjögren’s syndrome. Parotid lymphoepithelial lesions of HIV are generally bilateral, can be cystic, solid, or mixed, and are typically associated with hypertrophy of the Waldeyer’s ring lymphoid tissue (Fig. 3.10.55) and reactive cervical lymphadenopathy. Parotitis/Calculus Disease Parotitis is characterized by parotid enlargement, edema, increased enhancement, and inflammatory stranding of

Table 3.10.7  Common lesions of the parotid space

Fig. 3.10.50  a Schematic representation of the strategic location of the parapharyngeal space (PPS) in relation to the other spaces around it. b Correlative axial T1-weighted image shows normal high signal fat in the PPS. The parotid space (PS) is labeled

Congenital/ developmental

Hemangioma Venolymphatic malformation First branchial cleft cyst

Inflammatory/ infectious

Parotitis/parotid abscess Reactive lymphadenopathy Lymphoepithelial cysts/lesions

Neoplasm, benign

Pleomorphic adenoma Warthin’s tumor Lipoma Facial nerve schwannomas Oncocytoma

Neoplasm,   malignant

Mucoepidermoid carcinoma Adenoid cystic carcinoma Acinic cell carcinoma Carcinoma ex pleomorphic adenoma Salivary ductal carcinoma Squamous cell carcinoma Extranodal or nodal nonHodgkin’s lymphoma Nodal metastases, especially SCC of face and scalp

3.10  Head and Neck

Fig. 3.10.51  The pattern of displacement of the fat-filled parapharyngeal space (PPS) relative to the surrounding spaces helps discern the space of origin of lesions and aids differential diagnosis of suprahyoid neck masses. a Tumor arising from the pharyngeal mucosal space (PMS) displaces the PPS fat postero-

laterally. b Masticator space (MS) tumors displace the PPS posteromedially. c Parotid space (PS) tumors displace the PPS fat anteromedially. d Carotid space (CS) tumors displace the PPS fat anterolaterally

515

516

3  Brain, Head, and Neck

Fig. 3.10.52  Axial T1-weighted image shows a left oropharyngeal PMS tumor (mass) that extends into and obliterates the PPS fat. This tumor does not respect fascial boundaries and turned out to be a lymphoma of the palatine tonsil. Note the normal right tonsil (T) and right parapharyngeal space (PPS). The medial pterygoid (MP) and masseter (Ma) muscles, and the parotid gland (P) are labeled

adjacent fat. Progression to abscess results in a rim-enhancing mass. Dilatation of the intra- or extra-parotid duct system is indicative of ductal obstruction but calculus is not easy to identify on MR, and may require thin section (1–3 mm) non-contrast CT. Pleomorphic Adenoma Pleomorphic adenoma is the most common neoplasm of the parotid and generally presents as a well-circumscribed, round, or ovoid mass. Cystic degeneration or mucoid matrix within the mass appears as high signal intensity on T2-weighted image (Fig. 3.10.56) with dense homogeneous contrast enhancement that is particularly prominent on delayed imaging. Malignant transformation (carcinoma ex pleomorphic adenoma) can occur in long-standing pleomorphic adenoma or after multiple recurrences. Pleomorphic adenoma can be FDG avid so that PET scan is not entirely reliable in differentiating it from malignancy. Warthin’s Tumor Warthin’s tumors (papillary cystadenoma lymphomatosum) are more commonly seen in older males and are not infrequently bilateral. They are most commonly located in the parotid tail and appear as a well-circumscribed,

Fig. 3.10.53  a Axial precontrast T1-weighted image shows a large well-circumscribed soft tissue mass that is isointense to muscle and replaces the entire right parotid gland with multiple linear flow voids (arrowheads). The normal parotid gland (P) in children is not as hyperintense on T1-weighted image as in adults because it is less fatty in the young patient. b Coronal T2-weighted image shows a uniform high signal except for the flow voids. This tumor was intensely homogeneously enhancing on post-contrast T1-weighted image (not shown), and these imaging features are consistent with hemangioma

3.10  Head and Neck

multilobulated, cystic, and/or solid mass (Fig. 3.10.57). They can also have areas of hemorrhage. A characteristic feature is intense activity on FDG-PET scan so that they may be mistaken for malignancy.

Fig. 3.10.54  Axial FSE T2-weighted image with FS shows a well defined, homogeneous, high signal lesion (arrowheads) in the left parotid gland (P). The lesion had features consistent with a cyst on additional sequences (not shown). The patient had drainage from the external ear and surgery confirmed the diagnosis of a first branchial cleft cyst

Fig. 3.10.55  This HIV-positive patient presented with bilateral parotid enlargement. Coronal T1-weighted image shows multiple cystic lesions (arrowheads) throughout both parotid glands consistent with lymphoepithelial cysts. Adenoidal hypertrophy (A) is commonly seen in HIV patients and help distinguish these parotid lesions from those seen in Sjögren’s syndrome

Parotid Malignant Tumors The most common malignant tumors of the parotid include mucoepidermoid carcinoma and adenoid cystic carcinoma, followed by other rarer types such as carcinoma ex pleomorphic adenoma, acinic cell carcinoma, polymorphous low-grade adenocarcinoma, and lymphoma. T1-weighted image outlines the intermediate signal of most parotid tumors against the fatty glandular parenchyma and is often the most reliable sequence for identification of a parotid mass (Fig. 3.10.58). As a general rule, malignant tumors tend to have slightly lower T2-weighted signal intensity than benign lesions, but the radiologic appearance of a low-grade parotid malignant tumor is virtually indistinguishable from that of more common benign tumors such as pleomorphic adenoma. Poor margination and invasion of adjacent structures such as fat, muscles of mastication and temporal bone, and perineural spread along the facial nerve (Fig. 3.10.59) are obvious signs of malignancy. Masticator Space (MS) The MS extends from the skull base to the inferior border of the mandible, and contains the ascending ramus and posterior body of the mandible with the muscles of mastication (masseter, temporalis, medial and lateral pterygoid), motor, and sensory branches of the third division of the trigeminal nerve, and the inferior alveolar vein and artery enclosed between the split layers of the superficial layer of the deep cervical fascia. The suprazygomatic part of the MS extends from just medial to the foramen ovale medially to the zygomatic arch laterally on to the lateral surface of the temporalis muscle (Fig. 3.10.60). The MS blends into the buccal space anteriorly while the PPS is located posteromedially. The most common lesions of the MS are infection or inflammation, usually of odontogenic origin (Table 3.10.8), and CT is the modality of choice should imaging be required. Benign hypertrophy of the masseter is associated with bruxism, and may be unilateral creating a diagnostic dilemma. As described above, accessory parotid tissue lies adjacent to Stenson’s duct lateral to the masseter muscle and has the same signal characteristics as normal parotid gland. Injury or pathology of V3 can result in denervation atrophy of the muscles of mastication, which appear swollen, bright on T2-weighted image and enhancing during the acute phase. With time, fatty atrophy of the muscles sets in along with resolution of the bright signal on T2-weighted image, and the asymmetry may cause the unaffected side to appear pathologic. Infiltration of the muscles of the MS can occur by di-

517

518

3  Brain, Head, and Neck

Fig. 3.10.56  a Axial T1-weighted image shows a well-defined hypointense lesion (asterisk) in the right parotid gland distinctive from the normal surrounding fatty parotid tissue. The retromandibular vein (arrowhead) is a useful radiologic indicator for the plane of the facial nerve. b The lesion is uniformly hyper­ intense on T2-weighted image. c The lesion enhances homoge-

neously on post-contrast T1-weighted image. These findings are classic for pleomorphic adenoma. Note that on delayed imaging after gadolinium administration, pleomorphic adenomas become increasingly densely enhancing and may be difficult to differentiate from surrounding normal parotid tissue

rect invasion from adjacent sites such as the oral cavity, oropharynx, and maxillary sinus. Trismus is a reliable clinical indicator but occurs late, and radiologic imaging can delineate MS involvement before the patient becomes symptomatic. A notable exception is lymphoma, which may not produce trismus in spite of significant infiltration of the muscles of mastication. The branches of V3 are at risk for perineural spread if there is malignant involvement of the MS, and both retrograde as well as antegrade spread can occur either contiguously or as skip areas of involvement. Early perineural spread may not be radiologically evident, but abnormal enlargement must be present along with asymmetric enhancement of the nerve

for the diagnosis of perineural spread. Indirect signs of perineural involvement of V3 include enlargement of the foramen ovale (Fig. 3.10.61), obliteration of normal fat at the extracranial aperture of the foramen ovale, abnormal soft tissue that replaces CSF in Meckel’s cave, and denervation changes in the muscles of mastication. The buccal space lies anterior to the MS and its contents include the buccinator muscle, the buccal fat pad, distal parotid duct, and facial artery and vein. There is no fascial barrier between the buccal space and the MS and this allows spread of infection or tumor into it from the MS. The buccal space is not uncommonly involved by venous malformations of the head and neck (Fig. 3.10.62).

Table 3.10.8  Common lesions of the masticator space Congenital/ developmental

Hemangioma Venolymphatic malformation

Inflammatory/  infectious

Odontogenic infection: abscess, cellulitis Myositis

Neoplasm,   benign

Benign tumor of muscle or bone Nerve sheath tumor

Neoplasm,   malignant

Osteosarcoma Rhabdomyosarcoma Non-Hodgkin’s lymphoma Deep extension of mucosal SCC Metastatic disease

Odontogenic Infection CT is superior to MR in the evaluation of odontogenic infection because calculi, foreign bodies, and gas formation are more easily recognized on CT. Radiologic features associated with infection include periodontal disease, osteomyelitis, cellulitis, phlegmon, or abscess. Rhabdomyosarcoma The presence of a solid mass in the MS in a child should be considered suspicious for rhabdomyosarcoma unless proved otherwise. These are aggressive tumors that can be multicompartmental, causing destruction of mandible or involvement of the skull base but may remain well-circumscribed within the MS on imaging. They are generally isointense to muscle on T1-weighted image, inter­mediate on T2-weighted image, tend to enhance homogeneously or heterogeneously if necrosis is present (Fig. 3.10.63).

3.10  Head and Neck

Carotid Space (CS) The CS lies within the carotid sheath and is formed by contributions from all three layers of the deep cervical fascia. It extends from the skull base to the aortic arch, spanning the suprahyoid and infrahyoid neck in the process. Superiorly, the CS communicates directly with the carotid canal and jugular foramen at the skull base. The PPS is located anteriorly, the retropharyngeal space medially, the parotid space laterally, and the vertebral bodies posteriorly to the suprahyoid CS (Fig. 3.10.64). Its contents include the common and internal carotid arter-

ies, internal jugular vein, cranial nerves IX–XII, and the sympathetic chain. The tenth nerve descends in the neck within the CS and traverses its entire extent. The other cranial nerves exit the CS above the hyoid bone and are therefore relevant only to the nasopharyngeal part of the CS. The sympathetic chain lies posterior and medial to the carotid artery while the tenth nerve is located posteriorly and laterally to it, between the carotid and the internal jugular vein. Most lesions of the CS are either vascular or neurogenic in origin (Table 3.10.9). Asymmetry or tortuosity

Fig. 3.10.57  a Coronal T1-weighted image shows a well-defined heterogeneous lesion (arrowheads) arising from the tail of the right parotid gland with hyperintense foci that represent intratumoral hemorrhage or proteinaceous cysts. The normal parotid tissue is labeled (P). b On T2-weighted image, the lesion is heterogeneous but hyperintense. Heterogeneity and intrinsic T1 shortening are classic features suggestive of Warthin’s tumor

Fig. 3.10.58  a Axial FSE T2-weighted image with FS shows asymmetric enlargement of the right parotid gland without a distinctly identifiable lesion. b A large lesion (arrows) is easily distinguishable from the normal fatty parenchyma of the parotid gland (P) on T1-weighted image. Histopathological examination confirmed the diagnosis of squamous cell carcinoma

519

520

3  Brain, Head, and Neck

Fig. 3.10.59  This patient presented with new onset right facial paralysis following previous parotidectomy for carcinoma ex pleomorphic adenoma of the right parotid gland. The abnormal enlargement and enhancement of the mastoid segment of the right facial nerve (arrowheads) on this coronal post-contrast T1-weighted image with FS is consistent with perineural spread. The normal left parotid gland (P) is labeled

Fig. 3.10.61  This patient had new onset numbness in the distribution of V3 in the setting of carcinoma involving the left masticator space. Coronal T1-weighted image shows obvious widening of the left foramen ovale (arrowheads) compared with the right side (asterisk). There is obvious enlargement of left V3 (m) consistent with perineural spread of tumor

of the carotid artery, or asymmetric internal jugular veins may be confused with a true lesion, and it is important to examine contiguous slices or obtain a neurovascular imaging study such as MRA or CTA if appropriate. Paraganglioma Paragangliomas are neuroendocrine tumors of neural crest origin. There are four common sites within the head and neck where these tumors can be located: carotid body tumor (at the bifurcation of the common carotid artery), glomus vagale (from paraganglionic tisTable 3.10.9  Common lesions of the carotid space

Fig. 3.10.60  Normal anatomy of the masticator space (MS) on coronal T1-weighted image. The MS extends from the superior attachment of temporalis muscle (upper white arrowhead) to the inferior border of the mandible (lower white arrowhead). The zygoma (white arrow) divides the MS into suprazygomatic and infrazygomatic compartments. Its contents include the muscles of mastication: temporalis (T), lateral (LP) and medial pterygoid muscles (MP), masseter (Ma), as well as the mandible (M)

Vascular

Internal jugular vein thrombosis Carotid artery thrombosis Carotid artery aneurysm or pseudoaneurysm

Inflammatory/  infectious

Abscess

Neoplasm,   benign

Paraganglioma Schwannoma Meningioma (from posterior fossa via JF)

Neoplasm,   malignant

Neuroblastoma Non-Hodgkin’s lymphoma Direct extension of mucosal SCC Nodal metastases

3.10  Head and Neck

Fig. 3.10.62  a Axial T1-weighted image shows a well-defined soft tissue mass (arrowheads) that is isointense to muscle and fills the left buccal space. b The lesion is lobulated and very hyper­intense on axial FSE T2-weighted image with FS. Note the parotid ducts (arrowheads) coursing over the masseter muscles toward the buccal space. c The two rounded hypointense areas

(arrowheads) within the lesion, which is intensely and homogeneously enhancing on coronal post-contrast T1-weighted image with FS represent phleboliths which confirm the diagnosis. CT scan would more easily identify phleboliths and may be a useful adjunct

sue in the perineurium of the tenth nerve, most commonly the nodose ganglion), glomus jugulare (from the jugular bulb), and glomus tympanicum (from the posterior auricular branch of the vagus [Arnold’s] nerve, or from the tympanic branch of the ninth nerve [Jacobsen’s] nerve). These lesions are often asymptomatic and are increasingly detected incidentally on imaging performed for investigation of other problems. They may present as a neck mass which may be pulsatile, as a submucosal bulge of the lateral pharyngeal wall, with Horner’s syndrome or rarely with lower cranial neuropathy. Paragangliomas are highly vascular tumors and therefore show intense post-contrast enhancement on CT. Glomus jugulare can be associated with permeative, destructive changes in the adjacent skull base unlike meningioma, which is associated with permeative, sclerotic changes. Flow voids are seen on MR along with intense enhancement. The origin of these tumors can be surmised by their relationship to the carotid artery; a carotid body tumor is located in the carotid bifurcation and classically splays the internal and external carotid arteries (Fig. 3.10.64), while a glomus vagale generally displaces the carotid artery anteriorly (Fig. 3.10.65). Carotid body tumors generally do not have identifiable major feeding vessels and derive their vascular supply from multiple adventitial vessels from the adjacent carotid arteries. MRA, CTA, or catheter angiography does not provide additional information on the blood supply in these cases. In contrast, other paragangliomas such as glomus vagale

often have demonstrable feeder vessels, most commonly from the ascending pharyngeal artery. Schwannoma Schwannomas or neurilemomas are well-encapsulated tumors that arise from the Schwann cells of the peripheral nerve sheath. The most commonly involved nerves in the head and neck include the eighth nerve (acoustic neuroma), the tenth nerve, and the sympathetic chain. On imaging, schwannoma is seen as a well-circumscribed round or ovoid mass (Fig. 3.10.66). Schwannomas are isointense to muscle on T1-weighted image, have high signal intensity on T2-weighted image and enhance. When they are homogeneous, they are difficult to differentiate from neurofibroma but they can often be heterogeneous due to cystic change or hemorrhage. Unlike paragangliomas, schwannomas typically do not have flow voids even when they are large and they regressively remodel adjacent bone without permeative destruction. The nerve of origin of the schwannoma is generally difficult to pinpoint since both vagal and sympathetic chain schwannomas tend to displace the carotid artery anteriorly, and may be asymptomatic. Squamous Cell Carcinoma Invasion of the CS by SCC can occur either from a locally advanced primary tumor such as the palatine tonsil or from extracapsular extension of nodal metastasis (Fig. 3.10.67 here). The degree of encasement of the carotid artery on imaging is useful in predicting surgical resectabil-

521

522

3  Brain, Head, and Neck

Fig. 3.10.63  a Axial T1-weighted image shows a large, soft tissue mass occupying the left MS engulfing the muscles of mastication and the mandible. b Coronal post-contrast T1-weighted image with FS shows irregular focal necrosis (N) with bony invasion of the floor of the left middle cranial fossa but no intracranial extension. Histologic evaluation confirmed the diagnosis of rhabdomyosarcoma in this pediatric patient

3.10  Head and Neck 9 Fig. 3.10.64  a Coronal T1-weighted image shows soft tissue masses in the bifurcation of both carotid arteries. The masses splay the external carotid artery (ECA) and the internal carotid artery (ICA), and contain flow voids. b The lesions are hyperintense, have flow voids, and displace ECA (arrowhead) anteriorly and the ICA (arrow) posteriorly on axial T2-weighted image. c There is intense homogeneous enhancement of the lesions on axial post-contrast T1-weighted image. This constellation of findings is pathognomonic of carotid body tumor

Fig. 3.10.65  a Post-contrast axial T1-weighted image shows a densely enhancing left carotid space lesion (arrow) displacing the vessels (arrowheads) anteriorly rather than splaying them. b Axial T2-weighted image shows the characteristic flow voids of paraganglioma, and the diagnosis of glomus vagale was confirmed at surgery

ity of these tumors. Tumors that encase the carotid artery by more than 270° are generally considered unresectable. Retropharyngeal Space (RPS) The RPS is a potential space that lies posterior to the PMS enclosed between the middle and deep layers of the deep cervical fascia, and extends from the skull base to the level of T4 (Fig. 3.10.68). It is separated from the more posteriorly located “danger space” by a slip of deep cervical fascia that is not identifiable radiologically, so that these spaces are not considered distinct from each other for practical purposes. The danger space provides disease processes of the neck, especially infection, access to the mediastinum. The RPS is bound laterally by the CS on either side and the danger space and prevertebral spaces posteriorly. Since it contains only fat and lymph nodes, disease processes in the RPS are limited to lipoma, inflammatory or metastatic lymphadenopathy, and spread of infection or tumor. The lateral retropharyngeal nodes (LRPN) of Rouviere are located lateral to the prevertebral muscles and medial to the internal carotid artery at the level of the nasopharynx and upper oropharynx (Fig. 3.10.69). The medial retropharyngeal nodes (MRPN) are located closer to the midline anterior to the prevertebral muscles in the paramedian RPS above, but not below the level of the hyoid bone. The normal LRPN in adults measure >8 mm in long axis, but are more prominent in children. Normal MRPN, on the other hand, are generally not demonstrated radiologically. RP nodes commonly become enlarged in children with pharyngitis and adults with spine infection. Cellulitis of the RPS fat may be evident as stranding on radiologic imaging. These nodes may suppurate if the infection progresses and may result in a RPS abscess if they rupture into the RPS (Fig. 3.10.70). It is important to image the neck from the skull base to the mediastinum in order to assess the RPS in its full extent. Differential diagnosis from RPS edema is crucial since RPS abscess requires aggressive management including surgical drainage and intravenous antibiotics. Edema of the RPS can occur from jugular vein or lymphatic obstruction, previous radiation therapy, or non-infectious inflammatory processes. Non-Hodgkin’s lymphoma is a common cause for neoplastic enlargement of the LRPN. Metastatic involvement of the LRPN occurs most frequently from nasopharyngeal carcinoma, SCC of the tonsil, soft palate, posterior pharyngeal wall, or hypopharynx, and thyroid carcinoma. LRPN can also be involved by direct extension from locally advanced tumors of the PMS, the CS, or the spine and prevertebral space. Perivertebral Space (PVS) The PVS surrounds the vertebral column, and consists of prevertebral and paraspinal portions. The preverte-

523

524

3  Brain, Head, and Neck

Fig. 3.10.66  a Axial T1-weighted image shows a homogeneous soft tissue mass (M) in the right carotid space that displaces the internal carotid artery (arrowhead) anteriorly and the internal jugular vein (arrow) posterolaterally. b Axial post-contrast T1-weighted image with FS shows an enhancing mass with focal low-signal areas. c The lesion shows heterogeneous high T2 signal on axial FSE T2-weighted image with FS. At surgery, this schwannoma was found to arise from the sympathetic chain

bral space lies deep to the deep layer of the deep cervical fascia as it extends across the front of the vertebral body from one transverse process to the other. It therefore contains the vertebral body, prevertebral musculature, and the vertebral artery and vein. The paraspinal space is enclosed within the deep layer of the deep cervical fascia as it extends dorsally from the transverse process to attach to the nuchal ligament in the posterior midline. This space contains the posterior vertebral elements, paraspinal muscles, and fat. The RPS lies anterior and the CS

anterolateral to the prevertebral portion of the PVS. The RPS is therefore displaced anteriorly by a mass lesion of the prevertebral PVS and a tumor of the vertebral body will displace the prevertebral muscles anteriorly. Disease processes of the RPS do not transgress into the PVS or vice versa, because the deep layer of the deep cervical fascia provides an effective barrier. The most common diseases involving the PVS are infectious processes and primary, metastatic, or hematologic malignancy of the spine.

3.10  Head and Neck

Fig. 3.10.67  Axial T1-weighted image shows abnormal soft tissue mass (arrow) replacing the fat in the left carotid space and insinuating between the thrombosed internal jugular vein (J) and the internal carotid artery (C). Note the normal fat (arrowhead) in the contralateral carotid space. FNA confirmed the suspicion for recurrent squamous cell carcinoma in this patient who presented with new onset cranial neuropathy

Posterior Cervical Space (PCS) The PCS is the fascial space enclosed within the superficial and deep layers of the deep cervical fascia overlying the posterior triangle of the neck. Its apex lies at the mastoid tip where the posterior border of the sternocleidomastoid and anterior border of the trapezius muscles converge, while its base is located along the superior border of the clavicle. The CS forms its anterior limit, the PVS lies medially, and the sternocleidomastoid muscle and subcutaneous fat lie laterally. The plane of the hyoid bone divides it into a smaller suprahyoid and larger infrahyoid component. Its contents include the spinal accessory nerve, fat, and lymph nodes. The most common benign neoplasms include lipoma and cystic hygroma (Fig. 3.10.71) while the majority of malignant lesions are metastatic lymphadenopathy.

Fig. 3.10.68  Normal anatomy of the retropharyngeal space. a Paramedian sagittal T1-weighted image shows the normal hyper­intense stripe of retropharyngeal fat (arrowheads). b The retropharyngeal space with its normal fat (arrowheads) is located anterior to the prevertebral musculature (the longus colli [LC] muscle at this level)

3.10.3.1.2  Infrahyoid Neck The infrahyoid neck contains five distinct spaces, four of which traverse the suprahyoid neck (CS, PVS, PCS, and RPS) and have already been described above. Unlike the suprahyoid portion, the infrahyoid portion of the RPS contains only fat and no lymph nodes. The visceral

space (VS) is unique to the infrahyoid neck and contains the larynx, trachea, hypopharynx, esophagus, recurrent laryngeal nerves, level VI lymph nodes and thyroid and parathyroid glands. The aerodigestive component of the visceral compartment has also been discussed above so

525

526

3  Brain, Head, and Neck

Fig. 3.10.69  Axial FSE T2-weighted image with FS shows normal but asymmetric lateral retropharyngeal lymph nodes (arrowheads) in a child

that this section will be limited to description of the thyroid and parathyroid glands. 3.10.3.2  Thyroid and Parathyroid Glands 3.10.3.2.1  Thyroid Gland The thyroid gland consists of right and left lobes connected by an isthmus across the anterior aspect of the upper tracheal rings. An upward projection from the isthmus, the pyramidal lobe is frequently present. The thyroid is supplied by branches of the superior (branch of external carotid artery) and inferior thyroid (branch of thyrocervical trunk) arteries. The normal thyroid gland is intrinsically hyperdense on non-contrast CT because of its iodine content and enhances densely. MR does not delineate the thyroid as clearly as CT, but the normal gland enhances homogeneously following administration of gadolinium (Fig. 3.10.72). Ultrasonography is the modality of choice for the diagnosis of thyroid nodules. CT and MR are useful for the evaluation of the local extent of disease and regional lymphadenopathy. For the evaluation of well-differentiated thyroid cancer, it is important to avoid iodinated contrast since this can interfere with the ability of thyroid cancer cells to concentrate radioactive iodine (RAI) and can delay postoperative adjuvant treatment by several months. MR may be indicated in these cases, but CT with contrast can be safely used in patients with anaplastic thyroid carcinoma or medullary carcinoma where RAI is not part of the treatment plan.

Fig. 3.10.70  a Axial post-contrast T1-weighted image with FS shows features of a diffuse infective process in a diabetic patient involving the right masticator space, oral cavity and parapharyngeal space. Focal abscesses (A) are evident, notably in the retropharyngeal space (RPA). b Axial post-contrast T1-weighted image shows the inferior extent of the retropharyngeal space abscess (arrowheads) into the superior mediastinum

Asymptomatic thyroid nodules may be discovered incidentally on cross-sectional imaging studies performed to investigate other pathology of the head and neck. The risk for malignancy in such lesions is difficult to estimate, but certain features such as ill-defined borders with or without extrathyroid extension, stippled calcification

3.10  Head and Neck Fig. 3.10.71  a Sagittal T1-weighted image shows a well-defined low signal lesion in the posterior cervical space (PCS) which is bound by the sternocleidomastoid (SCM) anteriorly and the trapezius muscle (T) posteriorly. b Axial T2-weighted image with FS shows a unilocular, lobulated mass that insinuates along tissue planes in the PCS without invasion of adjacent structures. c The lesion does not enhance on axial post-contrast T1-weighted image with FS. These imaging characteristics are typical of cystic hygroma and the diagnosis was confirmed at surgical resection

within a nodule, vocal cord paralysis, or lymphadenopathy should be considered suspicious for malignancy until proved otherwise. For all other incidentally diagnosed dominant nodules, ultrasonography and ultrasoundguided aspiration biopsy followed by surgical resection if appropriate is recommended for tissue diagnosis. Goiter, colloid cyst, and adenoma are the most common benign lesions of the thyroid. Goiter is characterized by diffuse or multinodular enlargement of the gland, which can be large enough to cause displacement and/or compression of the central compartment viscera such as the trachea (Fig. 3.10.73). A colloid cyst appears as a wellcircumscribed bright lesion on T1-weighted image because of increased protein content or hemorrhage within it. An adenoma is also well circumscribed, but is largely solid and may have areas of calcification, cystic degeneration, or hemorrhage. As discussed above, there are no reliable imaging indicators for low-grade, well-differentiated thyroid cancer that is intraparenchymal and has not metastasized to re-

gional lymph nodes. Locally advanced well-differentiated cancer, poorly differentiated or anaplastic tumors can involve the contents of the VS. Invasion of the trachea and esophagus are important considerations in surgical planning and are generally more easily recognized on MR than CT (Fig. 3.10.74). Nodal metastases occur frequently from papillary thyroid carcinoma, especially in younger patients, and commonly appear cystic. Therefore, cystic cervical nodal metastases should prompt consideration of thyroid carcinoma as the primary source but other tumors such as SCC of the palatine tonsil and base of tongue can also produce cystic metastases. On T1-weighted image, thyroid metastases can have high signal intensity from hemorrhage or thyroglobulin (protein) content (Fig. 3.10.75) FDG-PET scan is being increasingly used in evaluation of thyroid carcinoma. Differentiated thyroid cancers are generally not FDG-avid ,while poorly differentiated and anaplastic carcinomas are. These FDG-avid lesions tend not to concentrate RAI, so that PET scan may be a

527

528

3  Brain, Head, and Neck Fig. 3.10.72  MR does not delineate the thyroid as clearly as CT. a Axial T1-weighted image shows a normal thyroid (T) that is of similar intensity to surrounding musculature. b The gland (T) enhances homogeneously on post-contrast axial T1weighted image with FS

Fig. 3.10.73  a Axial T1-weighted image shows bilateral nodular enlargement of the thyroid gland (arrows) with a hyperintense focus (arrowhead) in the left lobe representing colloid or blood. b The enlarged thyroid is well defined, without invasion of adjacent structures or lymphadenopathy on axial T2-weighted image. Total thyroidectomy confirmed the diagnosis of multinodular goiter

useful imaging modality in the presence of elevated serum thyroglobulin and negative RAI scan. Parathyroid Glands The parathyroid glands are most frequently located along the posterior border of the thyroid gland and are about 6 mm in size. Generally, there are two superior and two inferior glands, but there can be considerable variability

in location, size, and number of the parathyroids. Conventional CT and MR are not able to identify normal parathyroid glands. Ultrasonography and sestamibi scans are most commonly used in the assessment of parathyroid pathology. Enlarged parathyroid glands may be difficult to differentiate from lymphadenopathy on CT, but high signal intensity on T2-weighted image is suggestive of a parathyroid adenoma.

3.10  Head and Neck

3.10.3.3  The Pediatric Neck

Fig. 3.10.74  Coronal T1-weighted image shows large left thyroid mass invading the left lateral tracheal wall with intraluminal extension (arrowhead) into the trachea

Pathology of the neck in the pediatric age group is generally associated with developmental, infectious or inflammatory, and some neoplastic processes (Table 3.10.10). CT is the modality of choice for evaluation of infection and inflammation, but MR is used to evaluate complications such as intracranial or spinal extension. Congenital and neoplastic conditions, on the other hand, are typically assessed with MR which can provide clues to a particular diagnosis. Monitored anesthesia or general anesthesia is required for children younger than age 8 or 10 in order to obtain a good quality MR study. Fibromatosis colli presents as a palpable neck mass or torticollis in neonates and young infants. It is thought to represent fibroinflammatory response within the sternocleidomastoid muscle from traumatic delivery. Ultrasonography is generally adequate for diagnosis. On MR, there is fusiform enlargement of the belly of the sternocleidomastoid muscle which is intermediate signal on T1-weighted image, heterogeneous on T2-weighted image and enhances diffusely (Fig. 3.10.76). The differential diagnosis should include rhabdomyosarcoma which has been described above. Certain cystic masses that are seen in the pediatric age group can also present in older patients, and are described under this section. These include lymphatic malformation, dermoid/epidermoid, thyroglossal duct cyst, Thornwaldt cyst, and branchial cleft cyst. Vascular lesions of the head and neck include hemangiomas and vascular malformations, which can be transspatial. Although there is some confusion regarding terminology, the term hemangioma should be reserved for rapidly growing vascular lesions of infancy which

Table 3.10.10  Common lesions of the neck in the pediatric age group

Fig. 3.10.75  Axial T1-weighted image shows a high signal enlarged left level II lymph node (arrowheads) compared to a normal right sided lymph node (N). The abnormal lymph node was confirmed metastatic papillary carcinoma of the thyroid on FNA. The internal jugular vein (J) and internal carotid artery (C) are labeled

Congenital/  developmental

Lymphatic malformation Venous malformation Branchial cleft cyst Thyroglossal duct cyst Epidermoid/dermoid

Infectious/  inflammatory

Suppurative lymphadenopathy Abscess

Benign neoplastic or neoplasm-like

Fibromatosis colli Neurofibroma Hemangioma

Malignant neoplastic

Neuroblastoma Rhabdomyosarcoma Lymphoma

529

530

3  Brain, Head, and Neck 9 Fig. 3.10.76  a Axial T1-weighted image in a newborn with torticollis shows diffuse enlargement of the right sternocleidomastoid (T) compared with the contralateral normal muscle (M). b Axial post-contrast T1-weighted image with FS shows diffuse enhancement of the “mass” with c heterogeneous high signal on axial T2-weighted image with FS. Although rhabdomyosarcoma must be considered in the differential, these imaging features are most consistent with fibromatosis colli in a newborn (Images courtesy of Michelle S. Bradbury, MD, PhD)

ultimately involute by adolescence. Hemangiomas are generally intermediate signal on T1-weighted image, high signal on T2-weighted image and enhance intensely (see Fig. 3.10.53). Larger lesions may be associated with flow voids, and feeding vessels may be demonstrable. Fatty replacement occurs as the hemangioma involutes, and the lesion may be bright on T1-weighted image. Vascular malformations arise because of abnormal morphogenesis of blood vessels or lymphatics. The type of vessel that is predominantly involved serves to classify these lesions into capillary, venous, lymphatic, or arteriovenous malformations. However, in reality, vascular malformations often consist of a combination of these various types constituting a spectrum rather than discrete anomalies. Venous malformations are low-flow lesions that are typically multilobulated, contain venous lakes, and may have phleboliths (see Fig. 3.10.62). Arteriovenous malformations have serpiginous flow voids without a dominant mass (Fig. 3.10.77). Lymphatic malformations (lymphangioma or cystic hygroma) occur because of abnormal lymphatic vessel development so that these abnormal vessels fail to drain into normal lymphatic channel, producing multiloculated fluid-filled masses. They are most commonly located in the PCS and are frequently transspatial. On T1-weighted image, these lesions are primarily hypointense but high protein content or previous hemorrhage can cause high signal intensity. They are hyperintense on T2-weighted image and generally do not enhance. A characteristic feature is fluid-fluid levels due to hemorrhage (Fig. 3.10.78). A thyroglossal duct cyst (TGDC) is typically seen as a midline or paramedian cystic mass located in intimate relation to the infrahyoid strap muscles. TGDC arises from thyroid remnants along the path of descent of the thyroid anlage from the foramen cecum of the tongue to its normal location in the infrahyoid neck. Carcinoma may arise rarely within a TGDC, and should be suspected when there is radiologically demonstrable soft tissue within the cyst cavity. Branchial cleft cysts result from failure of obliteration of the cervical sinus during development. The second branchial apparatus is the most commonly affected and

3.10  Head and Neck Fig. 3.10.77  Axial T1-weighted image in a child with a submental vascular malformation shows vascular flow voids (arrowheads) in bilateral submandibular and sublingual spaces, without abnormal soft tissue. Note the bilaterally enlarged external carotid arteries (E). The diagnosis of arteriovenous malformation was confirmed at angiography (not shown)

Fig. 3.10.78  a Axial FSE T2-weighted image with FS shows a large, loculated, soft-tissue mass of the left face with multiple fluidfluid levels (arrowheads), consistent with hemorrhage within the multiple cystic spaces. b Axial post-contrast T1-weighted image with FS shows linear enhancement of the fibrous septa of cystic spaces in the anterior part of the lesion. These imaging features are characteristic of lymphangioma

its developmental defects are related to the vast majority of all branchial cleft anomalies. Imaging reveals a unilocular cystic mass that displaces the submandibular gland anteromedially and the sternocleidomastoid muscle posterolaterally (Fig. 3.10.79). Active infection can result in stranding of adjacent soft tissue as well as enhancement of the cyst wall which can persist after the infection has subsided. The diagnosis is easier if a “beak” is identifiable insinuating between the internal and external carotid arteries, or if a sinus tract or fistula can be identified tracking to the tonsillar fossa.

Dermoid and epidermoid cysts of the head and neck are slow-growing lesions that are typically midline. They are most commonly located in the floor of the mouth (Fig. 3.10.80). They occur from sequestration of skin along embryologic lines of fusion so that both are lined by squamous epithelium. Sequestration of skin appendages such as sebaceous glands and hair follicles gives rise to a dermoid cyst. While both lesions are unilocular and can look similar on radiologic imaging, the presence of fatty material which is high signal on T1-weighted image points to a diagnosis of dermoid.

531

532

3  Brain, Head, and Neck

Fig. 3.10.79  Axial contrast-enhanced CT scan of the neck shows a unilocular, well-defined cyst (BCC) located between the sternocleidomastoid muscle (SCM), and the submandibular gland (SMG). These are the typical imaging features of a second branchial cleft cyst

Fig. 3.10.80  a Coronal T1-weighted image shows an ovoid, well-defined mass in the submental space between the anterior bellies of the digastric muscles (D). The high signal may be from fat, blood, or protein within the mass. b The uniform low signal on coronal post-contrast T1-weighted image with FS is indicative of fat content rather than blood or protein, pointing to the diagnosis of dermoid cyst which was confirmed on histologic examination of the surgical specimen

3.10  Head and Neck

Suggested Reading 1. 2.

Harnsberger HR et al (eds) (2004) Diagnostic imaging: head and neck. Amirsys, Salt Lake City Kirchner JA (1998) Atlas on the surgical anatomy of laryngeal cancer. Singular, San Diego

3.

4. 5.

Som PM, Curtin HD, Mancuso AA (2000) Imaging-based nodal classification for evaluation of neck metastatic adenopathy. AJR Am J Roentgenol 174:837 Som PM et al (eds) (2003) Head and neck imaging, 4th edn. Mosby, St. Louis Stambuk HE, Karimi S, Lee N, Patel SG (2007) Oral cavity and oropharynx tumors. Radiol Clin N Am 1:1–20

533

Chapter 4

Spine and Spinal Canal

4.1 

Extradural Diseases of the Spine  .. . . . . .   536 C.S. Poon, J. Doumanian, G. Sze, M. Johnson, and C.E. Johnson

4.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   536

4.1.2 

Degenerative Spine Disease  . . . . . . . . . . . .   536

4.1.2.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   536 4.1.2.2  Imaging Techniques  .. . . . . . . . . . . . . . . . . .   537 4.1.2.3  Normal Anatomy of the Intervertebral Disk   .. . . . . . . . . . . . .   538

4

4.2.3 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   591

4.2.4 

Pathological Findings  .. . . . . . . . . . . . . . . . .   592

4.2.4.1  Neoplasms  . . . . . . . . . . . . . . . . . . . . . . . . . . .   592 4.2.4.2  Congenital Lesions  .. . . . . . . . . . . . . . . . . . .   603 4.2.4.3  Inflammatory Conditions  .. . . . . . . . . . . . .   607 4.2.4.4  Trauma  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   610 4.2.4.5  Vascular Malformations  . . . . . . . . . . . . . . .   611 4.2.5 

Acknowledgement  . . . . . . . . . . . . . . . . . . . .   614

4.1.2.4  Imaging Findings of Degenerative Spine Disease  .. . . . . . . . .   540 4.1.3 

Extradural Spine Infection  .. . . . . . . . . . . .   555

4.1.3.2  Imaging Techniques  .. . . . . . . . . . . . . . . . . .   556

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   614 4.3 

Intramedullary Diseases of the Spinal Cord  . . . . . . . . . . . . . . . . . . . .   617 P. Pawha, C. Shen, J. Doumanian, F. Lin, M. Johnson, R. Ashton, and G. Sze

4.3.1 

MRI Techniques   for Spinal Cord Imaging  . . . . . . . . . . . . . . .   617

4.1.3.3  Infectious Spondylodiskitis  . . . . . . . . . . . .   556 4.1.3.4  Spinal Epidural Abscess  .. . . . . . . . . . . . . . .   559 4.1.3.5  Septic Facet Joint Arthritis  .. . . . . . . . . . . .   562 4.1.4 

Extradural Spine Tumors  . . . . . . . . . . . . . .   562

Indications and Value of MRI  . . . . . . . . . .   613

4.3.1.1  Normal MRI Anatomy  . . . . . . . . . . . . . . . .   618

4.1.4.1  Imaging Techniques  .. . . . . . . . . . . . . . . . . .   562

4.3.2 

4.1.4.2  Benign Tumors  . . . . . . . . . . . . . . . . . . . . . . .   563

4.3.2.1  Ependymoma  .. . . . . . . . . . . . . . . . . . . . . . . .   619

4.1.4.3  Malignant Tumors  . . . . . . . . . . . . . . . . . . . .   570

4.3.2.2  Astrocytoma  .. . . . . . . . . . . . . . . . . . . . . . . . .   621

Vertebral Column Trauma  .. . . . . . . . . . . .   574

4.3.2.3  Hemangioblastoma  . . . . . . . . . . . . . . . . . . .   622

4.1.5.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   574

4.3.2.4  Metastasis  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   623

4.1.5.2  Imaging Techniques  .. . . . . . . . . . . . . . . . . .   575

4.3.3 

4.1.5.3  Biomechanics  .. . . . . . . . . . . . . . . . . . . . . . . .   576

4.3.3.1  Vascular Anatomy of the Spinal Cord  .. .   625

4.1.5.4  Imaging Findings  . . . . . . . . . . . . . . . . . . . . .   578

4.3.3.2  Spinal Cord Infarction  .. . . . . . . . . . . . . . . .   626

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   587

4.3.3.3  Vertebral Body Infarction  .. . . . . . . . . . . . .   630

4.2 

Intradural Extramedullary Spine   .. . . . .   590 D. Lin

4.3.3.4  Venous Infarction  .. . . . . . . . . . . . . . . . . . . .   630

4.2.1 

Introduction   . . . . . . . . . . . . . . . . . . . . . . . . .   590

4.2.2 

Examination Techniques  .. . . . . . . . . . . . . .   590

4.1.5 

Intramedullary Neoplasms  .. . . . . . . . . . . .   618

Vascular Diseases of the Spinal Cord  .. . .   625

4.3.3.5  Differential Diagnosis  . . . . . . . . . . . . . . . . .   631 4.3.3.6  Vascular Malformations  . . . . . . . . . . . . . . .   631 4.3.4 

Demyelinating Disease  . . . . . . . . . . . . . . . .   640

536

4  Spine and Spinal Canal 4.3.4.1  Multiple Sclerosis  . . . . . . . . . . . . . . . . . . . . .   641

4.3.6.4  Granulomatous Processes  .. . . . . . . . . . . . .   652

4.3.4.2  Devic’s Disease  .. . . . . . . . . . . . . . . . . . . . . . .   643

4.3.6.5  Differential Diagnosis  . . . . . . . . . . . . . . . . .   653

4.3.4.3  Acute Disseminated Encephalomyelitis     645

4.3.7 

4.3.4.4  Transverse Myelitis  .. . . . . . . . . . . . . . . . . . .   646

4.3.7.1  Clinical Manifestations  . . . . . . . . . . . . . . . .   653

Radiation Myelopathy  .. . . . . . . . . . . . . . . .   647

4.3.7.2  Classification  . . . . . . . . . . . . . . . . . . . . . . . . .   654

4.3.5.1  Clinical Presentation  . . . . . . . . . . . . . . . . . .   648

4.3.7.3  Pathophysiology  . . . . . . . . . . . . . . . . . . . . . .   654

4.3.5.2  Pathology  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   648

4.3.7.4  Technique  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   655

4.3.5.3  Imaging Findings  . . . . . . . . . . . . . . . . . . . . .   648

4.3.7.5  Imaging Findings  . . . . . . . . . . . . . . . . . . . . .   655

4.3.6 

4.3.7.6  Acute Findings  .. . . . . . . . . . . . . . . . . . . . . . .   656

4.3.5 

Intramedullary Infectious and Inflammatory Diseases  . . . . . . . . . . . .   649

4.3.6.1  Viral Myelitis  . . . . . . . . . . . . . . . . . . . . . . . . .   649 4.3.6.2  HIV-Associated Vacuolar Myelopathy  . .   650 4.3.6.3  Non-Viral Spinal Cord Infections  .. . . . . .   651

4.1  Extradural Diseases of the Spine C.S. Poon, J. Doumanian, G. Sze, M. Johnson, and C.E. Johnson 4.1.1  Introduction MRI has revolutionized the imaging of the spine, providing diagnostic information not available previously with other imaging modalities. Although CT and X-ray plain film still provide better details of the osseous structures and calcifications, these imaging modalities provide little information about the soft tissues that are often the anatomical structures most relevant to a patient’s clinical problem. MRI can provide excellent details of all the anatomical structures of the spine, including the bone, bone marrow, and soft tissues. CT and X-ray plain film remain the preferred choice for initial evaluation of spinal trauma, but for most other spine diseases, MRI has become the preferred imaging modality. Spine diseases that are best evaluated with MRI include degenerative spine disease, spine infection, and neoplasm. In addition, MRI is often required for evaluation of acute traumatic spine injury, particularly when injury to the spinal cord, ligaments or vascular structures is suspected. This section covers the application of MRI for evaluation of extradural spine diseases. These diseases are categorized under the sections of degenerative spine disease, infection, neoplasm, and trauma. Diseases involving primarily the other compartments of the spine will be discussed in the next sections.

Intramedullary Traumatic Injury  . . . . . . .   653

4.3.7.7  Chronic Findings  . . . . . . . . . . . . . . . . . . . . .   657 4.3.7.8  Pediatric Spinal Cord Injury  . . . . . . . . . . .   658 4.3.4.9  Subacute Combined Degeneration  .. . . . .   659 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   659

4.1.2  Degenerative Spine Disease 4.1.2.1  Introduction Degenerative disease of the spine is most commonly seen in the lumbar, followed by the cervical spine. The thoracic spine is least frequently associated with degenerative changes and clinical complaints that lead to radiological evaluation. Degenerative disease affects various anatomical components of the spine: the vertebral bodies, intervertebral disks, ligaments, and facet joints. In the cervical spine, the uncovertebral joints are also subject to degenerative disease. The sequelae of degenerative changes are narrowing of the neural foramina and spinal canal, which may further lead to compression of the spinal cord or nerve roots. The most common clinical complaint related to degenerative disease is pain. However, degenerative disease may also lead to radiculopathy and myelopathy which often manifests as radiating pain, paresthesia, spasticity, and weakness corresponding to the affected level of the spinal cord or peripheral nerve roots. In more severe cases, there may be bowel or bladder dysfunction. Previous studies have shown poor correlation between the morphology of degenerative disease and clinical symptoms, particularly the most common complaint of pain (Cavanaugh et al. 1997; Jensen et al. 1994). Degenerative changes are commonly seen in the older population, but many of these patients do not have significant clinical complaints. On the other hand, some patients may complain of severe pain and neurological deficits that are poorly correlated with imaging findings. This discrepancy may be due to inflammatory response in the surrounding

4.1  Extradural Diseases of the Spine

soft tissues as a major mechanism in eliciting back pain, rather than direct mass effect. In addition, degenerative changes may exert indirect compression on the nerve roots by distorting their normal surrounding soft tissue structures, such as epidural fat, rather than compressing the nerves directly. Despite all these shortcomings, MRI remains the imaging method of choice in most patients for evaluation of this common disease. MRI provides superior contrast resolution for evaluation of soft tissues of the spine, including spinal cord, nerve roots, disks, joints, ligaments, and marrow space. CT provides better depiction of osseous structures. Calcification and gas associated with degenerative disease are also better seen on CT. In the evaluation of postoperative spine, both MRI and CT are subject to image artifacts generated by hardware. The image artifacts on postoperative MRI and CT are different because of the completely different physical mechanisms of their generation. Hardware placement is usually better assessed on CT. However, with the popularity of new hardware materials such as titanium screws that often replace use of steel screws, and careful consideration of imaging protocols, MR image artifacts can be reduced significantly. In some cases, MRI and CT may provide supplemental information. 4.1.2.2  Imaging Techniques There are variations of spine imaging protocols in different institutions, depending on the preference of the inter­ preting radiologists and imaging equipment available. In this section, we outline a basic protocol designed for imaging of degenerative spine disease. For the best signal-to-noise ratio, spine imaging is generally performed using dedicated surface coils. The standard protocol for the evaluation of degenerative spine disease includes two orthogonal planes, usually sagittal and axial. Coverage of the sagittal images should include the bilateral neural foramina. In imaging of the lumbar spine, the conus medullaris should also be covered. The optimal strategy for the acquisition of axial images is a subject of controversy. Thin sections angled parallel to individual disks, covering the whole disks from pedicle to pedicle, are best for evaluation of disk bulging and herniation. However, angled axial images, with skipped areas of coverage between disks, may lead to missed diagnoses including migrated disk fragments and spondylolysis (Fiorella et al. 1999). These findings are easier to demonstrate on non-angled contiguous axial image acquisition, i.e., in a stack fashion. For the cervical and thoracic spine, thinner slices are needed to allow evaluation of the smaller structures. Ideally, the slice thickness should be no more than 3 mm with a minimal inter-slice gap of less than 1 mm. Higher resolution can be obtained with 3D acquisition tech-

niques. In the lumbar spine, the structures are larger. Thicker slices can be used if necessary. However, the slice thickness should be no more than 5 mm. Sagittal images are typically acquired using T1weighted and T2-weighted fast spin-echo (FSE) sequences. A fat-suppressed inversion recovery sequence (STIR) or fat-suppressed fast spin-echo sequence may be added. These fat-suppressed sequences are very sensitive for the evaluation of the marrow space of the vertebral bodies, paraspinal soft tissues, and the spinal cord. For the cervical spine and thoracic spine, axial 3D gradient-echo techniques with low flip angle and gradient moment nulling can be used instead of, or in addition to an axial T2-weighted FSE sequence (White 2000). The 3D gradient-echo sequences generate high signal intensity from CSF, intermediate-to-high signal from disk, and low signal intensity from osteophytes. They may provide better differentiation between herniated disk and osteophyte. In addition, their higher spatial resolution compared to 2D fast spin-echo techniques provides better evaluation of the smaller anatomic structures in the cervical and thoracic spine. One limitation of gradient-echo techniques is their sensitivity to susceptibility, leading to artifactual signal loss around bony structures, which can exaggerate the severity of disease. For routine imaging of degenerative spine, intravenous contrast is usually not necessary. For the lumbar spine, the axial images are usually acquired using T1-weighted and T2-weighted fast spin-echo sequences. Proton density weighted imaging sequences can provide good image contrast for the evaluation of the intervertebral disks. Although each pulse sequence provides unique advantages and limitations, the T2-weighted fast spin-echo sequences in the sagittal and axial planes probably provide the most diagnostic information because of their high signal-to-noise and contrast-to-noise ratios (Ruggieri 1999). T2-weighted images are sensitive for the evaluation of lesions within the spinal cord. In addition, the high signal intensity from cerebrospinal fluid generates a myelographic effect that provides excellent delineation of the anatomic relationship between the disks, osseous structures, nerve roots, and spinal cord. Traditionally, foraminal disease in the lumbar spine was thought to be best evaluated using T1-weighted imaging sequences because of the bright fat signal surrounding the darker nerve roots. However, fast spin-echo sequences also generate bright fat signal. With the availability of fast spin-echo sequences, this advantage of T1-weighted imaging sequences has become less important. The imaging of the postoperative spine requires some modification of the above protocol. To differentiate disk material from postsurgical scar, intravenous contrast material has been shown to be advantageous (Ross et al. 1990). It is important to start imaging immediately after administration of intravenous contrast, since the contrast agent may diffuse into the disk after a delay of 10 min-

537

538

4  Spine and Spinal Canal

utes or more, potentially eliminating the advantage of a contrast-enhanced study. A postoperative spine protocol should include sagittal and axial contrast-enhanced T1-weighted images with fat suppression. In the presence of orthopedic hardware, the gradient-echo sequences used for cervical spine should be replaced with T2-weighted axial FSE sequences. The echo spacing should be minimized and the frequency encoding direction should be directed along the long axis of the screws. This FSE sequence has been shown to be less sensitive to susceptibility artifacts from orthopedic hardware (Frazzini et al. 1997; Petersilge et al. 1996). Other measures that may help to reduce image artifacts from orthopedic hardware include increasing the field-of-view, increasing the readout bandwidth, reducing the voxel size, or imaging with a lower field MR system. Standard protocols used for imaging of degenerative spine disease at Yale University are summarized in Table 4.1.1.

4.1.2.3  Normal Anatomy of the Intervertebral Disk The three major components of a disk include the nucleus pulposus, annulus fibrosis, and the cartilaginous endplate. In the core of the disk is the nucleus pulposus (Fig. 4.1.1). It is composed of fibrocartilage. The ground substance in the nucleus pulposus consists of hyaluronic acid and glycosaminoglycans, which absorb water and therefore demonstrate high signal intensity on T2-weighted images. Surrounding the nucleus pulposus circumferentially is the annulus fibrosis, which is arranged as layers of lamellar structures. The annulus fibrosis consists of an inner ring, which immediately surrounds the nucleus pulposus, and an outer ring. The inner ring contains fibrocartilage and has signal intensity similar to the nucleus pulposus. The outer ring of the annulus fibrosis is thicker anteriorly. It contains dense fibrous lamellae that originate and attach to the compact cortical bone of the

Table 4.1.1  Standard protocols for imaging of degenerative spine disease at Yale University. For axial images, stack mode includes contiguous acquisition in a plane orthogonal to the sagittal images, while disk mode includes images angled parallel to individual disks. Variations to the protocols may be made for individual cases. Slice thickness (mm)

Inter-slice spacing  (mm)

• T1-weighted FSE

3

4

400

13

• T2-weighted FSE

3

4

3,000

110

• STIR

3

4

3,500

60

• T2-weighted FSE (stack mode)

3

4

3,500

110

• Proton density–weighted FSE (disk mode)

3

3.5

2,000

22

• 3D T2*-weighted GRE (Flip angle, 5°)

2

2

32

10

• T1-weighted FSE

3

4

400

9

• T2-weighted FSE

3

4

4,000

120

• Proton density weighted

3

4

2,000

15

• T2-weighted FSE (stack mode)

4

5

4,000

120

• Proton density–weighted FSE (disk mode)

3

4

3,000

13

TR  (ms)

TE (ms)

Cervical and thoracic spine Sagittal

Axial

Lumbar spine Sagittal

Axial

4.1  Extradural Diseases of the Spine

ring apophysis, thus helping to anchor the intervertebral disk to the vertebral bodies. These fibrous lamellae are often called Sharpey’s fibers. Because the outer layer of the annulus fibrosis contains denser collagen compared with other parts of the disk, it typically demonstrates low signal intensity on both T1-weighted and T2-weighted images. At the superior and inferior margins of the disk are the cartilaginous endplates. The cartilaginous endplates contain hyaline cartilage surrounded by an outer ring of dense bone called ring apophysis, which is fused to the vertebrae by the second decade of life. At the center of a normal disk, a cleft is often seen containing denser collagen and thus demonstrates lower signal intensity on T2-weighted images compared with the nucleus pulposus. The anterior and posterior longitudinal ligaments are attached to the anterior and posterior margins of the annular fibrosis respectively. Because of

the dense collagen content of the ligaments, they demonstrate low signal intensity on T1- and T2-weighted images and are difficult to differentiate from the outer ring of the annulus fibrosis. The ligamentum flavum lines the posterolateral margin of the spinal canal. It connects the laminae of consecutive vertebral levels. The facet joint is also called the zygapophyseal joint. It is a true synovial joint. On axial images, the superior articular process of a vertebral body can be seen on the anterior aspect of the facet joint, apposing posteriorly the inferior articular process of the vertebral body located at the level above. Each articular process includes cortical bone covered by a layer of cartilage. In young patients, a meniscus may be present in the joint. The facet joint is lined by synovium that extends to the space posterior to the ligamentum flavum. This is the location where synovial cysts may be seen and are discussed in more detail later.

Fig. 4.1.1  Normal anatomy of the lumbar spine. a Sagittal T2-weighted image in a central plane. All the disks demonstrate normal high signal intensity at the central core representing nucleus pulposus. This is surrounded circumferentially by dark annulus fibrosis. Superiorly and inferiorly, the disks are demarcated by the dark endplates (small white arrows). At the equator of the disks are dark intranuclear clefts (large white arrow). b Sagittal T2-weighted image in a lateral location shows the normal neural foramina with “keyhole” configuration. The L4 nerve root (white arrow) exiting through the L4–L5 neural foramen is seen located under the L4 pedicle, and surrounded by high signal intensity representing fat. c Axial T2-weighted image at

the L4–L5 level shows the disk with the high signal intensity nucleus pulposus at the center, surrounded by dark annulus fibrosis circumferentially. Note the small central concavity of the posterior disk margin. Loss of this concavity may suggest mild disk disease. Superior facets of L5 are located anterior to inferior facets of L4. Note the normal round shape of the thecal sac surrounded by disk anteriorly, the facets and ligamentum flavum posterolaterally, and the epidural fat posteriorly. The right L4 nerve root (between small arrows) is seen exiting the right L4–L5 neural foramen. N nucleus pulposus, A annulus fibrosis, L ligamentum flavum, S superior facet, I inferior facet, E epidural fat

539

540

4  Spine and Spinal Canal

4.1.2.4  Imaging Findings of Degenerative Spine Disease 4.1.2.4.1  Vertebral Bodies Degenerative disease may lead to signal intensity changes in the marrow immediately adjacent to the intervertebral disks, which are often referred to as diskogenic vertebral or endplate changes. There are three types of endplate changes previously classified by Modic et al. (1988). Histologically, type 1 change represents proliferation of fibrovascular tissues which replace the normal marrow. This can be seen as hypointensity on T1-weighted images, and hyperintensity on T2-weighted images. Type 2 change represents the proliferation of fatty marrow, and is seen as hyperintensity on both T1-weighted and T2-weighted fast spin-echo sequences. Type 3 change represents the replacement of marrow by dense bone, and is therefore hypointense on both T1 and T2-weighted images. Type 1 change may progress to type 2 change, while type 2 change can remain stable. Although three distinct Modic types were described, in some cases, mixed Modic types are present. Type 1 change may enhance with gadolinium and can be confused with infection or even neoplasm. The evidence of ancillary findings can usually help to differentiate among these considerations, although sometimes this may be difficult. The presence of degenerative changes in the intervertebral disk, further described in the following section, will help to support the diagnosis of Modic type changes. On the other hand, findings such as increased T2 prolongation in the intervertebral disks, which is unusual in the degenerative disk, inflammatory changes in the paraspinal soft tissues, or the presence of a paraspinal/epidural mass would suggest a more worrisome diagnosis. The disk itself is often hyperintense and enhances in diskitis, but maintains low signal intensity and lack of enhancement in degenerative disk disease. Osteophytes are commonly present near the endplates of the vertebral bodies and may be situated anteriorly, laterally or posteriorly. Osteophytes in the anterior

and lateral aspects of the spine are common in the older population. These changes are often called spondylosis deformans. They are thought to be an attempt to increase the surface area to help reduce the stress of axial loading. Patients with these types of osteophytes are often asymptomatic. On the other hand, posterior osteophytes are often associated with disk disease. The separation of osteophytes and displaced disk material can be difficult, especially in the cervical spine. A gradient-echo sequence may be helpful if distinction is important, as disk often demonstrates intermediate signal intensity, while osteophyte is hypointense. Very often, disk disease and osteophyte coexist and together are simply called disk–osteophyte complex. Regardless, the result is narrowing of the spinal canal or neural foramina. 4.1.2.4.2  Disk Disease Historically, the description of degenerative disk disease has been confusing, with non-standardized terminology and nomenclature. In 2001, the Combined Task Forces of the North American Spine Society, American Society of Spine Radiology, and American Society of Neuroradiology published recommendations in the standardization of nomenclature and classification of lumbar disk pathology (Fardon and Milette 2001). Further revisions of the recommendations are expected in the future, but the 2001 recommendations represent the most widely accepted consensus on nomenclature to date. Although the recommendations were originally intended for lumbar disk disease, the terminology can also be generalized to cervical and thoracic disk disease. In this section, we will follow these recommendations in the description of degenerative spine disease. Disk disease is probably the most important component of degenerative disease that may lead to clinical complaints. In normal young patients, the disk demonstrates hyperintensity on T2-weighted images. With aging, the annulus fibrosis thickens. The core of the disk will contain increased fibrous content, leading to de-

Type 1

Type 2

Histology: replacement of marrow by fibrovascular tissues (see Fig. 4.1.2a–c)

Histology: fatty marrow proliferation  (see Fig. 4.1.2d–f)

T1-weighted image: hypointense (see Fig. 4.1.2a)

T1-weighted image: hyperintense (see Fig. 4.1.2d)

T2-weighted image: hyperintense  (see Fig. 4.1.2b)

T2-weighted image: hyperintense  (see Fig. 4.1.2f)

Contrast enhancement: yes (see Fig. 4.1.2c)

Contrast enhancement: no (see Fig. 4.1.2f)

4.1  Extradural Diseases of the Spine

Fig. 4.1.2  Degenerative endplate changes previously described by Modic et al. (1988). Each type of degenerative endplate change is illustrated by a set of three images, acquired respectively from T1-weighted, T2-weighted and fat-suppressed contrast-enhanced imaging sequences. a–c Type 1 change at endplates adjacent to the L4–L5 disk in a 47-year-old woman with history of breast cancer. Note the hypointensity on T1-weighted image a, hyperintensity on T2-weighted image b, and contrast enhancement c. Metastatic disease can demonstrate similar signal abnormality. However, the location and extent of the abnormality is immediately adjacent to a severely degenerated disk, giving a typical appearance of degenerative endplate change. Note the

severe loss of height, loss of normal signal intensity, and diffuse bulging of the L4–L5 disk (arrow). d–f Type 2 change at endplates of T7–T8 disk (white arrow). Note the severe degenerative changes of the T7–T8 disk with loss of disk height and normal hyperintensity on T2-weighted images. There is mild enhancement in T7–T8 disk, probably due to presence of granulation tissues. Mild type 1 change is seen at T6–T7 level in the same patient. There is abnormal signal intensity at the T6–T7 level of the spinal cord (black arrow) from severe cord compression by a left paracentral disk herniation (not shown on this image). This patient had a previous laminectomy. There is enhancement at the surgical bed (white arrow, f) from granulation tissue

541

542

4  Spine and Spinal Canal

creased signal intensity on T2-weighted images. However, the disk height should be preserved. Manifestations of disk degeneration include loss of both disk height and the high signal intensity on T2-weighted images. Although this loss of high T2 signal intensity is often described as disk desiccation, previous studies suggest that this may actually be due to change in proteoglycan composition, rather than to absolute change in water content. Degenerative disks may calcify. In addition, disk degeneration may lead to negative pressure within the disk that leads to trapping of gas, often nitrogen. This is often described as the vacuum phenomenon. Although the vacuum phenomenon is better demonstrated on CT or radiography, it can also be seen on MR images (Fig. 4.1.3) as hypointensity on both T1- and T2-weighted images and is best demonstrated with a gradient-echo sequence due to the increased susceptibility artifacts. Occasionally, the vacuum effect traps fluid instead of gas, leading to increased signal intensity on T2-weighted image (Schweitzer and El-Nouneam 1998) (Fig. 4.1.4). Disk degeneration can lead to mechanical weakening and eventual failure of the annulus fibrosis. Annular disruption is also commonly referred to as an annular tear, an annular fissure, or a high intensity zone in the posterior annulus. It represents a disruption in the annular fibers. Three types have been described: concentric disruption, transverse disruption, and radial disruption (Yu et al. 1988). Concentric disruption represents the accumulation of mucoid material between the layers of lamellae of the annulus fibrosis. This mucoid material can be seen as increased signal intensity on T2-weighted images. It is often asymptomatic and is of limited clinical significance.

Transverse annular disruption represents horizontal disruption of lamellar fibers at the outer ring of the annular fibrosis. It is also likely to be asymptomatic and of no clinical significance. In radial disruption, all fibers of the annulus fibrosis are disrupted. There may be invasion by granulation tissues, resulting in pain. On T2-weighted images, annular disruption is seen as hyperintensity (Fig. 4.1.5). If inflammation or granulation tissue is present, annular disruption may also show enhancement. The hyperintensity and contrast enhancement may persist for months, making the differentiation of acute versus chronic disruption impossible. Disk degeneration may lead to displacement of the disk material beyond the normal intervertebral disk space, often through radial disruption of the annulus fibrosis. Normal disk space is defined craniocaudally by the vertebral body endplates, and circumferentially by the ring apophysis of the vertebral bodies. Disk displacement may be described as a bulge or herniation, depending on the percentage of circumference involved (Fig. 4.1.6). Disk bulging refers to diffuse displacement of disk material beyond the normal disk space, and covers more than 50% of the normal disk space circumference (i.e., more than 180° of the circumference) (Fig. 4.1.7). Disk displacement covering 50% or less of the circumference is called herniation, which may be further subcategorized into broad-based herniation (covering 25–50% of disk space circumference) and focal herniation (covering less than 25% of the circumference). Disk protrusion and extrusion are terms that describe the shapes of herniation. They can be applied to either focal or broad based herniation to further characterize the morphol-

Fig. 4.1.3  Vacuum disk phenomenon at L4–L5 and L5–S1 levels. a T1-weighted. b T2-weighted. c Lateral radiograph. The gas seen on the lateral radiograph (black arrows in c) correlates with the linear hypointensity on both T1- and T2-weighted images (white arrows in a). On the T2-weighted image, the star-burst shaped

hyperintensity at the superior endplate of S1 is caused by image artifact (b, white arrow). This patient also has grade 1 anterolisthesis of L4–L5 secondary to degenerative facet disease, and degenerative endplate changes at L4–L5 and L5–S1

4.1  Extradural Diseases of the Spine

ogy. When the width of the base of the disk herniation is greater than any other measurements in the same plane of the herniation, it is called a protrusion (Fig. 4.1.8). When any measurement of the herniation is greater than the measurement at its base, the herniation is described as an extrusion (Fig. 4.1.9). Migration refers to displacement of disk material away from the site of extrusion, regardless of continuity. Very often, disk extrusion is seen associated with superior or inferior migration of disk material. In reporting, it is important to make a note of disk migration since failure to recognize migrated disk fragments may lead to surgical failure. When the disk extrusion is separated from the parent disk, it may be further categorized as a sequestration (Fig. 4.1.10). Sequestered disk often demonstrates T2 prolongation compared to its disk of origin. This may be secondary to the presence of granulation tissue, immune response, or inflammation (Masaryk et al. 1988). Most disk

Fig. 4.1.4  Vacuum disk with fluid. a Sagittal T2-weighted image demonstrates high signal intensity within the L4–L5 disk, consistent with vacuum disk containing fluid (white arrow). The L5– S1 disk demonstrates protrusion and a small annular disruption at the disk margin (black arrow). Vacuum disk containing air is seen as linear hypointensity within the L5–S1 disk. Note the loss of height and normal signal of both L4–L5 and L5–S1 disks. b Axial T2-weighted image of L4–L5 disk showing fluid within the disk (white arrows)

Fig. 4.1.5  Annular disruption. On this T2-weighted image, there is hyperintensity at the margin of the L4–L5 disk (arrow), consistent with annular disruption. Note the loss of normal signal intensity in all the visualized disks secondary to degeneration. There is disk bulging at L3–L4 and disk protrusion at L4–L5

543

544

4  Spine and Spinal Canal Fig. 4.1.6  Schematic diagram illustrating the nomenclature and classification of disk disease as recommended by the Combined Task Forces of the North American Spine Society, American Society of Spine Radiology, and American Society of Neuroradiology (Fardon and Milette 2001)

Fig. 4.1.7  Disk bulging. On this proton density–weighted image, there is diffuse and circumferential extension of disk material beyond the margin of the ring apophysis (white arrowheads). Small asymmetry of the bulging is noted at the right foraminal region (see Fig. 4.1.12 for location classification), which can also be described as broad-based protrusion. In conjunction with the bilateral degenerative facet changes and hypertrophy of the ligamentum flavum (white stars), the disk bulging is causing moderate narrowing of the left neural foramen and severe narrowing of the right neural foramen

Fig. 4.1.8  Disk protrusion. Axial T2-weighted image demonstrates broad-based disk protrusion at the region of the left neural foramen (white arrows). There are bilateral facet arthropathy and narrowing of the neural foramina. A small annular disruption is also noted at the posterior margin of the disk (black arrow)

4.1  Extradural Diseases of the Spine

Fig. 4.1.9  Disk extrusion. a Sagittal T2-weighted image. There is a C5–C6 disk extrusion with inferior migration to the C6 level (white arrow). This has a “toothpaste” appearance, with disk herniated through a narrow base at the disk level. The cervical spinal cord is indented at the level of C5–C6. Note that the signal of herniated disk may be different from the parent disk.

b Axial T2-weighed image. Extruded disk is indicated by the white arrow. c Disk extrusion at the same level on 3D gradientecho sequence (black arrow). Note the typical hyperintensity of disk on the gradient-echo image contrasting with the hypointensity of the osseous structures, making it easier to differentiate disk herniation from osteophyte

sequestrations are seen in the epidural space, but rarely, they may migrate into the intradural space. In the absence of surgical history, infection or neoplasm, evaluation of degenerative disk disease rarely requires the use of contrast material. On contrast-enhanced images, disk herniation normally shows no enhancement within the disk. However, peripheral enhancement is common, probably due to the presence of granulation tissue or displacement of the epidural veins (Fig. 4.1.10b). Intuitively, one may think of disk herniation as a progressive process. However, it has been shown that the morphology of disk herniation actually improves with time in a large proportion of patients (Fig. 4.1.11). In a study by Bozzao (1992), up to 63% of patients had reduction of disk herniation by more than 30% over a followup period of 6–15 months. The position of disk herniation in the plane of the disk can be further specified (Fig. 4.1.12). Disk herniation within the spinal canal can be specified as central or paracentral, depending on its predominant position in the spinal canal. Herniation with the major component in the neural foramen is specified as foraminal. If the disk material is far laterally beyond the margin of neural foramina, it may be called extraforaminal. Disk herniation implies disruption of the annulus. However, the posterior longitudinal ligament may or may not be intact. When the posterior longitudinal ligament is still intact, a disk herniation may further be specified as subligamentous.

One problem that may arise is the differentiation between acute versus chronic herniation, especially when a patient presents with acute symptoms. On imaging, the differentiation can be difficult. However, the presence of calcification, ossification, or gas will suggest chronicity. On the contrary, increased disk signal intensity on T2-weighted images suggests acute disk herniation. Disks may also herniate superiorly or inferiorly through the vertebral endplates into the vertebral bodies. Intravertebral disk herniation is often called Schmorl’s node (Fig. 4.1.13). It is usually hypointense on T1-weighted images and hyperintense on T2-weighted images. Occasionally, adjacent bone marrow edema and enhancement attributed to vascularization may be seen. Schmorl’s nodes are very common and most cases are asymptomatic. Intravertebral disk herniation demonstrating enhancement or bone marrow edema is more likely to be symptomatic, possibly because of its inflammatory nature (Stäbler et al. 1997). This type of intravertebral disk herniation can mimic infection or neoplasm and may pose a diagnostic difficulty (Fig. 4.1.14). 4.1.2.4.3  Facet Joints The facet joints of the spine are true synovial joints. They may demonstrate degenerative changes similar to other synovial joints in the rest of the body (Fig. 4.1.15). The bone may demonstrate hypertrophy and develop osteo-

545

546

4  Spine and Spinal Canal

Fig. 4.1.10  Disk extrusion with sequestration (free fragment). This 57-year-old male presented with acute back pain and urinary retention. a Proton density weighted image demonstrates large disk herniation (white plus sign) at the left lateral recess. The base of the herniation is narrow, consistent with the definition of extrusion. The thecal sac (S) is compressed and displaced to the right. The left S1 exiting nerve root is compressed and effaced. However, the right S1 nerve root is well seen as a small rounded structure (white arrow). b Sagittal T1-weighted image with fat suppression and contrast enhancement demonstrates the disk herniation (black arrow) with inferior migration. The hypointense thecal sac (white arrow) is compressed and displaced. There is thickened enhancing tissue surrounding the disk extrusion, probably representing granulation tissues and displaced epidural veins. The herniated disk is mostly separated from the parent disk, suggesting sequestration

phytes as a result of mechanical stress. Osteoarthritic changes typically seen in other joints can be seen in the facet joints, and include subchondral sclerosis, thinning of the articular cartilage, joint space narrowing, and subchondral cyst formation. The joints may contain fluid or gas as a result of vacuum phenomenon similar to that seen in the intervertebral disks. The anterior aspect of the facet joints and the laminae are covered by ligamentum flavum. Anterior buckling or hypertrophy of the ligamentum flavum is often seen associated with degenerative changes of the facet joints. The uncovertebral joints are additional joints seen in the lower five cervical vertebral bodies that may be associated with degenerative disease. These joints have features of both cartilaginous and synovial joints. The uncinate process may undergo hypertrophy, leading to narrowing of the neural foramina (Fig. 4.1.16). Synovitis may be seen in degenerative facet disease. The inflammatory changes give rise to increased signal intensity on T2-weighted images and contrast enhancement. Involvement may include osseous structures of the facet, adjacent lamina, and surrounding soft tissues. The imaging appearance can simulate infection (Fig. 4.1.15b, Fig. 4.1.17). Juxta-articular cysts may form at degenerative facet joints. These may include synovial cysts and ganglion cysts. Pathologically, synovial cysts are lined by synovial tissue and communicate directly with the joint space. Ganglion cysts do not have synovial lining but have connective tissue capsules. They do not communicate with the joint space and are often filled with myxoid material. In practice, synovial cysts and ganglion cysts are difficult to differentiate by imaging. These cysts are most common in the lower lumbar spine, particularly at the L4–L5 level, possibly due to the higher degree of mobility. They may also occur in the cervical spine, but are rare in the thoracic spine. Juxta-articular cysts often present as posterolateral epidural lesions, but may also be seen outside the spinal canal in the paraspinal region (Fig. 4.1.18). The MR signal intensity is variable, depending on whether they contain proteinaceous material or hemorrhage. Gas may be present in synovial cysts, as they communicate with facet joints that may contain gas due to the vacuum phenomenon. The cyst walls may contain hemorrhage or calcification. There may be contrast enhancement in the wall or surrounding soft tissues if inflammatory response is present. Juxta-articular cysts can mimic disk extrusion. However, the continuity of a lesion with facet joint that demonstrates significant degenerative changes can help to differentiate a Juxta-articular cyst from disk herniation (Fig. 4.1.19). 4.1.2.4.4  Ligaments Buckling and hypertrophy of the ligamentum flavum are commonly seen in degenerative spine disease (Fig. 4.1.7).

4.1  Extradural Diseases of the Spine Fig. 4.1.11  Resolution of disk extrusion. a Initial sagittal T2-weighted image shows a large disk extrusion at the L4–L5 level. Degenerative changes at the L3–L4 and L5–S1 levels include loss of disk height and disk signal intensity. b Sagittal T2-weighted image at 6-week follow-up shows dramatic reduction of the disk extrusion, with only a small residual disk herniation

Fig. 4.1.12  Location classification of disk disease on an axial image. C central, P paracentral, F foraminal, E extraforaminal (far lateral)

Fig. 4.1.13  A round lesion is seen at the superior endplate of T12. On the STIR image (a), this lesion is slightly hyperintense. On the contrast-enhanced T1-weighted image (b), the component immediately adjacent to the T11–T12 disk does not enhance, but there is an enhancing component deeper in the vertebra (arrow). This lesion is stable on a follow-up study 15 months later, and is compatible with an intravertebral herniation (Schmorl’s node)

547

548

4  Spine and Spinal Canal

The ligamentum flavum is normally less than 4 mm in thickness. Ossification of the ligamentum flavum may also be present and can represent a normal variant. Ossification of the posterior longitudinal ligament (OPLL) occurs most commonly in older patients aged

50–70 years, and more often in men than in women, with a ratio of approximately 2:1. The exact etiology is unknown, but some believe the calcification and ossification are induced by repetitive stress on the ligament. OPLL has a strong association with diffuse idiopathic skeletal Fig. 4.1.14  Schmorl’s node with abnormal enhancement. a Sagittal proton density weighted image demonstrates a Schmorl’s node at the inferior endplate of T12. b Sagittal post-gadolinium T1-weighted image. The presence of granulation tissues gives rise to enhancement extending into the T12–L1 disk space, simulating early diskitis

Fig. 4.1.15  Facet arthropathy. Bilateral L5–S1 facet arthropathy (white arrows) on a axial T2-weighted image, b post-gadolinium contrast-enhanced image, and c CT image obtained one day after the MR examination. Because of differences in angulation, the laminas and spinous process are not seen on this CT image. The degenerative changes include facet hypertrophy, subchondral sclerosis and cyst formation. Inflammatory enhancement is commonly seen associated with facet arthropathy (b) and does

not necessarily indicate infection. There is a vacuum phenomenon, which is better seen on CT. Increased fluid may also be present in the synovial joint. The facet hypertrophy can sometimes become so large that it results in severe compression of the neural foramen or spinal canal, as illustrated in d and e (arrow) from a different patient. There is severe hypertrophy of the right T5–T6 facets compressing the thoracic spinal cord

4.1  Extradural Diseases of the Spine

hyperostosis (DISH), and calcification/ossification of the ligamentum flavum, suggesting they may also share similar pathophysiology. Ossification of the posterior longitudinal ligament is most often found in the cervical spine, particularly in the C3–C5 region. Thoracic and lumbar

involvement may also occur, and in these cases are most often seen at the T4–T7 levels and the L1–L2 levels, respectively. Ossification of the posterior longitudinal ligament is well demonstrated on lateral radiographs and CT images (Fig. 4.1.20a). On MR imaging, ossification of the posterior longitudinal ligament is seen as a linear vertical band of hypointensity between the vertebral bone marrow and the dural sac on all imaging sequences (Fig. 4.120b,c). The hypointensity is exaggerated on gradient-echo imaging sequences due to increased susceptibility effect and therefore generates very high contrast against the bright CSF signal on these types of sequences. Within the area of ossification, increased or intermediate signal intensity corresponding to marrow fat may be seen on T1-weighted and fast spin-echo imaging sequences. Three patterns of OPLL have been described: segmental, continuous, and mixed. Segmental distribution represents involvement of one or several vertebral levels without involvement of the intervening levels; continuous distribution represents involvement in an uninterrupted fashion. These three patterns occur with similar frequency. 4.1.2.4.5  Malalignment Secondary to Degenerative Disease

Fig. 4.1.16  Degenerative changes of the right uncovertebral joint with hypertrophy of the uncinate process and posterior osteophyte (white arrows), resulting in narrowing of the right neural foramen

Fig. 4.1.17  Facet arthropathy associated with inflammatory changes. Axial gadolinium-enhanced image with fat suppression demonstrates abnormal enhancement surrounding the right L3–L4 facet joint

Spondylolisthesis represents subluxation of a vertebral body in relation to the next inferior vertebral body. Anterolisthesis refers to anterior subluxation of a vertebral body in relation to the one below; retrolisthesis refers to posterior subluxation of a vertebral body in relation to the one below. The severity of spondylolisthesis can be classified into five grades. Grade I refers to subluxation meas­uring less than 25% of the anterior-posterior diameter of a vertebral body, grade II refers to 25–50% subluxation, grade III represents subluxation between 50 and 75%, grade IV represents subluxation of 75–100%, and grade V represents complete displacement greater than 100%. Anterolisthesis may be secondary to congenital or acquired/traumatic defects of the pars interarticularis, often referred to as spondylolysis (Jinkins et al. 1992). Severe degenerative disease of the intervertebral disk and related posterior facet joints may also lead to anterolisthesis. The anterolisthesis caused by facet degeneration is probably due to joint laxity. Anterolisthesis resulting from degenerative disk and facet disease is usually limited to grade I (Fig. 4.4.3), as opposed to anterolisthesis secondary to spondylolysis which can be associated with all grades. Other features that help to support degeneration as a cause of anterolisthesis include the presence of degenerative changes in the facet joints and disk, lack of evidence for pars interarticularis defects, spinal canal narrowing as opposed to widening at the level of subluxation (Ulmer et al. 1994), and involvement of multiple intervertebral levels. Pars defects occur most often at the

549

550

4  Spine and Spinal Canal

Fig. 4.1.18  Synovial cyst. a,b Two contiguous axial T2-weighted images at the L3–L4 level demonstrate a cystic epidural lesion (between white arrowheads) at the right facet joint. The continuity with the synovial space of the facet joint is better seen in b. Note the increase of fluid in the synovial space (white arrow

in a) associated with facet arthropathy. c Sagittal T2-weighted image. d Sagittal post-gadolinium fat suppressed image. Peripheral enhancement of the cyst is noted (arrow), suggesting inflammatory response

4.1  Extradural Diseases of the Spine

Fig. 4.1.19  A calcified synovial cyst. a–c Sagittal and axial T2-weighted images demonstrate a hypointense structure in the spinal canal. This lesion is located at the L4–L5 disk level. Although this lesion is adjacent to the right facet joint, which demonstrates significant degenerative changes, based on MR imaging alone, it is difficult to differentiate confidently synovial cyst from disk herniation. For consideration of a synovial cyst,

one may be puzzled by the almost complete absence of signal intensity on T2-weighted images. CT images (d,e), however, demonstrate a calcified cystic lesion arising from a severely degenerative right L4–L5 facet (black arrows), supporting the diagnosis of a synovial cyst. Posterior synovial cysts containing air are also demonstrated on CT (f, white arrows) but not visualized on the MR study, highlighting the complementary role of CT

level of L4–L5. When anterolisthesis is present, the pars interarticularis should be carefully examined to avoid missing a diagnosis of pars interarticularis defects. This is especially true when the anterolisthesis is more severe, as the higher grades make degenerative disease an unlikely cause (Fig. 4.1.21). Pars defects can be subtle on spinecho or fast spin-echo images. Gradient-echo imaging may provide a better depiction of pars defects because of the higher contrast between the actual defect and the adjacent dark bony cortices. If in doubt, CT should be considered as it usually provides better depiction of osseous structures that help to identify the pars defect. Retrolisthesis is usually secondary to degenerative disk disease. It is usually limited to grade I.

Disk bulging or pseudo-bulging often accompanies spondylolisthesis. Pseudo-bulging is caused by the thecal sac draping over nondisplaced disk at the level of slippage. Both disk bulging or pseudo-bulging may result in narrowing of the neural foramen. 4.1.2.4.6  Spinal Canal and Neural Foraminal Stenosis The presence of disk bulging/herniation, degenerative changes of the facet and uncovertebral joints, degenerative hypertrophy of the ligamentum flavum and malalignment can all lead to narrowing of the spinal canal or

551

552

4  Spine and Spinal Canal

Fig. 4.1.20  Ossification of the posterior longitudinal ligament. a CT demonstrating ossification of the posterior longitudinal ligament from C2 to C3 levels (white arrows). The ossification correlates with linear hypointensity at the same levels on sagittal and axial T2-weighted MR images (black arrows, b,c). Note

that on the sagittal MR image, there is severe degenerative disk disease with facet degenerative change and mild retrolisthesis at the C3–C4 intervertebral level, resulting in chronic spinal canal stenosis and abnormal signal intensity in the spinal cord. Similar findings are seen at the C5–C6 and C6–C7 levels

Fig. 4.1.21  Anterolisthesis secondary to spondylolysis. a Sagit­ tal proton density weighted image demonstrates grades II to III anterolisthesis. The L5 vertebral body is subluxed anteriorly and inferiorly relative to S1 (large white arrow). Because of the severity, this is unlikely to be solely secondary to degenerative

disease. Careful evaluation of the other images reveals defects of the bilateral L5 pars interarticularis (black arrows, b,c). For comparison, white arrow in b demonstrates a normal-appearing pars interarticularis at L4. A small disk protrusion is also noted at L4–L5 (small white arrow in a, white arrow in c)

4.1  Extradural Diseases of the Spine

neural foramina. Spinal canal stenosis can be assessed on an axial image (Fig. 4.1.22). The central spinal canal can be narrowed anteriorly by disk bulging/herniation and vertebral osteophytes. Posteriorly, it may be narrowed by facet disease and ligamentum flavum hypertrophy. Epidural lipomatosis can also lead to spinal canal stenosis due to an over-abundance of epidural fat (Hierholzer et al. 1996) (Fig. 4.1.22b,c). It may be caused by excessive exogenous or endogenous steroids, or idiopathic. Grading of spinal canal stenosis can be performed according to the recommendation of the Combined Task Forces of the North American Spine Society, American Society of Spine Radiology, and American Society of Neuroradiology (Fardon and Milette 2001). Spinal canal compromise of less than one third of the normal canal is graded as “mild”; between one and two thirds is “moderate”, and over two thirds is “severe.” An alternative way of looking at spinal canal stenosis is to assess the distortion of the normal round–oval shape of the spinal canal and thecal sac. With worsening stenosis, the spinal canal and thecal sac may be flattened and become triangular. There may be effacement of the cerebrospinal fluid space located between the degenerative processes causing the stenosis and the spinal cord or nerve roots. The lateral recess is best assessed on axial images. It is defined anteriorly by the vertebral body and interverte-

bral disk, posteriorly by the horizontal segment of the superior articular facet and ligamentum flavum, and laterally by the pedicle. The lateral recess may be narrowed by degenerative hypertrophy of the facet joint (Fig. 4.1.22a). Neural foraminal stenosis can be assessed on axial images and lateral sagittal images, using a semi-quantitative grading scheme similar to that used for the central spinal canal. Stenosis of the neural foramen can be caused by disk bulging/herniation, osteophytes of facet joints and uncovertebral joints, and ligamentum flavum hypertrophy/buckling (Figs. 4.1.22a, 4.1.23). In some patients, the spinal canal is narrowed congenitally, as seen in short pedicles or achondroplasia. This is often demonstrated as increasingly short pedicles and spinal canal narrowing as one goes from the superior to the inferior lumbar spine. Congenital spinal stenosis rarely leads to symptoms on its own. However, degenerative changes can exacerbate the stenosis, leading to an increased risk of symptomatic spinal stenosis caused by spondylosis. Severe spinal canal stenosis can lead to mass effect and ischemia of the spinal cord. Acute mass effect and ischemia may lead to edema in the spinal cord. Irreversible injury may result from the spinal canal stenosis, eventually leading to myelomalacia (Figs. 4.1.2e, 4.1.20b). Although atrophy of the affected segment of the spinal cord

Fig. 4.1.22  Spinal canal stenosis. a Axial T2-weighted image. There are disk extrusion (white arrow) at the right lateral recess and diffuse disk bulging seen as a dark rim surrounding the posterior and lateral circumference of the ring apophysis. Thickening of the ligamentum flavum (white arrowhead) and hypertrophy of the facet joints (white open arrow) are shown. These degenerative changes narrow the spinal canal anteriorly and posterolaterally. The spinal canal is now seen flattened later-

ally. The bilateral neural foramina are also severely narrowed. b At the L3–L4 level of a different patient (T2-weighted image), prominent epidural fat is seen (L). Degenerative hypertrophy of the bilateral facet joints (black arrow) and diffuse disk bulging are present. These processes act synergistically to narrow the spinal canal (note the small thecal sac) and the bilateral neural foramina (white arrows). c Sagittal T2-weighted image demonstrates prominent epidural fat narrowing the thecal sac (white arrows)

553

554

4  Spine and Spinal Canal

suggests myelomalacia, both acute and chronic processes may manifest as T2 prolongation and therefore can be difficult to differentiate based on imaging alone.

Among patients who have undergone lumbar spine surgery, 10–40% fail to have relief of symptoms (Fritsch et al. 1996). The causes of pain in the postoperative spine can be categorized into immediate complications of surgery, delayed complications of surgery, residual disease, and recurrent disease. Immediate complications include hematoma, infection, and pseudomeningocele. The MR features of hematoma and infection in the postoperative spine are similar

to those resulting from other causes, and have been discussed in other sections. Pseudomeningocele results from iatrogenic tear of the dura, resulting in the extradural collection of cerebrospinal fluid. This is often seen as a cerebrospinal fluid posterior to the thecal sac at the level of surgery (Fig. 4.1.24). A fluid–fluid level may be seen from layering of debris or blood products. A word of caution is that a small collection of fluid is common immediately after surgery and may bear little clinical significance. A fluid collection may demonstrate signal intensities similar to those of cerebrospinal fluid if it is serous, or behave like hemorrhage if it contains hemorrhagic products. Delayed complications include arachnoiditis, epidural fibrosis, and stenosis resulting from postoperative bony overgrowth. The pathogenesis of arachnoiditis is not well

Fig. 4.1.23  Right foraminal stenosis. The sagittal T1-weighted image in a right lateral location shows an L4–L5 disk protrusion narrowing the middle and inferior aspect of the right neural foramen (white arrow). The disk herniation displaces the right L4 nerve root superiorly. Note the loss of the normal keyhole appearance of the neural foramen at this level, compared to the normal configuration demonstrated at the L3–L4 level (black arrow)

Fig. 4.1.24  Pseudomeningocele. Sagittal STIR image shows a large fluid collection developed posterior to the C6–C7 level. This patient has a history of traumatic cervical spine injury and C6–C7 anterior fusion

4.1.2.4.7  Postoperative Spine

4.1  Extradural Diseases of the Spine

understood. It is an inflammatory process that causes the nerve roots of the cauda equina to adhere to each other or to the meninges. On MRI, one may see central clumping of nerve roots, or adherence of nerve roots to the wall of the thecal sac resulting in an apparent “empty thecal sac” (Fig. 4.1.25). In end-stage arachnoiditis, an inflammatory mass filing the thecal sac may be present. Contrast enhancement of the nerve roots is variable, but may only be minimal or mild. Immediately after surgery, it is common to see central clumping of nerve roots of the cauda equina. However, this reversible central clumping should resolve by 1 to 6 weeks after surgery. In imaging of patients with suspected residual or recurrent disease, the main issue is to differentiate residual or recurrent disk herniation from postsurgical epidural fibrosis (scar). Although both can give rise to recurrent symptoms, residual or recurrent disk disease is more amenable to further surgical treatment and therefore the distinction is still important. If imaging is performed shortly after the intravenous administration of gadolinium contrast, disk material will not enhance, while scar tissue in the epidural space will enhance homogeneously (Fig. 4.1.26). However, if imaging is delayed, contrast material will diffuse into the disk and the disk material may also enhance, making it difficult to differentiate the two entities. Both contrast-enhanced CT and contrast-en-

hanced MR may be performed to differentiate scar tissue from disk material. However, MR is by far the most accurate technique. Accuracy for non-enhanced CT is approximately 43%, while contrast-enhanced CT increases diagnostic accuracy to approximately 74% (Braun et al. 1985). Contrast-enhanced MR increases the accuracy further to over 96% (Ross et al. 1990). Spinal fusion is often a component of spinal surgery for degenerative disk disease. This may result in alteration of the biomechanics, leading to an increased risk of degenerative disease at other levels. The evaluation of the postoperative spine should therefore include the evaluation of proliferation of disease at other vertebral levels.

Fig. 4.1.25  Arachnoiditis. a Axial T2-weighted image demonstrates clumping of the nerve roots (white arrow). b Sagittal T2-weighted image illustrates the clumping throughout the course of the nerve roots, resulting in a “rat-tail” appearance of

the cauda equina and “empty” thecal sac. c Sagittal T1-weighted contrast-enhanced image with fat suppression illustrates mild enhancement of all the nerve roots throughout their course (small white arrows)

4.1.3  Extradural Spine Infection 4.1.3.1  Introduction Extradural infection of the spine may involve the vertebrae, intervertebral disks, epidural space, paravertebral soft tissues, or facet joints. Vertebral osteomyelitis or disk space infection usually arises from hematogenous spread. Epidural abscess may occur in the absence of disk space infection from hematogenous spread or direct extension from instrumentation. Epidural or paravertebral infec-

555

556

4  Spine and Spinal Canal

Fig. 4.1.26  Postoperative epidural fibrosis and residual–recurrent disk herniation. This patient had a L4–L5 right hemi-laminectomy and diskectomy for disk extrusion at the right lateral recess. Follow-up study demonstrates a lesion at a location similar to the preoperative study. a On the sagittal T2-weighted image, it is difficult to determine whether the lesion (white arrow) represents residual or recurrent disk herniation versus scar tissue. b Axial contrast-enhanced T1-weighted image demon-

strates enhancing soft tissues extending from the surgical track to the right paracentral epidural space, indicating the lesion consists of primarily epidural fibrosis (arrows). c The next image superiorly from the same imaging sequence shows a very small non-enhancing component at the posterior margin of the vertebral body (arrowhead). This is at the same location as the previous disk herniation, consistent with a small residual–recurrent disk herniation

tion may arise from extension of disk space infection or vertebral osteomyelitis. Septic facet joint arthritis may also arise from hematogenous spread, instrumentation, or extension of contiguous infection (Hadjipavlon et al. 2000; Mader et al. 1999). Risk factors of extradural infection include diabetes mellitus or other chronic medical illness, elderly patients, immunocompromised hosts, chronic steroid use, rheumatoid arthritis, urinary tract instrumentation, and trauma (Ledermann et al. 2003; Sharif 1992).

In general, extradural infection leads to inflammatory changes in the bone marrow and soft tissues and these can be readily demonstrated on fluid-sensitive MR pulse sequences and contrast-enhanced imaging series. Imaging sequences in a typical spine infection MR protocol include sagittal T1-weighted spin-echo and fast spin-echo T2-weighted images with fat suppression or STIR images to better assess the marrow edema. Axial T1-weighted and fast spin-echo T2-weighted images are routinely obtained. Post gadolinium sagittal and axial T1-weighted fat-suppressed images are useful for the evaluation of disk enhancement and epidural or paraspinal collections. Fat suppression is very useful to increase the conspicuity of enhancing abnormality.

4.1.3.2  Imaging Techniques MR is the imaging modality of choice for diagnosing spondylodiskitis, having greater sensitivity than SPECT gallium and bone scanning, and comparable specificity and accuracy (Love et al. 2000). Plain radiographs, which are usually normal for 2–8 weeks after symptom onset, may show vertebral demineralization and endplate destruction followed by increased bone density, disk space narrowing and paraspinal soft tissue density. In late stages, the disk may fuse. CT findings include paraspinal soft tissue swelling, endplate erosion or osteosclerotic changes, and occasionally bony sequestra. Contrast enhancement of the disk, marrow, and paravertebral soft tissues can be seen. Reformatted sagittal and coronal images allow for assessment of spinal deformity.

4.1.3.3  Infectious Spondylodiskitis The usual pattern of spinal infection is spondylodiskitis (vertebral osteomyelitis/disk space infection). Most commonly, vertebral osteomyelitis/diskitis is secondary to pyogenic infection or tuberculosis. The lumbar spine is most commonly involved, followed by the thoracic and cervical spine (Hadjipavlon et al. 2000; Weinberg and Silber 2004). Clinically, vertebral osteomyelitis usually presents with back pain, focal tenderness, and spasm. Fever is variable. Motor or sensory deficits may occur in some patients (Maiuri et al. 1997; Thrush and Enzman 1990). The

4.1  Extradural Diseases of the Spine

erythrocyte sedimentation rate is frequently increased in more than 90% of patients and C-reactive protein is also commonly elevated. The white blood cell count is variable. Confirming the presence of spondylodiskitis is usually accomplished with imaging, although determination of a causative organism is required to guide appropriate antibiotic treatment. Blood cultures are positive in 25–60% of cases, but biopsy to determine the causative organism is usually required for direction of appropriate antimicrobial therapy. Staphylococcus aureus is the most common causative organism and contributes to greater than 50% of cases. Pseudomonas may be encountered in drug abusers. In patients with sickle cell disease, salmonella is the classic infection although staphylococcus aureus is still the most common causative organism. 4.1.3.3.1  Vertebral Osteomyelitis/Diskitis In adults, vertebral osteomyelitis–diskitis (Fig. 4.1.27) most often arises from arterial inoculation of the endplate of one vertebra, with subsequent spread to the disk and adjacent vertebrae. Infection arises in the vertebral endplate owing to the better arterial vascular supply to the endplate as opposed to the equator, and the presence of end metaphyseal arteries, making it susceptible to bacterial seeding. Segmental arteries typically supply the disk and adjacent vertebrae (Rattcliffe 1980). In infants and young children, though not in adults, the disk is vascularized and can be initially infected. Characteristic MR findings of osteomyelitis/diskitis include (1) disk space narrowing with hypointense signal on T1-weighted images and variable but usually hyperintense signal on T2-weighted images, and (2) hypointense marrow signal on T1-weighted images in the portions of the vertebral abutting the disk with hyperintense signal on T2-weighted images secondary to marrow edema. Typically there is loss of the vertebral endplate cortex. There may be loss of the normal intranuclear cleft (Dagirmanjian et al. 1996; Modic et al. 1985). This is a useful finding when present in the lumbar spine but cannot be applied to the cervical or thoracic spine (Ledermann et al. 2003). The disk shows variable to diffuse post gadolinium enhancement while the vertebral endplates characteristically enhance avidly. Paraspinal and epidural phlegmon or abscess may be present, typically isointense to muscle on T1-weighted images and hyperintense on T2-weighted images. Phlegmon enhances diffusely, while abscess typically demonstrates the peripheral enhancement of a fluid collection. Imaging findings are not consistent, however, and may lag behind the onset of symptoms. In the case of spondylodiskitis, disks can be isointense to adjacent disks on T1-weighted images. On T2-weighted images, the disks may show fluid-equivalent signal intensity to adjacent

disks or less commonly iso- to hypointense to adjacent disks. With contrast, rim or diffuse disk enhancement is commonly found although enhancement may also be focal or absent. While disk height is usually decreased, it may be normal or occasionally increased. Vertebral body enhancement, though most often diffuse, can be variable. Occasionally, marrow signal alteration may be found in only one endplate, i.e., on only one side of the infected disk. Destruction of the vertebral endplates may be absent in a significant proportion of patients. Pseudo-sparing of the endplate related to chemical shift artifact may occur but can be avoided by selecting the phase encoding axis in the craniocaudal direction. Multilevel involvement sometimes occurs in pyogenic infection but is more commonly found in tuberculosis (Ledermann et al. 2003). The most sensitive criteria for disk space infection include (1) paraspinal or epidural inflammatory tissue (2) hyperintensity or fluid equivalent signal in the disk on T2-weighted images, (3) erosion or destruction of vertebral endplates on T1-weighted images, and (4) disk enhancement, particularly when found in combination. Many conditions can mimic infectious spondylodiskitis. Disk space infection/osteomyelitis must be distinguished from degenerative disk disease with edematous (type 1) marrow endplate changes (Modic et al. 1988). In degenerative disk disease, the disk characteristically remains hypointense on T2-weighted images with no or only mild linear disk enhancement, and there is absence of destruction of the vertebral endplates. In the unusual case of a T2 hyperintense degenerative disk with endplate irregularity, biopsy may be necessary. Spinal neuropathic arthropathy (Charcot spine) is the result of repeated trauma in the setting of diminished sensation. It is found most commonly in association with diabetes mellitus or spinal cord injury. It has imaging findings similar to those of disk space infection, including hyperintensity on T2-weighted images. Findings more common in spinal neuropathic arthropathy include spondylolisthesis, a vacuum disk, debris, and disorganization, associated facet joint disease, a diffuse signal intensity pattern in the vertebral bodies, and a rim of gadolinium disk enhancement (Wagner et al. 2000). Spondyloarthropathy associated with chronic hemodialysis has disk narrowing and destructive endplate changes, which mimic those of infectious diskitis. However, in addition to the appropriate clinical history of chronic dialysis, the disk remains intermediate to low in signal intensity on T2-weighted images, and the vertebral marrow is usually hypointense on both T1 and T2-weighted sequences. Amyloid deposits are found on disk biopsy (Filipo et al. 1996). Tuberculous spondylitis may be indistinguishable from pyogenic infection. Some distinguishing features of tuberculous spondylitis that may be present include vertebral collapse with gibbus deformity, subligamentous spread of infection, posterior element involvement, and

557

558

4  Spine and Spinal Canal

a large paraspinous abscesses out of proportion to vertebral involvement. In vertebral metastatic disease, the disk is typically spared. The vertebral involvement is usually noncontiguous, and the posterior elements are commonly affected. When clinical symptoms have improved or resolved following institution of adequate antibiotic therapy, MR findings indicative of active spondylodiskitis may still persist or progress during the healing process. Reduction in inflammatory soft tissue changes is an expected early finding in healing disk space infection. Gadolinium enhancement may persist or increase for months following resolution of the soft tissue changes and symptom resolution (Gillams et al. 1996). In healed disk infection, there

is persistent disk narrowing, although the marrow edema resolves and the disk is hypointense on T2-weighted images. Fatty marrow deposition may occur in the endplates. Normal postoperative MR findings following diskectomy may be difficult to distinguish from postoperative diskitis. Gadolinium enhancement of the marrow is more common in the setting of disk space infection. Disk enhancement and enhancement of the posterior annulus fibrosis also occur more commonly in the setting of disk space infection, although there is a large overlap with asymptomatic patients. The presence of all three of these findings is highly suggestive of disk space infection (Boden et al. 1992).

Fig. 4.1.27  Pseudomonas spondylodiskitis. a Sagittal T2weighted image shows increased signal at the center of the T9–T10 disk (solid arrow). Subtle hyperintensity is seen in the T9 and T10 vertebral bodies. The hyperintensity at the T7 inferior endplate (open arrow) represents Modic type 1 degenerative change. b Post-gadolinium T1-weighted image. The abnormality of the T9 and T10 vertebral bodies is much more obvious. Note the mild epidural extension. There is strong enhancement in the center of the disk. c Axial post-gadolinium T1-weighted image shows paraspinal lesions (white open arrows) and epidural lesions (black arrows). There is extension into the neural foramina bilaterally

4.1  Extradural Diseases of the Spine

4.1.3.3.2  Tuberculous Spondylitis Tuberculous spondylitis, caused by mycobacterium tuberculosis, accounts for more than half of all musculoskeletal tuberculosis (Okada et al. 2005). Clinically, patients may present with chronic back pain and focal tenderness. Fever is less common than with pyogenic osteomyelitis. Symptom onset is more insidious than with pyogenic infection. The diagnosis is often delayed for months following onset of infection. Neurologic complications from epidural abscess or phlegmon and vertebral collapse are more common than with pyogenic osteomyelitis. Neurologic complications may include paraparesis or paralysis, sensory disturbances, or bowel and bladder dysfunction. As with pyogenic infection, hypointense T1 and hyper­ intense T2 or STIR signal abnormality involving two adjacent vertebrae and the disk is common (Fig. 4.1.28). Vertebral, disk and paravertebral enhancement are seen. Epidural or paraspinous enhancing phlegmon or granulation tissue are found, and there may be peripheral enhancement of epidural or paraspinous abscess. Imaging findings are often indistinguishable from those of pyogenic infection, although there are often helpful distinguishing characteristics. Tuberculous osteomyelitis most commonly arises from hematogenous spread to the anterior vertebral body from a primary infection of the lungs or genitourinary tract. As with bacterial infection, two contiguous vertebral bodies are characteristically involved, although there may be relative sparing of the disk owing to the absence of proteolytic enzymes in tuberculosis. The thoracolumbar junction or thoracic spine is more commonly involved than the lumbar or cervical spine in tuberculous spondylitis, as opposed to bacterial infection. Simultaneous involvement of several vertebral bodies is more common in tuberculous infection. Skipped vertebral involvement may occur from subligamentous spread of infection beneath the anterior longitudinal ligament (Jung et al. 2004). Periosteal stripping and subligamentous infection leads to the vertebrae becoming avascular and susceptible to infection. Pressure of the subligamentous infection and ischemia can produce anterior vertebral scalloping. Although anterior vertebral infection is common, isolated posterior element tuberculous infection may occur. Differentiation of isolated tuberculous vertebral involvement from metastasis or lymphoma may require biopsy. Vertebral collapse with gibbus deformity accompanied by relative disk sparing is characteristic of tuberculous infection. Other imaging findings found more often in tuberculous spondylitis include a large paraspinous abscess or phlegmon, which may extend over multiple vertebral levels. Calcification of chronic paraspinous abscesses can be seen on CT. Thin, smooth rim enhancement of abscess walls is found, as opposed to thick, irregular abscess wall enhancement of pyogenic osteomyelitis. Well-defined paraspinous abnormal signal is found in tuberculospon-

dylitis while ill-defined signal abnormality is usually seen in bacterial infection. In addition to tuberculous granulomatous infection, granulomatous infection of the spine may be caused by brucellosis, a gram-negative bacteria of the genus Brucella. This is uncommon though endemic in some parts of the world, including the Middle East. The lumbar spine is most often involved. Spinal infection may be focal or diffuse. In focal vertebral involvement, the disks and soft tissues are spared. Focal erosions are seen at the diskovertebral junction, and there may be endplate sclerosis. In diffuse brucellosis spondylitis, there is involvement of two adjacent vertebrae via subligamentous and vascular spread of infection, with subsequent endplate erosion and disk infection. There may be accompanying epidural granulation tissue. Vertebral destruction, however, is limited to the endplates (Sharif et al. 1989). 4.1.3.3.3  Fungal Spondylodiskitis Fungal osteomyelitis and disk space infection is usually found in immunocompromised hosts. It may be difficult to distinguish from pyogenic or tuberculous infection (Munk et al. 1997). Some distinguishing features include the relative lack of T2 hyperintensity in the vertebrae, although this is not a reliable finding. There may be absence of T2 hyperintensity in the disks, with preservation of the intranuclear cleft. Posterior element involvement is also more common in fungal than in pyogenic infection (Williams et al. 1999). Isolated vertebral involvement may be indistinguishable from neoplastic disease or tuberculosis. 4.1.3.4  Spinal Epidural Abscess Spinal epidural abscess occurs with an incidence ranging from 0.2 to 2 cases per 10,000. It has significant associated neurologic morbidity and mortality. With improved diagnosis and treatment, mortality has dropped from approximately 30% of patients from 1950 to 1960, to approximately 15% of patients from 1990 through 1997 (Reihaus et al. 2000). Treatment usually requires emergent surgical decompression and long-term antibiotic therapy (Soehl and Wallenfang 2002). Medical therapy alone is sometimes used when there is extensive multilevel involvement, significant postoperative risk, or when paralysis has been present for more than three days. Patients typically present with fever, spinal pain, and tenderness. Erythrocyte sedimentation rate and leukocyte count are usually elevated. Neurologic findings include radiculopathy, extremity weakness, paralysis, and loss of bowel and bladder control. Neurologic deficit most often arises from direct mechanical cord compression, but may also result from secondary cord ischemia.

559

560

4  Spine and Spinal Canal

Fig. 4.1.28  Tuberculous spondylodiskitis. a Sagittal T2-weighted image shows increased signal intensity in the T11 and T12 vertebral bodies. There is complete destruction of the disk and collapse of the disk space (long arrow). Small collections of fluid within the disk space are consistent with necrosis. There is an epidural lesion at the posterior margin of the T11 and T12 vertebral body, resulting in spinal canal stenosis. b Sagittal postgadolinium fat-suppressed T1-weighted image shows diffuse enhancement in T11 and T12 vertebral bodies. Abnormal enhancement is also seen in T10 vertebra and the posterior ele-

ments. There is a paraspinal lesion extending along the anterior margin of the vertebral column. Note the “skipping” of the T10– T11 disk, with enhancing lesions going around its anterior and posterior margins. The non-enhancing component surrounding the T11–T12 disk space (arrows) correlates with debris and necrosis seen in a. c Axial post-gadolinium fat-suppressed T1weighted image. Paraspinal lesion (white arrows), epidural lesion (black arrow), and central necrosis (white open arrow) are seen. d Axial CT image shows fragmentation of the vertebral endplate and paraspinal lesions (white arrows), correlating well with c

4.1  Extradural Diseases of the Spine

Epidural abscess may occur at any age, but it is most commonly found in patients aged 30–60 years and is more common in females. The most common risk factors associated with epidural abscess are diabetes mellitus, trauma, intravenous drug abuse, and alcohol abuse. Epidural abscess may arise from hematogenous dissemination of infection, direct extension of adjacent infection, or instrumentation. Any spinal segment can be involved, although epidural abscess is more common in the lower thoracic and lumbar region. Anterior spinal epidural infection usually results from adjacent diskitis/osteomyelitis, while posterior epidural abscess most commonly arises through hematogenous spread. Cutaneous infection is the most frequently encountered source. Iatro-

genic introduction of infection from spinal or surgical procedures may occur. Epidural anesthesia is associated with about a 5% risk of epidural infection. Staphylococcus aureus is the most common causative organism. Tuberculosis is the next most frequently encountered causative organism, usually occurring secondary to vertebral osteomyelitis/diskitis. MR is the imaging modality of choice. On MR, spinal epidural abscess is typically iso- to hypointense to the spinal cord on T1-weighted images, and hyperintense on T2-weighted or STIR images (Fig. 4.1.29). Gradientecho images usually show iso- to hyperintense signal. The signal intensity may be similar to cerebrospinal fluid, although on proton density weighted images, the epidural

Fig. 4.1.29  Vertebral osteomyelitis–diskitis with epidural abscess. This patient had streptococcus viridian septicemia. a,b Sagittal STIR and post-gadolinium T1-weighted images show diffuse hyperintensity and enhancement in L4 to L5 vertebral bodies. There is destruction of the L4–L5 disk, with increased STIR signal intensity in the narrowed disk space. There is paraspinal phlegmon (open arrow). Epidural abscesses at the L4 and S1 levels demonstrate increased STIR signal intensity and rim enhancement (arrows). Note the dural enhancement extending from L3 to the sacral canal. The STIR hyperintensity in the L3–L4 and L5–S1 disks are abnormal and probably represents diskitis. The STIR hyperintensity at the inferior endplate of L2 is secondary to degenerative disease. c Sagittal diffusion-weighted image shows restricted diffusion in the epidural abscess (arrow). There is mild hyperintensity in L4 and L5 vertebral bodies. d Axial post-gadolinium T1-weighted image at L4 level shows the epidural abscess (arrow) indenting the thecal sac

561

562

4  Spine and Spinal Canal

abscess is typically hyperintense relative to cerebrospinal fluid. Diffusion-weighted imaging, if available, shows increased signal intensity in vertebral and epidural abscesses with decreased signal intensity on apparent diffusion coefficient maps (Eastwood et al. 2002). Contrast-enhanced MR imaging is valuable, showing homogeneously or heterogeneously enhancing phlegmon and peripherally enhancing abscess. Dural enhancement and prominent epidural venous plexus enhancement above and below the epidural abscess may be present. Associated cord signal abnormality may be related to compression with edema or ischemia. 4.1.3.5  Septic Facet Joint Arthritis Pyogenic facet joint infection is considered an uncommon entity, although in a recent series, it accounted for about 20% of all pyogenic spinal infections (Narvaez et al. 2006). Infection may arise from hematogenous spread, direct extension, or as a result of spinal interventional or diagnostic procedures, surgical intervention, and penetrating trauma. Risk factors are similar to those for vertebral osteomyelitis/diskitis and epidural abscess. Clinical findings include acute or chronic pain, focal tenderness, and fever. As with other infections, erythrocyte sedimentation rate, C-reactive protein, and white blood cell count are often elevated. Neurologic impairment can occur secondary to associated phlegmon or abscess (Okada et al. 2005). The lumbar spine is involved more than 90% of the time. Infection is usually unilateral, and staphylococcus aureus is the most common causative organism. MR imaging findings include marrow edema with T1 hypointensity, and T2 or STIR hyperintensity, as well as T1 hypointense/T2 hyperintense fluid in an often widened facet joint. There may be accompanying periarticular edema, phlegmon, or abscess. Diffuse or rim enhancement of the facet joint is seen following gadolinium administration. There is also enhancement of the facet joint and marrow. Enhancement of the adjacent soft tissues and peripheral enhancement of any associated abscess collection can be seen. Similar findings may be found at CT. Also, CT may disclose facet lytic osseous destruction or sclerotic change and bony debris. Differential considerations include osteoarthritis and synovial cysts. Marrow changes are less common in facet osteoarthritis and are usually absent in synovial cysts. Also, there is no accompanying soft tissue edema or abscess in either of these conditions. Rheumatoid arthritis may be difficult to distinguish from pyogenic infection, although the cervical spine is more commonly affected in rheumatoid arthritis, and there is often associated dens erosion. Metastatic disease typically has a different clinical presentation and is usually characterized by greater osseous destruction, as well as multiple vertebral lesions.

4.1.4  Extradural Spine Tumors Spine tumors are categorized traditionally into extradural, intradural extramedullary, and intramedullary based on their locations. Although this categorization is oversimplified, as some lesions may occupy more than one compartment, it is helpful in formulating a differential diagnosis for the more common tumors. In this chapter, we will consider only extradural tumors. Intradural tumors will be discussed in the sections on intradural and intramedullary spine disease. By definition, extradural tumors are located either within the vertebral column (i.e., the osseous elements of the spine) or within the epidural space. Further categorization of extradural tumors into vertebral lesions and epidural lesions sometimes may be helpful. In addition, the consideration of multiplicity may also help one to narrow down the differential diagnoses. Lesions that may present as multiple vertebral lesions include metastasis, hemangioma, multiple myeloma, lymphoma, and leukemia. Obviously, these lesions may also present solitarily. Other diseases that are usually solitary include aneurysmal bone cyst, giant cell tumor, osteoid osteoma, osteoblastoma, chordoma, and different types of sarcoma. Eosinophilic granuloma is actually a nonneoplastic condition but its imaging features simulate that of neoplastic diseases. It is therefore included in this discussion. Eosinophilic granuloma presents more commonly as solitary lesions, but presentations with multiple lesions are also seen. The objectives for imaging of extradural spine tumor include (1) diagnosis, (2) staging, (3) evaluation of complications, and (4) follow-up of response to treatment. To provide a better differential diagnosis of spine tumors, correlation of different imaging modalities, such as CT, MR, and nuclear bone scan, may be necessary. Some lesions demonstrate characteristic features that help one narrow down the differential diagnoses to one or two. Very often, especially when a solitary spine tumor is seen, image guided biopsy is necessary to establish the diagnosis. Complications of extradural spine tumors include pathological fractures and mass effect on the spinal cord or nerve roots. Pathological fractures can be the first clinical presentation of extradural spine tumor. In this case, the distinction between pathological fractures and benign fractures due to insufficiency or trauma is important. Exclusion of spinal cord compression is a common emergent indication for MR imaging of the spine. 4.1.4.1  Imaging Techniques While plain films and CT remain better in depicting details of bony structures, MR imaging is superior in demonstrating an infiltrative process in bone marrow, epi-

4.1  Extradural Diseases of the Spine

dural involvement, or mass effect on the spinal cord or nerve roots. Similar to MR imaging protocols for other extradural spine disease, a protocol for imaging of extradural spine tumor may include T1- and T2-weighted fast spin-echo sequences in the sagittal and axial planes. However, two other imaging techniques are very useful in the imaging of extradural spinal neoplasm. The first is the STIR (short tau inversion recovery) imaging sequence, which is often acquired in the sagittal plane. Because of the predominance of fatty marrow in the adult spine, a neoplastic process may be obscured by the normally high marrow signal intensity on T2-weighted fast spin-echo imaging sequences. The STIR imaging sequence provides reliable fat signal suppression, making tumors stand out from the fatty marrow background. The second important technique is the contrast-enhanced T1-weighted imaging sequence with fat suppression. The fat suppression prevents the normally high background marrow signal from obscuring enhancing lesions on T1-weighted images. The use of fat suppression, either with the STIR imaging sequence or other techniques, such as frequency selective radiofrequency pulses or Dixon methods, are particularly important in the evaluation of neoplasms that may present as diffuse replacement of normal bone marrow, including multiple myeloma, lymphoma, leukemia, and some types of metastasis. A technique that has been relatively recently applied to the spine and that may provide additional diagnostic information for imaging of spine neoplasms is diffusion weighted imaging. This technique is still controversial. However, some authors have suggested that diffusion weighted imaging can differentiate between pathological compression fractures and benign compression fractures (Baur et al. 1998). Pathological fractures tend to demonstrate restricted diffusion, while benign fractures show increased diffusion. This is believed caused by the presence of densely packed tumor cells, leading to reduction of extracellular free water and molecular movement. In benign fractures, there may be expansion of extracellular volume due to hemorrhage and edema, and therefore increased apparent diffusion. However, other authors have disputed this work (Castillo et al. 2000). 4.1.4.2  Benign Tumors

sacrum (5%). Hemangiomas are most often seen within the vertebral body, but involvement of the pedicles is also common. They are usually asymptomatic, but hemorrhage, epidural extension, or compression fractures may lead to symptoms, possibly as a result of compression of the nerve roots or spinal cord. Hemangiomas are commonly diagnosed as incidental findings. Occasionally, they may have an aggressive appearance. Histopathologically, hemangiomas represent enlarged slow-flowing vascular channels that may be thrombosed. The lesions may destroy some bony trabeculae, thus resulting in compensatory thickening of the remaining trabeculae. There may be fatty involution. These histopathological findings explain their typical imaging appearance, On MRI, these lesions are typically hyperintense on both T1 and T2-weighted images because of their high fat content (Ross et al. 1987). They tend to enhance with contrast (Fig. 4.1.30). More aggressive hemangiomas tend to have a lower fat content (Fig. 4.1.31). 4.1.4.2.2  Osteoid Osteoma Osteoid osteoma is usually seen in younger patients less than 30 years of age. Patients with osteoid osteoma classically present with nocturnal pain that is relieved with aspirin, presumably due to production of prostaglandins by the tumor. The lesion may be associated with painful scoliosis. Approximately 10% of osteoid osteomas are seen in the spine, most often in the lumbar spine (59%), followed by the cervical (27%), thoracic (12%), and sacral (2%) regions (Gamba et al. 1984). They primarily involve the posterior elements including the pedicles, articular facets, and laminae, with up to 90% seen at these locations. Osteoid osteoma involving the vertebral body is uncommon. Osteoid osteoma is better evaluated with CT, where a small lucent nidus (<2.0 cm) with a sclerotic rim is typically seen. Central calcifications may be seen in the nidus. Nuclear bone scan shows intense uptake of the lesion. On MRI, the nidus of the lesion is hypointense on T1-weighted images and hyperintense on T2-weighted images. The sclerotic rim is typically dark on both T1- and T2-weighted images. Marked bone marrow edema may be seen associated with osteoid osteoma, giving a more aggressive appearance on MR imaging in some cases.

4.1.4.2.1  Hemangioma

4.1.4.2.3  Osteoblastoma

Vertebral hemangiomas are probably the most common primary spine tumors. These are benign vascular tumors that can be solitary (in 66% of cases) or multiple (34% of cases) (Schmorl and Junghanns 1971). They are most common in the thoracic spine (34%), followed by the lumbar spine (29%), the cervical spine (6%), and the

Osteoblastoma is often considered as giant osteoid osteoma. However, the clinical presentation of osteoblastoma is generally different from osteoid osteoma, giving dull localized pain as opposed to the nocturnal pain of osteoid osteoma. Like osteoid osteoma, it typically arises from posterior elements, but may extend into the ver-

563

564

4  Spine and Spinal Canal

Fig. 4.1.30  Hemangioma. a Sagittal T1-weighted, b STIR, and c post-gadolinium T1-weighted images show a round lesion in T8 vertebral body. The presence of fat in the lesion gives the hyperintensity on T1-weighted image. The vascular component leads to high signal intensity on STIR image and enhancement. d On CT, prominent lucencies, representing vascular channels are present. There are dense, thickened bone trabeculae (asterisk)

tebral body (see Fig. 4.1.32) or cross the intervertebral space to involve an adjacent vertebra. There is an equal distribution among the cervical, thoracic, and lumbar spine (Kroon and Schurmans 1990). Epidural extension may also be present (Myles and MacRae 1988). The imaging features of osteoblastoma may be non-aggressive. On the other end of the spectrum, more extensive bone destruction with a wide zone of transition and soft-tissue abnormality can be seen around the lesion, giving an aggressive appearance. The aggressive appearance of osteoblastoma can simulate malignancy, posing a serious challenge in radiological diagnosis. On CT, the typical appearance is of a large lytic ex-

pansile lesion surrounded by thinned cortex. On MR imaging, it is hypointense on T1-weighted images and may have variable signal intensity on T2-weighted images, especially in the presence of hemorrhage or calcifications. Fluid–fluid levels may be seen, particularly in the presence of a co-existing aneurysmal bone cyst component. Associated soft tissue mass may be present. The lesion usually shows enhancement, which may extend to the adjacent soft tissues. Extensive peritumoral inflammatory response may be present, giving an aggressive flare phenomenon (Fig. 4.1.33) (Crim et al. 1990). The differential diagnosis for osteoblastoma includes osteoid osteoma, which is generally smaller (<2 cm), and

4.1  Extradural Diseases of the Spine

Fig. 4.1.31  Vertebral hemangioma with associated epidural mass. a,b Sagittal T1- and T2-weighted images show a large mass occupying most of the L2 vertebral body. There is a large epidural component, causing impingement upon the cauda equina. c Post-gadolinium T1-weighted image shows intense enhancement. d Axial CT shows central frank bone destruction. However, the remaining bone trabeculae show a spiculated pattern characteristic of vertebral hemangioma. (From: Magnetic resonance imaging of the brain and spine, 3rd edn., Lippincott Williams & Wilkins, Philadelphia © 2002)

has a less expansile appearance. Aneurysmal bone cyst may have a similar appearance to osteoblastoma and is an important differential consideration for lesions arising in the posterior elements of the spine. Approximately 10–15% of osteoblastomas actually contain coexisting aneurysmal bone cysts. 4.1.4.2.4  Aneurysmal Bone Cysts Most aneurysmal bone cysts are seen in young patients, with approximately 80% of cases seen in patients less than 20 years of age (Kransdorf and Sweet 1995). The

exact etiology is uncertain. The lesion may be primary (de novo) or secondary. Secondary aneurysmal bone cyst may be associated with other primary bone tumors, including osteoblastoma, giant cell tumor, chondroblastoma, chondromyxoid fibroma, fibrous dysplasia and other fibro-osseous lesions, fibromyxoma, osteosarcoma and other primary bone sarcomas, brown tumors of hyperparathyroidism, and metastases. It was postulated that these other primary tumors cause venous obstruction or arteriovenous fistulae, giving rise to secondary aneurysmal bone cyst (Kransdorf and Sweet 1995). The neural arch is the typical location of involvement, with about 60% in the posterior elements, and 40% arising from the

565

566

4  Spine and Spinal Canal Fig. 4.1.32  Osteoblastoma of the left L4 pedicle. a,b AP and lateral lumbar spine X-ray. c Axial CT. d Axial T2-weighted image. e Axial contrast-enhanced T1-weighted image with fat suppression. Plain films demonstrate an expansile lytic lesion centered at the L4 pedicle, extending into the vertebral body. The lesion is well-circumscribed with a narrow transition and sclerotic rim, giving a benign appearance (arrows in a,b). The findings are confirmed on CT. Axial T2-weighted MR image demonstrates fluid-fluid level within the lesion (arrow in d). In this case, the osteoblastoma is localized within the spine. There is no evidence of associated soft tissue mass or peritumoral edema and inflammatory changes (contrast this with Fig. 4.1.33)

vertebral bodies. Like osteoblastoma and giant cell tumor, aneurysmal bone cysts may cross the intervertebral space to involve an adjacent vertebra. Aneurysmal bone cysts are typically seen as bubbly expansile lytic lesions surrounded by thinned cortex (Fig. 4.1.34). There may be internal septations and lobulations. Aneurysmal bone cysts contain multiloculated bloodfilled spaces, and layering of fluids can be seen if the patients stay a sufficient amount of time in the supine position before imaging (usually 10 min). On MR images, a dark rim may be seen on both T1- and T2-weighted images. This dark rim represents a thickened intact periosteal membrane and may be seen even when bony cortex is not visualized on CT. Enhancement of the internal septae may be seen. Aneurysmal bone cysts may contain a variable amount of solid components, with 5–7.5% presenting as predominantly solid lesions (Bertoni et al. 1993).

4.1.4.2.5  Giant Cell Tumor Other than hemangioma, giant cell tumor is the most common benign tumor involving the spine. It is also the most common primary tumor arising in the sacrum. The lesions are more frequently seen in women in the second to fourth decades of life, and may respond to hormonal stimulation leading to a dramatic increase in size during pregnancy (Murphy et al. 1996). As opposed to osteoblastoma, osteoid osteoma, and aneurysmal bone cyst, giant cell tumor occurring in the spine superior to the sacrum usually favors the vertebral body rather than the posterior elements. There may be extension to the posterior elements or crossing of the intervertebral space to involve an adjacent vertebra. Paraspinal soft tissues and collapse of vertebral body may be present. Giant cell tumor is usually hypervascular. Areas of hemorrhage or necrosis may be present.

4.1  Extradural Diseases of the Spine

Fig. 4.1.33  Aggressive osteoblastoma at right superior facet of C7. a Plain film in anteroposterior view shows an expansile lytic lesion involving the right posterior element of C7 (black arrow). b 99mTc MDP bone scan shows increased uptake. c,d Axial CT in bone and soft tissue window shows a lytic lesion with an ill-defined border and wide transition. There is mineralized matrix within the lesion. As opposed to the case of Fig. 4.1.32, there is the appearance of an adjacent soft tissue mass (white arrow, d). Axial fat-suppressed T2-weighted images (e,f) and sagittal fat-suppressed contrast-enhanced T1-weighted image (g) reveal extensive T2 prolongation and enhancement involving multiple levels (open arrowheads, g). The soft-tissue abnormality encases the right vertebral artery (arrow, e,f) and widens the C7–T1 right neural foramen. Because of the aggressive appearance, more aggressive tumors, such as sarcomas, were also considered initially. Hydromyelia of the cervical spinal cord is noted incidentally on the axial MR images (hyperintense lesion at central spinal cord). O osteoblastoma, N right C7–T1 neural foramen

567

568

4  Spine and Spinal Canal

Fig. 4.1.34  Aneurysmal bone cyst at left lateral mass of C1. a CT shows an expansile heterogeneous lytic lesion with thinned cortex and focal cortical destruction (arrows). T2-weighted axial b and coronal c MR images demonstrate a multilobulated mass

with well-circumscribed hypointense rim (arrows). The lesion contains cystic components, but fluid–fluid levels are not demonstrated. d Axial contrast-enhanced T1-weighted image shows enhancement of the wall and septae

4.1  Extradural Diseases of the Spine

Although giant cell tumor is generally classified as a benign neoplasm for the purpose of discussion, it has a poorer prognosis compared to other benign extradural spine tumors. Giant cell tumors are locally aggressive and recurrence has been estimated at a rate of 40–60%. Approximately 7% of cases have malignant features, with metastasis most common to the lungs. On CT, giant cell tumor usually appears as a lytic expansile mass without evidence of mineralized matrix. The MR signal intensity characteristics on T1- and T2-weighted images are variable. Most lesions demonstrate low to intermediate signal intensity with T1 and T2 weighting. This is in contrast to most other neoplasms that generally show increased signal intensity with T2 weighting. Enhancement is usually seen in the soft tissue component of the mass.

it is generally hypointense on both T1- and T2-weighted images. There should be no evidence of contrast enhancement, and the lesion should be surrounded by normal marrow. On MR and CT, enostosis and blastic metastasis can simulate one another. Enostosis therefore may pose a diagnostic difficulty in a patient suspected of metastatic bone disease. In general, enostosis does not demonstrate radioactivity uptake, whereas this is not true for blastic metastasis. Nuclear bone scan may therefore be helpful to differentiate the two diagnoses. Although most bone islands are stable, interval growth or decrease in size has been reported. When the lesion increases in diameter by more than 25% within a 6-month period or by 50% within 1 year, then biopsy should be considered to exclude blastic malignancy (Murphy et al. 1996).

4.1.4.5.6  Osteochondroma

4.1.4.5.8  Eosinophilic Granuloma

Although osteochondroma is a common benign bone lesion, spinal osteochondroma accounts for only 1–4% of solitary lesions and is seen in only 1–9% of patients with hereditary multiple exostosis (Murphy et al. 2000). Lesions in the spine most often arise in the posterior elements, with a predilection for the spinous processes. Osteochondroma appears as a sessile or pedunculated lesion protruding from normal bone. It is composed of normal bone, with its cortex and marrow in continuity with the cortex and marrow of the underlying bone. A hyaline cartilage cap can be identified from which growth of the lesion occurs. These features can be seen on CT and MR and are the hallmarks for diagnosis. On MRI, the cortex and marrow demonstrate signal intensity similar to that of normal bone. The cartilaginous cap of osteochondroma demonstrates low to intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Calcification may be present in the cartilage cap. The cartilage cap should be thin. A thickened cap measuring greater than 1–2 cm is suspicious for malignant degeneration.

Eosinophilic granuloma is actually a non-neoplastic condition, but is included in this discussion because of its tumor-like appearance. Eosinophilic granuloma represents a form of Langerhans cell histiocytosis. It is usually seen in young patients under the age of 19. Involvement of the spine is not common and usually is limited to the vertebral body with sparing of the posterior elements. The classic imaging presentation is vertebra plana seen as severe collapse of a vertebral body. The lesion itself is lytic on plain films or CT. MR signal characteristics are nonspecific, with low signal intensity on T1-weighted images and intermediate to high signal intensity on T2-weighted images. There is no or very little paraspinal or epidural soft tissue component, distinguishing it from malignant neoplasm.

4.1.4.5.7  Enostosis Enostosis is also called bone island. It is common in the axial skeleton, often seen just beneath the cortex. Histologically, it represents lamellar compact bone with a Haversian system embedded within the medullary canal. Enostosis is better seen on plain films and CT. On these studies, the lesion appears as a blastic lesion with irregular, radiating spicules at the periphery, but is still relatively well-defined. It can reach a large size greater than 2 cm. Because the lesion contains mostly compact bone,

4.1.4.5.9  Angiolipomas These are rare lesions that may be seen in extradural space. An angiolipoma represents an extradural lipoma that also contains vascular elements. Most lesions are seen in adults. Generally, they are seen in the posterior and lateral aspects of the epidural space, but occasionally, infiltrative extension into adjacent soft tissues and the vertebral bodies may be seen. Extension over several vertebral levels is common. On T1-weighted images, the fatty component of the lesions is hyperintense. Within the lesions, there is inhomogeneous signal thought to represent capillary and venous channels. Contrast-enhanced T1-weighted imaging should be done with fat suppression and will show heterogenous enhancement (Provenzale et al. 1996).

569

570

4  Spine and Spinal Canal

4.1.4.3  Malignant Tumors

4.1.4.3.2  Chordoma

4.1.4.3.1  Sacrococcygeal Teratomas

Chordoma is a malignant lesion which arises from notochordal remnants, and therefore may occur anywhere from the skull base to the sacrum. It is most commonly seen in the sacrococcygeal region (50%), followed by the clival/sphenooccipital region (35%). Involvement of the vertebrae is much less common (15%) (Sze et al.1988). Chordoma is rare in children. It is most commonly seen in adults, with a peak age between the fifth and sixth decades of life. Chordoma is usually seen as a midline lytic lesion associated with a large paraspinal or epidural soft tissue mass. On CT, amorphous calcifications may be present in approximately 50–70% of cases. On MRI, the lesion is usually hypo- or isointense on T1-weighted images. On T2-weighted images, it is usually hyperintense due to the high water content of the tumor. Internal fibrous septations, hemorrhage, and cystic necrosis may be present. Contrast enhancement is variable, but is intense in most cases.

Sacrococcygeal teratomas (also known as germ cell tumors of coccyx) are seen in young children. Although they are rare tumors, they are actually the most common tumors of the spine diagnosed in newborns. The tumors arise from multipotential cells of Hensen’s node that migrate to the coccyx. They may contain hair, teeth, cartilage, and fat. They often extend from the coccyx and may grow both internally and externally. The tumors can be classified into four types based on location (Schey et al. 1997). Type 1 tumors are almost completely external and distort the buttocks. Type 2 tumors have a mostly external component and a small internal component. Type 3 tumors are predominantly intrapelvic with significant displacement or invasion of surrounding pelvic structures. Type 4 tumors are entirely internal with no external portions. Most tumors are benign, but approximately 17% have malignant features. The higher numbered types tend to occur in older patients and are more likely to be malignant, presumably due to delay in diagnosis. On imaging, the tumor is typically seen as a presacral soft tissue mass (Fig. 4.1.35). Cystic components, calcifications, and hemorrhage may be present. There may be erosion of the coccyx.

Fig. 4.1.35  Sacrococcygeal teratoma. Sagittal T2-weighted image with fat suppression shows a large hyperintense presacral tumor (T) extending to the postsacral region. There is extension superiorly to the S3 level (large arrow). The tumor surrounds the sacrum but there is no bone destruction. It compresses the rectum (R), but does not invade it. The intact posterior rectal wall is clearly delineated (small arrows). B bladder

4.1.4.3.3  Osteosarcoma Primary osteosarcoma arising in the spine accounts for only 0.6–3.2% of all osteosarcomas and 5% of all primary malignant tumors of the spine (Murphy et al. 1996). Secondary osteosarcoma is much more common. Both vertebral body and posterior elements can be involved. On CT, osteoblastic activity is often seen but osteolytic bone changes may also be present. Up to 20% of osteosarcomas may present as purely lytic lesions without evident bone matrix. An associated soft tissue mass is often present in the paraspinal and epidural regions. CT allows visualization of the bone matrix and is the best imaging tool for evaluation and characterization of the primary lesions. MR is superior for the evaluation of the effects on the spinal cord and nerve roots. The MR signal intensity pattern is variable and depends on the degree of matrix mineralization. Within the spinal column, lesions with a predominantly osteoid matrix may have low signal intensity on all imaging sequences. However, paraspinal and epidural soft tissue components of the tumor generally show hyperintensity on T2-weighted images and enhancement, allowing it to be well delineated. Hemorrhage may be seen in teleangiectatic osteosarcoma and gives rise to fluid/fluid levels. These lesions may mimic aneurysmal bone cyst. 4.1.4.3.4  Chondrosarcoma Chondrosarcoma is a malignant tumor arising from cartilage. Most lesions in the spine are primary lesions, but some arise from malignant transformation of osteochon-

4.1  Extradural Diseases of the Spine

droma. Chondrosarcoma tends to be of lower grade compared to osteosarcoma, and patients have a better prognosis. The lesion may involve the vertebral body and the posterior elements. On CT, chondrosarcoma shows lytic destruction of the bones. There is often a calcified chondroid matrix with a “rings and arcs” pattern. On MR, the soft tissue component of the lesion shows increased signal intensity on T2-weighted images because of the high water content. The mineralized component may be hypointense on all imaging sequences. Hemorrhage may be present. Contrast enhancement is usually seen. 4.1.4.3.5  Ewing Sarcoma Ewing sarcoma generally occurs in children and young adults. It is a small round cell tumor. Primary spine lesions are much less common than secondary spine lesions.

Fig. 4.1.36  Ewing sarcoma in the sacrum. CT images in soft tissue (a) and bone window (b) demonstrate presacral soft tissue mass lesion (M) and permeative lytic changes in the sacrum (open arrow). There is erosion of the sacral cortex (solid arrow). B bladder. The soft tissue mass and lytic changes in the bone have progressed significantly over 2 weeks. Sagittal MR images from fat-suppressed T1-weighted sequence c and fat-suppressed

CT may show permeative lytic changes of the bone with possible soft tissue extension (Fig. 4.1.36). Sclerotic changes may also be present but are rare. The MR signal characteristics are nonspecific. The lesion is usually hypointense on T1-weighted images, hyperintense on T2-weighted images, and demonstrates enhancement. There may be marked peritumoral edema and enhancement that extend beyond the true margin of tumor. 4.1.4.3.6  Multiple Myeloma Multiple myeloma is the most common primary bone tumor in adults, accounting for 45% of vertebral column tumors and 34% of malignant bone tumors (Bazan 1993). It is most commonly seen in the fifth to seventh decades of life. Multiple myeloma is caused by the multifocal malignant proliferation of monoclonal plasma cells within

T2-weighted sequence d demonstrate the lesion in the sacrum (arrow) extending to the presacral region (M). Note the peritumoral edema (curved arrows). e,f Coronal images from a postgadolinium fat-suppressed T1-weighted sequence demonstrate the enhancing mass (M). There is enhancement around the tumor that correlates with the peritumoral edema seen in d

571

572

4  Spine and Spinal Canal

bone marrow. When there is only a solitary lesion, the disease process is called plasmacytoma. Because multiple myeloma involves primarily the red marrow, it is most commonly seen in the axial skeleton, with the spine, skull, and ribs being the most common sites. Long bones of the extremities may also be involved. The lesions may appear in three patterns: localized, diffuse, and variegated. The localized pattern describes relatively well circumscribed lesions. In the diffuse pattern, there is diffuse involvement of the marrow spaces. The variegated pattern describes patchy heterogeneous

involvement. On plain films and CT, myeloma lesions are seen classically as punched-out lytic lesions (Fig. 4.1.37). On MR (Figs. 4.1.37 4.1.38), the lesions are low to isointense on T1-weighted images and hyperintense on T2-weighted images. Contrast enhancement is often demonstrated, but may be decreased in treated lesions. MRI is particularly useful in depiction of the epidural extension of myeloma lesions. Because multiple myeloma lesions are usually negative on nuclear bone scan, MR becomes an important tool for diagnosis, staging, and follow-up of this common disease.

Fig. 4.1.37  Pathological fracture secondary to multiple myeloma. a Axial CT image in bone window. b Sagittal T1-weighted image. c Sagittal STIR image. d Sagittal diffusion-weighted image. e Sagittal contrast-enhanced T1-weighted image with fatsuppression. f Axial T2-weighted image. The CT image shows a heterogeneous lytic lesion in the T5 vertebral body. There is paraspinal soft-tissue thickening. The MR images demonstrate replacement of normal fatty marrow in T3 and T5 vertebral bodies, leading to hypointensity on the T1-weighted image, and hyperintensity on the STIR image. There is abnormal contrast

enhancement in T3 and T5 vertebrae. A wedge compression fracture of T5 can be seen on the sagittal images. Increased signal intensity on the diffusion-weighted image suggests that this is a pathological fracture. Epidural extension is noted, causing indention of the thecal sac and impingement on the thoracic spinal cord (white arrow in c, black arrows in f). Note the central anchoring of the epidural lesion by the median raphe, which is commonly seen in epidural metastasis. Paraspinal extension is seen on the CT and MR images (white arrowheads in e,f)

4.1  Extradural Diseases of the Spine

4.1.4.3.7  Leukemia Leukemia is a systemic disease which may involve the spine. Leukemia is the most common malignancy in children. Acute lymphocytic leukemia constitutes 80% of all childhood leukemia. Acute myelogenous leukemia accounts for approximately 10% while the remaining 10% represent less common histological types. In adults, chronic lymphocytic leukemia and chronic myelogenous leukemia are the major histological types. On imaging, leukemia may manifest as diffuse osteopenia with multiple vertebral fractures. Lytic bone lesions may also be present. There may be leptomeningeal involvement which is best demonstrated by MRI. On MR images, the lesions are usually hypointense on T1-weighted images and hyperintense on T2-weighted images and STIR images. Contrast enhancement is usually present. 4.1.4.3.8  Lymphoma Lymphoma (Figs. 4.1.39, 4.1.40) in the spine may be primary or secondary. Secondary lymphoma is much more common than primary lymphoma. Primary lymphoma of the spine may occur as intramedullary, epidural, leptomeningeal, or osseous lesions. Epidural lymphoma typically presents as a soft tissue mass, often but not necessarily associated with bone destruction. On plain films or

Fig. 4.1.38  Diffuse multiple myeloma. a,b Sagittal T1-weighted images of the cervical, thoracic and lumbar spine show numerous lesions involving virtually every level of the spine. Both vertebral bodies and posterior elements are involved. Note the background marrow signal intensity is dark compared with the

CT, osseous lesions may demonstrate permeative lytic destruction of bones. Rarely, osseous lymphoma may show up as diffusely dense lesions, giving the appearance of ivory vertebrae. On MR images, lytic osseous or extradural lesions are generally iso- to hypointense on T1-weighted images, and iso- to hyperintense on T2-weighted images. There is generally homogeneous enhancement. The lesions may cross the intervertebral disks and involve multiple levels. Dense osseous lesions are usually hypointense on both T1-weighted and T2-weighted images. 4.1.4.3.9  Metastasis Metastasis is the most common tumor in the spine. In adults, common primary tumors that often metastasize to the spine include lung, breast, prostate, renal, melanoma, and colon cancer. In children, spinal metastasis may be seen in neuroblastoma, sarcoma, germ cell tumor, and Wilms’ tumor. Although lymphoma may present as a primary spine tumor, spinal lymphoma more often represents metastasis from systemic disease. On MR images, the imaging appearance is variable depending on whether the lesions are lytic or blastic. Lytic metastasis is usually hypointense on T1-weighted images and hyperintense on T2-weighted images. Contrast enhancement is usually present (Fig. 4.1.41). Sclerotic metastasis, such as seen in prostate cancer, may demonstrate hypointensity on both T1 and T2-weighted images

intervertebral disks. This is an abnormal finding in adults, and suggests an underlying diffuse infiltrative process of the spine. c,d Post-gadolinium T1-weighted images show enhancement of the multiple myeloma lesions

573

574

4  Spine and Spinal Canal

(Fig. 4.1.42). Occasionally, sclerotic metastasis may be seen as a well-defined nodule, mimicking a benign lesion such as an enostosis (a bone island) on both CT and MRI

(Fig. 4.1.43). Nuclear bone scan may be helpful to differentiate the two entities, as increased uptake is usually present for sclerotic metastasis but not for enostosis. Metastasis may involve the osseous structures or epidural space. Assessment of epidural lesions is important not only for staging but also to exclude mass effect on the spinal cord and nerve roots (Fig. 4.1.44). This is one of the most important indications for MR imaging of this disease in the spine. 4.1.5  Vertebral Column Trauma 4.1.5.1  Introduction

Fig. 4.1.39  Large B-cell lymphoma in a patient with HIV. a,b Post-gadolinium fat-suppressed T1-weighted images in sagittal and axial planes. There is a diffusely- enhancing lesion in the L4 vertebral body with epidural extension (white arrows). The epidural enhancement extends from L2 to the sacral canal. Bilateral L4–L5 neural foramina are involved (b). There are paraspinal lesions all around the vertebral column, with infiltration on the bilateral psoas muscles (black arrows) and in the left erector spinae muscle (open white arrow). In spite of the extensive involvement, the intervertebral disks are essentially preserved, distinguishing this from infection which is also an important consideration in an immunocompromised patient

Injury to the vertebral column may have devastating consequences, particularly if there is involvement or compression of the spinal cord. Plain films and CT continue to play a vital role in the evaluation of injury or fracture to the vertebral column. Plain films or CT are generally the first line imaging study in patients with a history of trauma to the spine. CT is excellent for evaluating the osseous structures and detecting subtle fractures not seen on plain films. MRI, however, continues to play an increasingly important role in the evaluation of spinal trauma. Indications for MRI of the vertebral column should include those patients who are to undergo operative fixation regardless of neurologic status (Saifuddin 2001). It can be argued that MRI has a role to play in all cases of spinal trauma as it accurately depicts injury to the vertebral bodies, the ligaments, the intervertebral disks, the extramedullary space, and the spinal cord (Saifuddin 2001). MRI is also excellent at demonstrating unsuspected injury to the vertebral bodies in the absence of radiographic findings (Qaiyum et al. 2001). In addition, in those patients with documented injury to the spinal cord or vertebral column, it has been proven that MRI can disclose unsuspected injury at contiguous and non-contiguous levels of the spinal column (Qaiyum et al. 2001). The sensitivity and specificity of MRI with regard to the evaluation and diagnosis of ligamentous rupture and injury continues to be a topic of study and research (Saifuddin 2001). MRI is the modality of choice for depicting patency or injury to the ligaments of the spine (Hassankhani and Bencardino 2003). However, the distinction between a ligamentous fragment and cortical bone fragment is often difficult to make even with MRI (Atlas 2002). MRI tends to be insensitive for evaluating fractures involving the posterior elements, due to the smaller size and more complex geometry of the posterior elements, as well as the decreased proportion of medullary space relative to the vertebral bodies (Atlas 2002). A recent study suggested that MRI has sensitivity of 36.7% for anterior column fractures and 11.5% for posterior column fractures in comparison to CT (Atlas 2002).

4.1  Extradural Diseases of the Spine

In the following section, the technical aspects of MR imaging of the traumatic vertebral column is discussed. In addition, the biomechanics of vertebral column trauma are depicted to better understand the specific injury patterns. Finally, the MR imaging findings are described in detail with regard to the vertebral bodies, the ligamentous structures, and the intervertebral disks. Included in the MR imaging findings section is a discussion of both traumatic epidural hematoma formation and vascular dissection injury.

A typical imaging protocol for imaging the spinal cord and vertebral column to evaluate for possible traumatic injury includes sagittal T1, sagittal T2, sagittal STIR, and sagittal T2*-weighted gradient recalled echo (GRE) images. In addition, axial T2- and axial T2*-weighted GRE imaging may be performed as well. For the cervical spine, axial STIR imaging may be added. STIR images are ideal for detecting soft tissue injury, while the T2* GRE are

Fig. 4.1.40  Lymphoma. a,b Reformatted sagittal CT images of the thoracic and lumbar spine show a blastic lesion in the T6 vertebral body associated with a severe compression fracture and a mixed lytic–blastic lesion in L5. c,d On sagittal STIR im-

ages, there is increased signal intensity in the mixed lytic–blastic lesion of L5, but not in the blastic lesion of T6. e,f Post-gadolinium T1-weighted sagittal images show enhancement in the L5 lesion, but not in the T6 blastic lesion

4.1.5.2  Imaging Techniques

575

576

4  Spine and Spinal Canal

excellent for depicting subtle bleeding and blood products, which demonstrate low signal “blooming” artifact. For evaluating potential carotid or vertebral artery dissection in the cervical spine, 2D time of flight and/or contrast-enhanced MRA are suggested. In addition, T1weighted fat-saturation axial images through the neck are vital for the evaluation of possible intramural throm-

bus. Fat-saturation images are also recommended with regard to evaluation of the interspinous ligaments, as they tend to be surrounded by fat on either side (Ruggieri 1999). With regard to MR imaging of the vertebral column and spinal cord status post injury, special considerations must be taken into consideration. Any and all devices, including life support systems, monitoring machines, external fixators, cervical casts, halos, body casts, and traction devices must be MRI compatible (Atlas 2002). External spinal fixators composed of ferromagnetic material may destroy the static magnetic field, resulting in artifact and image degradation (Atlas 2002). The actual transfer of the patient to the scanning table should be performed only by properly trained individuals to minimize the risk of further injury (Atlas 2002). Sedation may be required to decrease patient motion, whether it is voluntary or involuntary (Atlas 2002). Selection of the proper surface coils is determined by the location of injury, the accessibility of the area of interest, and the type of coils available (Atlas 2002). Whenever possible, a phased-array coil system should be utilized to maximize the area of coverage, and optimize the MR signal (Atlas 2002). Metallic fragments from penetrating trauma require additional consideration, as associated artifact may cause significant image degradation (Atlas 2002). In addition, metallic fragments may pose a significant safety hazard to the patient with regard to potential migration or dislodging of the fragment in question (Atlas 2002). Correlation with CT or plain films is required whenever there is potential for metallic fragments, and if need be CT myelography should be performed if MRI is contraindicated (Ruggieri 1999). 4.1.5.3  Biomechanics

Fig. 4.1.41  Lytic metastasis from adenocarcinoma of the lungs. a Sagittal T2-weighted and b post-gadolinium fat suppressed T1-weighted image show T2 prolongation and enhancement at the superior endplate of the T12 vertebral body (solid white arrow) and at the right pedicle of L1 (white open arrow). The enhancing T12 lesion correlates with lytic change on the CT image (c) from a PET CT study, which also shows intense glucose uptake

Discussion of the biomechanics of vertebral column trauma is important to gain a better understanding of the MRI findings associated with vertebral column injury. The biomechanical framework of the spine consists of rigid components, which include the vertebral bodies and posterior elements, and flexible components, which include the disks and ligaments (Atlas 2002). The rigid and flexible components work in concert to allow normal movement of the spinal column, following the elastic modulus with regard to stretching and compression (Atlas 2002). Dissipation of forces is well tolerated if the force is applied gradually over time, or if the resultant motion does not exceed the translational design of the vertebral column; if either of these rules is violated, then the elastic modulus may be compromised and exceeded, resulting in tissue injury (Atlas 2002). To aid in the diagnosis, prognosis, and treatment of spinal column trauma, a simplified classification system has been described depicting the specific mechanism of injury, including hyperflexion,

4.1  Extradural Diseases of the Spine

hyperextension, flexion-rotation, extension-rotation, axial loading, and lateral translation (Table 4.1.2) (Atlas 2002). However, this model is somewhat limited, as few injuries to the spinal column demonstrate any pure mechanism (Atlas 2002). The distribution of trauma is another important consideration with regard to trauma biomechanics and pathophysiology. Due to its proximity to the skull, the pattern of injury of the upper cervical spine differ from that of the lower cervical spine, especially in pediatric patients, whose head size is proportionately larger as compared to adults (Atlas 2002). The biomechanics and structural differences of the thoracic spine from T1 through T10 differs from the cervical spine, lower

thoracic spine, and lumbar spine, as the thoracic cage provides additional support, protection, and energy absorbing capacity (Atlas 2002). In addition, the facet joints of the rib-bearing thoracic vertebral bodies are coronally as opposed to sagittally oriented, which provides additional protection against anterior translation (Atlas 2002). The three-column model (Table 4.1.3), proposed by Denis and utilized to assess for spinal stability, divides the spinal column into the anterior, middle, and posterior column (Takhtani and Melhem 2000). The anterior column consists of the anterior longitudinal ligament, the anterior half of the vertebral body, and the anterior annulus fibrosis (Takhtani and Melhem 2000). The middle

Fig. 4.1.42  Diffuse blastic metastasis from prostate carcinoma. a Sagittal T1-weighted, b STIR and c post-gadolinium T1-weighted images. Lesions are seen at all vertebral levels. Most of these lesions are hypointense on all imaging sequences, roughly correlating with the blastic lesions seen on CT images (d). An epidural lesion is seen at the T9 level (open arrow). There is diffuse meningeal metastasis (solid arrows)

577

578

4  Spine and Spinal Canal Table 4.1.2  Functional classification of cervical spine fractures and dislocations (Takhtani and Melhem 2000) Mechanism of Injury

Type

Hyperflexion

Anterior subluxation (sprain) Bilateral interfacetal dislocation Simple wedge fracture Clay-shoveler’s fracture Teardrop fracture

Hyperextension

Dislocation (sprain or strain) Avulsion fracture of the anterior arch of C1 Fracture of the posterior arch of C1 Teardrop fracture of C2 Laminar fracture Hangman’s fracture Fracture or dislocation

Vertebral compression

Jefferson fracture Burst fracture

Hyperflexion with rotation

Unilateral interfacetal dislocation

Hyperextension with rotation

Pillar fracture

Lateral flexion

Uncinate process fracture

Indeterminate

Atlanto-occipital disassociation Odontoid fractures

Table 4.1.3  Components of the three columns of the spine (Takhtani and Melhem 2000) Column

Components

Anterior

Anterior longitudinal ligament Anterior annulus fibrosis Anterior vertebral body

Middle

Posterior vertebral body Posterior annulus fibrosis Posterior longitudinal ligament

Posterior

Posterior elements Facet capsule Interlaminar ligaments (flava) Supra or interspinous ligaments

column consists of the posterior longitudinal ligament, the posterior half of the vertebral body, and the posterior annulus fibrosis (Takhtani and Melhem 2000). The posterior column consists of the posterior elements, the ligamentum flavum, and the interspinous ligaments (Takhtani and Melhem 2000). Involvement or injury of any two or all three columns is indicative of spinal instability, defined as the inability of the vertebral column to maintain normal alignment under normal physiologic load, potentially resulting in loss of structural integrity and damage to the spinal cord/nerve roots, and incapacitating deformity of the spinal column (Takhtani and Melhem 2000). 4.1.5.4  Imaging Findings 4.1.5.4.1  Normal Anatomy The normal MR appearance (Fig. 4.1.45) of the vertebral bodies is determined by the ratio of fatty yellow marrow to the hematopoietic red marrow (Volger and Murphy 1988). With increasing age, the amount of fatty yellow marrow increases. Fatty marrow is depicted as hyperintense signal on T1-weighted images, high signal on fast spin-echo T2-weighted images, and hypointense signal on STIR images (Volger and Murphy 1988). However, this pattern is variable, as the bone marrow of the vertebral bodies tends to have more hematopoeitic red marrow than the appendicular skeleton, even in adults (Volger and Murphy 1988). Pediatric patients have an increased amount of red marrow, seen as lower signal intensity on T1-weighted images and higher signal intensity on T2-weighted and STIR images (Volger and Murphy 1988). The bony cortex of the vertebral bodies demonstrates low signal on all MR imaging sequences (Saifuddin 2001). The ligaments of the spine include the anterior longitudinal ligament, the posterior longitudinal ligament, and the posterior ligamentous complex (Saifuddin 2001). The anterior longitudinal ligament is located ventral to the vertebral bodies, and extends from the skull base through the sacrum, providing elasticity during flexion, extension, and rotation (Atlas 2002). Injuries to the anterior longitudinal ligament are typically the result of hyperextension injuries (Atlas 2002). The posterior longitudinal ligament is located between the ventral dural sac and posterior margins of the vertebral bodies and disks, extending from the skull base through the sacrum (Atlas 2002). Posterior longitudinal ligamentous injury may occur with hyperflexion and hyperextension injuries (Atlas 2002). The posterior ligamentous complex is comprised of the ligamentum flavum, the interspinous ligament, the supraspinous ligament, and the facet joint capsule (Atlas 2002). The ligamentum flavum is a continuous strip of fi-

4.1  Extradural Diseases of the Spine

Fig. 4.1.43  Osteoblastic metastasis from high-grade ovarian carcinoma. a On CT, a dominant sclerotic lesion is seen at left lateral aspect of the L5 vertebral body. This lesion is well circumscribed, with a spiculated border, simulating a bone island (enostosis). There is a small blastic lesion more medially (small arrow) and similar lesions at multiple levels of the spine.

Sagit­tal MR images obtained at the same time demonstrate the dominant lesion (arrow) to be hypointense on both T1- (b) and T2-weighted images (c). A smaller lesion is seen in the sacrum (arrowhead). There is no evidence of enhancement (d). Followup at 6 months (e,f) demonstrates an increase of the dominant lesion associated with surrounding enhancement (arrows)

broelastic tissue that bridges the lamina (Atlas 2002). The ligamentum flavum opposes hyperflexion and distraction of the posterior elements in order to maintain normal alignment (Atlas 2002). The interspinous ligaments oppose hyperflexion and distraction of the posterior elements as well, and are best depicted on midsagittal images (Atlas 2002). The facet joint complexes are dynamic structures which limit compression and distraction of the posterior elements during flexion and extension, while resisting rotation and translation (Atlas 2002). The facet joints of the lumbar and cervical spine are oriented in the sagittal plane, as opposed to the thoracic spine, in which the facet joints are oriented in the coronal plane (Atlas 2002).

4.1.5.4.2  Vertebral Column Injury to the vertebral column may be manifested as changes in the alignment of the normal vertebral column, often noted on plain films and CT, and best seen on sagittal MR images (Saifuddin 2001). Osseous injuries to the vertebral column include fracture deformities and compressive injuries (Atlas 2002). Vertebral body fractures are depicted as breaks in the continuity of the signal void representing the vertebral body bony cortex (Saifuddin 2001). In particular, T2* GRE images may depict non-displaced vertebral body fractures through the vertebral bodies and/or posterior elements as hyperintense linear irregularity, which may be difficult to depict on

579

580

4  Spine and Spinal Canal

other spin-echo imaging sequences (Atlas 2002). Trauma can also manifest as signal changes within the vertebral body related to hemorrhage and edema (Saifuddin 2001). When present, vertebral body edema is often manifested as hyperintense signal on STIR or T2-weighted images (Fig. 4.1.46), and iso- to hypointense on T1-weighted images (Qaiyum et al. 2001). Wedge compression deformities may appear as bands of T2 hyperintensity paralleling the anterior vertebral body cortex (Saifuddin 2001). While acute and chronic compression deformities may have similar appearances on plain films or CT, they can often be differentiated on MRI (Fig. 4.1.47). Acute or subacute compression fractures are associated with bone marrow edema and inflammatory changes, while chronic compression deformities are unlikely to demonstrate these findings. Burst fractures (Fig. 4.1.48) occur with axial loading, and demonstrate abnormal signal changes in the vertebral bodies with associated disruption of the posterior vertebral body cortex and retropulsion of fracture fragments, potentially resulting in spinal cord compression (Saifuddin 2001). Because of the retropulsion of fracture fragments and possible spinal canal involvement, burst

fractures have a high propensity for neurological deficits (Atlas 2002). Edema with evidence of fracture or cortical break, representing injury and fractures of the bony trabeculae, may be seen within vertebral bodies associated with axial loading injuries (Saifuddin 2001). Chance type fractures typically involve the anterior/superior corner of the involved vertebral body, resulting in anterior wedge compression deformity, or corner fracture of the anterior/ superior end plate (Grove et al. 2005). Chance fractures also involve disruption and distraction of the posterior elements, which may involve the osseous structures, ligaments, or both, depicted as abnormal increased T2 signal involving the posterior column (Grove et al. 2005). Chance fractures also have an increased association with additional visceral organ damage (Grove et al. 2005). Qaiyum et al. (2001) have demonstrated that invisible, unsuspected injury to the vertebral column may be seen in patients with known acute spinal trauma. These additional injuries predominantly include vertebral body edema at both contiguous and non-contiguous levels with regard to the site of the initial injury, suggesting that the entire spine be examined when dealing with a single injured segment of the spinal column and spinal cord. As

Fig. 4.1.44  Epidural metastasis causing spinal cord compression and nerve root compression. a Axial T2-weighted image in a patient with metastatic renal cell carcinoma shows pathological fracture and epidural mass squeezing the spinal cord. b Sagittal STIR image is misleading since it shows that the spinal cord is only displaced posteriorly but there is preservation of the anteroposterior diameter of the spinal cord. This highlights the importance of including an imaging plane orthogonal to the long axis of the spinal cord for evaluation of spinal cord

compression. Note the severe compression of the T9 vertebral body. c Sagittal STIR image in a different patient with metastatic prostate carcinoma and lower extremity weakness shows soft tissue lesions in the neural foramina of L3–L4 to L5–S1 levels (white arrows). There is expansion of the neural foramina and compressing on the exiting nerve roots (dark signal intensity). Metastasis is seen in the prevertebral space from L1 to L5 and in the vertebral bodies of L2 to S2

4.1  Extradural Diseases of the Spine

was previously discussed, MRI tends to be insensitive to fractures involving the posterior elements (Fig. 4.1.49), due to the smaller size and more complex geometry of the posterior elements, as well as the decreased proportion of medullary space relative to the vertebral bodies (Atlas 2002). The vertebral bodies of the cervical spine deserve special notice with regard to the biomechanics and specific vector forces causing injuries in predictive patterns. Atlanto-axial dissociation, depicted as distraction of the atlas and skull in relation to the axis, often manifests as ligamentous disruption, soft tissue edema, and hemorrhage (Takhtani and Melhem 2000). When seen on MRI, the skull will be displaced superiorly, and a significant amount of T2-weighted hyperintensity will be noted in the adjacent soft tissues and within the affected vertebral bodies (Takhtani and Melhem 2000). If acute hemorrhage is present, then hypointense signal on T2-weighted images may be seen (Takhtani and Melhem 2000). Hyperflexion injuries of the cervical spine result from forward rotation and translation of the cervical spine in the sagittal plane, resulting in anterior subluxation of the vertebral body, reversal of the normal lordotic curvature

of the cervical spine, anterior narrowing and posterior widening of the affected disk spaces, and fanning of the spinous processes (Takhtani and Melhem 2000). These fractures are secondary to disruption of the posterior ligamentous complex, the posterior longitudinal ligaments, the posterior annulus, and the interlaminar ligaments (Takhtani and Melhem 2000). MRI will demonstrate adjacent T2 hyperintensity correlating to soft tissue and ligamentous injury, with associated contour irregularities, loss of vertebral body height, end plate injuries, and bone marrow edema related to vertebral body fractures (Takhtani and Melhem 2000). Hyperextension injuries of the cervical spine may result in a Hangman’s fracture, which is a traumatic spondylolisthesis at the level of the C2 vertebral body, with resultant fractures through the pars interarticularis and adjacent structures (Takhtani and Melhem 2000). On MRI, this will be seen as hyperintense signal on T2-weighted images in both the C2 vertebral body and the adjacent soft tissue, secondary to edema (Takhtani and Melhem 2000). The actual fracture may not be seen, although the deformity to the vertebral column should be recognized (Takhtani and Melhem 2000). Hyperexten-

Fig. 4.1.45  Normal cervical spine. Midline sagittal T1-weighted (a) and T2-weighted (b) images show normal vertebral body height, alignment, and signal intensity. All intervertebral disks demonstrate normal high signal intensity on the T2-weighted image, with the exception of T1–T2 disk showing mild degenerative signal loss. The anterior longitudinal ligament (white open arrow) and posterior longitudinal ligament (white arrowhead) are seen as dark bands along the anterior and posterior margins of the vertebral column. The superior extension of the posterior longitudinal ligament from C2 to the anterior rim of the foramen magnum is called the tectorial membrane. The apical ligament connects the superior tip of the dens to the basion (white

arrow). The ligamentum flavum is seen as a dark band extending along the posterior margin of the spinal canal, and extends to the opisthion (small black arrow). Interspinous ligaments (white asterisk) are seen between each consecutive spinous process. Note that all spinous processes are spaced and angulated evenly. c Off midline, the facets are aligned normally as a stack of shingles (anterior to white dashed line). The vertebral artery (curved white arrow) is seen as flow void within the foramina transversarium. For this reason, vertebral artery injury needs to be considered in the presence of cervical spine fracture that extends into the foramen transversarium

581

582

4  Spine and Spinal Canal Fig. 4.1.46  a Sagittal and b axial T2-weighted images of lumbar spine demonstrate edema involving the left aspect of the L3 vertebral body extending to the left-sided posterior elements (arrows). Note the hematoma formation of the adjacent paraspinal muscles (H)

Fig. 4.1.47  Acute and chronic compression in an elderly woman. a Sagittal T1-weighted, b STIR, and c post-gadolinium images. Compression deformities are seen in the T5, T9, and T12 vertebral bodies. However, only the T5 and T9 compression deformities are associated with bone marrow edema and enhancement, consistent with acute or subacute fractures. The T12 compression deformity is chronic

4.1  Extradural Diseases of the Spine

Fig. 4.1.48  L2 burst fracture. a Sagittal proton density weighted image and b axial T2-weighted image. There is retropulsion of fracture fragments (arrows), causing narrowing of the spinal canal. Mild compression deformity at the L1 vertebral body is also noted

sion dislocation injuries result predominantly from ligamentous injury and distraction of the anterior column, buckling of the middle column, and transient posterior dislocation of the intervertebral disk with immediate reduction after impact (Takhtani and Melhem 2000). MRI can depict the hyperintense T2-weighted signal changes in the adjacent soft tissue and the vertebral bodies related to the soft tissue injury, ligamentous disruption, bone marrow edema, and vertebral body fracture (Takhtani and Melhem 2000). Cervical vertebral compression injuries, including Jefferson fractures and burst fractures, result from an axial loading force on the cervical spine (Takhtani and Melhem 2000). Jefferson fractures occur with simultaneous disruption of the anterior and posterior arches of C1, with or without disruption of the transverse atlanto-axial ligament (Takhtani and Melhem 2000). Burst fractures in the cervical spine resemble burst fractures elsewhere,

resulting in compression of the vertebral body with fracture of the posterior cortex and retropulsion of fracture fragments, with possible compression of the spinal cord (Takhtani and Melhem 2000). MRI may not be as sensitive as CT for depicting actual fracture lines through the vertebral bodies and posterior elements, but MRI continues to be an invaluable resource depicting edema of the vertebral bodies, as well as edema, hemorrhage, and soft tissue swelling of the adjacent soft tissues (Takhtani and Melhem 2000). 4.1.5.4.3  Ligaments On a normal MRI study, the ligamentous structures are seen as thin, dark signal on all pulse sequences, due to their dense, fibro-elastic composition (Hassankhani and Bencardino 2003). A gap in the normally dark ligamen-

583

584

4  Spine and Spinal Canal

tous signal with increased signal of the adjacent soft tissues (Figs. 4.1.49, 4.1.50), particularly involving the posterior paraspinal region, is often indicative of ligamentous injury (Hassankhani and Bencardino 2003). Ligamentous injuries may include rupture, depicted as discontinuity of the normal dark ligamentous signal on all imaging sequences, stripping of an intact ligament, partial avulsion, or attenuation of the ligament, and combined osseoligamentous injury (Saifuddin 2001). These injuries are best depicted on STIR or T2-weighted images, in which the associated edema is depicted as hyperintense signal (Saifuddin 2001). Fat-saturation images are recommended, particularly with regard to the interspinous ligaments, as they tend to be surrounded by fat on either side (Saifuddin 2001). Disruption of the interspinous ligaments (Fig. 4.1.50) may appear as hyperintense signal on STIR or T2-weighted

images within the interspinous space, with or without an increase in the interspinous distance (Saifuddin 2001). Fat-saturation techniques increase imaging sensitivity, as the normally hyperintense fat typically seen adjacent to the interspinous ligaments may be a source of confusion (Saifuddin 2001). Prevertebral soft tissue swelling and edema are important secondary signs of anterior longitudinal ligament rupture and injury, which are particularly common in hyperextension injuries of the cervical spine (Saifuddin 2001). It is interesting to note that the anterior longitudinal ligament may be absent on T1-weighted spin-echo images at the level of degenerated intervertebral disks in patients with no history or clinical manifestations of trauma (Saifuddin 2001). The imaging appearance of altered facet joint alignment and subluxation (Fig. 4.1.51) is similar to the plain film and CT appearance (Atlas 2002). Facet joint capsule rupture is indicated by

Fig. 4.1.49  Subtle comminuted fracture of T11 left lamina secondary to gunshot wound. a Sagittal T1-weighted image shows subtle fracture line (black arrow) and small epidural hematoma (white open arrow). b Sagittal STIR image shows only mildly increased signal intensity in posterior element. The small hematoma is hyperintense (white open arrow). The ligamentum flavum has been disrupted (white arrow). c The fracture is difficult

to see on the axial T2-weighted image. The small epidural hematoma is seen as moderate signal intensity (open white arrow) slightly indenting the thecal sac. d,e Axial CT images in bone and soft tissue window confirm the fracture (black arrow) and an epidural hematoma (black open arrow) indenting the thecal sac (black arrowheads). The punctuate hypodensity in the spinal canal represents air

4.1  Extradural Diseases of the Spine

separation and distraction of the articular surfaces of the facet, with the presence of excess fluid between articular facets (Saifuddin 2001). 4.1.5.4.4  Intervertebral Disks Traumatic changes to the intervertebral disks include disk injury and post traumatic disk herniation (Atlas 2002). Identification and classification of disk injury and herniation is an important factor in determining the timing and type of surgical decompression and stabilization (Atlas 2002). Unrecognized disk herniation secondary to trauma may result in neurologic deterioration following stabilization (Hassankhani and Bencardino 2003). Disk injuries in the setting of acute trauma include intradisk hemorrhage and edema, annular rupture, end-

plate avulsions, and herniations (Saifuddin 2001). In the acute stage of injury, hemorrhage, and edema will have the appearance of hyperintense signal on T2-weighted images, while in the subacute stage hemorrhage will be depicted as hyperintense signal on T1-weighted images (Saifuddin 2001). Disk injury is implied when there is asymmetric narrowing or widening of the intervertebral disk, or if there is focal T2 hyperintensity within or adjacent to the disk (Hassankhani and Bencardino 2003). An injured disk is typically brighter in signal on T2-weighted images as compared to normal adjacent disks, which demonstrate a hypointense disk annulus on all pulse sequences (Hassankhani and Bencardino 2003). Rupture of the normally hypointense annulus, which commonly accompanies flexion-distraction and hyperextension injuries, can be identified as discontinuity of the typical dark signal of the annulus (Saifuddin 2001). Avulsion of the disk from the end plate can be seen in hyperextension injuries of the cervical spine, often identified as an area of hyperintensity on T2-weighted images in the anterior disk space paralleling the vertebral body end plate (Saifuddin 2001). Disk herniation is seen most commonly with hyperextension injuries and fracture-dislocations, particularly involving the cervical spine (Saifuddin 2001). Intraosseous herniations are uncommon, but may occasionally occur with a thoracolumbar burst fracture (Saifuddin 2001). There is a high association of post traumatic disk herniation with facet dislocation and subluxation (Fig. 4.1.51) (Hassankhani and Bencardino 2003). The extruded disk may lie behind the upper anteriorly subluxed vertebral body, and at the time of reduction the extruded disk may be displaced further posteriorly, resulting in cord compression (Hassankhani and Bencardino 2003). MRI is recommended in the preoperative evaluation of facet subluxation reduction, as it may alter the surgical approach (Hassankhani and Bencardino 2003). In addition, MRI is strongly recommended in patients with worsening neurological symptoms following closed reduction (Hassankhani and Bencardino 2003). 4.1.5.4.5  Traumatic Epidural Hematoma

Fig. 4.1.50  Sagittal T2-weighted image through the cervical spine demonstrates injury to the posterior longitudinal ligament (small arrow) and the ligamentum flavum (open arrow) at the level of C5–C6. Note the loss of the normal low signal intensity of the involved ligaments, consistent with disruption of the ligaments. There is disruption of the interspinous ligament at C5–C6 and extensive soft tissue contusion and edema in the surrounding soft tissues. Linear hyperintensity is seen tracking along the anterior longitudinal ligament from the C3–C4 to the C6–C7 levels (arrowheads), but the anterior longitudinal ligament remains intact

Epidural hematomas of the spine are not as uncommon as once thought, especially in the setting of trauma (Sklar et al. 1999). Epidural hematomas may produce spinal cord and/or cauda equina compression with neck and back pain, sensory and motor deficits, and progression to paraplegia and quadriplegia (Sklar et al. 1999). Epidural hematomas may have a variable appearance on MRI, depending on the age of the hematoma, including iso- to slightly hyperintense signal on T1-weighted images, and hyperacute hematomas are hyperintense on T2-weighted images (Sklar et al. 1999). They can be overlooked since their signal characteristics are not that different from ce-

585

586

4  Spine and Spinal Canal

rebrospinal fluid. Older hematomas follow the expected appearance of blood products and are easier to detect. On T2* GRE images, epidural hematomas are depicted as strikingly low signal intensity, due to the blooming artifact from the associated blood products (Sklar et al. 1999). In addition, a low signal curvilinear line may be noted surrounding the dural sac, best seen on T2*-weighted GRE images, also related to blood products (Sklar et al. 1999).

The epidural hematoma may show direct continuity with the adjacent osseous spine, and is often located posteriorly/laterally within the spinal canal (Sklar et al. 1999). Capping of the epidural fat and compression of the ligamentum flavum may be demonstrated as well (Sklar et al. 1999). Epidural hematomas may compress the thecal sac without compressing the cord, or may compress both or neither (Sklar et al. 1999).

Fig. 4.1.51  Right C6–C7 unilateral perched facet with comminuted fracture. a Sagittal CT image shows fracture at C6 inferior facet (arrow), with the small fracture fragment displaced posteriorly. There is superior and posterior subluxation of the superior facet of C7 (open arrow). b,c Sagittal T1-weighted fast spinecho and T2*-weighted gradient-echo images show almost the same findings as CT. The fracture is better seen on the gradientecho image because of the higher contrast between bones and soft tissue structures. d Sagittal proton density weighted image

shows disruption of the anterior longitudinal ligament (long arrow). There is anterolisthesis of C6 over C7, and disk herniation (short arrow). e Parasagittal STIR image confirms that the posterior longitudinal ligament is also disrupted (short arrow). Note the edema at interspinous ligaments and posterior paraspinal soft tissues. f Axial T2*-weighted gradient-echo image confirms the disk herniation (short arrow). There is rotary subluxation of the C6 vertebral body (clockwise rotation) because of the right unilateral perched facet

4.1  Extradural Diseases of the Spine

4.1.5.4.6  Vascular Injury MRA in association with MRI has been shown to be an excellent noninvasive method for both initial and followup evaluation of cervical artery dissection (Fig. 4.1.52) (Levy et al. 1994). Vascular injuries associated with vertebral column injury occur most commonly in the cervical spine, due to the close proximity of the vessels to the spine (Hassankhani and Bencardino 2003). Vascular dissection is a growing cause of infarction in young adults (Levy et al. 1994). It is the cause of infarction in 2% of adults aged 40–60 years old (Levy et al. 1994). The vertebral arteries are more commonly affected than the carotid arteries (Hassankhani and Bencardino 2003). Fracture of the posterior elements of the cervical spinal vertebral bodies may compress or stretch the ipsilateral vertebral artery (Hassankhani and Bencardino 2003). The pathophysiology of vascular dissection is related to an intimal tear or ruptured vasa vasorum, resulting in intramural hematoma formation, with subsequent stenosis and pseudoaneurysm formation (Ross et al. 2004). Vascular dissection is depicted on fat-saturated T1-weighted images as a crescentic focus of hyperintensity surrounding a signal flow void, commonly seen in the intracranial portion of the vessel, which may extend for variable length (Ross et al. 2004). On diffusion-weighted images, reduced diffusion indicative of acute ischemia may be noted in either the anterior or posterior circulation, depending on whether there is involvement of the carotid artery or vertebral artery respectively (Ross et al. 2004). MRA may demonstrate a Kantor sign or “string sign,” resulting from luminal narrowing caused by the dissection (Ross et al. 2004). MRA may also demonstrate lack of flow distal to the dissection, and apparent increase in the external artery diameter (Levy et al. 1994). Intermediate signal intramural hematoma or a thin, curvilinear, hypointense intimal flap may be evident as well (Ross et al. 2004). References 1.

2.

Fig. 4.1.52  Vascular dissection secondary to traumatic cervical spine injury. a Maximum intensity projection from contrastenhanced MRA shows multiple segments of narrowing in left vertebral artery and bilateral internal carotid arteries (arrows). b Axial fat-suppressed T1-weighted image at the level of the largest right internal carotid artery lesion (long arrow on a). There is prominent intramural hematoma (intraluminal hyperintensity) causing severe narrowing of the true lumen of the internal carotid artery (white arrows)

3. 4. 5.

Atlas SW (2002) Magnetic resonance imaging of the brain and spine, 3rd edn. Lippincott Williams & Wilkins, Philadelphia Baur A et al (1998) Diffusion-weighted MR imaging of bone marrow: differentiation of benign versus pathologic compression fractures. Radiology 207:349–356 Bazan C (1993) Imaging of lumbosacral spine neoplasms Neuroimag Clin N Am 3:591–608 Bertoni F, Bacchini P, Capanna R et al (1993) Solid variant of aneurysmal bone cyst. Cancer 71:729–734 Boden SD, Davies DO, Dina TS et al (1992) Postoperative diskitis: distinguishing early MR imaging findings from normal postoperative disk space changes. Radiology 184:765–771

587

588

4  Spine and Spinal Canal 6.

7.

8.

9.

10.

11.

12.

13. 14.

15.

16.

17.

18.

19.

20.

21.

Bozzao A, Gallucci M, Masciocchi C, et al (1992) Lumbar disk herniation: MR imaging assessment of natural history in patients treated without surgery. Radiology 185:135–141 Braun IF, Hoffman JCJ, Davis PC, et al (1985) Contrast enhancement in CT differentiation between recurrent disc herniation and postoperative scar: prospective study. Am J Roentgenol 145:785 Castillo M, Arbelaez A, Smith JK, Fisher LL (2000) Diffusion-weighted MR imaging offers no advantage over routine nonconrast MR imaging in the detection of vertebral metastases. Am J Neuroradiol 21:948–953 Cavanaugh JM, Ozaktay AC, Yamashita T et al (1997) Mechanisms of low back pain: a neurophysiologic and neuroanatomic study. Clin Orthop Rel Res 335:166–180 Crim JR, Mirra JM, Eckardt JJ, Seeger LL (1990) Widespread inflammatory response to osteoblastoma: the flare phenomenon. Radiology 177:835–836 Dagirmanjian A, Schils J, Mettency M, Modic MT (1996) MR imaging of vertebral osteomyeltis revisited. Am J Roentgenol 167:1539–1543 Eastwood JD, Vollmer RT, Provenzale JM (2002) Diffusion-weighted imaging: a patient with vertebral and epidural abscesses. Am J Neuroradiol 23:496–498 Fardon DF, Milette PC (2001) Nomenclature and classification of lumbar disc pathology. Spine 26:E93–E113 Filipo RM, Cotten A, Chastanet P et al (1996) Evaluation of destructive spondyloarthropathy with hemodialysis by computerized tomographic scan and magnetic resonance imaging. J Rheumatol 23:869–873 Fiorella D, Jones T, Helms CA et al (1999) Angled axial MR images in the lumbar spine: a potential source of diagnostic error. Radiology 213(P):157 Frazzini VI, Kagetsu NJ, Johnson CE, Destian S (1997) Internally stabilized spine: optimal choice of frequency-encoding gradient direction during MR imaging minimizes susceptibility artifact from titanium vertebral body screws. Radiology 204:268–272 Fritsch EW, Heisel J, Rupp S (1996) The failed back surgery syndrome: reasons, intraoperative findings, and longterm results: a report of 182 operative treatments Spine 21:626–633 Gamba JL, Martinez S, Apple J, Harrelson JM, Nunley JA (1984) Computed tomography of axial skeletal osteoid osteomas. AJR Am J Roentgenol 142:769–772 Gillams AR, Chaddha B, Carter AP (1996) MR Appearances of the Temporal Evolution and Resolution of Infectious Spondylitis. American Journal of Radiology; AJR Am J Roentgenol 166:903–907 Grove CJ, Cassar-Pullicino VN, Tins BJ, Tyrell PNM, McCall IW (2005) Chance-type flexion-distraction injuries in the thoracolumbar spine: MR imaging characteristics. Radiology 236:601–608 Hadjipavlon AG, Mader JT, Necessary JT, Muffoletto AJ (2000) Hematogenous pyogenic spinal infection and their surgical management. Spine 25:1668–1679

22. Hassankhani A, Bencardino JT (2003) Magnetic resonance imaging of sports injuries of the spine. Top Magn Reson Imag 14:87–102 23. Hierholzer J, Benndorf G, Lehmann T et al (1996) Epidural lipomatosis: case report and literature review. Neuroradiology 38:343–348 24. Jensen MC, Brant-Zawasdzki MN, Obuchowski N et al (1994) Magnetic resonance imaging of the lumbar spine in people without back pain. NEJM 331:69–73 25. Jinkins JR, Matthes JC, Sener RN, et al (1992) Spondylolysis, spondylolisthesis, and associated nerve root entrapment in the lumbosacral spine: MR evaluation. AJR Am J Roentgenol 159:799–803 26. Jung N, Jee W, Ha K et al (2004) Discrimination of tuberculous spondylitis from pyogenic spondylitis on MRI. AJR Am J Roentgenol 182:1405–1410 27. Kransdorf MJ, Sweet DE (1995) Aneurysmal bone cyst: concept, controversy, clinical presentation, and imaging. AJR Am J Roentgenol 164:573–580 28. Kroon HM, Schurmans J (1990) Osteoblastoma: clinical and radiologic findings in 98 new cases. Radiology 175:783–790 29. Ledermann HP, Schweitzer ME, Morrison WB, Carrino JA (2003) MR Imaging findings in spinal infections: rules or myths? Radiology 228:506–514 30. Levy C, Laissy JP, Raveau V, Amarenco P, Servois V, Bousser MG, Tubiana JM (1994) Carotid and vertebral artery dissections: three-dimensional time-of-flight MR angiography and MR imaging versus conventional angiography. Radiology 190:97–103 31. Love C, Patel M, Lonner BS et al (2000) Diagnosing spinal osteomyelitis: a comparison of bone and Ga-67 scintigraphy and magnetic resonance imaging. Clin Nucl Med 25:963–977 32. Mader JT, Shirtliff ME, Bergquist SC, Calhoun J (1999) Antimicrobial treatment of chronic osteomyelitis. Clin Orthop Related Res 360:47–65 33. Maiuri F, Iaconetta G, Gallicchio B et al (1997) Spondylodiscitis: clinical and magnetic resonance diagnosis. Spine 22:1741–1746 34. Masaryk TJ, Ross JS, Modic MT et al (1988) High-resolution MR imaging of sequestered lumbar intervertebral disks. AJNR Am J Neuroradiol 9:351–358 35. Modic MT, Feiglin DH, Pirano DW et al (1985) Vertebral osteomyelitis and assessment using MR. Radiology 157:157–166 36. Modic MT, Steinberg PM, Ross JS et al (1988) Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging. Radiology 166:193–199 37. Moorthy S, Prabhu NK (2002) Spectrum of MR imaging findings in spinal tuberculosis. AJR Am J Roentgenol 179:979–983 38. Munk PL, Lee MJ, Poon PY et al (1997) Candida osteomyelitis and disc space infections of the lumbar spine. Skeletal Radiol 26:42–46

4.1  Extradural Diseases of the Spine 39. Murphy MD, Andrews CL, Flemming DJ et al (1996) Primary tumors of the spine: radiologic-pathologic correlation. Radiographics 16:1131–1158 40. Murphy MD, Choi JJ, Kransdorf MJ et al (2000) Imaging of osteochondroma: variants and complications with radiologic – pathologic correlation. Radiographics 20:1407–1434 41. Myles ST, MacRae ME (1988) Benign osteoblastoma of the spine in childhood. J Neuro Surg 68:884–888 42. Narvaez J, Nolla JM, Narvaez JA et al (2006) Spontaneous pyogenic facet joint infection. Semin Arthritis Rheumatol 35:272–283 43. Okada F, Takayama H, Doita M et al (2005) Lumbar facet joint infections associated with epidural and paraspinous abscess: a case report with review of the literature. J Spinal Disord Tech 18:458–61 44. Petersilge CA, Lewin JS, Duerk JL et al (1996) Optimizing imaging parameters for MR evaluation of the spine with titanium pedicle screws. AJR Am J Roentgenol 166:1213–1218 45. Provenzale JM et al (1996) Spinal angiolipomas: MR features. AJNR Am J Neuroradiol 17:713–719 46. Qaiyum M, Tyrell PNM, McCall IW, Cassar-Pullicino VN (2001) MRI detection of unsuspected vertebral injury in acute spinal trauma: incidence and significance. Skeletal Radiol 30:299–304 47. Rattcliffe JF (1980) The arterial anatomy of the adult human vertebral body. J Anat 131:57–79 48. Reihsaus E, Waldbauer H, Seling W (2000) Spinal epidural abscess: a meta analysis of 915 patients. Neurosurg Rev 23:175–204 49. Ross JS et al (2004) Diagnostic imaging – spine. AMIRSYS, Salt Lake City 50. Ross JS, Masaryk TJ, Modic MT (1987) Vertebral hemangiomas: MR imaging. Radiology 165:165–169 51. Ross JS, Masaryk TJ, Schrader M et al (1990) MR imaging of the postoperative lumbar spine: assessment with gadopenetate dimeglumine. AJR Am J Roentgenol 155:867–872 52. Ruggieri PM (1999) Pulse sequences in lumbar spine imaging. MRI Clin N Am 7:425–437 53. Saifuddin A (2001) MRI of acute spinal trauma. Skeletal Radiol 30:237–246 54. Schey WL, Shkolnik A, White H (1997) Clinical and radiographic considerations of sacrococcygeal teratomas: an analysis of 26 new cases and review of the literature. Radiology 125:189–195 55. Schmorl G, Junghanns H (1971) The human spine in health and disease, 2nd edn. Grune and Stratton, New York, p 325

56. Schweitzer ME, El-Noueam KI (1998) Vacuum disc: frequency of high signal intensity on T2-weighted MR images. Skeletal Radiol 27:83–86 57. Sharif HS (1992) Role of MR imaging in management of spinal infections. AJR Am J Roentgenol 158:1333–1345 58. Sharif HS, Hideyan OP, Clark DC et al (1989) Brucellar and tuberculous spondylitis: comparative imaging features. Radiology 171:419–425 59. Sklar EM, Donovan Post JM, Falcone S (1999) MRI of acute spinal epidural hematomas. J Comput Assist Tomogr 23:238–243 60. Soehle M, Wallenfang T (2002) Spinal epidural abscess: clinical manifestation, prognostic factors, and outcomes. Neurosurgery 51:79–87 61. Stäbler A, Bellan M, Weiss M et al (1997) MR imaging of enhancing intraosseous disk herniation (Schmorl’s nodes). AJR Am J Roentgenol 168:933–938 62. Sze G, Uichanco LS, Brant-Zawadzki M et al (1988) Chordomas: MR imaging. Radiology 166:187–191 63. Takhtani D, Melhem ER (2000) MR imaging in cervical spinal trauma. MRI Clin N Am 8:615–633 64. Teplick JG, Laffey PA, Berman A et al (1986) Diagnosis and evaluation of spondylolisthesis and/or spondylolysis on axial CT. AJNR Am J Neuroradiol 7:479–481 65. Thrush A, Enzman D (1990) MR imaging of infectious spondylitis. AJNR Am J Neuroradiol 11:1171–1180 66. Ulmer JL, Elster AD, Mathews VP, King JC (1994) Distinction between degenerative and isthmic spondylolisthesis on sagittal MR images: importance of increased anteroposterior diameter of the spinal canal (“wide canal sign”). AJR Am J Roentgenol 163:411–416 67. Volger JA, Murphy WA (1988) Bone marrow imaging. Radiology 168:679–693 68. Wagner SC, Schweitzer ME, Morrison WB et al (2000) Can imaging findings help differentiate spinal neuropathic arthropathy from disk space infection? Initial experience. Radiology 214:693–699 69. Weinberg J, Silber JS (2004) Infections of the spine: what the orthopedist needs to know. Am J Orthop 33:13–17 70. White ML (2000) Cervical spine: MR imaging techniques, and anatomy. MRI Clin N Am 8:453–469 71. Williams RL, Fuki MB, Meltzer CC et al (1999) Fungal spinal osteomyelitis in the immunocompromised patient: MR findings in three cases. Am J Neuroradiol 20:381–385 72. Yu S, Haughton VM, Sether LA, Wagner M (1988) Annular fibrosis in bulging intervertebral disks. Radiology 169:761–763

589

590

4  Spine and Spinal Canal

4.2  Intradural Extramedullary Spine D. Lin 4.2.1  Introduction MRI is fundamental in the diagnostic evaluation of spinal lesions, providing exquisite anatomic delineation and lesion characterization in various compartments. While the classification of lesions into discrete compartments, including extradural, intradural extramedullary, and intramedullary, is perhaps an oversimplification (since many lesions can occur in several compartments), it is nevertheless a convenient location-based scheme for generating differential diagnoses. In the intradural extramedullary (IDEM) spinal compartment, the primary differential consideration of a space-occupying lesion includes neoplasm (overwhelmingly dominated by nerve sheath tumors and meningiomas) and congenital, often cystic, lesions such as epidermoid, dermoid, and arachnoid cysts. These mass lesions, when reaching a certain size, may compress and displace the spinal cord or nerve roots, while widening the ipsilateral subarachnoid spaces above and below and maintaining a sharp angle with the adjacent dura and neural element, forming a so-called meniscus sign. These are classic features important for locating a lesion in the IDEM compartment on myelography, but can also be very useful on MRI. In addition to displaying these indirect signs, MRI has the advantage of allowing direct visualization of the lesion itself and providing information about its signal characteristics and enhancement pattern, which is critical for reaching a diagnosis. Furthermore, contrast-enhanced MRI provides a sensitive means of depicting leptomeningeal disease, which may be a manifestation of infectious, inflammatory, or neoplastic processes. At times leptomeningeal disease presents as focal soft tissue thickening and modularity, particularly in neoplasia, but myriad pathological processes cause smooth, thin enhancement along the pial covering that can only be depicted by a contrast-enhanced MR examination. In this section, IDEM spinal lesions will be described according to the general categories of neoplastic, congenital, vascular, and infectious or inflammatory processes. The lumbar spine is unique in that it contains the cellular distal terminus of the spinal cord (conus medullaris), the cauda equina, and the filum terminale (fibrous continuation of the pia mater). Certain neoplastic lesions preferentially arise from these anatomical structures, with histology overlapping what is typically considered intramedullary in origin. For the sake of completeness, lesions associated with the cauda equina and filum terminale will be discussed in this section along with other intradural extramedullary pathologies.

4.2.2  Examination Techniques For MRI of the cervical and thoracic spine, patients are positioned supine with head first. Imaging of the lumbar and sacral spine, on the other hand, is best performed with patients positioned supine with feet first, so as to minimize possible feelings of claustrophobia. A knee support can be used for comfort, and this has the additional advantage of flattening the lumbar lordosis, thus increasing signal to noise by reducing the distance between the spinal canal and the radiofrequency (RF) receiver coil (Van Goethem et al. 2004). When the thoracic spine alone is the site of interest, either the cervical or the lumbar spine is included in a series of large field-of-view sagittal images for localization purposes. It is often helpful to include one or more markers applied to the skin on the patient’s back. MRI of the spine is typically performed with phasedarray receiver coils, since they combine the good sensitivity of surface coils with the coverage available from larger body coils. The use of multiple coil elements also offers the possibility of scan time reductions using parallel imaging techniques such as SENSE or SMASH. Usually, multiple selectable coil elements covering the cervical, lumbar, or thoracic spine, respectively, are available, and combinations can be chosen depending on the desired anatomical coverage. Phased-array coils are usually receive-only, with a single, large-volume coil used for transmitting RF pulses; however, more recently, particularly for high-field MRI systems, transmit phased-array coil designs have started to become available. Algorithms for optimal combination of multiple coil elements and corrections for coil inhomogeneities are also available and should be used routinely. An anterior saturation band is applied in thoracic and lumbosacral spine imaging, in order to reduce pulsation artifacts related to respiratory and cardiac motion. A typical examination consists of both sagittal and axial images, using T1-weighted spin-echo and T2-weighted fast spin-echo (FSE) sequences. T1-weighted imaging is most sensitive for detection of bone marrow signal abnormality such as that due to vertebral metastases. Short tau inversion recovery (STIR) or other fat suppressed T2-weighted imaging also highlights marrow signal abnormality related to edema or cellular infiltration, and is often included in the sagittal plane. Slice thickness is usually 3–4 mm in both sagittal and axial images. An uneven number of slices on the sagittal sequence are preferred, so that the midpoint of the cord and spinal canal is depicted on the middle slice. Axial images can be obtained through areas of interest rather than the entire spine. For the evaluation of intradural disease, post–gadolinium-diethylenetriaminepentaacetate (Gd-DTPA) sagittal and axial T1-weighted imaging should always be performed for better detection and characterization of lesions (Sze 1993).

4.2  Intradural Extramedullary Spine

The typical spine protocol is outlined in Table 4.2.1. Using this protocol, examples of normal appearance of the spine are illustrated in Figs. 4.2.1, 4.2.2, 4.2.3, and 4.2.4. Other sequences may sometimes be of use in spinal MRI as well: for instance, diffusion-weighted MRI (DWI) is helpful for differentiating cystic or necrotic tumors from abscesses, which usually show restricted diffusion; FLAIR (fluid attenuated inversion recovery) MRI, which suppresses cerebrospinal fluid (CSF) signal, may offer superior detection of intramedullary lesions, and screening sagittal gradient recalled echo (GRE) out-ofphase sequences are useful for visualizing vertebral pathology. MR myelography, a very long echo time (TE) FSE T2-weighted sequence, depicts CSF spaces and provides similar information to conventional X-ray myelography, without the need for contrast injection (Fig. 4.2.5). In addition, high-resolution fast acquisition sequences such as steady state free procession (SSFP, also known as FISP, or FIESTA), the contrast of which depends on both T1 and T2, provide superior discrimination between the neural elements and CSF. Finally, Gd-DTPA contrast-enhanced MR angiography (Bowen et al. 2003) has been used to delineate dural arteriovenous fistula (AVF). 4.2.3  Normal Anatomy By definition, the IDEM space lies between the dura mater and the spinal cord. The meningeal coverings in the

spine include three layers: (1) the outer-most, thickest and toughest layer called dura mater; (2) the avascular, thin, fragile, and spider web–like middle layer called the arachnoid mater, internal to which is the subarachnoid space containing CSF; and (3) a very thin, vascular membrane called the pia mater that closely invests the spinal cord, with thin, outward extensions at regular intervals that attach to the dura forming delicate denticulate ligaments. The dura mater is also called pachymeninx, while the pia and arachnoid layers constitute the leptomeninges. In adults, the spinal cord usually terminates at the L1–L2 disk level and assumes a tapered, coned configuration, therefore named conus medullaris. Caudal to the conus, the lumbar, sacral, and coccygeal spinal nerve roots descend and extend obliquely to exit in the respective neural foramina. The appearance of the collection of intradural spinal nerve roots from their attachment at the conus yields the name of cauda equina, or horse’s tail (Fig. 4.2.5). The filum terminale is a fibrovascular filament continuing from the pia covering the apex of the conus medullaris. The upper part, termed filum terminale internum, is a thin strand enclosed within the dural sac and surrounded by the cauda equina fibers which may reach as far as the lower border of S2. The lower part, filum terminale externa, represents the portion where it pierces the dura and attaches to the first segment of the coccyx. While the filum terminale is mostly made of connective tissue, several cellular elements are found within

Fig. 4.2.1  Normal appearance of the cervicothoracic spine in a healthy 30-year-old male on (a) T1-weighted, (b) T2-weighted, and (c) STIR images. Note that a large field of view is used in this case to include both cervical and thoracic spine so that the vertebral level can be easily counted. An anterior saturation band has been applied to minimize respiratory pulsation artifact from the motion transmitted from the anterior chest and abdominal wall

591

592

4  Spine and Spinal Canal

it including neuronal components, glial cells, and ependymal cells (Choi et al. 1992). This explains the genesis of some neoplasms found in the conus, cauda equina, and filum terminale that should be more correctly considered intramedullary tumors, yet typically are localized to the IDEM compartment because of the gross morphologic findings. On MRI, the contents of the spinal canal, from outward in, including the dura, spinal cord and cauda equina nerve roots are best outlined on T2-weighted images against the bright signal from CSF (Fig. 4.2.5). 4.2.4  Pathological Findings 4.2.4.1  Neoplasms Spinal neoplasms are rare lesions in the general population. About 30% of all spinal tumors are located in the IDEM compartment (Table 4.2.2) (Van Goethem et al. 2004); among these the primary neoplasms are dominated by nerve sheath tumors and meningiomas. As mentioned above, the lumbar spine is unique in its cellular content; therefore the neoplasms arising in this region reflect its anatomy and cellular components, which overlap both intramedullary and extramedullary elements. The primary neoplasms found in the conus and filum include schwannoma, myxopapillary ependymoma, and paraganglioma. Finally, the spinal subarachnoid space is a site for metastases, from CSF seeding of intracranial Fig. 4.2.2  Normal T2-weighted MRI appearance, in the axial neoplasms such as medulloblastoma or ependymoma, or plane, of the spinal canal at (a) cervical and (b) thoracic level hematogenous dissemination along the leptomeninges of

Fig. 4.2.3  Normal appearance of the lumbar spine (a) T1weighted, (b) T2-weighted, and (c) STIR images

4.2  Intradural Extramedullary Spine

Fig. 4.2.5  MR myelography (TR 9,370/TE 1,100, 40-mm thick/ 0 spacing, FOV 25 × 25 cm) of normal lumbar spine showing the outline of the conus medullaris and cauda equina contrasted by bright CSF

other primary neoplasms, including lung and breast cancer. Systemic, hematologic disorders such as lymphoma, leukemia, and myeloma may also involve the subarachnoid space.

on T1-weighted imaging, and enhance with contrast. Both also tend to be T2 hyperintense, but schwannomas may show more heterogeneity and appear isointense or slightly dark, and have the propensity of developing cystic changes (Lee 2000). While they share similar characteristics on neuroimaging, histologically they have distinctly different patterns. Furthermore, the nerve sheath tumors in patients with neurofibromatosis type 1 (NF1) are typically neurofibromas, whereas those occurring sporadically or in patients with NF2 are predominantly schwannomas (Halliday et al. 1991).

Nerve Sheath Tumors Nerve sheath tumors are by far the most common IDEM tumor, but often occupy mixed compartments in the intradural and extradural spaces (15%), and are occasionally completely extradural in location (25%). Very rarely are they found intramedullary (Van Goethem et al. 2004). They are histologically benign and include schwannoma and neurofibroma. Both typically involve the dorsal sensory nerve roots (Sevick and Wallace 1999), and sensory changes were found to represent the third most common presentation (more frequently in the cervical spine involvement) in a series of 66 spinal nerve sheath tumors (el-Mahdy et al. 1999), following local tenderness and radicular pain (60% of cases) and motor symptoms (24%). Both appear circumscribed, usually isointense to muscle

Schwannoma Schwannoma is the most common tumor, comprising about 60–65% of all IDEM tumors (el-Mahdy et al. 1999). Affecting females and males at about the same frequency, it commonly involves young to middle-aged adults, presenting during the fourth decade of life, most often occurring sporadically as a solitary mass. It has increased incidence of occurrence and multiplicity associated with NF2 (Mautner et al. 1995; Patronas et al. 2001) (Fig. 4.2.6). In NF2 patients, multiple spinal tumors including schwannomas, meningiomas (extramedullary), and ependymomas (intramedullary) are found at about 75–89% frequency, and 25–35% of these patients reported symptoms (Mautner et al. 1995). Among the extramedullary tumors in NF2, schwannomas represent the

Fig. 4.2.4  Normal T2-weighted MRI appearance of the lumbar spine in the axial planes at the level of the (a) conus medullaris, (b) proximal and (c) distal cauda equina

593

594

4  Spine and Spinal Canal Table 4.2.1  Typical MR parameters for spine imaging at 1.5 T Plane

TR  (ms)

TE  (ms)

TI  (ms)

Slice thickness (mm)

Field of view  (mm)

 No. of excitations

T1 SE

Sagittal

617

9

4

280

1.5

T2 FSE

Sagittal

3,050

124

4

280

2

T1 SE

Axial

700

10

4

200

3

T2 FSE

Axial

3,700

90

4

200

4

STIRa

Sagittal

3,750

65

4

280

1.5

150

Optional. Most useful in the setting of trauma and in the evaluation of spinal cord signal abnormality TR relaxation time, TE echo time, TI inversion time, SE spin echo, FSE fast spin echo, STIR short tau inversion recovery a 

predominant tumor type in studies based on histology and MR appearance (Patronas et al. 2001). Schwannoma is a benign, circumscribed, and encapsulated neoplasm arising from the transitional zone between the central (oligodendrocyte) and peripheral myelin (Schwann cell). Histologically it is composed of tissue following two growth patterns: a cellular component consisting of spindle-shaped Schwann cells (Antoni type A) and a less

Fig. 4.2.6  NF2 in an 18-year-old female with back pain, left sided hearing loss, and left foot deformity due to high arch. Post-contrast MRI of the spine shows innumerable IDEM enhancing nodules along the surface of the spinal cord and the cauda equina, characteristic of multiple schwannomas

cellular component rich in myxoid and microcyst formations (Antoni type B). Grossly, it is a discrete, ovoid, or dumbbell-shaped mass that may cause scalloping of the posterior vertebral body, or widening of the neural foramen. On MRI it is T2 hyperintense relative to the cord, with avid contrast enhancement (Fig. 4.2.7), and not infrequently has cystic formation. Spinal schwannomas have a greater propensity to cystic changes compared to their intracranial counterpart (Sevick and Wallace 1999). On the other hand, heterogeneous T2 signals with areas of hypointensity may reflect hemorrhage, increased cellularity, or collagen deposition (Friedman et al. 1992). Neurofibroma Although neurofibroma also arises from the nerve sheath derived from Schwann cells, it is distinguished from schwannoma in many aspects. It is composed of perineural cells, fibroblasts, and mucopolysaccharide matrix, in addition to Schwann cells (Sevick and Wallace 1999). Unlike schwannoma, it is unencapsulated, arising from and enveloping the dorsal sensory nerve root and fascicles, making it impossible to separate from the adjacent nerves. Neurofibroma is rarely sporadic even when solitary, and usually occurs in the context of NF1. In a study of 54 patients with NF1, 65% of them had spinal tumors evident on MRI, among whom 66% had neurological deficits and 34% were asymptomatic (Sze 1993). In general, spinal tumors are less clinically significant in NF1 compared to NF2 (Sze 1993; Thakker et al. 1999), and only the presence of spinal neurological deficits in NF1 should prompt further evaluation. In Thakkar’s series, the extramedullary tumors in NF1 were most commonly intraforaminal (57%) and intradural extramedullary (33%) in location. In a subset of 12 patients who underwent surgical intervention, all extramedullary intraspinal tumors were proven to be neurofibromas (Sze 1993). The MRI appearance of neurofibroma is similar to that of schwannoma, being T2 hyperintense but often also containing central T2 hypointensity, most likely reflect-

Second most frequent

Most frequent intramedullary

Rare

T-spine

Conus, caudaequina

Conus, caudaequina

Meningioma

Ependymoma

Paraganglioma

30–50

40–60

30–40

Age (years)

Local > irradiating pain

Local, irradiating pain, cord symptoms late

Local, irradiating pain

Radicular pain, paresthesia, numbness

Clinical

Sometimes scalloping/erosion

Sometimes calcification

Scalloping

Plain radio­ graph/CT

Hyperintense

Hyperintense

Hypo- to isointense Isointense

Iso- to slightly hyperintense

Hypo- to isointense

T2 Very hyperintense

a

Isointense

T1

MR

Strong

Strong

Strong

Mild to moderate

Gd enhance

Yes

Yes

No

Yes

Cyst

Modified from Van Goethem et al. 2004 T1 T1-weighted images, T2 T2-weighted images, Gd enhance Gd-DTPA enhancement pattern, Cyst tendency of cyst formation, Ca calcification, Heme hemorrhage a MR appearance: signal compared to the spinal cord and nerve roots

Most frequent

T-spine

Nerve sheath

Inci­dence

Level

Lesion

Table 4.2.2  Common IDEM neoplasms

No

No

Yes

No

Ca

Yes

Yes

No

No

Heme

4.2  Intradural Extramedullary Spine 595

596

4  Spine and Spinal Canal

Fig. 4.2.7  Typical appearance of a schwannoma in a patient presenting with radiculopathy. a T2-weighted and b contrastenhanced T1-weighted images show a dumbbell-shaped, wellcircumscribed mass expanding the left neural foramen. It is T2 hyperintense with central heterogeneously T2 dark areas, and

shows avid enhancement. In this case the tumor is extradural in location. In a different patient with NF2 shown in c postcontrast T1-weighted image, in addition to bilateral extradural schwannomas, there are two enhancing schwannomas in the IDEM compartment (white arrows)

ing dense collagen evident on histopathology, while the periphery contains myxomatous tissue (Sevick and Wallace 1999) (Fig. 4.2.8). Plexiform neurofibroma is unique, occurring only in patients with NF1, and shows a diffuse, infiltrative pattern (Fig. 4.2.9). There is a life-time estimated risk of 5–15% of malignant transformation of the plexiform neurofibroma (Evans et al. 2002), often suggested on neuroimaging by irregularity, associated bony erosion, destruction, and infiltration. Meningioma This is the second most common neoplasm, accounting for about 35% of IDEM primary tumors. Meningioma arises from arachnoid cells, is typically well circumscribed, and grows slowly. In a series of 36 cases of spinal meningiomas, 83% of them were found to be completely IDEM in location (Gezen et al. 2000), while a small portion of spinal meningiomas can occur in the extradural or mixed compartments. Similar to, or more so, compared to its intracranial counterpart, spinal meningioma often affects women (F:M = 3:1 in Gezen’s series of 36 patients, but a higher female prevalence was found in other series; Roux et al. 1996), with a peak age of occurrence in the fifth to sixth decades. It occurs most frequently in the thoracic spine (80%), followed by cervical spine and rarely in the lumbar region (Van Goethem et al. 2004). Most patients present with localized or radicular pain (83%), weakness (83%), sensory loss (50%), and less commonly, bladder and bowel dysfunction (Gezen et al. 2000). On MRI, it has a broad dural attachment, is isointense to the cord on T1-weighted imaging, isointense to slightly hyperintense on T2-weighted imaging, and has a moderate to avid enhancement pattern, at times with dural tails (Fig. 4.2.10). Calcification may occur, particularly associated with the psammoma bodies on histology (Kleihues and Cavenee 2000). When heavily calcified, it may demonstrate dark signal on all MR pulse sequences with little contrast enhancement, and is well delineated on CT (Fig. 4.2.11). Adjacent bone reaction may be seen.

Fig. 4.2.8  Neurofibromas in a patient with NF1. T2-weighted sagittal images of the lumbar spine demonstrating multiple well-circumscribed neurofibromas occupying the IDEM compartment of the spine (a) and expanding the neural foramina (b). On contrast-enhanced axial T1-weighted image (c), the mass shows mild enhancement

4.2  Intradural Extramedullary Spine

In general, meningioma is amenable to surgical resection with good outcomes, although total resection becomes difficult if the tumor is located ventral to the cord and is calcified (Roux et al. 1996). 4.2.4.1.1  Neoplasms in the Cauda Equina and Filum Terminale

Fig. 4.2.9  22-year-old male with NF1 and left arm pain. Three axial T2-weighted MR images through the lower cervical and upper thoracic spine show multiple plexiform neurofibromas, predominantly extradural and extraspinal in location, involving the neural foramina, paraspinal, and prevertebral regions. The masses in the middle section fill the bilateral neural foramina and cause marked compression of the cord from both sides

Fig. 4.2.10  Meningioma in a patient with myelopathy. a Postcontrast sagittal and b axial T1-weighted images show an enhancing IDEM mass at the level of C2, with broad dural attachment and dural tails, mildly compressing the right ventral aspect of the cord

Neoplasms in the lumbar region include nerve sheath tumors, meningiomas, myxopapillary ependymomas, paragangliomas, and leptomeningeal metastases. While meningioma is a common intradural tumor, it is infrequently found in the lumbar spine (Gezen et al. 2000; Roux et al. 1996). Figure 4.2.12 shows an unusual, distal lumbosacral intradural metastatic meningioma that has recurred after previous surgical resection, with evidence of diffuse leptomeningeal enhancement due to reactive changes. An example of a cauda equine schwannoma is shown in Fig. 4.2.13, with morphology and signal characteristics similar to schwannoma occurring in other spinal locations. Of note, ependymoma at this site appears intradural extramedullary and may be classified as such based on morphology, but is considered an intramedullary tumor arising from ependymal cells that are also present in the filum. Also of interest is hemangioblastoma, which is a pial-based vascular tumor derived from endothelial cells, most often located intramedullary (60%) but at times easily confused with an IDEM mass because of its peripheral location. Hemangioblastoma can also occur in the IDEM (21%) or other compartments (Zee 1996). In individuals with von Hippel-Lindau (VHL) syndrome, there is in-

Fig. 4.2.11  Heavily calcified meningioma, well defined on a non-contrast-enhanced CT, in the mid thoracic spine in a 43year-old male who had 2-year history of progressive myelopathy of the lower extremities. Surgical pathology confirmed a meningioma with extensive fibrosis and calcification

597

598

4  Spine and Spinal Canal

creased incidence of developing hemangioblastoma, with a greater propensity for multiple lesions (Fig. 4.2.14). Ependymoma Ependymoma is common and accounts for 63% of intraspinal primary gliomas (Sevick and Wallace 1999). The peak incidence is in the fourth and fifth decades with mean age of diagnosis at approximately 35 years, without particular sex predilection (Sevick and Wallace 1999).

Ependymoma arises from ependymal cells, and hence is located centrally in the cord, causing expansion. On histopathology it is characterized by sheets of ependymal cells and perivascular pseudo-rosette (Kleihues and Cavenee 2000). While the cellular type ependymoma presents as an intramedullary mass predominantly in the cervical spine, the most common locations of occurrence are the conus medullaris and the filum terminale, where a particular subtype, namely myxopapillary ependymoma

Fig. 4.2.12  39-year-old female with epithelioid meningioma metastatic to the lungs and liver. There has been previous lumbar surgery. Sagittal a T1, b T2, and c post-contrast T1-weighted images of the lumbosacral spine show a large, sausage-shaped IDEM mass, isointense to the neural element with some peripheral and central T2 hypo­ intensity probably reflecting calcification. The mass enhances avidly. Note also intense enhancement of the cauda equina nerve roots and pial surface of the distal spinal cord (c). Repeated CSF sampling showed a reactive cytological profile with increased cellularity of mixed phenotype

Fig. 4.2.13  9-year-old female complaining of lower back and lower extremity pain.   a T1-weighted, b T2-weighted, and c contrast-enhanced T1weighted sagittal images of the lumbar spine show a sausageshaped, well circumscribed, and partially cystic mass in the spinal canal associated with the cauda equina. The solid portion of the mass shows intense contrast enhancement. This was surgically resected and proven to be a schwannoma

4.2  Intradural Extramedullary Spine

(30% of all ependymomas), is found almost exclusively (Lee 2000; Van Goethem et al. 2004; Wippold II et al. 1995). In fact the most common tumor in the conus and filum is myxopapillary ependymoma, which accounts for more than half of all tumors in these locations and has a distinct MRI appearance (Lee 2000). This is a slowgrowing, low-grade tumor that can cause low back pain and progressive leg pain and weakness, gait disturbances, and bowel and urinary incontinence (Van Goethem et al. 2004). It is well delineated, typically iso- to hypointense on T1-weighted imaging and hyperintense on T2weighted imaging, with intense contrast enhancement. It also has the propensity of cystic formation and hemorrhage (Fig. 4.2.15), sometimes with associated subarachnoid hemorrhage and superficial siderosis (Choi et al. 2002; Van Goethem et al. 2004). When large, myxopapillary ependymoma often expands the spinal canal, causing pressure erosion with scalloping of the vertebrae and neural foramina (Wippold II et al. 1995) (Fig. 4.2.16).

Fig. 4.2.14  30-year-old male with von Hippel-Lindau disease. a Post-contrast sagittal T1-weighted image shows multiple enhancing hemangioblastomas in the cerebellum (not shown) and within the cord. b Axial T2-weighted image shows a peripherally located intramedullary cystic hemangioblastoma expanding the cord. Multiple cystic masses are present in the kidneys. c Magnified view of the conus shows a cystic hemangioblastoma with a central enhancing solid nodule, and another small solidly enhancing one in the proximal filum

Paraganglioma Spinal paraganglioma is a rare neuroendocrine tumor that is almost exclusively found in the cauda equina or filum terminale. It arises from neural crest-derived paraganglionic cells. Spinal paraganglioma is predominantly of the sympathetic type, in contrast to the parasympathetic origins of more commonly encountered extra-adrenal paragangliomas in the head and neck, including carotid body tumor and glomus jugulare (Sundgren et al. 1999). The most common presenting symptoms are low back pain and sciatica, but other symptoms may include sensory or motor deficits in the lower extremity, or bladder, bowel and erectile dysfunction (Sundgren et al. 1999; Yang et al. 2005). Paraganglioma is a hypervascular tumor often associated with vascular pedicles, and appears well circumscribed, ovoid or sausage shaped, encapsulated, and soft, adherent to the filum or nerve roots. Microscopically, it is composed of uniform ganglion cells with mature and transitional neural differentiation arranged in clusters, in a characteristic organoid or Zellballen architecture. These cells are circumscribed by a vascular connective tissue stroma. While the tumor is of neuroendocrine origin, it is rarely endocrine active (Yang et al. 2005). Nevertheless, a few reports have documented flush-like attacks (probably related to elevated 5hydroxytryptamine levels in these tumors on biochemical immunohistochemical analysis) and hyperadrenergic symptoms (Sonneland et al. 1986). On MR imaging, paraganglioma is a well delineated and sharply defined mass demonstrating T1 hypointensity and T2 hyperintensity, and intense contrast enhancement (Fig. 4.2.17). While these features do not allow easy distinction from other common neoplasms such as ependymomas and schwannomas occurring in the same location, its hypervascularity with occasionally associated serpiginous flow voids, reflecting dilated vessels, can be an important clue to the diagnosis (Abe et al. 1999). More frequently, heterogeneous T2 signal can be found, particularly with hypointensity along the tumor margins, indicating hemosiderin deposits from previous bleeds (Araki et al. 1993; Yang et al. 2005). Metastasis A number of pathways for leptomeningeal metastases have been postulated, including direct spread through the CSF, hematogenous spread via the choroid plexus or leptomeningeal vessels, venous spread through the Batson’s plexus, or retrograde spread via the perineural lymphatics. The primary CNS neoplasms, including medulloblastoma (Fig. 4.2.18), ependymoma (Fig. 4.2.19), germ cell tumors, and high-grade gliomas have a greater propensity of dissemination via CSF pathways. Other systemic neoplasms such as lung (Fig. 4.2.20), breast, gastric carcinomas, and melanoma can also involve the CSF spaces, many probably via the hematogenous route (Table 4.2.3). Contrast-enhanced MR images are most sensitive in depicting involved areas along the surface of the spinal cord

599

600

4  Spine and Spinal Canal Fig. 4.2.15  Myxopapillary ependymoma in a 42-year-old male complaining of low back pain, left leg numbness, and weakness. a T1-weighted, b T2-weighted, and c contrast-enhanced T1-weighted sagittal images of the lumbar spine show an ovoid, avidly enhancing IDEM mass associated with the cauda equina, with central and peripheral areas of T2 hypointensity consistent with hemorrhage. Surgical pathology confirmed the diagnosis with evidence of dense fibrous tissue and hemosiderin

Fig. 4.2.16  Myxopapillary ependymoma in a 50-year-old female with dull mid-back pain that is worse with sitting, left foot numbness, and right knee pain.   a T1-weighted, b T2-weighted, and   c contrast-enhanced T1-weighted sagittal images of the lumbar spine show a large lobular IDEM mass filling the distal spinal canal and causing erosion of the posterior vertebral bodies of the sacrum

Fig. 4.2.17  Cauda equina paraganglioma. a Post-contrast sagittal and b axial T1-weighted images show an ovoid, well defined, and avidly enhancing mass associated with the nerve roots. (Courtesy of Dr. Bruce Wasserman)

4.2  Intradural Extramedullary Spine

and nerve roots. The enhancement may be smooth or nodular (examples in Figs. 4.2.18, 4.2.19, 4.2.20; see also Tables 4.2.4, 4.2.5 for various etiologies of nerve root enhancement and enlargement). The lumbar region along the caudal aspect of the thecal sac is the most common site of involvement, probably related to effects of gravity, with signs of nerve root enlargement, clumping, nodularity, and sometimes loss of CSF clarity and poor definition of the conus (Moulopolos et al. 1997). While CSF cytol-

ogy is often considered to be the gold standard in the evaluation of leptomeningeal disease, contrast-enhanced MRI has been found to depict disease involvement in 50% of cases where CSF sampling was negative (Gomori et al. 1998). In the evaluation of disseminated medulloblastoma, contrast-enhanced MRI had greater sensitivity and specificity compared to contemporaneous CSF sampling, although the latter was improved by analysis of multiple samples (Meyers et al. 2000).

Fig. 4.2.18  18-year-old female with medulloblastoma recurrence presenting with leptomeningeal dissemination in the spine. Post-contrast sagittal T1-weighted image of the cervical and upper thoracic spine shows a focal enhancing mass dorsal to the T2 cord (in addition to a number of nodules in the remaining spine). Note also deformity of the cerebellum related to previous medulloblastoma resection

Table 4.2.3  Common spinal neoplasms in the intradural compartment Pediatric age group

Adult age group

Ependymoma Primitive neuroectodermal tumor, retinoblastoma Pineal region tumor   (pineoblastoma, germinoma) Glioblastoma Leukemia Polyradiculoneuropathy

Lymphoma Melanoma Metastases from lung, breast, gastric cancer

Table 4.2.4  Differential diagnoses for nerve root enhancement Guillain-Barré syndrome Chronic inflammatory demyelinating polyradiculoneuropathy Infectious polyradiculopathy Neoplastic infiltration Radiation or chemotherapy induced polyradiculopathy Acute and chronic lumbar disk disease

Fig. 4.2.19  Recurrent ependymoma with CSF dissemination. Post-contrast images at a sagittal thoracic, b sagittal lumbar, c axial thoracic, and d axial lumbar levels show nodular enhancement along the surface of the cord and nerve roots, particularly dorsally in the thoracic region (arrows)

601

602

4  Spine and Spinal Canal Table 4.2.5  Differential diagnoses for nerve root enlargement Neurofibromatosis Chronic inflammatory demyelinating polyradiculoneuropathy Hereditary sensory and motor neuropathy Lymphoma (neurolymphomatosis) Amyloidosis Paraneoplastic syndrome Vascular malformation Arachnoiditis Disk disease Tethered cord

Fig. 4.2.20  46-year-old male patient with small cell lung cancer presenting with lower back pain. a Sagittal and b axial post-contrast T1-weighted images of the lumbar spine shows diffusely thickened, smooth leptomeningeal enhancement along the surface of the distal spinal cord and cauda equina nerve roots due to metastases (arrows). CSF confirmed small cell carcinoma

4.2.4.1.2  Hematological Malignancies Hematological malignancies can be divided into several groups, including lymphoma, leukemia, and plasma cell dyscrasia such as multiple myeloma. These can involve the spine in various forms by CSF dissemination and direct neoplastic infiltration of leptomeninges, perineural vasculature, or nerve roots. Paraprotein-related inflammatory demyelination, and paraneoplastic syndrome are additional manifestations with similar neuroimaging

findings and are further discussed under the section on inflammation. Lymphoma is grouped into Hodgkin’s lymphoma (HL) and non-Hodgkin’s lymphoma (NHL). HL is characterized by Reed-Sternberg cells, and involves the spine only in advanced disease, primarily by extending through the bone or dura to the epidural space. NHL, on the other hand, tends to have extranodal involvement, and more commonly extends to the CNS, mostly in the form of leptomeningeal disease. On contrast-enhanced MRI, the appearance is the same as that described for leptomeningeal carcinomatosis secondary to other systemic malignancies. In the setting of primary CNS lymphoma or systemic NHL, there is also a distinct yet uncommon manifestation of nerve parenchymal infiltration, called neurolymphomatosis. In this condition, patients present with nodular nerve root and plexus enlargement with or without enhancement on MRI, and symptoms of radicular and peripheral neuropathy (Baehring et al. 2003). Acute leukemia, particularly acute lymphocytic leukemia (ALL), has a high propensity of CNS involvement with leukemic infiltration of the meninges or spinal roots. Current standard treatment of these patients, therefore, includes CNS prophylaxis with intrathecal chemotherapy. Chronic leukemias include chronic myeloid leukemia (CML) and chronic lymphocytic leukemia (CLL). Neurological involvement by CML is uncommon, while CLL is a lymphoproliferative disorder that frequently results in various neuromuscular disorders in affected patients. Peripheral neuropathy may reflect leukemic infiltration of nerve roots and peripheral nerves, or may be related to paraneoplastic syndrome/immune-mediated inflammatory demyelination. Multiple myeloma most commonly affects the skeletal system, with diffuse infiltration of the spine by plasma cells, which can then be complicated by vertebral collapse as well as epidural extension of disease, resulting in compression of the spinal cord or nerve roots. Much more rarely, myelomatous meningitis can occur, manifesting with cloudiness of CSF, nerve root enhancement, and (at times) enlargement (Fig. 4.2.21). The pattern of hypertrophic, enhancing nerve roots with a similar clinical manifestation of peripheral neuropathy can, however, be related to paraproteinemia, or a complication of amyloidosis in the setting of multiple myeloma. 4.2.4.1.3  Unusual Leptomeningeal Neoplasia Diffuse Leptomeningeal Gliomatosis This is a rare condition with diffuse primary glioma involvement of the leptomeninges, thought to arise from heterotopic neuroglial cell rests present in the leptomeninges. Contrast-enhanced MRI shows diffuse leptomeningeal enhancement with nodularity (Fig. 4.2.22), but the pattern is indistinguishable from carcinomatous

4.2  Intradural Extramedullary Spine

Fig. 4.2.21  Post-contrast images sagittal (a) and axial (b) at the level of the conus and axial at the level of the cauda equina (c), showing diffusely thickened and enhancing nerve roots in a 50-year-old male with disseminated multiple myeloma. Enhancement is also present along the surface of the distal spinal cord. Note a pathological compression fracture of L1 vertebra (arrow), and multiple rounded areas of T1 marrow hypointensity reflecting myelomatous osseous involvement. (Courtesy of Dr. Dima Hammoud)

Fig. 4.2.22  Diffuse leptomeningeal gliomatosis in a 16-year-old male presenting with headache, nausea, and vomiting. CSF showed markedly elevated protein. Contrast- enhanced MRI showed thickened leptomeningeal enhancement throughout the neuraxis. The final diagnosis was established by pathology from open biopsy, which revealed diffuse infiltration of the leptomeninges by cytologically atypical S100positive round and oval cells with pale to clear cytoplasm, numerous apoptotic cells with a high proliferation index, and scattered lymphocytes

leptomeningeal metastasis, lymphoma, leukemia, or infectious/inflammatory conditions such as tuberculosis or sarcoidosis.

moid cysts. Many of these lesions, however, are not confined to this space, but may occur in the extradural or intramedullary compartments as well.

Diffuse Leptomeningeal Melanomatosis Melanocytes are derived from the neuroectodermal cells, and are normally found in the pia-arachnoid layers of the CNS. These melanocytes can proliferate and give rise to conditions ranging from diffuse melanosis, well-differentiated melanocytomas, to discrete or diffuse infiltrative malignant melanomas. Primary diffuse leptomeningeal melanomatosis is a rare variant of primary malignant melanoma, which spreads throughout the leptomeninges and perivascular spaces (Pirini et al. 2003). It is seen more frequently in adults. In children, it may be associated with extensive pigmented skin lesions as an autosomal dominant disorder. On MRI, melanin-containing tumors have a unique appearance, with T1-hyperintensity before contrast administration (Fig. 4.2.23).

Arachnoid Cyst While arachnoid adhesions and cysts often form as a sequela of trauma, surgery, hemorrhage, or infection/ inflammation, idiopathic arachnoid cyst is a rare lesion and occurs because of defects in the arachnoid space. It is most frequently found in the thoracic spine (81%) posterior to the cord, with variable communication with the CSF space, and may be under pressure, causing compression of the cord. In a series of 21 patients presenting with congenital arachnoid cysts that were unrelated to previous trauma or inflammatory insult, it was found that the dorsal cysts most often cause pain (93%), while ventral cysts frequently produce myelopathy (83%) and weakness (Wang et al. 2003). When small, many of these may be asymptomatic and found incidentally. On MR imaging, the arachnoid cyst follows CSF signal intensity on all pulse sequences and is very inconspicuous when small, as the cyst wall is typically thin and difficult to detect (Fig. 4.2.24). In fact, the diagnosis of arachnoid cyst may be suspected often because of secondary signs—mass effect upon the cord, which may be displaced or compressed against the

4.2.4.2  Congenital Lesions A number of congenital or developmental lesions can be found in the IDEM compartment, including arachnoid cysts, neuroenteric cysts, lipomas, and epidermoid/der-

603

604

4  Spine and Spinal Canal

spinal canal. CT myelography can be particularly helpful in establishing the diagnosis in conjunction with MR imaging (Fig. 4.2.25). Although most arachnoid cysts will fill with contrast given time, the initial appearance

on myelography is a well circumscribed intradural extramedullary filling defect. Contrast clearance from the cyst is also often much delayed compared to the remaining CSF space.

Fig. 4.2.23  Diffuse leptomeningeal melanomatosis in an 18-year-old college student with a history of odontoid fracture from a motor vehicle accident who complained of headache, intermittent numbness in the right leg and foot, and an episode of dizziness and vomiting. Initial head CT showed nodular hyperattenuation along the cortical surface throughout the cerebral hemispheres and mild hydrocephalus. MRI of the brain and entire spine further demonstrated innumerable leptomeningeal

T1 hyperintense nodular masses and diffuse leptomeningeal enhancement throughout the neuroaxis. a T1-weighted image before contrast, b T1-weighted image after contrast of the cervical spine, and c T1-weighted image after contrast of the lumbar spine showing diffuse leptomeningeal melanomas. CSF analysis documented malignant melanomas, while clinical examination failed to demonstrate a primary cutaneous or ocular source of tumor

Fig. 4.2.24  Arachnoid cyst in a 56-yearold male complaining of neck and shoulder pain. a Sagittal T1-weighted, b sagittal T2-weighted images of the thoracic spine show an ovoid lesion (arrows) posterior to the T7 cord causing moderate anterior displacement and compression. The lesion is isointense to CSF, but a very thin membrane can be vaguely discerned along the lesion’s superior margin on T2-weighted image (upper arrow). The presence of a cystic wall, as well as the mass effect are corroborated on axial T2-weighted image shown in c

4.2  Intradural Extramedullary Spine Fig. 4.2.27  Lipomyelomeningocele in a 7-year-old complaining of decreased sensation in the feet. Sagittal T1-weighted image shows spina bifida, patulous thecal sac, tethered cord, intradural lipoma associated with the conus, and a syrinx or dilated ventriculus terminalis

Fig. 4.2.25  Same patient as in Fig. 4.2.25. CT myelography a sagittal reformation and b axial image show the arachnoid cyst as a well-circumscribed IDEM filling defect dorsal to the thoracic cord, exerting mass effect on the cord

Neuroenteric Cyst Neuroenteric cyst, or enterogenous cyst, is a rare developmental lesion that derives from the endoderm. When it occurs in the spine, it is most frequently found in the IDEM compartment, and rarely in the intramedullary location. The most common sites include the cervicothoracic junction and lower lumbar region. Often it is associated with spinal dysraphism, including skin abnormality and vertebral segmentation anomaly (Fig. 4.2.26). On histological examination the cyst is lined by mucinsecreting, at times ciliated, columnar or cuboidal epithelium similar to that found in the respiratory or intestinal tract, therefore given various names including primitive foregut, bronchogenic, endodermal, and respiratory cyst (Kleihues and Cavenee 2000; Leech and Olafson 1977).

Fig. 4.2.26  Neuroenteric cyst in a young patient presenting with cervical myelopathy. Sagittal T2-weighted image shows a lobular extramedullary cystic lesion associated with segmentation anomaly of the adjacent cervical vertebrae. (Courtesy of Dr. David Yousem)

Lipoma Intradural lipoma can be classified as a congenital lesion, or as a histologically benign neoplasm. While it can be found anywhere along the spine, it is most frequently seen in the lumbosacral region, associated with spinal dysraphism in various forms including lipomyelomeningocele (Fig. 4.2.27), intraspinal lipoma (Fig. 4.2.28), and lipoma of the filum terminale (Fig. 4.2.29). Without associated spinal dysraphism, isolated intradural lipoma is very rare, but can also arise most often in the subpial juxtamedullary location. It is slow-growing, and may cause pain and spastic para- or quadraparesis by compression of the spinal cord or nerve roots. Surgical resection is controversial, as it is associated with significant risks of morbidity. In the management of 3 cases of intradural thoracic lipomas, Klekamp et al. (2001) reported that spinal decompression was sufficient to achieve neurological improvement. Finally, fibrolipoma or fibrofatty infiltration of the filum terminale is often an incidental finding, noted, e.g., in 4% of patients in a series of randomly selected MR examinations of the lumbar spine (Brown et al. 1994). A small accumulation of fat along the filum is demonstrated as a hyperintense dot on axial T1-weighted images, and as a hyperintense tubular shape on sagittal views (Fig. 4.2.30). This is rarely symptomatic and should be considered as a normal variant. However, larger and more extensive involvement may be correlated with neurological symptoms of cord tethering (Sevick and Wallace 1999).

605

606

4  Spine and Spinal Canal Fig. 4.2.28  Intradural lipoma in the distal thoracic-proximal lumbar spine, with a component of extradural lipomatous tissue infiltrating the lumbosacral paraspinal region. a Sagittal T1-weighted image and b T2-weighted image with fat suppression show this large intraspinal lipoma as very T1 hyperintense, with the same signal as the subcutaneous fat. Note also spinal dysraphism with a patulous distal thecal sac and segmentation anomaly of the distal lumbar vertebrae. On the axial T1-weighted image, the lipoma is multilobulated and T1 hyperintense, and its signal is completely suppressed on the sequence with fat saturation (T1+FS)

Fig. 4.2.29  30-year-old male with right foot numbness, right leg pain, low back pain, and right calf atrophy. a Sagittal T1-weighted image and b T2-weighted image with fat suppression reveal a sacral intradural lipoma associated with a tethered cord. An extradural component of lipoma is also present

Epidermoid Cyst Also called epidermoid sequestration cyst, this is a rare congenital lesion of the ectodermal origin arising from cells displaced during embryogenesis, and is usually associated with other developmental dysraphic anomalies such as hemivertebrae, spina bifida, and dermal sinus (Parenti et al. 1993). It is most often seen in the conus and cauda equina. However, about 40% of epidermoid cysts are acquired lesions, often as a late consequence of

Fig. 4.2.30  Fibrofatty infiltration of the filum terminale in an asymptomatic individual. In this case the T1-hyperintense fibrolipoma is quite prominent, measuring about 2 mm in the transverse dimension in a, and spanning along several vertebral segments on the sagittal T1-weighted image b

previous surgical procedures, including lumbar puncture (Machida et al. 1993). On MR imaging, an epidermoid cyst is typically hypo- to isointense to the neural element on T1-weighted images, and hyperintense on T2-weighted images, without associated enhancement except perhaps along the rim (Fig. 4.2.31). While it is a cystic lesion, it is more T1 hyperintense than CSF, probably reflecting a variable amount of lipid content. On a FLAIR sequence, the hyperintensity of an epidermoid cyst in the brain or

4.2  Intradural Extramedullary Spine

Fig. 4.2.31  Epidermoid in an asymptomatic 50-year-old female who denied any history of lumbar puncture. Two rounded, welldefined non-enhancing masses associated with the filum, showing iso- to minimally hyperintensity on T1-weighted image relative to the neural element (a), hyperintensity on T2-weighted image and not suppressed by fat (b). There is no associated enhancement on post-contrast T1-weighted image (c). Note also a low, tethered cord. (Courtesy of Dr. Deepak Takhtani)

skull will often allow it to be distinguished from an arachnoid cyst. In addition, an epidermoid cyst often shows a restricted diffusion pattern (being hyperintense on DWI) in contrast to hypointensity of free water in the arachnoid cyst. It is expected that DWI will also prove applicable to spine imaging for distinguishing epidermoid cysts from arachnoid cysts or other congenital cystic lesions. 4.2.4.3  Inflammatory Conditions This broad category includes both infectious and noninfectious inflammatory etiologies. Infection A number of infectious agents can affect the IDEM compartment predominantly involving the meninges, particularly leptomeninges. These include viral, bacterial, fungal, and tuberculous infections. The most common MR findings are diffuse enhancement along the surface of the cord and along the nerve roots, at times accompanied by thickening of the involved nerve roots. Radiculitis is often the result of viral infection (Fig. 4.2.32). Tuberculous infection most commonly manifests in the extradural compartment resulting in spondylodiskitis, and less frequently radiculitis, while extremely rarely involving the intramedullary space. Diffuse nerve root enhancement, however, is a nonspecific pattern and can be

Fig. 4.2.32  Cytomegalovirus (CMV) radiculitis. Sagittal T1weighted images of the lumbar spine before (a) and after (b) contrast administration show mild uniform thickening and enhancement of the nerve roots in a patient with AIDS

seen in a myriad of conditions other than infection (see Tables 4.2.3, 4.2.4). Noninfectious inflammatory conditions may have varied manifestation, including arachnoiditis, meningeal enhancement or mass, and nerve root thickening with or without enhancement. Sarcoidosis Sarcoidosis is an idiopathic multisystem granulomatous disorder which affects the central nervous system in about 5% of cases (Delaney 1979). Spinal involvement represents a small fraction (less than 10%) of neurosarcoid cases, typically with concurrent cranial disease. Spinal sarcoidosis is seen more frequently in young adults in their 20s and 30s, without particular sex predilection (Hamasaki et al. 2003). All spinal levels may be affected, and spinal involvement can occur in any compartment: intramedullary, intradural extramedullary, epidural spaces, and vertebral bodies (Hamasaki et al. 2003). However, IDEM sarcoidosis is much less common than intramedullary disease (Connor et al. 2001; Delaney 1979; Hamasaki et al. 2003). Cauda equina involvement is even rarer, presenting with smooth or nodular enhancement of the nerve roots (Fig. 4.2.33a,b) and may be associated with polyradiculopathy and back pain (Zajicek 2000). In the involved sites, there is accumulation of non-caseating granulomas, and these granulomas are found in a perivascular distribution, together with lymphocyte infiltra-

607

608

4  Spine and Spinal Canal

Inflammatory Demyelinating Poly (Radiculo) Neuro­pathy There is a spectrum of acquired immune-mediated inflammatory disorders of the nerve roots and peripheral nervous system that produce weakness and sensory dysfunction in the extremities. The acute variant of inflammatory demyelinating polyneuropathy, also known as Guillain-Barré syndrome, is often related to antecedent infection or immunization, and has a self-limited clinical course over a period less than 1 month (Koller et al.

2005; Pytel et al. 2003). MRI may show diffuse nerve root enhancement (Fig. 4.2.34). The chronic variant, chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), is characterized by symmetric weakness of distal and proximal muscles, concurrent sensory deficits, elevated CSF protein without pleocytosis, a demyelinating pattern on nerve conduction tests, and signs of demyelination, remyelination, and inflammation in nerve root biopsy specimens (Koller et al. 2005). In contrast to Guillain-Barré, CIDP has a chronic relapsing remitting or progressive course of motor and sensory dysfunction persisting for more than 2 months. CIDP may be associated with a number of concurrent disease processes including HIV infection, Hepatitis C, Sjögren’s disease, diabetes, melanoma, lymphoma, and monoclonal gammopathies (Koller et al. 2005). On MRI, it is most frequently seen as proximal nerve root enlargement and enhancement in the cauda equina (Fig. 4.2.35) or brachial plexus (Fig. 4.2.36). In about 50% of cases, the involved brachial plexus show irregular swelling and T2 hyperintensity, and may mimic NF1, or rarely, present with cord compression (Koller et al. 2005; Pytel et al. 2003). The MRI findings of hypertrophic nerve roots and plexus, and histopathological appearance of onion-bulb proliferation of Schwann cell processes around the axons with inflammatory changes are similarly found in the hereditary motor and sensory neuropathy (HMSN) subtypes, including Charcot-Marie-Tooth type I and Dejerine-Sottas syndrome (HMSN type III) (Fig. 4.2.37). HMSN, however, represents a group of genetically determined peripheral

Fig. 4.2.33  Two cases of spinal sarcoidosis. a Pre- and b postcontrast sagittal T1-weighted images of the lumbar spine show enhancing nerve roots in the cauda equina, due to leptomeningeal sarcoidosis, a rare manifestation. c Post-contrast sagittal T1-weighted image of the thoracic spine in another patient with neurosarcoid shows markedly thickened meningeal enhancement with nodular extension into the cord parenchyma

Fig. 4.2.34  Guillain-Barré syndrome. Post-contrast T1-weighted a sagittal and b axial images show thickening and enhancement of the nerve roots. (Courtesy of Dr. Deepak Takhtani)

tion of the arachnoid mater. Granulomatous involvement of the adventitia of meningeal vessels may extend along the perivascular space into the parenchyma, accounting for the characteristic appearance of thickened meninges, enhancing meningeal mass, sometimes with cord parenchymal extension (Fig. 4.2.33c). Junger et al. (1993) postulated four stages of sarcoid involvement. During the first phase, there is linear leptomeningeal enhancement along the surface of the cord, and parenchymal involvement represents a secondary phenomenon related to inflammatory extension. During the second phase, there is intramedullary, centripetal spreading of this inflammatory process. Consolidation occurs in the third phase. Finally, in the chronic stage, cord atrophy results from ischemia and disruption of neural pathways. MRI offers a reasonable sensitivity (up to 82%) but rather poor specificity for the diagnosis (Zajicek 2000). Spinal sarcoidosis may mimic meningioma, lymphoma, metastatic disease, infectious or other inflammatory processes.

4.2  Intradural Extramedullary Spine Fig. 4.2.35  52-year-old female with demyelinating inflammatory neuropathy of the cauda equina in the setting of immunoglobulin G (IgG) monoclonal gammopathy. a Sagittal T2-weighted image of the lumbar spine shows enlarged cauda equina nerve roots filling the dural sac. b Post-contrast T1-weighted image shows no significant enhancement. c Serial axial T2-weighted images show hypertrophic intrathecal nerve roots with near effacement of the CSF. Note also the multilocular, T2 hyperintense, and cystic appearance to the enlarged sacral plexus (arrows)

Fig. 4.2.36  Chronic inflammatory demyelinating polyneuropathy in a 52-year-old male presenting with progressive cervical myelopathy. a Coronal T2-weighted image shows diffuse fusiform enlargement and hyperintensity of the cervical nerve roots and plexus. b Axial T2-weighted image shows severe cord compression by the enlarged nerve roots. T1-weighted images c pre- and d post-contrast show moderate enhancement of the nerve roots

Fig. 4.2.37  Dejerine-Sottas disease in an 18-year-old male. There has been previous lumbar laminectomy decompression. a Coronal and b sagittal T2-weighted images show that the lumbar nerve roots are diffusely enlarged, filling the distal spinal canal. c,d Post-contrast axial T1-weighted images show moderate enhancement of these enlarged nerve roots

609

610

4  Spine and Spinal Canal

neuropathies with symmetric, predominantly distal motor and sensory deficits, also with gradual progression. While nerve root enhancement is not invariably found, the presence of enhancement suggests a breakdown of the blood–nerve barrier related to active demyelination and inflammation. Arachnoiditis Arachnoiditis may be idiopathic, or secondary to inflammation caused by surgery, trauma, hemorrhage, or meningitis. In arachnoiditis the intrathecal nerve roots

are distorted. A number of signs have been described: empty thecal sac due to nerve roots plastered against the periphery of the dura, and mass-like clumping of the nerve roots (Fig. 4.2.38). Aseptic arachnoiditis can also occur as a rare, late complication of ankylosing spondylitis, resulting in cauda equina syndrome often consisting of cutaneous sensory deficits in the lower extremities and perineum, and disturbance of sphincter function (Charlesworth et al. 1996). MRI shows striking dural ectasia and arachnoid diverticula, with erosion of the posterior vertebral bodies and posterior elements, as well as nerve root adhesion (Fig. 4.2.39). The underlying pathogenesis is unknown, but has been postulated to relate to demyelination, post-irradiation ischemia, or compression (Mitchell et al. 1990). 4.2.4.4  Trauma

Fig. 4.2.38  Arachnoiditis secondary to previous surgery. Laminectomy has been performed in the lumbar spine. a The intrathecal nerve roots show marked distortion and are clumped together. b Clumping of the nerve roots appears mass-like. c Empty-thecal-sac appearance due to peripherally adherent nerve roots

Trauma to the dura can be caused by penetrating injury, nerve root/sheath avulsion, and severe spinal dislocation (Zee 1996). Penetrating trauma as a result of lumbar spine surgery, needle puncture, stab, or gunshot wound may introduce air into the subarachnoid space (Figs. 4.2.40, 4.2.41). Intradural gas is dark on all pulse sequences and may produce magnetic susceptibility artifacts (Fig. 4.2.40). Hemorrhage in the subdural space is rare, but may result from blunt or penetrating trauma, or in the setting of anticoagulation or bleeding diathesis. Nerve root avulsion is often a complication of traumatic stretch injury during birth, causing brachial plexopathy and, rarely, may result from high-velocity motor vehicle accidents affecting the lumbosacral plexus. CT myelography has been the standard presurgical evaluation of cervical nerve root avulsion for depicting the completely or partially avulsed nerve root, dilated and empty nerve root sleeve, and formation of pseudomeningocele (Fig. 4.2.42). MRI is complementary, and high-resolution T2-weighted imaging such as SSFP sequences and MR myelography Fig. 4.2.39  Aseptic arachnoiditis due to complication of ankylosing spondylitis in a 54-year-old male presenting with cauda equina syndrome. a Sagittal T1- and b T2-weighted images show a patulous distal thecal sac with multiple arachnoid diverticulae. Ankylosis of the lumbar vertebrae (bamboo spine) is evident. c Axial image shows partial sclerosis of the sacroiliac joints. d,e Axial T2-weighted images demonstrating smooth asymmetric erosion of the lamina and enlargement of the thecal sac, with clumping of the thickened nerve roots and empty thecal sac indicating arachnoiditis

4.2  Intradural Extramedullary Spine

Fig. 4.2.40  27-year-old male who suffered from a stab wound to the back. a T2- and b T1-weighted sagittal images show a dot of intradural gas (white arrow) causing a magnetic susceptibility artifact. A penetrating wound entry site (black arrowheads) can be seen in the upper back extending from the paraspinal region through a disrupted dura into the thecal sac

are particularly helpful in aiding diagnosis, and demonstrating the extent of pseudomeningoceles (Carvalho et al. 1997; Hans et al. 2004). Formation of pseudomeningoceles is also a frequent sequela of spinal surgery such as laminectomy or disk surgery. These pseudomeningoceles typically appear as a bulge of the posterior thecal sac at the site of surgery, and, when large, as a lobular fluid collection extending outward to the paraspinal region. At times, a distinct dural defect, indicating the site of direct communication between the subarachnoid space and the pseudomeningocele, can be identified on MRI. Post-traumatic or post-surgical arachnoiditis has been described in the previous section. Arachnoid adhesions, arachnoid cysts, and epidermoid cysts can also be acquired following trauma or surgical procedures. Their appearance is similar to their congenital counterparts, with arachnoid cysts and adhesions (Fig. 4.2.43) demonstrating CSF signal intensity on all pulse sequences on MR, and variable conspicuity of membranous septations. Depending on the size of open communication with the subarachnoid space, there will be variable rate of contrast filling on myelography. Focal adhesion and tethering of the spinal cord at a site of dural rent (most often ventrally in the thoracic region) causes spinal cord herniation, which has a distinct appearance, characterized by C- or S-shaped distortion and kinking of the cord, and sometimes cord atrophy, that can be effectively imaged by CT myelography (Fig. 4.2.44) or MRI. This appearance may mimic a dorsal arachnoid cyst that compresses the cord. Finally, an example of an acquired epidermoid cyst following lumbar puncture is shown in Fig. 4.2.45.

Fig. 4.2.41  Moderate-sized intradural air-fluid level in a patient who has recently undergone lumbar surgery. Post-operative soft tissue distortion is also noted in the paraspinal region and intradural nerve roots

Fig. 4.2.42  Brachial plexus injury in a 4-month-old infant. Consecutive axial CT myelography images show absence of the ventral C6 root, associated with a dilated nerve root sleeve on the left side (arrow). This finding is corroborated on the coronal reformatted image, which also shows a proximal stump of avulsed nerve root (arrow)

4.2.4.5  Vascular Malformations Spinal vascular malformations are complex and heterogeneous lesions which may occur in the IDEM compartment; they include several types of malformations with

611

612

4  Spine and Spinal Canal

arteriovenous connections and, very rarely, cavernous malformation. Anson and Spetzler’s (Anson and Spetzler 1992) classification of spinal vascular malformations into 4 categories is probably the most frequently used scheme in considering arteriovenous lesions (Table 4.2.6). More recently, Spetzler et al. (2002) proposed a modified classification based on the location, pathophysiology, neuroanatomy, and imaging features, that divided the arteriovenous lesions into arteriovenous malformations (AVM) and arteriovenous fistulas (AVF). Each category was then further divided according to the spinal compartment involved. Spinal angiography is the standard diagnostic modality for all these lesions, and is capable of providing dynamic assessment of the involved vessels, site, flow characteristics, and fistulous connection. MRI can be helpful as the noninvasive initial evaluation, allowing depiction of abnormal vascular flow voids along the surface of the cord (best seen on T2-weighted imaging) and enhancement (on post-contrast T1-weighted imaging) due to dilated intradural vessels; however, the exact fistulous site cannot be delineated (Fig. 4.2.46). Nevertheless, MRI offers the advantage of evaluating the spinal cord for evidence of edema or hemorrhage, and by adding contrast-enhanced spinal MR angiography (MRA) to the standard MRI examination of dural AVF, Saraf-Lavi et al. found significant improvement in the localization of the fistula as well (Saraf-Lavi et al. 2002).

Fig. 4.2.43  Syrinx and extensive arachnoid adhesions in a 13-year-old boy post-operative for spine tumor. a T1- and b T2weighted sagittal images demonstrate marked distortion of the spinal canal and intradural contents with multiple septations, adhesive strands and tethering and thinning of the cord. There is also remodeling of the vertebral bodies

Table 4.2.6  Anson and Spetzler’s classification of spinal vascular malformations Type

Age affected

I

Dural AVF

50–60 years

II

Intramedullary, glomus AVM

20–30 years

III

Intramedullary and extramedullary, juvenile AVM

Children, adolescents

IV

Perimedullary (intradural extramedullary) AVF

The most common arteriovenous lesion is a type I malformation, dural AVF (or intradural dorsal AVF), which is considered an acquired lesion with anomalous connection between the radicular artery and medullary vein. Dural AVF occurs most commonly in the thoracolumbar spine in a slow, insidious manner, affecting middle aged individuals at a mean age of 50 years, with male predominance, and causing progressive myelopathy or radiculopathy. Arteriovenous shunting results in venous congestion and venous hypertension, and further induces hypoperfusion of the spinal cord. At times, T2 hyperintense cord signal abnormality can be identified

Fig. 4.2.44  Cord herniation shown on CT myelography. a Sagit­ tal reformatted image shows a characteristic C-shaped kinking and tethering of the cord to the ventral surface of the dura, accompanied by widening of the dorsal CSF spaces. b Axial image also shows evidence of focal cord atrophy

4.2  Intradural Extramedullary Spine Fig. 4.2.45  Acquired epidermoid cyst in a patient who has previously undergone lumbar puncture for myelographic examination of degenerative disk disease. a Sagittal T2- and b post-contrast T1-weighted images show a round, non-enhancing cystic mass (arrows) in the proximal cauda equina at the level of L2–L3, a typical entry site for lumbar puncture. Note that the mass shows T1 signal that is slightly hyperintense to CSF and nearly isointense to the neural element. c Axial T2-weighted image shows displacement of the intrathecal nerve roots

on MRI, reflecting edema or ischemia, corresponding to the Foix-Alajouanine syndrome. The other arteriovenous lesions (types II–IV) are considered congenital lesions and discovered at a younger age, sometimes presenting with spinal subarachnoid

hemorrhage. Aside from type II being the intramedullary AVM, most lesions have part or all components of the AVM/AVF in the IDEM compartment. An example of a type III perimedullary AVF is shown in Fig. 4.2.47, with characteristic vascular flow voids, enhancement, as well as a dilated filum vein. 4.2.5  Indications and Value of MRI

Fig. 4.2.46  12-year-old female with subarachnoid hemorrhage due to spinal peri-medullary AVM. Sagittal and axial T2-weighted image show serpiginous flow voids surrounding the distal cord (arrows). Spinal angiogram confirmed a perimedullary AVM with a 2-cm nidus at the T12 vertebral level

MRI is the imaging modality of choice in evaluating IDEM spine lesions. It has superior sensitivity for anatomic depiction and tissue characterization of lesions compared to CT and myelography, allowing definitive diagnosis of a myriad of pathologic processes. For some lesions, it is complementary to a CT myelographic examination. For example, CT offers a better osseous detail in a congenital lesion (such as a neuroenteric cyst) that may occur with concomitant vertebral anomalies, and in association with a neoplasm that produces erosion or hyperostosis. Associated calcification in a spinal meningioma is best revealed by CT. Furthermore, intradural arachnoid cysts are optimally evaluated by a combination of MRI and CT myelography since, on MR imaging alone, a small arachnoid cyst can be rather inconspicuous because of its isointensity to CSF on all pulse sequences. The nature of an arachnoid cyst—whether it has a wide communication with the adjacent CSF, or forms a more closed system presenting as a space-occupying lesion— and the distinction of it from spinal cord herniation, can be much better characterized by the dynamic distribution of intrathecally administered myelographic contrast. While all the lesions presenting as space occupying masses can be depicted by CT myelography, the nature of the lesion is better characterized by MR with contrast administration. On T2 weighted images, nerve sheath tumors often show increased signal intensity and may be

613

614

4  Spine and Spinal Canal Fig. 4.2.47  Type III perimedullary AVF. a T2-weighted image shows multiple punctate and serpiginous flow voids anterior and posterior to the surface of the cord (arrows) as well as the cauda equina nerve roots. Some of these foci are slightly hyperintense on T1-weighted imaging before contrast administration b, likely reflecting slower vascular flow. c Post-contrast T1-weighted image shows more extensive enhancement along the distal cord and nerve roots. The nerve roots are thickened, and there is a venous varix of the filum terminale (block arrow).   (Courtesy of Dr. Deepak Takhtani)

obscured by the surrounding CSF. Meningiomas tend to be isointense to the cord. Since all these neoplasms enhance, contrast-enhanced MRI significantly increases the conspicuity of these lesions and is superior to CT myelography (Sze 1993). When the tumors are small, or for leptomeningeal disease, contrast material becomes essential for adequate depiction. The difficulty of detecting leptomeningeal tumors can be attributed to several factors. Leptomeningeal metastases may be very small and simply spread along the surface of the cord and nerve roots without forming a discrete nodularity; partial volume averaging effect with adjacent CSF can obscure a small lesion; and CSF pulsation and motion artifacts further confound the problem. Contrast-enhanced MRI of the entire spine is particularly useful for screening of leptomeningeal dissemination of neoplastic disease, and can be performed with selected whole-spine post-contrast sequences to accompany brain MRI. In the pediatric age group, discovery of intracranial ependymoma, primitive neuroectodermal tumor (PNET, including medulloblastoma of the posterior fossa and retinoblastoma), pineal region tumors including pineoblastoma and germinoma, or high-grade glioma should prompt screening of the entire neuraxis. In the adult population, lymphoma, melanoma, metastasis from lung, breast, and gastric cancer have the propensity of leptomeningeal spread. Since CSF cytology can be falsely negative, contrast-enhanced MRI is an indispensable complementary tool in evaluating high-risk patients. Acknowledgement The author thanks Dr. Peter Barker for a critical review of the chapter.

References 1.

Abe H, Maeda M, Koshimoto Y, Baba H, Noriki S, Takeuchi H, Kubota T and Ishii Y (1999) Paraganglioma of the cauda equina: MR findings. Radiat Med 17:235–237 2. Anson JA, Spetzler RF (1992) Interventional neuroradiology for spinal pathology. Clin Neurosurg 39: 388–417 3. Araki Y, Ishida T, Ootani M, Yamamoto H, Yamamoto T, Tsukaguchi I, Nakamura H (1993) MRI of paraganglioma of the cauda equina. Neuroradiology 35:232–233 4. Baehring JM, Damek D, Martin EC, Betensky RA, Hochberg FH (2003) Neurolymphomatosis. J Neurooncol 5:104–115 5. Bowen BC, Saraf-Lavi E, Pattany P M (2003) MR angiography of the spine: update. Magn Reson Imaging Clin N Am 11:559–584 6. Brown E, Matthes JC, Bazan C III, Jinkins J R (1994) Prevalence of incidental intraspinal lipoma of the lumbosacral spine as determined by MRI. Spine 19:833–836 7. Carvalho GA, Nikkhah G, Matthies C, Penkert G, Samii M (1997) Diagnosis of root avulsions in traumatic brachial plexus injuries: value of computerized tomography myelography and magnetic resonance imaging. J Neurosurg 86:69–76 8. Charlesworth CH, Savy LE, Stevens J, Twomey B, Mitchell R (1996) MRI demonstration of arachnoiditis in cauda equina syndrome of ankylosing spondylitis. Neuroradiology 38:462–465 9. Choi BH, Kim RC, Suzuki M, Choe W (1992) The ventriculus terminalis and filum terminale of the human spinal cord. Hum Pathol 23:916–920 10. Choi JY, Chang KH, Yu IK, Kim KH, Kwon BJ, Han MH, Kim IO (2002) Intracranial and spinal ependymomas: review of MR images in 61 patients. Korean J Radiol 3:219–228

4.2  Intradural Extramedullary Spine 11. Connor SE, Marshman L, Al-Sarraj S and Ng V (2001) MRI of a spinal intradural extramedullary sarcoid mass. Neuroradiology 43:1079–1083 12. Delaney P (1979) Spinal sarcoidosis. South Med J 72:770 13. el-Mahdy W, Kane PJ, Powell MP, Crockard HA (1999) Spinal intradural tumours: Part I—Extramedullary. Br J Neurosurg 13:550–557 14. Evans DG, Baser ME, McGaughran J, Sharif S, Howard E, Moran A (2002) Malignant peripheral nerve sheath tumours in neurofibromatosis 1. J Med Genet 39:311–314 15. Friedman DP, Tartaglino LM, Flanders AE (1992) Intradural schwannomas of the spine: MR findings with emphasis on contrast-enhancement characteristics. AJR Am J Roentgenol 158:1347–1350 16. Gezen F, Kahraman S, Canakci Z, Beduk A (2000) Review of 36 cases of spinal cord meningioma. Spine 25:727–731 17. Gomori JM, Heching N, Siegal T (1998) Leptomeningeal metastases: evaluation by gadolinium enhanced spinal magnetic resonance imaging. J Neurooncol 36:55–60 18. Halliday AL, Sobel RA, Martuza RL (1991) Benign spinal nerve sheath tumors: their occurrence sporadically and in neurofibromatosis types 1 and 2. J Neurosurg 74:248–253 19. Hamasaki T, Noda M, Kamei N, Yamamoto S, Ochi M, Yasunaga Y (2003) Intradural extramedullary mass formation in spinal cord sarcoidosis: case report and literature review. Spine 28:E420–E423 20. Hans FJ, Reinges MH, Krings T (2004) Lumbar nerve root avulsion following trauma: balanced fast field-echo MRI. Neuroradiology 46:144–147 21. Junger SS, Stern BJ, Levine SR, Sipos E, Marti-Masso J F (1993) Intramedullary spinal sarcoidosis: clinical and magnetic resonance imaging characteristics. Neurology 43:333–337 22. Kleihues P, Cavenee WK (2000) World Health Organization classification of tumours. In: Kleihues P and Sobin LH (eds) Pathology and genetics of tumours of the nervous system. IARC Press, Lyon, p 314 23. Klekamp J, Fusco M, Samii M (2001) Thoracic intradural extramedullary lipomas. Report of three cases and review of the literature. Acta Neurochir (Wien) 143:767–773; discussion 773–764 24. Koller H, Kieseier BC, Jander S, Hartung H P (2005) Chronic inflammatory demyelinating polyneuropathy. N Engl J Med 352:1343–1356 25. Lee RR (2000) MR imaging of intradural tumors of the cervical spine. Magn Reson Imaging Clin N Am 8:529–540 26. Leech RW, Olafson RA (1977) Epithelial cysts of the neuraxis: presentation of three cases and a review of the origins and classification. Arch Pathol Lab Med 101:196–202 27. Machida T, Abe O, Sasaki Y, Shirouzu I, Aoki S, Hoshino Y, Seichi S, Maehara T (1993) Acquired epidermoid tumour in the thoracic spinal canal. Neuroradiology 35:316–318 28. Mautner VF, Tatagiba M, Lindenau M, Funsterer C, Pulst SM, Baser ME, Kluwe L, Zanella FE (1995) Spinal tumors in patients with neurofibromatosis type 2: MR imaging study of frequency, multiplicity, and variety. AJR Am J Roentgenol 165:951–955

29. Meyers SP, Wildenhain SL, Chang JK, Bourekas EC, Beattie PF, Korones DN, Davis D, Pollack IF, Zimmerman RA (2000) Postoperative evaluation for disseminated medulloblastoma involving the spine: contrast-enhanced MR findings, CSF cytologic analysis, timing of disease occurrence, and patient outcomes. AJNR Am J Neuroradiol 21:1757–1765 30. Mitchell MJ, Sartoris DJ, Moody D, Resnick D (1990) Cauda equina syndrome complicating ankylosing spondylitis. Radiology 175:521–525 31. Moulopoulos LA, Kumar AJ Leeds NE (1997) A second look at unenhanced spinal magnetic resonance imaging of malignant leptomeningeal disease. Clin Imaging 21:252–259 32. Parenti G, Fiori L, Marconi F, Tusini G (1993) Primary cauda equina tumors. J Neurosurg Sci 37:149–156 33. Patronas NJ, Courcoutsakis N, Bromley CM, Katzman GL, MacCollin M, Parry D M (2001) Intramedullary and spinal canal tumors in patients with neurofibromatosis 2: MR imaging findings and correlation with genotype. Radiology 218:434–442 34. Pirini MG, Mascalchi M, Salvi F, Tassinari CA, Zanella L, Bacchini P, Bertoni F, D’Errico A, Corti B, Grigioni W F (2003) Primary diffuse meningeal melanomatosis: radiologic-pathologic correlation. AJNR Am J Neuroradiol 24:115–118 35. Pytel P, Rezania K, Soliven B, Frank J, Wollmann R (2003) Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) with hypertrophic spinal radiculopathy mimicking neurofibromatosis. Acta Neuropathol (Berl) 105:185–188 36. Roux FX, Nataf F, Pinaudeau M, Borne G, Devaux B, Meder J F (1996) Intraspinal meningiomas: review of 54 cases with discussion of poor prognosis factors and modern therapeutic management. Surg Neurol 46:458–463; discussion 463–454 37. Saraf-Lavi E, Bowen BC, Quencer RM, Sklar EM, Holz A, Falcone S, Latchaw RE, Duncan R, Wakhloo A (2002) Detection of spinal dural arteriovenous fistulae with MR imaging and contrast-enhanced MR angiography: sensitivity, specificity, and prediction of vertebral level. AJNR Am J Neuroradiol 23:858–867 38. Sevick RJ, Wallace CJ (1999) MR imaging of neoplasms of the lumbar spine. Magn Reson Imaging Clin N Am 7:539– 553, ix 39. Sonneland PR, Scheithauer BW, LeChago J, Crawford BG, Onofrio BM (1986) Paraganglioma of the cauda equina region. Clinicopathologic study of 31 cases with special reference to immunocytology and ultrastructure. Cancer 58:1720–1735 40. Spetzler RF, Detwiler PW, Riina HA, Porter RW (2002) Modified classification of spinal cord vascular lesions. J Neurosurg 96:145–156 41. Sundgren P, Annertz M, Englund E, Stromblad LG, Holtas S (1999) Paragangliomas of the spinal canal. Neuroradiology 41:788–794

615

616

4  Spine and Spinal Canal 42. Sze GK (1993) Clinical experience with gadolinium contrast agents in spinal MR imaging. J Comput Assist Tomogr 17(Suppl 1):S8–S13 43. Thakkar SD, Feigen U, Mautner VF (1999) Spinal tumours in neurofibromatosis type 1: an MRI study of frequency, multiplicity and variety. Neuroradiology 41:625–629 44. Van Goethem JW, van den Hauwe L, Ozsarlak O, De Schepper AM, Parizel PM (2004) Spinal tumors. Eur J Radiol 50:159–176 45. Wang MY, Levi AD, Green BA (2003) Intradural spinal arachnoid cysts in adults. Surg Neurol 60:49–55; discussion 55–46

46. Wippold FJ II, Smirniotopoulos JG, Moran CJ, Suojanen JN, Vollmer DG (1995) MR imaging of myxopapillary ependymoma: findings and value to determine extent of tumor and its relation to intraspinal structures. AJR Am J Roentgenol 165:1263–1267 47. Yang SY, Jin YJ, Park SH, Jahng TA, Kim HJ, Chung CK (2005) Paragangliomas in the cauda equina region: clinicopathoradiologic findings in four cases. J Neurooncol 72:49–55 48. Zajicek J (1990) Sarcoidosis of the cauda equina: a report of three cases. J Neurol 237:424–426 49. Zajicek JP (2000) Neurosarcoidosis. Curr Opin Neurol 13:323–325 50. Zee CS (1996) Neuroradiolgy. McGraw-Hill, New York

4.3  Intramedullary Diseases of the Spinal Cord

4.3  Intramedullary Diseases of the Spinal Cord P. Pawha, C. Shen, J. Doumanian, F. Lin, M. Johnson, R. Ashton, and G. Sze

of arteriovenous malformations, but it has also been used in the preoperative localization of the artery of Adamkiewicz (Yoshioka et al. 1992). Currently spine MRA is typically performed using contrast-enhanced 3D GRE imag-

Magnetic resonance imaging (MRI) has revolutionized spinal cord evaluation. Prior to MRI, myelography allowed us to visualize the spinal cord contour but provided no information about intrinsic cord disease that did not alter the contour. The ability to evaluate tissue characteristics of the cord cannot be achieved by any modality other than MRI. Advances in MRI techniques now allow us to detect a wide spectrum of intramedullary diseases and in some cases make a specific diagnosis. In this section we begin with a brief discussion of imaging techniques and normal anatomy. We divide discussion of intramedullary disease processes into spinal cord neoplasms, vascular diseases, demyelinating processes, infectious/inflammatory processes, radiation myelopathy, and traumatic injury. 4.3.1  MRI Techniques for Spinal Cord Imaging MR imaging of the spinal cord can be performed in many different effective ways using various protocols. As with all MR imaging, the sequences used should be tailored to best serve the clinical situation. Routine spine imaging typically employs fast spin-echo T2-weighted and T1weighted imaging. FSE T2 has largely replaced conventional spin-echo and has become an essential part of the spine examination. Proton density–weighted imaging is frequently used in routine exams, and is sensitive for intrinsic cord signal abnormality (Fig. 4.3.1) Gradient recalled echo (GRE) imaging can serve many purposes in spine MR imaging. GRE imaging demonstrates significant magnetic susceptibility, resulting in dephasing or ‘blooming’ artifact, which depends on tissue specific T2* relaxation times. This sequence is especially useful when looking for hemorrhage, as in the setting of trauma or vascular malformation. Fast spin-echo inversion recovery (FSE-IR) or short tau inversion recovery (STIR) imaging uses T2 weighting with suppression of fat signal. This is useful for detection of cord parenchymal signal abnormality. This sequence is particularly useful in evaluation of trauma and demyelinating disease (Fig. 4.3.1) Gadolinium contrast administration is useful in the settings of infection, tumor, prior surgery, vascular malformations, demyelinating disease, or a leptomeningeal process. Fat saturation is frequently used in post-contrast imaging. Magnetization transfer technique can augment contrast enhancement. Spinal MR angiography (MRA) has made great advances in recent years. Its main use is in the evaluation

Fig. 4.3.1  A patient with multiple sclerosis demonstrates demyelinating plaques in the upper cervical spinal cord. These are most conspicuous on proton density–weighted (c) and STIR (d) imaging. The lesions are much more faintly seen on T2 FSE images (b) and are difficult to detect on T1-weighted imaging (a). Note that normal CSF is hypointense on T1-weighted images (a) and hyperintense on T2-weighted images (b). CSF signal is isointense to the cord on the proton density image (c), although it may be slightly hypo- or hyperintense depending on the parameters used

617

618

4  Spine and Spinal Canal

ing. This can be performed using a steady state technique or a fast (time-resolved) technique. Post-processing is performed using maximum intensity projections in sagittal and coronal planes (Pattany et al. 2003). Diffusion-weighted imaging (DWI) is based on the motion of water molecules, and in recent years has been used with increasing frequency in the spine. As in brain imaging, DWI has been found to be excellent in the detection of acute spinal cord infarction (Fujikawa et al. 2004). DWI has also been proposed as a way of distinguishing neoplastic from benign vertebral disease, but this has not been definitively demonstrated (White 2000). Diffusion tensor imaging of the spine can provide more quantitative information regarding the motion of water molecules with parameters such as fractional anisotropy and average diffusivity. This technique is still rapidly evolving and has many technical challenges, but it has been used in the setting of cord compression and traumatic cord injury (Facone et al. 2005) and has potential for use in numerous white matter processes.

Fig. 4.3.3  Axial proton density–weighted (a) and gradient-echo (b) images demonstrate the normal mildly hyperintense central gray matter and the surrounding hypointense white matter

4.3.1.1  Normal MRI Anatomy The normal spinal cord begins just inferior to the medulla and extends to the tapered conus medullaris, which typically terminates at the level of the L1 vertebral body in adults, and the level of L2–L3 in infants and young children. The spinal cord usually has two areas of mild relative enlargement, one in the cervical region and the other in the conus medullaris (Fig. 4.3.2). The cord gen-

Fig. 4.3.4  Normal Anatomy. Axial FIESTA image (a) beautifully demonstrates the dorsal and ventral nerve roots exiting the cord from each side. An axial T2-weighted image (b) demonstrates the ventral median sulcus (arrow) at the anterior surface of the spinal cord

erally has a smooth contour throughout its course. The spinal cord is isointense to the brainstem on all imaging sequences, while the surrounding cerebrospinal fluid demonstrates hypointensity on T1-weighted images, and hyperintensity on T2-weighted images. The spinal cord demonstrates uniform intensity on T1 images. On axial proton density, T2 GRE, STIR, and sometimes T2 FSE imaging of the spinal cord, the normal central gray matter can be seen as hyperintense, in distinction to the surrounding white matter (Fig. 4.3.3). On sagittal T2-weighted images, artifactual longitudinal thin linear hyperintensities are routinely seen, known as Gibbs artifact or truncation artifact. At each level, on axial T2-weighted images (particularly T2 GRE), dorsal and ventral nerve roots can often be seen exiting the dorsolateral and ventrolateral aspects of the cord surface on the right and left sides (Fig. 4.3.4). On the anterior surface of the cord, there is a midline groove called the ventral median sulcus (Fig. 4.3.4). 4.3.2  Intramedullary Neoplasms

Fig. 4.3.2  Sagittal T2-weighted (a) and T1-weighted (b) images of a pediatric patient’s spine demonstrate normal subtle enlargements of the spinal cord in the cervical region and the conus medullaris

Magnetic resonance (MR) of the spinal cord has revolutionized the way intramedullary lesions are diagnosed and characterized, but accurate distinction between his-

4.3  Intramedullary Diseases of the Spinal Cord

tological subtypes remains elusive due to overlapping appearances. T1-weighted sequences provide excellent morphologic detail of the spinal cord, while T2-weighted sequences allow characterization of the intramedullary lesion and surrounding edema. Associated intramedullary cysts and cavities are usually well visualized on both T1 and T2-weighted images. MRI evaluation is also very sensitive in characterizing hemorrhage. While non-contrast MR evaluation of the spinal cord can accurately detect intramedullary lesions, intravenous gadolinium contrast-enhanced MR increases specificity and allows for better characterization. In the setting of peritumoral edema, gadolinium contrast is helpful in delineating the location and extent of the lesion. This is most helpful in hemangioblastomas and metastases as both tend to be well circumscribed lesions. Precise characterization of intramedullary gliomas is more difficult as they may be ill defined and infiltrating. Additionally, enhancement in primary gliomas is often variable and incomplete. Further complicating matters, there have also been documented cases of non-enhancing gliomas (Sze 2002). Generally, intramedullary spinal cord tumors are rare and intracranial tumors are much more common. Like the brain, the spinal cord is composed of neurons, astrocytes, oligodendrocytes, and ependymal cells. While the cellular composition of the spinal cord may be similar to that found in the brain, the cell types of primary intramedullary spinal cord tumors occur with different frequency compared to those found in the brain. Gliomas, astrocytomas and ependymomas, account for nearly 90% of the intramedullary spinal cord tumors (Lowe 2000; Zimmerman and Bilaniuk 1988). Hemangioblastomas are much less frequent, representing less than 5% of all intramedullary tumors (Lowe 2000; Sze 2002; Zimmerman and Bilaniuk 1988). Intramedullary metastases are rarer than hemangioblastomas (Sze 2002; Zimmerman and Bilaniuk 1988). There have also been reported cases of intramedullary subependymoma, ganglioglioma, schwannoma, and other lesions, which are not covered due to their rarity. 4.3.2.1  Ependymoma 4.3.2.1.1  Epidemiology Intramedullary ependymomas usually present in the fourth to fifth decades of life, with a variable spectrum of age presentation, but are uncommon in children (Sze 2002). Ependymomas have a male predominance of approximately 3:2 (Sze 2002; Zimmerman and Bilaniuk 1988). They are the most common primary tumors of the lower cord, conus medullaris, and filum terminale. Myxopapillary ependymomas are most common in the conus and filum.

4.3.2.1.2  Clinical Presentation and Treatment Clinical presentation is nonspecific with patients typically presenting with radicular or focal pain in their neck or back. Weakness, gait disturbances, paresthesias, and bladder or bowel dysfunction may also be seen. Treatment is directed towards complete resection, as recurrence is rare with complete removal of an encapsulated tumor. Incomplete resection may result in CSF dissemination or metastasis. 4.3.2.1.3  Pathology Ependymal cells line the ventricular system of the brain and the central canal of the spine. As a result, ependymomas tend to be central in location and demonstrate centrifugal growth. These tumors generally grow slowly and can be rather large at presentation. Grossly, ependymomas are cylindrical, elongated masses that cause fusiform expansion of the spinal cord. They are often soft and friable, with a cleavage plane separating the tumor from the spinal cord. The connective stromata of ependymomas are vascular, often resulting in hemorrhage between the spinal cord and the tumor. Rarely, ependymomas may be pigmented, but the pigment is not melanin. Ependymomas may grow into the conus medullaris, adhering and engulfing adjacent nerve roots. Cyst formation is associated with approximately half of all ependymomas. Unlike intracranial ependymomas, calcification is uncommon in intramedullary ependymomas (Zimmerman and Bilaniuk 1988). The most common histology is of a cellular type with well-defined cuboidal or low columnar cells arranged in a papillary fashion. The most common lesion of the filum terminale is the myxopapillary type, which is prone to hemorrhage and may be responsible for unexplained subarachnoid hemorrhage. 4.3.2.1.4  Imaging Findings on plain radiography or computed tomography are often nonspecific. Plain radiographs of the spine may show erosion or scalloping of the pedicles or posterior vertebral bodies, suggestive of a lesion within the spinal cord. MR evaluation of intramedullary ependymomas often reflects their heterogeneous pathology. Non-contrast images often demonstrate symmetric widening of the spinal cord or a focal mass, frequently near the conus (Bydder et al. 1985). The greatest enlargement of the spinal cord is usually at the level of the solid component of the tumor. On T1-weighted images, the lesion is often hypointense to isointense to the spinal cord and demonstrates heterogeneity on T2-weighted sequences (Figs. 4.3.5–4.3.7).

619

620

4  Spine and Spinal Canal

Areas of hemorrhage are common. Hemosiderin deposition is often seen at the superior and inferior borders of the mass, appearing hypointense on both T2-weighted and T1-weighted images. Areas of hypercellularity will appear hypointense on T2-weighted images. Ependymomas often demonstrate variable and heterogeneous postcontrast enhancement (Bydder et al. 1985). While astrocytomas and ependymomas may be generalized into irregular and well-defined tumors respectively,

their imaging findings often overlap. Several secondary findings may suggest a particular tumor. Ependymomas occur more frequently in the lower spinal cord and conus medullaris than astrocytomas (Lowe 2000; Sze 2002). A small cleavage plane may also be seen with ependymomas. Astrocytomas arise eccentrically, often from the posterior spinal cord, while ependymomas are central and arise from the ependymal cells of the central canal (Lowe 2000; Sze 2002). Hemorrhage is more common within

Fig. 4.3.5  Ependymoma. Sagittal T1-weighted (a) and STIR (b) sequences demonstrate a focal intramedullary lesion at the craniocervical junction. The mass is nearly isointense to the spinal cord on T1-weighted images and demonstrates T2 hyperintensity. Note the two small intratumoral cysts (arrowheads) best seen on the sagittal STIR image (b). Sagittal (c) and axial (d) post-contrast T1-weighted images demonstrate heterogeneous and peripheral lesion enhancement. Sagittal proton density image of the thoracolumbar region (e) demonstrates two additional

intramedullary lesions (arrows), which are presumed to be additional ependymomas in this patient with neurofibromatosis type 2. An axial T2-weighted image (f) at this level demonstrates the characteristic central location seen in ependymomas. Sagittal (g) and axial (h) post-contrast images through these lesions demonstrate solid enhancement. Note that a small anterior meningioma (arrow) is also present in this patient with neurofibromatosis type 2 in the sagittal post-contrast image (g)

4.3  Intramedullary Diseases of the Spinal Cord Fig. 4.3.6  Ependymoma. Sagittal (a) and axial (d) T2-weighted images demonstrate extensive cyst formation cranial to the solid component (arrow) of the tumor located at the T3–T4 level. This briskly enhances (arrow) on post-contrast images (c, f). Diffuse expansion of the cord is present. Intramedullary hypointensity (arrow) on STIR (b) imaging and hyperintensity on T1-weighted imaging (e) re­pre­sent intratumoral hemorrhage

4.3.2.2  Astrocytoma 4.3.2.2.1  Epidemiology

Fig. 4.3.7  Ependymoma. Sagittal T2-weighted image (a) demonstrates a heterogeneous focally expansile intramedullary lesion in the thoracic cord. Sagittal proton density (b) and contrast-enhanced (c) images demonstrate both the intramedullary ependymoma (arrow) and a meningioma (arrowhead) at the cervicomedullary region in this patient with neurofibromatosis type 2

ependymomas than in astrocytomas, but this finding is nonspecific. Areas of low signal intensity on T2-weighted imaging, reflecting hypercellularity, are more common in ependymomas. Myxopapillary ependymomas are soft expansile masses found near the filum terminale. On histological exam, abundant mucin accumulation around the vessels and between the cells is characteristic. These lesions are generally hyperintense on both T1 and T2-weighted images.

Approximately one third to one half of all spinal cord gliomas are astrocytomas. Astrocytomas, however, are more common in children. The peak incidence of astrocytomas is in the third to fifth decade, with a slight male predilection (Zimmerman and Bilaniuk 1988). Astrocytomas are most commonly found in the cervical and thoracic cord, and their prevalence decreases in the lumbar spine. Most astrocytomas are intramedullary, but exophytic lesions extending into the intradural extramedullary space have been reported (Zimmerman and Bilaniuk 1988). 4.3.2.2.2  Clinical Presentation and Treatment Presenting symptoms are frequently nonspecific which may delay diagnosis (Erimer and Onofrio 1985). Local or referred pain, weakness, gait abnormalities, or bladder disturbances are common symptoms. Progressive motor and sensory changes may be seen on physical examination. Children may present with progressive scoliosis. Due to their low grade predominance, astrocytomas may be large at presentation and span several vertebral segments. Surgical resection followed by radiation is recommended in all cases. Prognosis is dependent on histological grade. Low-grade tumors may have a protracted

621

622

4  Spine and Spinal Canal

course while higher-grade tumors may disseminate into the CSF space. 4.3.2.2.3  Pathology Astrocytomas are neoplastically transformed astrocytes; three quarters are grade I or II. Astrocytomas generally cause fusiform expansion of the spinal cord. Gross appearances of astrocytoma often depend on the degree of hemorrhage. Aggressive tumors are more vascular and cystic change is visualized in a third of astrocytomas (Reimer and Onofrio 1985; Sze 2002; Zimmerman and Bilaniuk 1988). Often, astrocytomas are infiltrative and do not have a clear boundary with the spinal cord. The tumor is often eccentric and located posteriorly, involving the posterior columns. They may be multi-segmental over several vertebral levels. Uncommonly, holocord astrocytoma may involve the entire spinal cord. Astrocytomas are usually less hemorrhagic and less necrotic than ependymomas. The imaging appearance of astrocytomas and ependymomas commonly overlap, making histological distinction on MR difficult. 4.3.2.2.4  Imaging Plain radiographs of the spine may show widening of the spinal canal with bony erosion. With respect to MR, spinal cord enlargement is almost always present. Astrocytomas demonstrate isointense to hypointense signal on T1-weighted sequences. The lesion and associated surrounding edema demonstrates high signal on T2-weighted sequences (Figs. 4.3.8, 4.3.9). Postcontrast enhancement is almost always visualized but variable, ranging from homogeneous to heterogeneous. The tumor margins are usually irregular or poorly defined due to the infiltrative nature of the tumor in contrast to an ependymoma, where a capsule or cleavage plane may be seen. While enhancement is usually immediate, delayed enhancement may be seen in necrotic tumors. MR evaluation is valuable in differentiating tumor from associated cyst formation. Associated cysts may be intratumoral or peritumoral (rostral or caudal). Peritumoral cysts, filled with proteinaceous or hemorrhagic fluid, are often reactive, benign, and not lined with tumor cells. Both reactive and tumor cysts demonstrate decreased signal intensity on T1-weighted sequences and increased signal intensity on T2-weighted sequences. On non-contrast sequences, these cysts may be indistinguishable from solid tumor if there is tumoral necrosis. Hemorrhage or protein within the cyst cavity may also make the cyst appear isointense to the spinal cord. In contrast to tumor cysts, the walls of benign cysts do not enhance. Complex benign syrinxes may have a similar appearance to a cystic astrocytoma. Gliosis in the

Fig. 4.3.8  Cervical astrocytoma. Sagittal T1- (a) and T2-weighted (b) images demonstrate a focal intramedullary lesion. Note the small peritumoral cyst (arrow) at the superior aspect of the lesion. Sagittal (c) and axial (d) post-contrast T1-weighted images demonstrate lesion enhancement. Note the eccentric and posterior intramedullary location

cyst wall is secondary to chronic CSF pulsations and may demonstrate increased signal intensity on long TR images, mimicking tumor. 4.3.2.3  Hemangioblastoma 4.3.2.3.1  Epidemiology Hemangioblastomas rarely involve the spinal cord, constituting less that 5% of all intramedullary tumors (Baker et al. 2000), but are the third most common intramedullary tumor after ependymoma and astrocytoma. They typically present in the fourth decade, without gender predilection (Browne et al. 1976). Spinal hemangioblastomas may be either single or multifocal. Involvement of the thoracic cord is most

4.3  Intramedullary Diseases of the Spinal Cord Fig. 4.3.9  Thoracic astrocytoma. Sagittal T1- (a) and T2-weighted (b) images demonstrate an intramedullary mass spanning multiple vertebral levels. Note the large surrounding peritumoral cysts both cranial and caudal to the lesion. Sagittal post-contrast T1-weighted images (c) demonstrate solid heterogenous enhancement in the lesion, and lack of enhancement in the peritumoral cysts

common, followed by the cervical cord. While most hemangioblastomas are intramedullary, there have been reported cases of intradural extramedullary and extradural lesions. Approximately a third of patients with intramedullary hemangioblastomas have von Hippel Lindau (VHL) syndrome, an autosomal dominant disorder with near complete penetrance (Aminoff et al. 1974; Costigan and Winkelman 1985; Lowe 2000; Sze 2002). Typical findings of VHL include cerebellar and intramedullary hemangioblastomas, retinal angiomatosis, renal cell carcinoma, and pheochromocytomas. Clinical presentation of patients with VHL is often due to the extramedullary findings rather than the spinal cord lesion itself. All patients with hemangioblastomas are recommended to have complete MR evaluation of their neural axis as occult lesions may be discovered. 4.3.2.3.2  Treatment Complete surgical removal often results in cure as there is often a cleavage plane between the tumor and adjacent spinal cord. Radiation as adjunctive therapy has been used in instances of incomplete excision. 4.3.2.3.3  Pathology Hemangioblastomas are composed of endothelial cells mixed with stromal cells containing both fat and hemosiderin (Browne et al. 1976). The endothelial cells form masses, cords, and thin-walled vessels. Eventually, the tumor will consist of a collection of capillaries with both feeding arteries and draining veins. Associated cysts are usually reactive, without tumor cells. Associated cyst formation is seen in up to two thirds of intramedullary hemangioblastomas (Browne et al.

1976). Cystic fluid is often proteinaceous from transudation of fluid and hemorrhage. Meningeal varicosities on the dorsal surface of the cord are also associated with cord hemangioblastomas (Baker et al. 2000). 4.3.2.3.4  Imaging Plain radiographs may demonstrate widening of the spinal canal. Unlike ependymomas and astrocytomas, hemangioblastomas have characteristic MR imaging findings which may allow for specific identification. Most intramedullary hemangioblastomas are eccentrically located and over half are associated with cranial and caudal cysts (Baker et al. 2000), often out of proportion to the size of the primary lesion (Figs. 4.3.10, 4.3.11). The cysts vary in signal intensity, dependent on their protein content (Lowe 2000; Sze 2002). As a result, the cysts may occasionally demonstrate similar signal to normal unenhanced spinal cord. Dorsal flow voids, from feeding arteries or meningeal varicosities, may be seen and help in distinguishing hemangioblastomas from other lesions. Spinal hemangioblastomas are usually hypointense to isointense on T1-weighted images and isointense to hyperintense on T2-weighted images with associated marked edema. Post-contrast evaluation demonstrates a well-defined and markedly enhancing tumor nidus, allowing differentiation from the spinal cord and associated cyst (Browne et al. 1976; Sze et al. 1988). 4.3.2.4  Metastasis 4.3.2.4.1  Epidemiology While extradural metastases are relatively common, meta­ static intramedullary tumors are uncommon (Costigan and Winkelman 1985). Given the percentage weight of

623

624

4  Spine and Spinal Canal

Fig. 4.3.10  Cervical hemangioblastoma. An intramedullary nodule (arrow) at the C4 level demonstrates minimal hyperintensity on T1-weighted imaging (a), hypointensity on T2-weighted imaging (b), and intense enhancement on postcontrast imaging (c). Note the extensive cyst formation and edema cranial and caudal to the enhancing nodule. Axial contrast-enhanced CT image (d) nicely demonstrates the intensely enhancing nodule. Angiographic image (e) from a vertebral artery injection demonstrates the highly vascular nature of this lesion (arrowhead), as well as a feeding artery (black arrow) and draining vein (white arrow)

Fig. 4.3.11  Multiple hemangioblastomas. This patient with Von Hippel Lindau disease had already undergone partial resection of cerebellar hemangioblastomas. Sagittal T2-weighted image (a) demonstrates a rounded intramedullary mass (arrow) in the midcervical region which is similar in signal intensity to the spinal cord. Postsurgical changes and residual or recurrent tumor are seen in the cerebellum. Sagittal (b) and axial (c) post-contrast T1-weighted images demonstrate solid nodular enhancement both in the cervical cord and in the cerebellum. Associated cysts are seen cranial and caudal to the enhancing cervical nodule

the cord in relation to the entire neural axis, intramedullary metastases are rarer than expected (Zimmerman and Bilaniuk 1988). The incidence of intramedullary metastasis, however, may be increasing with improved survival. Intramedullary metastases are often accompanied by intracranial involvement. Of all intramedullary cord metastases, lung carcinoma is the most common (Jellinger et al. 1979; Lowe 2000; Sze 2002). Other non-neurogenic tumors such as breast carcinoma, melanoma, lymphoma, colon carcinoma, and renal cell carcinoma can also occur (Costigan and Winkelman 1985; Jellinger et al. 1979). Several postulated metastatic routes have been proposed (Costigan and Winkelman 1985; Jellinger et al. 1979): hematogenous arterial seeding, vertebral venous system (Batson’s plexus) seeding, and direct extension from the CSF or nerve roots. Intramedullary metastases are often associated with leptomeningeal disease. Thoracic cord involvement is most common, followed by the

cervical and lumbar cord. While metastatic involvement of the spinal cord may be multifocal, most reported cases are solitary. 4.3.2.4.2  Clinical Presentation and Treatment The most common clinical presentation of intramedullary metastasis is focal pain. Radicular pain, which is common in extradural tumors, may also be present with cord metastasis. Weakness, paresthesias, and bowel and bladder disturbances can also occur. Symptoms are often rapidly progressive with complete paraplegia within several months of clinical presentation. Radiation is the preferred treatment as surgery offers little advantage. Most patients demonstrate rapid progression of symptomatology and soon succumb to diffuse metastatic disease.

4.3  Intramedullary Diseases of the Spinal Cord

4.3.2.4.3  Imaging MR evaluation often demonstrates a widened cord, extending for a considerable length in relation to the true lesion size. The cord may demonstrate low signal intensity on T1-weighted sequences and high signal intensity on T2-weighted sequences. Central low T1 signal may be confused with a cyst; however, cysts are rarely associated with metastases. On T2-weighted sequences, the metastatic nidus may be seen as the focus of relatively lower intensity, surrounded by edema which appears disproportionally large in relation to the metastatic focus (Fig 4.3.12). Intramedullary metastatic deposits often demonstrate homogenous and marked enhancement (Sze et al. 1988), but their appearance may be variable. The primary differential consideration of these MR findings is hemangioblastoma. Intramedullary metastasis may rarely be hemorrhagic with varying MR appearance. 4.3.3  Vascular Diseases of the Spinal Cord A variety of vascular diseases can affect the spinal cord. Spinal cord infarction is an important entity whose diagnosis requires familiarity with the unique patterns and clinical circumstances associated with it. Arteriovenous

malformations encompass a variety of lesions, including dural fistulas and glomus type malformations. Both arteriovenous malformations and cavernous malformations are important causes of treatable myelopathy, for which early diagnosis and treatment are essential. To understand any of these vascular diseases of the spinal cord, we must first have a working knowledge of the spinal cord’s vascular anatomy. 4.3.3.1  Vascular Anatomy of the Spinal Cord 4.3.3.1.1  Arterial Supply The spinal cord has a superficial and a deep arterial system. Anteriorly the superficial or extrinsic system is comprised of the longitudinally oriented anterior spinal artery and the pial plexus. The deep arterial system of the spinal cord consists of radial perforating arteries and sulcal arteries. These are usually too small to be resolved even on catheter angiography (Pattany et al. 2003). The radial arteries, or vasacorona, supply the peripheral portion of the spinal cord by giving off small perforating branches which penetrate the cord parenchyma in a circumferential fashion. The sulcal, or central, arteries travel in the ventral median fissure and arborize to supply the

Fig. 4.3.12  Metastasis from non-small-cell lung carcinoma. Sagittal T2-weighted (a) and proton density–weighted (b) images demonstrate a focal slightly expansile area of low signal (arrow) with marked surrounding high signal (edema) extending cranially and caudally. Axial T2-weighted (d) and proton density–weighted (e) images also show a rounded low signal lesion with a hyperintense rim. The lesion is low signal on axial T1-weighted imaging (f). An axial T2-weighted image (g) just superior to image (d) shows the cord edema above the lesion. Sagittal post-contrast T1-weighted image (c) demonstrates peripheral enhancement of the lesion

625

626

4  Spine and Spinal Canal

central portion of the spinal cord (Lasjaunias and Berenstein 1987). The anterior spinal artery is formed superiorly by branches from each vertebral artery and is supplied by multiple radiculomedullary feeding arteries along its course. The anterior spinal artery lies in the anterior median sulcus, extending from the medulla to the filum terminale. This artery supplies the anterior two thirds or the spinal cord. Posteriorly, the superficial system is more variable and plexiform, consisting of the pial plexus and often two small longitudinally oriented posterolateral spinal arteries. These paired posterolateral spinal arteries may arise at their superior extent from the vertebral arteries or posterior inferior cerebellar arteries, and are supplied along the course of the spinal cord by radiculomedullary arteries. The posterior arterial system supplies the posterior one third of the spinal cord. The anterior spinal artery and two posterior spinal arteries join at the conus medullaris to form the cruciate anastomosis (Bemporad and Sze 2000). The anterior spinal artery and posterior spinal arteries/plexus are supplied by anterior and posterior radiculomedullary arteries, respectively. Segmental arteries are present at every spinal level to supply the nerve roots and dura, but only some of the levels also give rise to radiculomedullary branches to the cord. There are approximately 6–10 anterior radiculomedullary arteries and 10–20 posterior radiculomedullary arteries throughout the length of the cord (Bemporad and Sze 2000). The cervical radiculomedullary arteries arise from branches of the vertebral arteries, and from branches of the deep and ascending cervical arteries (Thron 1988). There is often a prominent radiculomedullary artery at the C5–C6 level, known as the artery of cervical enlargement. The thoracolumbar spinal cord is supplied by radiculomedullary branches from the intercostals and lumbar arteries. Lumbar and intercostal segmental arteries give rise to anterior and posterior divisions. The posterior division gives rise to a radicular artery to the nerve roots, a muscular branch, and a branch to the vertebral body and dura (Yoshioka et al. 2003) Again, at some levels there are radiculomedullary arteries which supply the cord. In the thoracic region, there is typically a paucity of anterior feeding medullary arteries. The most important radiculomedullary artery is the Artery of Adamkiewicz which supplies the lower third of the spinal cord. It is a dominant anterior radiculomedullary artery arising between T9 and T12 in 62–75% of cases. It arises from the left side in 68–80% of cases (Thron 1988; Yoshioka et al. 2003, 102) There is significant variability in the level of origin, and it may arise from the lower lumbar region (26%) or more superiorly between T6 and T8 (12%) (Thron 1988). This artery forms a hairpin, downward turn as it joins the anterior spinal artery (Yoshioka et al. 2003). The sacral cauda equina region is supplied by iliolumbar and sacral arteries arising from the internal iliac arteries (Thron 1988).

4.3.3.1.2  Venous Drainage Radial and sulcal intrinsic medullary veins drain into a complex superficial venous network. Longitudinally and obliquely oriented transmedullary veins are also present The superficial venous network is extensively anastomosed and relatively symmetrically arranged, without the anterior predominance seen in the arterial system (Lasjaunias and Berenstein 1987) There are dominant longitudinal venous channels in the midline both anteriorly and posteriorly, named the anterior median spinal vein and the posterior median spinal vein. Blood drains from these veins into anterior and posterior radiculomedullary veins, which join each other as they run along the spinal nerve roots (Thron 1988) The radiculomedullary vein communicates with the epidural venous system as it joins the intervertebral vein. There is a valve at this location preventing the epidural blood from refluxing into the intradural space, but there are no valves within the intradural venous system (Bemporad and Sze 2000). Posterior radiculomedullary veins may have significantly larger calibers than anterior radiculomedullary veins. Variably, a dominant large anterior or posterior great radicular vein may be present (less frequently present and not necessarily at the same level as its arterial counterpart, the artery of Adamkiewicz) (Thron 1988). 4.3.3.2  Spinal Cord Infarction Spinal cord infarction is a rare but important and potentially catastrophic clinical entity. It is important for the radiologist to be familiar with the characteristic imaging findings, typical distribution patterns, diagnostic pitfalls, and various clinical etiologies and predisposing circumstances. Infarction in the spinal cord is much less common than in the brain, likely due to collateral vascularity. Compared with infarction in the brain, spinal cord infarction is much less likely to be caused by atherosclerotic disease. In many cases, there is a specific predisposing clinical setting such as vascular surgery or dissection, among many others, as are discussed later. In the absence of an identifiable cause and presence of vascular risk factors such as smoking, diabetes mellitus, and hypertension, atherosclerotic disease is often a presumed etiology. The diagnosis carries a high morbidity and mortality, with one series reporting 13 of 28 patients having a poor outcome, including three deaths. Within months after cord infarction, pain often becomes a disabling symptom for surviving patients. Prognosis has been found to be best predicted by severity of neurological deficits at presentation (Masson et al. 2004). Treatment for spinal cord infarction is still generally supportive, with occasional use of anticoagulation and antiplatelet therapy (Mikulis et al. 1992).

4.3  Intramedullary Diseases of the Spinal Cord

4.3.3.2.1  Etiologies Relatively common causes of spinal cord infarction include vascular surgery, aortic aneurysm, vascular dissection, spinal surgery, cardioembolic disease, atherosclerotic disease, and hypotension (Masson et al. 2004, Weidauer et al. 2002; Yuh et al. 1992). One of the more frequent causes of infarction is vascular surgery, particularly repair of thoracic or abdominal aortic aneurysm. Aortic aneurysm repair involving the thoracoabdominal region is much more likely than repair limited to the abdominal region to cause spinal cord ischemia (Hurst 2001). In many cases of operative spinal cord vascular injury, the artery of Adamkiewicz is involved, potentially resulting in paraplegia. The incidence of spinal cord ischemia in thoracoabdominal aortic aneurysm repair has been reported to be 7% with the use of various spinal cord protective maneuvers and 16% without protection (Yoshioka et al. 2003). Neurosurgeons have found that reimplantation of intercostal arteries reduces the incidence of such complications. For this reason, MRA of the spine can be useful in identifying the artery of Adamkiewicz preoperatively. One series was able to identify this artery in 66.7 % of patients with aortic aneurysm (Yoshioka et al. 2003). Spinal cord ischemia has been reported in the setting of other vascular surgeries including aortoiliac bypass, aortofemoral bypass, and femoroiliac bypass (Yuh et al. 1992). Apart from the surgical setting, aortic aneurysm with thrombus can itself lead to cord infarction by direct involvement and thrombosis of spinal arteries, or by mural thrombus becoming a source of emboli. Aortic dissection can lead to hypoperfusion or occlusion of the artery of Adamkiewicz, as well as any of the intercostal and lumbar arteries. Dissection is more likely to affect the spinal cord when it involves the descending aorta than when it is confined to the ascending aorta and aortic arch. This is typically accompanied by the clinical presentation of chest pain with radiation to the back. Ischemic injury in this setting may involve radiculo­medullary arteries at multiple levels and may involve both the anterior and posterior spinal artery territories (Hurst 2001). Dissection of one or both vertebral arteries can lead to infarction of the cervical spinal cord. Unilateral infarctions occur as a result of ipsilateral vertebral artery dissection. Because a unilateral segmental artery often supplies both sides of the spinal cord, bilateral cord infarction can result from either unilateral or bilateral vertebral artery dissection. This may be accompanied by clinical symptoms of neck or occipital pain. Concomitant infarcts in the brainstem, cerebellum, or posterior cerebral artery distribution may be seen (Crum et al. 2000). Severe vertebral artery stenosis has also been reported to cause cervical spinal cord infarction (Suzuki et al. 1998). Cord infarct has also been noted in the setting of spinal decompression surgery. Hypotension,

either in the operative setting or otherwise, can result in spinal cord infarctions that are often in a watershed distribution. Other less common reported causes include vasculitis, complications from various angiographic procedures and percutaneous needle procedures, fibrocartilaginous emboli (intravasation of herniated disk material), and cocaine use (Masson et al. 2004; Mikulis et al. 1992; Weidauer et al. 2002; Yuh et al. 1992). 4.3.3.2.2  Imaging Findings and Patterns of Involvement MRI is the clear imaging modality of choice in suspected acute spinal cord infarction. MRI has the ability to assess the internal tissue characteristics of the cord as well as to identify spinal cord enlargement. One prospective series demonstrated MR imaging abnormalities in 24 of 28 patients with clinically determined cord infarction (Masson et al. 2004) The most consistent finding in cord infarction is intramedullary high signal intensity on T2-weighted imaging. This can be seen on sagittal images as longitudinally oriented T2 hyperintense signal within the cord. Signal abnormality is often central, ventral, or ventrolateral in location and has been described as “pencil-like” (Fig. 4.3.13) (Weidauer et al. 2002; Yuh et al. 1992). This T2 prolongation often extends over multiple levels, but may also be limited to a short segment. In the acute setting this T2 hyperintensity represents a combination of infarction and vasogenic edema, making it initially difficult to determine the exact level and extent of infarct (Mikulis et al. 1992). Cord expansion in the involved area is well demonstrated on sagittal T1- and T2-weighted imaging. T1-weighted images often demonstrate no signal change, but may occasionally reveal hypointense signal (Yuh et al. 1992) (Fig. 4.3.13). Axial T2-weighted images are important for confirmation and better delineation of the signal abnormality. Spinal cord infarcts usually involve the anterior spinal artery territory, as the dual posterior spinal arteries and posterior pial plexus have a more redundant circulation. As such, T2 prolongation in infarction is usually seen within the anterior two-thirds of the cord (Fig. 4.3.13). Gray matter is more vulnerable to ischemia as a result of its higher metabolic activity and requirements. Therefore, it is involved earlier and more often than white matter (Weidauer et al. 2002). Signal abnormality may be seen throughout the gray matter distribution or only in a portion of the gray matter (Figs. 4.3.13, 4.3.14). More extensive infarcts will not uncommonly involve white matter tracts as well and may involve the entire anterior spinal artery territory or even most of the cord at a given level (Table 4.3.1). Anterior spinal artery syndrome is characterized by paraparesis or quadraparesis, loss of temperature and

627

628

4  Spine and Spinal Canal

Fig. 4.3.13  Anterior spinal artery infarct. Early in the time course, sagittal T2-weighted image demonstrated subtle longitudinal pencil-like hyperintensity (arrows) within the anterior aspect of the cord. Imaging performed on the following day shows marked progression of the signal abnormality: Sagittal STIR (b) and axial T2-weighted images (c, d) show the signal abnormality occupying the anterior two-thirds (denoted by arrows on image b) of the cord in a classic pattern. The infarct is centered in the upper thoracic cord, compatible with a watershed distribution. Expansion of the cord is also present

Table 4.3.1  Acute spinal cord infarction: key MRI findings T2 hyperintensity Cord expansion Anterior > posterior Gray matter > white matter Central > peripheral Bilateral > unilateral Hyperintense on DWI

pain sense bilaterally below the affected level, pain at the level of the lesion, and bladder dysfunction. Paralysis is often initially flaccid and areflexic, evolving to spastic paralysis. In the cervical region, an ipsilateral Horner’s syndrome may be present (Hurst 2001; Suzuki et al. 1998; Weidauer et al. 2002). The anterior horns are frequently involved, often bilaterally, resulting in a classic “snake-eyes” or “owl-eyes” appearance (Masson et al. 2004; Weidauer et al. 2002).

The central territory of the anterior spinal artery, which includes most of the gray matter, is most commonly involved. This central territory is supplied by the sulcal arteries, which arborize from the center of the cord, as discussed earlier. In one series, Weidauer et al found 13 of 16 patients to have bilateral T2 hyperintensities, the majority of which were symmetric. The three remaining patients had unilateral paramedian infarctions which were ascribed to occlusion of a lateralized spinal sulcal artery (Weidauer et al. 2002). The periphery of the cord is rarely involved alone, owing to the plexiform nature of the vasacorona (Table 4.3.1). Posterior spinal artery territory infarcts are rare. The posterior spinal artery territory includes the posterior dorsal horns, the dorsal columns, and a variable portion of the spinothalamic and corticospinal tracts (Crum et al. 2000). Posterior territory infarcts have been reported to appear as triangular dorsal T2 hyperintensities and may be unilateral or bilateral (Weidauer et al. 2002). The rare posterior spinal artery syndrome clinically presents with deficits in proprioception, light touch, and vibration (Weidauer et al. 2002). Posterior spinal artery infarction can coexist with anterior spinal artery infarction, particularly in the settings of aortic dissection, aortic aneurysm/surgery, or hypotension. The watershed region between the anterior and posterior spinal artery territories includes the anterior dorsal horns and portions of the spinothalamic and corticospinal tracts (Crum et al. 2000). The anterior and posterior artery territories are relatively isolated from one another, without adequate anastomoses. There is also a potential watershed region between the central anterior spinal artery supplied by the sulcal arteries, and the peripheral territory supplied by the vasacorona (Weidauer et al. 2002). A minority of infarctions may develop areas of hemorrhagic transformation, demonstrating foci of hyperintense T1 signal and hypointense T2 signal (Fig. 4.3.14). T2* gradient-echo imaging is even more sensitive for the detection of small foci of hemorrhage. Venous infarction is more likely to be hemorrhagic (Mikulis et al. 1992). DWI has been shown to be very sensitive for spinal cord infarction, analogous to the imaging of cerebral infarction. Hyperintensity is seen in acute infarction, representing restricted diffusion of water molecules in the infarcted region (Fig. 4.3.14). Positive findings on DWI have been reported in the first few hours when no other MRI abnormalities were present (Fujikawa et al. 2004). 4.3.3.2.3  Enhancement and Evolution of Findings T1 and T2 imaging abnormalities are often not evident in the first four hours after onset of symptoms, analogous to brain imaging. After four hours, MRI is quite sensitive in demonstrating the above discussed signal abnormalities. Enhancement is usually, but not always, absent in

4.3  Intramedullary Diseases of the Spinal Cord

Fig. 4.3.14  Conus medullaris infarct. Sagittal T2-weighted (a) and proton density–weighted (b) images demonstrate central hyperintense signal in the conus medullaris with expansion. Axial proton density–weighted image (c) demonstrates abnormal signal hyperintensity within a classic gray matter distribution. Foci of intramedullary hyperintensity on non-contrast T1-weighted axial image (d) are consistent with petechial hemorrhage. Enhancement is seen on axial (e) and sagittal (f) postcontrast images. Hyperintense signal on diffusion weighted imaging (g) coupled with dark signal on the ADC map (h) indicates restricted diffusion (arrows)

the acute stage. Similar to cerebral infarction, enhancement usually develops in the subacute stage (Mikulis et al. 1992) (Fig. 4.3.14, 4.3.15). Edema may subside in the subacute to chronic stage, with the remaining areas of T2 prolongation providing a better representation of the infarcted region of spinal cord. In the chronic stage, this T2 prolongation may persist for months or years. Atrophy of the spinal cord at and below the level of infarction is usually seen in the chronic stage. 4.3.3.2.4  Patterns of Distribution by Level Most spinal cord infarctions are thought to occur in the upper thoracic region and in the thoracolumbar junction (Yuh et al. 1992). To understand better the frequency with which spinal cord ischemia occurs at different spi-

nal levels, let us take another brief look at the vascular anatomy. The anterior spinal artery is largest in caliber in the cervical level, where it has the largest number of radiculomedullary feeding arteries (usually two to four) (Crum et al. 2000). The thoracic segment of the anterior spinal artery often receives only one radiculomedullary feeder and is generally narrow in caliber, being smallest at the mid-thoracic level. There are also fewer sulcal arteries at this level (Yuh et al. 1992). As such, this becomes a watershed region between cervical radiculomedullary arteries and the artery of Adamkiewicz and is more vulnerable to hypoperfusion and ischemia (Fig. 4.3.13). The artery of Adamkiewicz, arising from T9–T12 in a majority of cases, is frequently the sole supply of the entire lower thoracic and lumbar region including the conus medullaris. Occlusion of this artery generally cannot be compensated for by other vessels. Compromise of the

629

630

4  Spine and Spinal Canal

Fig. 4.3.16  Axial T2-weighted image demonstrates a triangular region of hyperintensity (arrow) within the vertebral body in this patient with vertebral body infarction

Fig. 4.3.15  Conus medullaris infarct. Sagittal T2-weighted (a) and proton density–weighted (b) images demonstrate central hyperintense signal in the conus medullaris with mild expansion. Axial T2-weighted image (c) demonstrates the gray matter to be hyperintense and swollen. Axial post-contrast image (d) shows enhancement, also within the gray matter distribution

artery of Adamkiewicz with subsequent thoracolumbar spinal cord infarction is often a result of aneurysm, surgery, or dissection of the thoracoabdominal aorta or spinal surgery (Weidauer et al. 2002; Yuh et al. 1992). 4.3.3.3  Vertebral Body Infarction Vertebral body infarction can be seen in association with spinal cord infarction, as they share blood supply from segmental arteries. This can be a very useful confirmatory sign that an intramedullary signal abnormality is ischemic in nature (Masson et al. 2004; Mikulis et al. 1992; Weidauer et al. 2002; Yuh et al. 1992). The vertebral body is supplied by anterior and posterior central arteries. The anterior central arteries supply the anterior and anterolateral portions of the vertebral body. Posterior central artery branches begin in the four corners of the posterior vertebral body and converge in the center to supply its central and posterior portions. The central and posterior portions of the vertebral body thus have a redundant blood supply, and are relatively resistant to ischemia. The deep medullary and endplate

regions constitute the watershed and end artery regions where infarction is most likely to occur (Yuh et al. 1992). A characteristic triangular shaped region of bone marrow signal abnormality involving the endplate and deep medullary regions was described by Yuh et al in association with spinal cord infarcts (Yuh et al. 1992). This is seen as increased T2 signal and decreased T1 signal within the bone marrow (Fig. 4.3.16). Vertebral body infarctions were seen in 3 of 12 cases in the series by Yuh et al. and in 1 of 16 cases in another series (Weidauer et al. 2002; Yuh et al. 1992). The vertebral abnormalities may be at different levels than the spinal cord abnormalities, as the radiculomedullary arteries have an ascending course as they travel along nerve roots. Associated infarction of the well collateralized vertebral bodies is a sign of more extensive vascular compromise (Yuh et al. 1992). 4.3.3.4  Venous Infarction Venous infarction of the spinal cord is usually secondary to arteriovenous malformation, which is discussed later. Outside of this setting, venous infarction is less common but still important to consider. Venous infarction may be embolic, hemorrhagic, or nonhemorrhagic (Kim et al. 1984). Venous infarction is more likely than is arterial infarction to undergo hemorrhagic transformation. Hemorrhagic venous infarction has a poor prognosis. Hemorrhagic and embolic infarcts usually demonstrate a sudden clinical onset and rapid evolution, whereas nonhemorrhagic infarcts may have a slower course and longer survival period (Kim et al. 1984; Niino et al. 1999). Venous infarcts have been associated with thrombophlebitis, decompression sickness, lower extremity thrombosis, pulmonary emboli, polycythemia, and leukemia among others (Niino et al. 1999).

4.3  Intramedullary Diseases of the Spinal Cord

4.3.3.5  Differential Diagnosis Intramedullary T2 hyperintensity and an expansile process can be nonspecific findings. Clinical correlation with a presentation of sudden onset of characteristic neurologic deficits and clinical suspicion for spinal cord infarction are of paramount importance in making the diagnosis. Other diagnoses can often be revealed or excluded based on clinical presentation, clinical course, and laboratory analysis of CSF. If the signal abnormality is not in a characteristic gray matter distribution, the imaging diagnosis can be more challenging. Multiple sclerosis plaques will be more often peripheral as they involve white matter, are much more likely to be multiple, and will typically demonstrate a waxing and waning clinical and imaging course (Suzuki et al. 1998). Both infarction and multiple sclerosis can demonstrate enhancement at certain stages, just as in the brain. The differential diagnosis also includes myelitis, demyelinating processes and neoplasm. Neoplasm may sometimes have a similar appearance to infarction, but usually demonstrates more enhancement. Neoplasm will not typically present with an abrupt onset of symptoms. T2 hyperintensity in the distribution of gray matter can also be seen in poliomyelitis; however, this too can often be distinguished on clinical grounds. 4.3.3.6  Vascular Malformations 4.3.3.6.1  Arteriovenous Malformations Arteriovenous malformations (AVMs) have become much better understood in the past 20 years, with advances in angiography and MRI. Many different classification schemes have been proposed over the years. We will use the classification system described by Anson and Spetzler, which has been widely accepted and is used in much of the recent radiological literature (Table 4.3.2). This system is based on the angioarchitecture of the malformation as well as implications for treatment. It is important to be familiar with the MRI findings of AVMs, as early diagnosis and treatment can reverse the course of myelopathy in many cases. The classification scheme is divided into four main categories with a few subcategories. The term arteriovenous malformations is used here to describe both arteriovenous fistulas (types I and IV) and true glomus type malformations (types II and III). Type I AVMs are dural AV fistulas, whereas type IV AVMs are perimedullary intradural AV fistulas. Type II AVMs are glomus type malformations with a compact intramedullary nidus. Type III AVMs are called juvenile malformations and have a larger less compact nidus. They involve the intramedullary, intradural, and sometimes extradural spaces often

including the vertebral body. The Anson and Spetzler classification scheme did not take into account metameric malformations, which involve all the tissues of a single metamere, as they do not necessarily share the same arterial supply and venous drainage as the other spinal AVMs (Anson and Spetzler 1992). Type I Arteriovenous Malformations Epidemiology Type I AVMs, also known as dural AV fistulas, are the most common type. These account for up to 80% of spinal AVMs. They typically occur in the lower thoracic or conus medullaris region. They are thought to be gravity dependent, perhaps explaining their typical location. Rarely do type I AVMs occur in the cervical region, some of which may have intracranial venous drainage (Bemporad and Sze 2000). Conversely, some cranial AVMs have venous drainage to the cervical spinal region, and may result in cervical myelopathy. These lesions tend to occur in males between their fifth and eighth decades. One series of 66 patients with dural AVFs found an average age of 62 years and a male predominance of 3.4:1 (Gilbertson et al. 1995). Presentation, Pathophysiology, and Sequelae Patients typically present with myelopathy and chronic progressive neurological deterioration. Common presenting symptoms include weakness, numbness, and pain (Gilbertson et al. 1995). Symptoms may be exacerbated by certain postures, exercise, and pregnancy. These effects may be due to mechanically impaired venous return and gravitational forces resulting in increased venous pressure (Aminoff et al. 1974). Up to 80% of patients eventually develop bowel, bladder, or sexual dysfunction (Bemporad and Sze 2000). Symptoms are largely the result of venous hypertension and resulting chronic venous ischemia. There are no valves in the intradural venous system. This allows for easy transmission of backpressure from the arterialized draining veins to the coronal venous plexus, and subsequently to the intramedullary veins and the spinal cord. This transmitted backpressure reduces the arteriovenous pressure gradient, decreasing arterial intramedullary flow and causing chronic ischemia. Aminoff et al. (1974) found ischemic changes in the spinal cord corresponding to the distribution of the draining intramedullary veins. They also noted sparing of the anteromedial regions, attributed to their separate venous drainage. Because of the progressive clinical deterioration of untreated patients, early diagnosis and intervention is of paramount importance. Foix-Alajouanine syndrome, first described in 1926, is a subacute necrotizing myelopathy associated with type I AV fistulas. Findings of necrosis, fibrosis, and progressive destruction of the spinal cord were initially described. These were later postulated to be the sequelae of throm-

631

632

4  Spine and Spinal Canal Table 4.3.2  Classification and common features of spinal arteriovenous malformations (Anson and Spetzler 1992; Bemporad and Sze 2000)   Type

  Location

Fistula vs. nidus type malformation

 Common vertebral levels

  Age (decade)

 Gender predilection

  Presentation

I

Nerve root sleeve

Fistula

Thoracolumbar

5th and beyond

Male

Slowly progressive

II

Intramedullary

Nidus type malformation

Slightly more common in cervical region

3rd to 5th

None

Rapidly progressive

III

Intramedullary and extramedullary

Nidus type malformation

Entire spine

1st to 3rd

None

Variable

IV

Perimedullary

Fistula

Conus medullaris and thoracic

2nd to 5th

None

Progressive conus medullaris and cauda equine syndrome

bosed spinal veins. Thrombosis is not common, however, on angiography or on pathological analysis. This syndrome is likely the result of chronic venous ischemia or possibly recurrent subarachnoid hemorrhage with local arachnoid fibrosis. This myelopathy is usually found in the thoracic cord where the arterial supply is most sparse, with greater involvement of the gray matter (Criscuolo et al. 1989; Mishra and Kaw 2005). Angioarchitecture Type I AVMs are dural arteriovenous fistulas which typically occur in the region of the dural root sleeve. A feeding radicular artery enters the dural root sleeve dorsolaterally and fistulizes to a medullary vein. The vein becomes arterialized under the high pressures, although it has low flow. The vein drains intradurally connecting with and transmitting the increased pressures to the coronal venous plexus (Anson and Spetzler 1992). This results in dilated veins that often extend cranially and caudally from the site of the fistula along the dorsal aspect of the cord. Type I AVMs frequently have a single feeding artery (called type I-A by Anson and Spetzler), but may also have multiple feeders (type I-B). It is important to recognize additional vessels for treatment purposes (Anson and Spetzler 1992). Imaging Findings While angiography remains the most sensitive and detailed imaging modality for arteriovenous malformations, MRI has become extremely useful as a noninvasive way of identifying the presence of an arteriovenous malformation and for guiding angiography and therapy of these lesions. Gilbertson et al. (1995) demonstrated MRI to be as sensitive as and more specific than my-

elography in identifying AVMs. Saraf Lavi et al. (2002) demonstrated a sensitivity of 85–90% and a specificity of 92–100% in the detection of dural arteriovenous fistula. As type 1 AVMs are by far the most common and most studied class, we spend more time discussing their imaging characteristics. MRI findings in type I AVMs include hyperintense T2 signal and hypointense T1signal within the spinal cord parenchyma, expansion of the spinal cord, and cord enhancement after contrast administration. The dilated veins are seen as serpiginous flow voids on T2- and T1-weighted imaging and enhancing dilated vessels after contrast administration. Scalloping of the spinal cord contour can also be noted as a result of the dilated venous system (Bemporad and Sze 2000) (Figs. 4.3.17, 4.3.18, 4.3.19). Fig. 4.3.17  Dural AV fistula. Sagittal post-contrast T1-weighted image demonstrates prominent enhancing intradural vessels which are more numerous posteriorly. An arrow demonstrates one prominent anterior vessel as an example

4.3  Intramedullary Diseases of the Spinal Cord Fig. 4.3.18  Cranial dural AV fistula resulting in spinal cord congestion and myelopathy. Sagittal (a) and axial (d) T2-weighted images demonstrate a long segment of prominent spinal cord hyperintensity extending from the midcervical to the midthoracic region. Numerous flow voids (arrowheads) are seen posterior to the cord. The abnormalities are difficult to detect on the non-contrast sagittal T1-weighted image (b). After contrast administration (c) a few enhancing vessels can be seen posterior to the cord. Angiography (e) revealed that the AV fistula was actually intracranial. e Occipital artery (white arrowhead) selective injection with reflux into the ascending pharyngeal artery (long black arrow) which supplies a dural AV fistula (black arrowhead). A short black arrow demonstrates the draining vein. White arrows indicate the dilated perimedullary veins within the cervical spinal canal, resulting in the spinal cord venous congestion seen on MR

Fig. 4.3.19  Thoracolumbar dural AV fistula. T2weighted (a,e) and STIR (b) images demonstrate parenchymal hyperintensity involving the lower thoracic cord and conus medullaris. A few mildly prominent vessels can be seen anterior to the cord on STIR, T1-weighted (c) and post-contrast T1-weighted images (d,f) (arrows denote vessels on images d and f)

633

634

4  Spine and Spinal Canal

The most sensitive MRI finding is hyperintense signal within the cord parenchyma on T2-weighted imaging. A number of series demonstrated this finding to be present in all cases (Bemporad and Sze 2000; Gilbertson et al. 1995). This increased signal is the result of edema related to venous congestion and chronic ischemia. The absence of T2 prolongation within the cord makes the presence of a type I AVM very unlikely. Decreased signal on T1 imaging is less reliably seen. Venous congestion also can result in cord enlargement, which is seen in a majority of cases. As a result of the chronic venous ischemia, over time there is a breakdown of the blood–spinal cord barrier. As a result, patchy contrast enhancement can be seen, often in the area of abnormal T2 signal (Fig. 4.3.20). Gilbertson et al. found this in 9/25 cases, but others have noted this finding more often (Bemporad and Sze 2000; Gilbertson et al. 1995). Prominent flow voids and serpentine enhancement are also noted in a majority of cases, and are a more specific finding. These are mostly seen posterior to the cord and represent the dilated coronal venous plexus and posterior median spinal vein (Fig. 4.3.17, Costigan and Winkelman 1985; Crum et al. 2000). Saraf-Lavi and Bowen et al. (2002) found that flow voids and serpentine enhancement which extended over three contiguous vertebral segments were strongly associated with the presence of a dural AVF, while the absence of flow voids had the strongest association with the normal control cases. They found the addition of MRA did not increase the sensitivity of MRI for detecting AVFs, but did significantly improve the ability to predict the vertebral level of the fistula. The dilated venous plexus may scallop the surface of the spinal cord, giving the cord a serrated contour on T1 images.

After treatment of spinal dural arteriovenous fistulas, most of the MRI findings above described either resolve or diminish. In a study of 10 cases, Willinsky et al. (1995) found spinal cord swelling to resolve in nine patients, parenchymal T2 hyperintensity to resolve in six patients and decrease in three, and prominent subarachnoid vessels to resolve in seven patients and decrease in three. No findings correlated to clinical outcome, but persistence of abnormal findings indicated failure of treatment. Enhancement of the cord may persist as a result of irreversible breakdown of the blood–cord barrier from chronic ischemia, even if T2 signal abnormality resolves (Bemporad and Sze 2000; Gilbertson et al. 1995). Type II Arteriovenous Malformations Type II AVMs are intramedullary glomus type malformations. They have a compact nidus within the spinal cord parenchyma without a normal intervening capillary bed. They may extend to the pial surface or may be partially exophytic. They can occur in any part of the spine, but may be more common in the cervicothoracic region (Anson and Spetzler 1992). Epidemiology Type II malformations have no gender predilection and typically occur from the third to fifth decades. Associations with Osler-Weber-Rendu syndrome and KlippelTrenaunay-Weber syndrome have been described (Bemporad and Sze 2000). Clinical Presentation Unlike the dural AVFs discussed earlier, type II AVMs tend to have a much more sudden onset of symptoms and tend to present at a younger age. Symptoms often occur as a result of subarachnoid or intramedullary hemorrhage

Fig. 4.3.20  Enhancement in a thoracolumbar dural AV fistula. Precontrast sagittal T1-weighted images (a,b) do not demonstrate any obvious abnormality. Post-contrast imaging (c,d) shows a long segment of enhancement in the anterior aspect of the cord (arrows). Enhancing vessels (arrowheads) can also be seen posterior to the cord in this region

4.3  Intramedullary Diseases of the Spinal Cord

(Anson and Spetzler 1992). Neurologic symptoms may be accompanied by acute back pain. A more gradual presentation may be the result of vascular steal away from normal cord parenchyma resulting in chronic ischemia. Venous hypertension and cord compression from spinal aneurysms or venous varices may also result in a more gradual onset of symptoms. As with type I AVMs, symptoms may worsen with pregnancy (Bemporad and Sze 2000). Angioarchitecture The arterial supply is from the anterior spinal artery more often than from the posterior spinal artery, but can be from either. There are usually multiple feeders. Venous drainage is via intramedullary veins and the coronal venous plexus. Type II AVMs have high pressure and high flow, with rapid filling and early venous drainage demonstrated on angiography (Anson and Spetzler 1992). The high pressure predisposes to aneurysm formation, which is seen in up to 50% of cases. High pressures also result in pressurized venous varices which can compress the cord (Bemporad and Sze 2000).

Fig. 4.3.21  Intramedullary AVM. Non-contrast sagittal (a,b) and axial (c,d) T1-weighted images demonstrate an intramedullary nidus of flow voids (arrow) which focally expands the spinal cord. Prominent flow voids can also be seen anterior and posterior to the cord, cranial to the nidus. Large intramedullary flow void seen at the C4 level on image (b) likely represents a dilated intramedullary vein

Imaging Findings The most characteristic and direct MRI finding is an intra­ medullary tangle of flow voids, representing the nidus (Figs. 4.2.21, 4.2.22). The cord is often focally expanded in this region. As with type 1 AVMs, hyperintense signal can be seen within the cord parenchyma on T2-weighted imaging. This is usually seen in the region of the nidus and may represent edema, cord infarction, or gliosis. As a result of these chronic processes, spinal cord atrophy is often seen (Bemporad and Sze 2000). Dilated flow voids and enhancing vessels in the intradural space represent the dilated coronal venous plexus and draining medullary veins. These are often more prominent posteriorly. As these are under high pressure, scalloping or even compression of the spinal cord may be noted.

Fig. 4.3.22  Intramedullary AVM. Sagittal T2-weighted images (a,b) demonstrate an intramedullary nidus of abnormal vessels (flow voids), which expand the cord. On image (a), a small amount of intramedullary T2 hyperintensity (arrow) is present, best seen at the superior aspect of the nidus. Sagittal T1-weighted images (c,d) demonstrate a small amount of hyperintense signal (d) compatible with cord hemorrhage (arrow). The hemorrhage is seen as dephasing artifact (dark signal) (arrow) on the sagittal T2* gradient-echo image.

635

636

4  Spine and Spinal Canal

Hemosiderin from prior hemorrhage is often seen as hypointense signal, particularly on T2-weighted imaging. T2* GRE sequences are very sensitive for detection of blood products within the cord, and can be performed in sagittal or axial planes (Fig. 4.2.22). Subarachnoid hemorrhage can be difficult to visualize within the spinal canal. Type III Arteriovenous Malformations Type III malformations are also usually intramedullary lesions, but often have extramedullary and even extradural components. They are also called juvenile arteriovenous malformations. It is not uncommon for these lesions to involve the vertebral body and surrounding soft tissues. They can occur at any level in the spine (Anson and Spetzler 1992). Although not included in Anson and Spetzler’s classification scheme, metameric arteriovenous malformations deserve mention in this section. These are extensive malformations that involve all the tissues of a single metamere, including intraspinal contents, bone, soft tissue and musculature and even skin. When skin involvement is present, this is known as Cobb’s syndrome. These rare lesions have different treatment strategies as they do not share the same supply and drainage as the true spinal AVMs (Bemporad and Sze 2000). Clinical Presentation Type III AVMs are rare congenital lesions, comprising 7% of spinal AVMs. There is no gender predilection. These lesions typically present with symptoms in patients less than 30 years of age. Symptoms may be either acute as a result of subarachnoid hemorrhage or hematomyelia, or chronic and progressive as a result of ischemia. Ischemia may be arterial as a result of arterial steal away from cord parenchyma, or venous ischemia as a result of venous hyper­tension. This backpressure and hypertension may be exacerbated by venous thrombosis (Anson and Spetzler 1992; Bemporad and Sze 2000). Angioarchitecture There are usually multiple arterial feeders. Supplying arteries can arise from the anterior spinal artery, the posterior spinal artery, or the vasacorona. These are high-flow malformations. Venous drainage is through intramedullary veins, the coronal venous plexus, extramedullary veins, and the epidural venous plexus. Treatment is very difficult and the goal is often palliation. Imaging Findings Imaging findings are very similar to those of type II AVMs, as they are both intramedullary malformations, but type III AVMs are more extensive. They have a less compact nidus than the type II lesions and usually have an extramedullary component. MRI is very useful in differentiating between the two lesions in a noninvasive way, and in determining the extent of involvement of the

extradural tissues. Findings include a tangle of flow voids or enhancing vessels which involve the cord and extend into surrounding tissues, enhancing hyperintense T2 signal within the cord, cord atrophy, and evidence of prior hemorrhage within the cord parenchyma. Type IV Arteriovenous Malformations Type IV malformations are intradural perimedullary arteriovenous fistulas. They are typically located anterior or anterolateral to the spinal cord and are most commonly encountered in the region of the conus medullaris (Anson and Spetzler 1992). They are divided into three subclassifications, which have implications for treatment options. Epidemiology Type IV AVMs are found in patients between their second and fifth decades. There is no gender predilection. They comprise 10–20% of spinal AVMs. These are generally congenital malformations, but can occasionally be posttraumatic. An association with Osler-Weber-Rendu syndrome has been described (Andersson et al. 2003; Bemporad and Sze 2000). Clinical Presentation Symptoms are usually attributable to progressive ascending myelopathy and sphincter dysfunction. This is usually the result of venous hypertension, particularly with the small low flow type fistulas with a single feeder. In larger fistulas with multiple feeders, higher blood flow and increasing venous dilatation can also cause symptoms by vascular steal or cord compression, respectively. Subarachnoid hemorrhage can occasionally cause acute symptoms in these larger fistulas (Andersson et al. 2003; Bemporad and Sze 2000). Angioarchitecture The arterial supply is usually from the anterior spinal artery, although it can less commonly be from the posterior spinal artery. The feeding artery fistulizes to a perimedullary vein, which becomes dilated. These lesions are true fistulas without an intervening vascular network. Type IV malformations are subdivided into subtypes A, B and C, based on number of arterial feeders, degree of venous dilatation, and flow. These subclassifications are clinically useful, as they have implications for treatment. The arterial supply in type IV-A malformations is a single vessel, often the artery of Adamkiewicz. These lesions have slow flow and moderate venous dilatation. These are small fistulas which are typically treated surgically (Anson and Spetzler 1992). Type IV-B AVMs may have multiple feeding vessels and rapid flow. These are intermediate sized lesions which can be treated with embolization, surgery or a combination of both (Anson and Spetzler 1992) Unlike types A and C, which are more common in the conus medullaris region, type B malformations are more common in the thoracic region (Bem-

4.3  Intramedullary Diseases of the Spinal Cord

porad and Sze 2000). Type IV-C AVMs are giant fistulas with multiple feeders and very high flow. Because of their size and surgical risk, these lesions are treated endovascularly (Anson and Spetzler 1992). Imaging Findings MRI findings include hyperintense signal within the spinal cord and enlarged tortuous perimedullary flow voids and enhancing vessels. As with type I AV fistulas, the parenchymal T2 prolongation represents venous congestion (Andersson et al. 2003). In type IV-A fistulas, perimedullary vessels may be variably seen due to their small size. In type IV-C fistulas, the perimedullary flow voids,

and enhancing vessels are quite prominent and tortuous. Other MRI findings are similar to those found in type I AVMs, as described above (Figs. 4.2.23, 4.2.24). Magnetic Resonance Angiography in Evaluation of AVM MR angiography of the spinal cord has been evolving in its technique over the last decade and is increasingly used in conjunction with MRI in evaluation of arteriovenous malformations. The prevailing technique is 3D contrast-enhanced GRE imaging (Pattany et al. 2003). Bowen et al. demonstrated that only the largest vessels are seen on the MIP images, typically the normal anterior

Fig. 4.3.23  Type IV AVM (perimedullary fistula) in a 10-year-old male. Sagittal (a,b) T2-weighted images, coronal (c) T2-weighted image, and sagittal T1-weighted image (d) demonstrate serpiginous intradural flow voids (white arrow) as well as a large intramedullary venous varix (black arrow) which expands the conus medullaris in this surgically proven type IV AVM. The intramedullary varix is well demonstrated on axial T1-weighted (g) and proton density–weighted (h) images. Note the hyperintense intramedullary signal superior to the varix on sagittal (a) and axial (f) T2-weighted imaging. Axial T2-weighted images (f, i) also demonstrate prominent intradural flow voids (arrows). Post-contrast T1-weighted image (e) demonstrates heterogeneous enhancement within the large varix and a prominent pulsation artifact at this level. Gadolinium-enhanced spinal MRA (j) shows an enlarged artery of Adamkiewicz (arrowhead) leading to the point of fistulization in the region of the conus, where the large venous varix (black arrow) and a prominent draining intradural vein (white arrow) are seen

637

638

4  Spine and Spinal Canal

Determining the level of a malformation noninvasively can be very helpful, as digital subtraction (DS) angiography can be applied using a targeted approach. After successful surgical or endovascular treatment, the abnormal appearance of intradural vessels often resolves. This usually accompanies resolution of T2 hyperintensity of the cord parenchyma. The persistence of large tortuous vessels has been associated with failure of treatment (Pattany et al. 2003). Mascalchi et al. (2001) found MR angiography to be more sensitive than MRI in depicting flow in perimedullary or intramedullary vessels, indicating failure of treatment. Seven patients with dural AV fistula who demonstrated residual flow after embolization were all subsequently confirmed to have treatment failure on DS angiography. 4.3.3.6.2  Cavernous Malformations

Fig. 4.3.24  Type IV AVM in a patient with Klippel-TrenaunyWeber syndrome. Sagittal T2-weighted images (a–c) and sagittal T1-weighted image (d) demonstrate innumerable intradural flow voids extending from the mid thoracic region to the cauda equina. On an axial T1-weighted image (e), abnormal intradural vessels are seen pushing the cord to the right. Spinal MRA coronal MIP image (f) demonstrates numerous abnormal predominantly left sided intradural vessels extending into multiple left sided neural foramina. At angiography, at least 4 anterior spinal arterial feeders were present

and posterior median spinal veins. Veins are preferentially demonstrated likely due to their larger size (Bowen et al. 1996). The major spinal arteries are generally poorly demonstrated, although the artery of Adamkiewicz was successfully demonstrated prior to aortic surgery in 66.7% of patients in a series of 30 patients by Yoshioka et al. (2003). In the presence of a spinal dural arteriovenous fistula, Saraf-Lavi et al. (2002) demonstrated increased tortuosity, length, size and number or the dominant intradural vessels. In this series, the addition of MRA to conventional MRI improved the ability to predict the vertebral level of the fistula. In intramedullary AVMs, 3D contrastenhanced MRA can depict the nidus and draining veins.

These vascular malformations are also known as cavernous angiomas, cavernomas, and cryptic vascular malformations. These are small discrete round or oval vascular malformations, which may have a lobulated margin. Cavernous malformations have slow flow and no arteriovenous shunting. In this regard they are quite different from AVMs and therefore are not included in Anson and Spetzler’s classification scheme. Cavernous malformations are composed of sinusoidal vascular spaces and dilated capillaries with very slow flow. They have a variable endothelial lining which may be as thin as one cell layer. The immediately surrounding spinal cord parenchyma is typically gliotic and laden with hemosiderin, as a result of prior hemorrhages. On gross pathology, they are purple to red in color and are described as having a “mulberry-” or “raspberry-like” appearance (Krings et al. 2005). Spinal cavernous malformations are pathologically indistinguishable from their intracranial counterparts, and often coexist (Hurst 2001). These lesions range in size from millimeters to a few centimeters (Santoro et al. 2004). They can occur anywhere in the neural axis. These lesions can also occur intradurally and extradurally, however the majority of spinal cavernous malformations are intramedullary. For the purposes of this section, we focus on intramedullary malformations. Epidemiology Spinal cavernous malformations were once considered rare, but as a result of MRI, they have been diagnosed with increasing frequency. Cavernous malformations involving the central nervous system have an estimated incidence of 0.2–4%, but the incidence of spinal cavernous angiomas is not well established (Hurst 2001) Cavernous malformations make up 5–12% of all spinal vascular malformations. They may be multiple in 14.6–25% of cases (Krings et al. 2005; Santoro et al. 2004).

4.3  Intramedullary Diseases of the Spinal Cord

Cavernous malformations are thought to be congenital lesions, with sporadic forms and inherited autosomal dominant forms with variable penetrance. These lesions have also been known to occur de novo after radiation and trauma (Bemporad and Sze 2000) A female predominance of 2:1 has been described, although more recently some authors have suggested that there is little or no female bias (Krings et al. 2005; Santoro et al. 2004). Clinical Presentation Symptomatic presentation is variable and can occur at any age, but is more frequently in the fourth decade (Hurst 2001). Patients can present with acute symptoms as a result of hemorrhage. One review of symptomatic patients suggested a hemorrhage rate of 1.6% per year (Bemporad and Sze 2000). When this happens, hemorrhage can be internal and limited to the lesion, usually enlarging it, or hemorrhage may extend into the surrounding cord parenchyma. Hemorrhage is usually accompanied by sudden back pain, and may be followed by sensorimotor deficits several hours later. Multiple such episodes may cause a stepwise neurological deterioration. Patients may alternatively experience either a rapid or chronic gradual decline after an acute presentation. Patients may also present with only gradual decline and no acute presentation (Bemporad and Sze 2000). Patients usually develop a progressive myelopathy after symptoms begin (Krings et al. 2005). This process can be often be reversed with surgical intervention. Symptomatic cavernous malformations are resected, often with microsurgical techniques, and generally with good results. The treatment of asymptomatic lesions is controversial, but some have advocated resection of enlarging lesions in surgically amenable locations (Santoro et al. 2004). Imaging Findings MRI is the clear imaging modality of choice for cavernous malformations, as these lesions are angiographically occult. The imaging features in the spine are similar to those in the brain. They have a classic imaging appearance, although not pathognomonic. Additionally, after acute hemorrhage the appearance may be much less specific. Cavernous malformations are rounded or ovoid lesions which can vary from a few millimeters to a few centimeters in size. They may have a lobulated margin. They classically have a hypointense peripheral ring on T2-weighted imaging, representing hemosiderin staining of the immediately surrounding tissue (Fig. 4.2.25). These lesions demonstrate a “blooming” phenomenon on T2*-weighted gradient-echo imaging, due to their magnetic susceptibility. This appears as rounded hypointense signal, typically larger than the size of the lesion itself. Small malformations which are not seen at all on other sequences may be revealed on the T2* gradient-echo sequences as punctuate hypointense foci. T2*-weighted im-

Fig. 4.3.25  Cavernous malformation. Sagittal T1-weighted (a), T2-weighted (b), STIR (c), and post-contrast (d) images demonstrate a small lesion in the anterior aspect of the thoracic cord, which demonstrates a complete hemosiderin (dark signal) ring. Note the internal hyperintense signal on all sequences. Axial (e) T2-weighted image also demonstrates the hemosiderin ring (arrow) and central hyperintensity. Note the lack of surrounding edema

aging is an important sequence to perform in the search for multiple lesions and may help in suggesting the diagnosis in atypical cases. Calcification is occasionally present, which may be better seen on CT (Krings et al. 2005). These lesions may enlarge over time, or may enlarge acutely secondary to acute hemorrhage (Fig. 4.2.26). Internally, these lesions often demonstrate heterogeneous hyperintense signal on T2-weighted imaging. Classically the central portion demonstrates a reticulated, stippled or septated internal appearance, with blood products at various stages of evolution. Faint enhance-

639

640

4  Spine and Spinal Canal

Fig. 4.3.26  Cavernous malformation with hemorrhage. Sagittal STIR image (a) and axial T2-weighted images (d,e) demonstrate an intramedullary lesion with a hypointense hemosiderin ring (arrowhead in b, arrow in e) and central hyperintensity. There is also extensive surrounding T2 hyperintensity (arrow in a) representing edema and hemorrhage, and expansion of the cord. On

T1-weighted imaging (b), T1 hyperintensity surrounding the lesion represents methemoglobin. Post-contrast T1-weighted imaging (c) demonstrates a blush of enhancement (arrow) within the lesion, but not within the surrounding areas of edema and hemorrhage

ment may be present. Cavernous malformations may cause focal enlargement of the spinal cord, which may become more pronounced after acute hemorrhage. They may also focally bulge one contour of the cord if they are peripherally located (Santoro et al. 2004) The surrounding cord parenchyma may demonstrate edema (hyperintense T2 signal) or sometimes hemorrhage in various stages of evolution. Hemorrhagic neoplasms such as ependymoma may have a similar appearance, although they typically demonstrate more enhancement. After acute hemorrhage, the imaging appearance may be less characteristic and specific. Hemorrhage and increased surrounding edema may obscure the underlying malformation. In these cases, differentiation from arteriovenous malformation, tumor, or inflammatory lesion may be difficult. Angiography is sometimes performed to exclude a small arteriovenous malformation (Krings et al. 2005).

drocytes in the central nervous system and Schwann cells in the peripheral nervous system, insulates axons. With its high electrical resistance and low capacitance, myelin speeds the propagation of action potentials. Demyelination causes a decrease in conduction velocity or, in severe cases, a complete conduction block, which can be measured by evoked potentials. Chronic demyelination can lead to axonal degeneration and changes in cytoskeletal structure, and subsequent cord atrophy. Clinical remission can occur with remyelination, or an upregulation of sodium channels along the axon. Demyelination within the cord is best assessed with MRI, particularly with T2-weighted sequences. Brain lesions demonstrating T2 prolongation are commonly seen and reflect a multitude of diseases. Accompanying signal abnormalities in the spinal cord are less common, resulting in an increased specificity for demyelinating disease. Unfortunately, there is only a weak correlation between T2 signal abnormalities within the cord and clinical symptomatology (Lycklama à Nijeholt et al 2003). Cord cross-sectional area, which decreases with neuronal loss and cord atrophy, can complement qualitative assessments, and correlates well with disability as measured by Kurtzke’s Expanded Disability Status Scale (Losseff et al. 1996). Magnetization transfer ratios (Bot et al. 2004a) and MR spectroscopy are newer MR techniques that may increase diagnostic specificity and better assess disease progression.

4.3.4  Demyelinating Disease Spinal cord demyelination may result from one of a variety of causes or predisposing factors or may be idiopathic. Known etiologies include multiple sclerosis (MS), parainfectious and post-vaccinal states, systemic autoimmune diseases, paraneoplastic syndromes, trauma, or vascular causes. Myelin, which is produced by oligoden-

4.3  Intramedullary Diseases of the Spinal Cord

4.3.4.1  Multiple Sclerosis 4.3.4.1.1  Epidemiology In the United States, approximately 300,000 people are afflicted with multiple sclerosis (MS). The incidence of MS is 0.5 to 1 per 1,000 people per year, and there is an overall risk of 0.2% of developing MS during one’s lifetime. Nearly 70% of patients present between the ages of 21 and 40, but MS has been diagnosed in patients from ages 3 to 67. Adult women are approximately twice as likely to develop MS as men. Although the precise cause of MS is not known, it is likely due to a combination of genetic and environmental factors. People with class II major histocompatibility complex HLA-DR2, which is commonly found in Caucasians of northern European dissent, are particularly susceptible. Environmental agents, such as viral infections or bacterial antigens, can then trigger various immune responses. The subsequent cascade of immunologic events may lead to different clinical patterns and different responses to therapy. Myelin basic protein and proteolipid protein are thought to be primary targets of aberrant Th1 helper T cells, which can be increased in MS (Mouzaki et al. 2004). The incidence of MS varies widely between different geographic locales. The highest incidence is found in the northernmost latitudes of the northern and southern hemispheres. This environmental influence occurs during the first 15 years of life. For example, there is a higher incidence of MS among Caucasians who emigrate from Great Britain to South Africa, compared with native-born Caucasian South Africans. Clustered cases of MS have also been described, lending further evidence for environmental influences. 4.3.4.1.2  Clinical Information Multiple sclerosis is characterized by inflammation and scarring of the central nervous system, with multiple occurrences that are separated by space (location) and time. MS has been divided into four common clinical patterns, including relapsing remitting, secondary progressive, progressive relapsing, and primary progressive (Lublin and Reingold 1996). The relapsing remitting pattern is characterized by clearly defined relapses, separated by quiescent time periods of full or partial recovery. The primary progressive pattern entails continual worsening of clinical symptoms, albeit at a variable rate progression, with occasional plateaus and possibly brief improvements. The secondary progressive pattern is a sequential combination of the first two. Most patients with relapsing remitting MS convert to steady disease progression, as their baseline neurologic function declines. The fourth pattern, progressive relapsing, is characterized by initial

progressive disease. Progression is exacerbated by distinct relapses, with or without recovery. Rather than intervening plateaus or partial recovery, symptoms worsen between relapses. Other conditions represent a spectrum of demyelinating disorders, and controversy continues today regarding the proper classification of these disorders as MS variants or separate diseases (e.g., neuromyelitis optica). 4.3.4.1.3  Pathology The acute stage of MS is characterized by perivenular inflammation, with infiltration of macrophages and lymphocytes in areas of well-defined demyelination. Inflammation occurs through both cell-mediated and immune complex-mediated pathways. In individual patients with MS, demyelination may be mediated by T cells and macrophages, antibody and complement, or by primary injury to the oligodendrocytes. Fibrillary gliosis accompanies demyelination in chronic multiple sclerosis. Cross-sectional area is reduced in multiple sclerosis, reflecting glial, and neuronal loss. Cord atrophy correlates with clinical disability. Axonal loss can be identified in both spinal lesions and in normal appearing cord tissues, but does not lead directly to cord atrophy. Transverse myelitis occurring in MS is often partial, rather than complete. Lesions tend to occur in the posterior and lateral white matter columns of the spinal cord (Tartaglino et al. 1995). Other variants of MS include Devic’s neuromyelitis optica, tumefactive MS, Schilder’s diffuse sclerosis, Baló’s concentric sclerosis, and Marburg’s variant of multiple sclerosis. Devic’s disease is characterized by antibody– complement-mediated demyelination. Schilder’s diffuse sclerosis is a particularly aggressive form of MS, usually found in childhood. Baló’s concentric sclerosis shows oligodendrocyte dystrophy. Marburg’s acute or fulminant MS involves an aberrant myelin basic protein. 4.3.4.1.4  Imaging Findings MRI is sensitive for detection of spinal cord lesions in MS. Diagnostic criteria for MS specify the number, location, size, and enhancement characteristics of brain lesions (McDonald et al. 2001). However, brain lesions lack specificity, since many diseases can produce white matter lesions with similar appearances. Imaging of the spinal cord has been suggested to increase the diagnostic accuracy for MS (Bot et al. 2004). At autopsy, spinal cord lesions are seen in more than 90% of patients with MS. The cervical cord is involved most commonly. Lesions are typically elongated, longitudinally oriented, and vary in length from a few mm to several vertebral bodies in length. However, they rarely span more

641

642

4  Spine and Spinal Canal

Fig. 4.3.27  Axial T2-weighted (a,b) and axial proton density– weighted (c) images demonstrate bilateral peripherally located expansile hyperintense lesions in this patient with multiple sclerosis. Post-contrast axial image (d) demonstrates peripheral enhancement of the MS plaques bilaterally

than two vertebral bodies. Lesions are often peripheral in location, as they involve primarily white matter tracts (Figs. 4.2.27, 4.2.28). Multiple cord lesions may be present simultaneously. Cord swelling is sometimes seen in acute lesions (Fig. 4.2.29). As in the brain, a solitary expansile, mass-like cord lesion is sometimes termed tumefactive MS (Fig. 4.2.30). Lesion enhancement is variable and can be patchy or ring like. Enhancement usually occurs in the active phase (Figs. 4.2.27, 4.2.29). Hyperintense T2 lesions are seen in about 80% of patients with MS. Conventional T2-weighted images and dual-echo T2 sequences are more sensitive for plaque detection than fast spin-echo T2 (FSE T2) images. Short tau inversion recovery and proton density–weighted sequences may improve sensitivity (Fig. 4.3.1). In the brain, MS plaques are often hypointense to white matter on precontrast T1-weighted images, but this is less helpful in the spinal cord, where lesions are more commonly isointense (Figs. 4.3.1, 4.2.31). Hypointense cord lesions on T1-weighted imaging correlate better with functional disability than brain lesions, possibly because cord lesions are more likely to involve motor Fig. 4.3.28  Sagittal (a) and axial (b) T2-weighted images demonstrate a focal MS plaque (arrow) restricted to the left dorsal column white matter. The same patient had a simultaneous large expansile enhancing plaque in the thoracic cord, as seen in sagittal (c) and axial (d) T2-weighted and pre-contrast (e) and post-contrast (f) sagittal T1-weighted images (arrowhead demonstrates enhancement)

4.3  Intramedullary Diseases of the Spinal Cord Fig. 4.3.29  A solitary mildly expansile lesion is seen in this patient with multiple sclerosis. Axial (d) and sagittal proton density–weighted (b) and sagittal T2-weighted (a) images demonstrate an expansile focus of hyperintense signal at the C7/T1 level. The lesion (arrow) is hypointense on axial precontrast T1-weighted imaging (e), and demonstrates ring enhancement on axial (f) and sagittal (c) post-contrast images

tracts. T1 hypointensity also correlates with cord atrophy. The incidence of new lesions within the cord is one-tenth that of the brain, probably reflecting differences in tissue volume. However, new spinal lesions are more likely to be symptomatic. More recently, newer MRI magnets and coils have improved the resolution of spinal cord imaging, particularly within the cervical cord. Advanced imaging techniques, including T1 relaxation time, 1H-magnetic resonance spectroscopy, magnetization transfer ratio, diffusion tensor imaging, and functional MRI, have also been utilized to study spinal cord lesions in MS (Tartaglia and Arnold 2006; Tench et al. 2005). The goal of these techniques is to find better correlates between imaging and disability, and help identify underlying pathologic processes. In one study, magnetization transfer gradient-echo sequences were 35% more sensitive than FSE- T2, and fast short tau inversion recovery (STIR) sequences were 66% more sensitive than FSE T2 (Rocca et al. 1999). 4.3.4.2  Devic’s Disease Neuromyelitis optica (NMO), or Devic’s disease, is a demyelinating process that affects the optic nerves and

the spinal cord, sparing the brain (Fig. 4.2.32). It can be monophasic, or follow a relapsing and remitting course like MS. Indeed, some patients with NMO fulfill the diagnostic criteria for MS. Nonetheless, recent evidence suggests that NMO is distinct from typical MS (Yoshioka et al. 2003). About 90% of patients with NMO are female. While oligoclonal bands are often found in the cerebrospinal fluid of MS, they are less common in NMO. Unlike MS, cervical lesions with T2 prolongation in NMO often span three or more vertebral segments. Relapses are usually severe. Since there is a partial overlap of symptoms and imaging findings between MS and NMO, over time some cases of presumed NMO will ultimately be diagnosed with MS. The imaging characteristics of NMO frequently differ from multiple sclerosis (Filippi et al. 1999). Many patients with NMO have normal MRI scans of their brains. Those with brain imaging abnormalities (often age related) do not show the progression of these signal abnormalities commonly seen with MS. In the spine, acute changes in NMO often lead to cord swelling, while this is less common in MS. While MS plaques often involve the lateral and posterior portions of the spinal cord, the central and posterior portions are typically involved in NMO (Fig. 4.2.32).

643

644

4  Spine and Spinal Canal

Fig. 4.3.30  Tumefactive multiple sclerosis. Sagittal T2-weighted image demonstrates a single large expansile intramedullary hyper­intense lesion centered at the C3 level in this patient with multiple sclerosis

Fig. 4.3.32  Devic’s disease. Sagittal T2-weighted image (a) demonstrates a long segment of hyperintense signal, which is mildly expansile, in the upper thoracic cord in a patient with Devic’s disease. Axial post-contrast T1-weighted image (b) demonstrates an area of solid enhancement within the T2 signal abnormality. Sagittal STIR (c) and axial T2-weighted (d) images obtained one month later show increased extent of the hyperintense signal and increased spinal cord expansion. Multiple serial brain MRIs of the patient over many years were negative Fig. 4.3.31  This patient’s sagittal T1-weighted image (a) of the cervical spinal cord looked essentially normal, but the sagittal T2-weighted image (b) revealed numerous patchy areas of signal abnormalit

4.3  Intramedullary Diseases of the Spinal Cord

4.3.4.3  Acute Disseminated Encephalomyelitis Acute disseminated encephalomyelitis (ADEM) is a monophasic demyelinating process of the central nervous system. It is an immune-mediated reaction that is seen after a prodromal infection or a vaccination. The cerebrum, cerebellum, brainstem, optic nerves, and spinal cord may all be involved in the disease process. 4.3.4.3.1  Epidemiology The exact incidence of ADEM is unknown. Spinal cord involvement in ADEM is less common than is brain involvement. Concomitant brain lesions are usually present, but isolated cord involvement can also be seen. This entity is seen mostly in children or young adults (Khong et al. 2002). There is no clear gender predilection (Singh et al. 1999). ADEM is more prevalent in tropical countries, particularly in India. In these countries, multiple sclerosis (a common differential diagnosis) is significantly less common than in colder climates (Singh et al. 1999, 2000). 4.3.4.3.2  Pathology ADEM is thought to be an autoimmune-mediated inflammatory demyelinating process. An antibody–antigen reaction to myelin is triggered by an infection or vaccination. Common pathogens in preceding infections include mycoplasma pneumoniae and viral infections. Nowadays a preceding viral infection is often nonspecific, but formerly included measles, mumps, varicella, and herpes, among others (Garg 2003). Pathologic findings demonstrate perivenous inflammation and demyelination. There is infiltration of vessel walls and perivascular cuffing with lymphocytes, macrophages, and edema. Vessels of both white matter and gray matter are affected (Feydy et al. 1997 Garg 2003). 4.3.4.3.3  Clinical Presentation and Course Symptoms typically begin 1–3 weeks following a systemic infection (often an upper respiratory illness) or a vaccination. In some cases, no clear inciting event is identified (Caldemeyer et al. 1994; Garg 2003). There is usually a rapid onset of symptoms, over the course of a few days (Singh et al. 1999). Symptoms usually last for 2–4 weeks, but can occasionally resolve in a few days or up to a few months. ADEM is classically a monophasic condition, but recurrences have been reported, in which case reevaluation for chronic processes like multiple sclerosis may be warranted. Recurrences within a few months of presentation are thought to be part of the same episode. This

has been referred to as a multiphasic variant of ADEM (Caldemeyer et al. 1994; Garg 2003). Symptoms can be very variable, ranging from subclinical to seizures, coma, and rarely death (Khong et al. 2002). Spinal cord involvement is suspected when the patient has myelopathic symptoms. These commonly include areflexia, lower extremity weakness, or urinary retention (Singh et al. 1999). Other common presenting symptoms in ADEM involving the brain and optic nerves include headache, change in mental status, aphasia, dysarthria, ataxia, seizures, and visual symptoms (Khong et al. 2002). The prognosis for ADEM is generally favorable, often with little or no permanent neurological sequelae after resolution of the acute episode. There is, however, a wide spectrum of clinical courses. Permanent deficits and death have been reported (Khong et al. 2002; Singh et al. 1999). Treatment includes corticosteroids, immunoglobulin therapy, and plasmapheresis, which often result in dramatic improvement (Marchioni et al. 2005). Diagnosis is made by a combination of clinical presentation, CSF analysis, electrophysiological studies, evoked potential studies, and MR imaging (Singh et al. 1999). The patient should have negative CSF titers and cultures to exclude acute infection when considering the diagnosis. Distinction from an initial episode of multiple sclerosis may be difficult on both clinical and imaging grounds. Indeed, some patients thought to have ADEM will go on to develop multiple sclerosis. Some believe that these two diseases are part of the same spectrum. 4.3.4.3.4  Imaging Findings MR is the imaging modality of choice and has a high sensitivity for lesion detection. There are, however, uncommon cases where the clinical diagnosis is made but MRI does not demonstrate an abnormality (Marchioni et al. 2005). MRI is also useful in excluding other diagnoses, in particular evaluating for the presence of characteristic brain lesions of multiple sclerosis. Overall, MRI has contributed to the increased diagnosis of ADEM by its sensitive detection of lesions (Caldemeyer et al. 1995). ADEM may affect the brain and spinal cord in a widespread manner, or may be isolated to a focal process involving a single lesion in the cord, brain, or involvement of the optic nerve (Singh et al. 1999). Spinal cord involvement may be quite extensive with some cases involving the entire length of the spinal cord in a confluent fashion (Fig. 4.2.33). Alternatively there may be multiple lesions at different levels of the cord or the process may be limited to a single short-segment focus. In the case of a solitary tumefactive lesion, biopsy is sometimes performed to exclude a neoplasm. Spinal cord lesions in ADEM are highly variable in appearance, number, and size. White matter is classically involved more than is gray matter, although this is not

645

646

4  Spine and Spinal Canal

always the case. The lesions are best seen on long TR sequences. Lesions are hyperintense on T2-weighted and proton density–weighted images. T1-weighted images may demonstrate variable hypointensity or may demonstrate no obvious abnormality. Cord swelling may be present and can be focal or diffuse (Fig. 4.2.33). Hemorrhage is rare but occasionally can be seen. Contrast enhancement can be seen in some lesions as a result of breakdown of the blood–spinal cord barrier. Patterns of enhancement are not specific and can be solid, patchy, or ring-like. Pial enhancement can rarely be present (Singh et al. 1999) Lesion enhancement is more common in acute lesions and often resolves with treatment (Caldemeyer et al. 1994; Feydy et al. 1997). The imaging characteristics of ADEM are not specific and by themselves can be indistinguishable from those of multiple sclerosis. Although there is significant overlap in the imaging appearances of these two entities, lesions of ADEM tend to be larger, extend over a longer segment of the cord, and have less defined margins than those of multiple sclerosis. ADEM more commonly involves the thoracic cord, whereas cervical involvement is more common in multiple sclerosis. Both processes, however, can be seen anywhere along the spinal cord (Singh et al. 2000) Follow-up imaging can help confirm diagnosis, as new lesions are not expected in ADEM (Feydy et al. 1997). Although the extent and severity of findings has not been found to correlate with clinical course or outcome, MRI is particularly useful for following patients (Khong et al. 2002; Marchioni et al. 2005). Lesions of ADEM markedly improve with therapy. Improvement on MR imaging often follows clinical improvement (Caldemeyer et al. 1994; Singh et al. 1999). Follow-up imaging generally demonstrates resolution of enhancement and cord swelling, and improvement of signal abnormality. Some residual cord signal abnormality may be present, although complete resolution is not uncommon. Subsequent cord atrophy has been reported but is not typical (Singh et al. 1999). 4.3.4.4  Transverse Myelitis Fig. 4.3.33  ADEM. This 7-year-old male with a recent prior viral infection developed leg weakness, ataxia, inability to ambulate, and urinary retention. Sagittal T1-weighted image of the cervical and upper thoracic spine (a) demonstrates mild cord hypointensity and cord expansion. The abnormality was present throughout the entire spinal cord. Sagittal T2-weighted (b,d) and proton density–weighted (c) images demonstrate diffuse hyperintense signal within the cord. Sagittal post-contrast T1-weighted imaging (e) shows no enhancement. The patient recovered clinically and follow-up MRI (sagittal T2-weighted) (f,g) obtained 6 months later demonstrates no residual abnormality in this patient with an episode of ADEM

Acute transverse myelitis (ATM) is a monophasic monofocal episode of acute spinal cord inflammation with both motor and sensory pathways affected below the level of involvement. Frequently encountered presenting symptoms include leg/arm weakness, pain, sensory symptoms, and bowel/bladder dysfunction. Transverse myelitis is more common in middle aged adults (Choi et al. 1006). The incidence has been estimated between 1.34 and 4.6 new cases per 1,000,000 per year (Scotti and Gerevini 2001). Prognosis is variable; some make a full recovery within a few months, while a smaller percentage have poor outcomes including some reported deaths (Savvas et al. 2003).

4.3  Intramedullary Diseases of the Spinal Cord

Transverse myelitis is not well understood, and has many causes including post-infectious and post-vaccinal. ATM may also be associated with multiple sclerosis (a first presentation) or with autoimmune disease, but is most often idiopathic (Choi et al. 1996; Scotti et al. 2001). There is also an acute necrotizing form of transverse myelitis. Pathologic analysis in ATM typically demonstrates perivenular demyelination and inflammation, although necrosis involving both gray and white matter can also be seen (Choi et al. 1996; Savvas et al. 2003). Transverse myelitis is often a diagnosis of exclusion, and some have used it as a wastebasket term for myelopathy of a yet unidentified cause. Some cases which are initially called transverse myelitis are later discovered to be due to multiple sclerosis, vasculitis, viral myelitis, ischemia, vascular malformation, and other diagnoses. 4.3.4.4.1 

Imaging Findings

Imaging findings demonstrate significant overlap with those of MS and ADEM and can be nonspecific. A possible distinguishing factor in some cases is that transverse myelitis is a monofocal spinal cord lesion. The thoracic cord is most commonly involved in transverse myelitis, whereas the cervical cord is more often involved in MS. Lesions are hyperintense on T2-weighted imaging and iso to hypointense on T1-weighted imaging, and often extend over a few vertebral segments (although much longer segment involvement can also be seen). Usually both halves of the cord are involved, including both sensory and motor tracts (Choi et al. 1996; Savvas et al. 2003) (Fig. 4.2.34). In one series, the T2 signal hyperintensity in most cases was well marginated, centrally located, and occupied greater than two thirds of the cross-sectional area of the cord (Choi et al. 1996). There may be focal cord swelling in the area of signal ab-

normality. Enhancement is variable and can be patchy, nodular, or diffuse. The enhancement often occupies a relatively small portion of the area of T2 signal abnormality (Choi et al. 1996; Savvas et al. 2003) (Fig. 4.2.34). The extent of the imaging abnormalities have not been shown to correlate with clinical severity or outcome (Savvas et al. 2003). 4.3.5  Radiation Myelopathy Radiation myelopathy, also called radiation myelitis, is a rare but very serious consequence of therapeutic irradiation. This entity remains a diagnosis of exclusion, as findings are nonspecific and primary pathology must be the first consideration. Pallis et al. (2000) described three criteria for the diagnosis of radiation myelopathy: the spinal cord was included in the radiation field, the neurologic signs correspond to the segment of irradiated spinal cord, and lastly, other causes of myelopathy must have been excluded. The CSF should be normal and there should be no evidence of metastatic disease (Wang et al. 1992). History and time course of symptoms are important in making the diagnosis. Radiation exposure to the spinal cord occurs when either the spine itself is being treated or when neighboring tissues are treated and inclusion of the spinal cord in the radiation field is unavoidable. Of patients undergoing radiation therapy for nasopharyngeal carcinoma, 1–10% will develop radiation myelopathy involving the cervical spinal cord. Wang et al. found that the incidence of radiation myelopathy correlates positively with the total radiation dose, dose per fraction and length of spinal cord irradiated (Wang et al. 1992). They reported an incidence of 5% with doses of 57–61 Gy and 50% with doses of 68–73 Gy. There also is increased risk with shorter treatment times (Zweig and Russell 1990). Fig. 4.3.34  Transverse myelitis. This patient with a monoepisodic focal myelopathy was presumed to have transverse myelitis. The abnormality is difficult to detect on a T1-weighted sagittal image (a). Sagittal T2-weighted (b) and proton density–weighted (c) images demonstrate a focal hyperintense cord lesion at the C3 level. Axial T2-weighted image (d) shows that the lesion is centrally located and occupies a majority of the cord cross-sectional area, including both right and left sides. An axial post-contrast image (e) reveals a small focus of enhancement within the lesion posteriorly

647

648

4  Spine and Spinal Canal

4.3.5.1  Clinical Presentation Early, or transient, radiation myelopathy is generally a self-limited clinical entity that often manifests as intermittent electric shock-like sensations radiating from the neck to the extremities. When this phenomenon is brought about by neck flexion, it is referred to as Lhermitte’s sign. This may begin 6 weeks to 6 months after radiation treatment has started, and resolves 2–9 months later. This is thought to be a result of transient demyelination, and these patients have spontaneous complete recovery (Rampling and Symonds 1998). Delayed radiation myelopathy, or chronic progressive radiation myelopathy, is a much more serious condition that occurs months to years after radiation. This is the entity we are primarily concerned with in MR imaging, and will concentrate on from here on. There is often a subacute onset of symptoms following a long, asymptomatic period (Wang and Shen 1991). Latency periods between radiotherapy and onset of symptoms are variable, but generally follow a bimodal distribution. There is an early latency peak at 12–14 months, and the late latency peak is at 24–28 months. Wang et al. (1992) used a cutoff of 18 months to distinguish early and late latency periods. The late latency period can be up to many years. Clinical manifestations include sensory deficits (paresthesias, decreased pain, and temperature sensation), motor deficits (paraplegia, quadriplegia), bladder and bowel dysfunction, and possibly diaphragmatic dysfunction. Complete transection or hemitransection (BrownSequard syndrome) at the level of irradiation are serious potential sequelae. Symptom onset can be sudden or gradual (Rampling and Symonds 1998; Wang et al. 1992; Zweig and Russell 1990). 4.3.5.2  Pathology Pathologic changes in radiation myelopathy comprise white matter changes and vascular changes. The combination of these is characteristic of radiation myelopathy (Rampling and Symonds 1998). Gray matter is rarely involved. White matter changes include axonal loss, demyelination, and necrosis. Vascular changes tend to involve the small vessels and may have a predilection for the venous side (Schultheiss and Stephens 1992). These include hyaline degeneration, small-vessel perivascular fibrosis, capillary proliferation, endothelial swelling, fibrinoid necrosis, telangiectasias, and vasculitis (Komachi et al. 1995; Rampling and Symonds 1998). Zweig et al. described biopsy of a focally enhancing region on MRI showing vacuolation and gliosis of white matter with fibrinoid damage of small vessels and capillaries (Zweig and Russell 1990). Clinically it has been found that cases of white matter changes and combined white matter and

vascular changes tend to have a shorter latency period, while cases that involve only vascular lesions have a longer latency period (Wang et al. 1992). 4.3.5.3  Imaging Findings The diagnosis of radiation myelopathy is largely clinical. Imaging findings are nonspecific and the entity remains a diagnosis of exclusion. MRI is the imaging modality of choice. The most important role of MRI is exclusion of other pathology, including metastatic disease and recurrent tumor. Myelography has been used in the past in evaluation for radiation myelopathy, primarily to exclude other pathology. In cases of radiation myelopathy, myelography can be normal or may demonstrate cord swelling. Wang et al. studied 10 patients who developed radiation myelitis after radiation therapy for nasopharyngeal carcinoma. Three patients had a shorter latency period (7–18 months) and 7 patients had a longer latency period (22–144 months). There was no significant correlation between latency period and MRI findings. There was, however, a correlation between MRI findings and amount of time after onset of symptoms that MRI was performed (Wang et al. 1992). Patients imaged up to 8 months after symptom onset demonstrated T2 hyperintensity and T1 hypointensity, and some demonstrated cord swelling and focal enhancement (Fig. 4.2.35). All patients in the series who were imaged early demonstrated T1 hypointensity and T2 hyperintensity over a long segment of cord parenchyma. This is of course a very nonspecific finding, seen in myelopathy of any cause. Six of eight patients who were imaged early demonstrated cord swelling. Four of five cases that received contrast showed a focal area of enhancement. This may appear as ring enhancement. The enhancement may be due to breakdown of the blood cord barrier and increased vascular permeability. The enhancing focus eventually becomes an area of focal atrophy. These imaging findings may persist for many months (Wang et al. 1992). Abnormalities may later involve segments above and below the radiated segment secondary to Wallerian degeneration (Medonca 2001). In the chronic stage, cord atrophy is often seen and signal abnormality may resolve, as was the case in the two patients imaged at least 3 years after symptom onset in Wang’s series (Wang et al. 1992). Again, the imaging findings are not specific and must be correlated with location and time course of radiation, as well as neurologic physical findings. Differential diagnosis includes tumor recurrence, chemotherapy induced myelitis, and other causes of myelopathy. Seeing bright signal fatty replacement of bone marrow in vertebral bodies at the level of myelopathy may be a helpful secondary sign that radiation injury is the cause.

4.3  Intramedullary Diseases of the Spinal Cord

Fig. 4.3.35  Radiation myelopathy. Sagittal T1-weighted image (a) demonstrates mild hypointensity and swelling of the cervical spinal cord. Note the radiation therapy related changes in the vertebral bodies, manifested by diffuse markedly hyper-

intense marrow signal. Use this as a clue to the diagnosis! T2weighted imaging (b) demonstrates a long segment of diffuse cord hyperintensity. Contrast administration (c) reveals a focus of enhancement in the upper cervical cord (arrow)

4.3.6  Intramedullary Infectious and Inflammatory Diseases

4.3.6.1.1  Infective Agents

Many infectious agents have been reported to cause myelopathy; however, infectious myelopathy is still a somewhat rare entity. Since many infectious processes are treatable, timely and accurate diagnosis is essential to improving outcome. Due to the ability of MR to evaluate tissue characteristics of the spinal cord, MR is the imaging modality of choice in evaluation for possible cord infection or inflammatory disease. MRI is useful to confirm clinical suspicion, show subclinical disease, and evaluate treatment response. Routine spine imaging using T1, T2 and contrast-enhanced T1-weighted images are usually adequate for evaluation. STIR may also be helpful to identify lesions and edema. This section addresses the MR appearance of infectious and inflammatory myelopathies according to the following categories: viral myelitis, HIV-associated vacuolar myelopathy, opportunistic infection, and granulomatous disease (Table 4.3.3). 4.3.6.1  Viral Myelitis Viral myelitis describes acute inflammatory injury of the spinal cord secondary to viral infection. Viral myelitis is quite rare, with 1/100,000 incidence per year (Chong et al. 1999). It is more commonly seen in immunocompromised patients. The actual incidence is probably higher than reported due to the difficulty of making a specific diagnosis of viral myelitis. Diagnosis is usually made by polymerase chain reaction (PCR) for positive virus in CSF. Some cases of actual viral myelitis are surely labeled as “idiopathic” transverse myelitis due to insufficient workup.

Herpes Virus Myelitis Herpes virus myelitis, also called zoster myelitis, is the most prevalent type of viral myelitis. One series found herpes virus myelitis to make up approximately 29% of all probable and confirmed cases in Finland (Koskiniemi and Rantalaiho 2001). This entity is relatively easy to diagnose with classic cutaneous findings of chickenpox and shingles in adults. Its mechanism of spinal cord invasion is also better understood compared to other viruses. When the thoracic dermatomes are involved by the Varicella zoster virus, the virus may occasionally spread centripetally, instead of the more common route of centrifugally along the corresponding nerves following reactivation, and result in a necrotizing myelopathy (Chong et al. 1999). Others Other viral agents in the herpes virus group, including herpes simplex virus 2 (HSV-2, the etiological agent of genital herpes), Epstein-Barr virus (the etiological agent of infectious mononucleosis), and cytomegalovirus (CMV) may also cause myelitis at the time of primary infection. A large number of other viruses also can cause myelopathy and are listed in Table 4.3.4. 4.3.6.1.2  Imaging Findings The MR appearances of various viral myelitides are quite similar. They typically demonstrate fusiform cord enlargement at the involved levels. The spinal cord parenchyma typically demonstrates diffuse hyperintense T2 signal. T1 signal may be normal or mildly hypointense. There may be variable, nonfocal enhancement on post-

649

650

4  Spine and Spinal Canal Table 4.3.3  MR appearances of spinal cord infection associated with various pathogens intramedullary traumatic injury    Infective agents

  Most common location

  Spinal cord morphology

Pre-contrast T1 signal relative to spinal cord gray matter

 Post-contrast T1 enhancement pattern

T2 signal relative to spinal cord gray matter

Viral (non-HIV)

Cervical and thoracic cord

Enlarged

Hypointense but higher than CSF

Variable

Hyperintense

HIV

Usually start in mid-to-low thoracic cord, with increasing rostral involvement

Atrophic

May be normal

No enhancement or infrequent patchy enhancement

Two- to fivesegment diffuse hyperintensity

Toxoplasma

Cervical and thoracic cord

Enlarged

Isointense

Surface or nodular

Hyperintense

Bacterial (e.g., Staphylococcus or Streptococcus)

Cervical and thoracic cord

May be enlarged

Hypointense

Variable

Hyperintense area larger than enhancing lesion

TB (meningitis)

Cervical and upper thoracic

May be enlarged

Increased CSF signal; cord signal may be normal or hypointense

Linear

Hyperintense with irregular cord surface

TB (tuberculoma)

Evenly divided in cervical, thoracic and lumbar cord

Enlarged

Hypointense to isointense

Rim or nodular

Variable: smaller lesions often hyperintense and larger lesions often hypointense

Sarcoidosis

Cervical and upper thoracic

Enlarged

Hypointense to isointense

Patchy peripheral parenchymal and leptomeningeal enhancement

Hyperintense

Table 4.3.4  Viral causes of myelopathy (Berger and Sabet 2002; Koskiniemi and Rantalaiho 2001) Epstein-Barr Measles Varicella zoster Borrelia burgdorferi (Lyme Disease) Cytomegalovirus Parainfluenzae viruses Herpes simplex Influenza A Herpes virus 6 Echovirus

Herpes B (monkey virus) Hepatitis HTLV-I and -II (human T-cell lymphotropic virus) Rabies HIV Coxsackie Rubella West Nile Mumps

contrast images. The enhancing region may represent the site directly affected by the viral attack. The area of hyperintense T2 signal is usually larger than the enhancing area and represents adjacent tissue edema (Fig. 4.2.36). 4.3.6.2  HIV-Associated Vacuolar Myelopathy HIV-associated vacuolar myelopathy, also known as AIDS-associated myelopathy, is the most common form of myelopathy in HIV patients. The diagnosis is one of exclusion based on clinical, laboratory and radiologic findings. There is no specific treatment to date. It does not appear to be related to direct infection of the spinal cord, suggesting that its pathogenesis may be caused by a retroviral-induced metabolic disorder. It may also be associated with abnormal Vitamin B12 utilization. HIV-

4.3  Intramedullary Diseases of the Spinal Cord

Fig. 4.3.36  Zoster myelitis. T1-weighted imaging (a) demonstrates cord swelling in the upper cervical and in the upper thoracic regions. These two areas of fusiform swelling demonstrate hyperintensity (arrows) on T2-weighted imaging (b) and

enhancement (arrows) on post-contrast T1-weighted imaging (c). Post-contrast sagittal image of the thoracic spine (d) demonstrates multiple additional areas of patchy mild enhancement (arrows)

4.3.6.2.1  Imaging Findings

Fig. 4.3.37  HIV-associated vacuolar myelopathy. Sagittal T1-weighted image (a) demonstrates subtle hypointensity in the posterior aspect of the cord. T2 hyperintensity (b) in this region is much more conspicuous and extends over a few segments in the upper cervical cord (arrows). Axial T2-weighted image (c) demonstrates focal hyperintensity in the posterior columns. Note the similar appearance of this case to Fig. 4.3.44, which demonstrates subacute combined degeneration

associated vacuolar myelopathy can be indistinguishable radiologically and pathologically from Vitamin B12 deficient myelopathy (Fig. 4.2.37). Distinction may be made on a laboratory basis, as vitamin B12 deficient myelopathy will typically demonstrate a negative HIV test and decreased plasma B12 levels, unlike HIV-associated vacuolar myelopathy.

MR evaluation most commonly demonstrates spinal cord atrophy, although normal cord morphology can also be seen. The involved segment often starts in the mid to lower thoracic cord, with progressive rostral involvement up toward the cervical spinal cord. T1 signal is usually normal. T2 imaging demonstrates diffuse hyperintensity, typically over two to five vertebral levels. The classic appearance shows involvement of white matter tracts laterally and symmetrically (Fig. 4.2.37). Contrast-enhanced images demonstrate no enhancement in most cases, although patchy enhancement has been described. Patients with advanced myelopathy tend to demonstrate more extensive MRI abnormalities. 4.3.6.3  Non-Viral Spinal Cord Infections A large number of other non-viral pathogens have been reported to cause myelopathy as listed below in Table 4.3.5. Many of these non-viral infections are opportunistic, often occurring in the immunocompromised. Opportunistic spinal cord infections are frequently associated with HIV or immunocompromised states, although they also may occur in immunocompetent patients. A history of prior trauma or surgery, systemic bacteremia, meningitis, or congenital abnormality such as dermal sinus predisposes a patient to spinal cord in­ fection.

651

652

4  Spine and Spinal Canal Table 4.3.5  Non-viral infective causes of myelopathy (Berger and Sabet 2002; Koskiniemi and Rantalaiho 2001 44) Fungi Toxoplasma TB Atypical mycobacteria Bacterial Mycoplasma pneumoniae Chlamydia pneumoniae Listeria organisms Cystercercus Schistosoma mansoni

4.3.6.3.1  Imaging Findings Imaging findings are again nonspecific and similar to viral myelitis, demonstrating fusiform cord swelling, normal to hypointense T1 signal and hyperintense T2 signal. The enhancement pattern is usually nonspecific and can vary from marginal to nodular to solid enhancement. T2 hyperintense signal is usually larger (several segments above and below) than the area of enhancement. 4.3.6.4  Granulomatous Processes 4.3.6.4.1  Tuberculosis Although tuberculosis is primarily an illness of the respiratory system, it can spread to involve many organ systems. Tuberculous involvement of central nervous system is uncommon compared to involvement of other systems and accounts for approximately 1–10% of all case of tuberculosis. Like its intracranial counterpart, spinal tuberculosis has two related pathologic process: tuberculous meningitis and the intraspinal tuberculoma. Wadia and Dastur (1969) described the most common form of intraspinal tuberculosis as a primary spinal cord lesion associated with spinal meningitis. The second most common form is secondary downward extension of intracranial tuberculosis meningitis. The least common form of intraspinal tuberculosis is extension from vertebral tuberculous infection. Imaging Findings The MR imaging features of TB meningitis often include increased CSF signal intensity on T1-weighted images, which may lead to complete loss of the cord-CSF interface and a shaggy cord outline. There is increased signal intensity within the spinal cord on T2-weighted images, an irregular cord surface, and linear enhancement. Associated thickened and clumped cauda equina nerve roots (arachnoiditis), nodules in the subarachnoid space,

arachnoid cyst formation, loculation of the cerebrospinal fluid, and secondary syrinx formation can also be seen. Frequently, but not always, intramedullary tuberculoma is associated with tuberculous meningitis/arachnoiditis of the spine. MR imaging of intramedullary tuberculoma usually demonstrates a single lesion with imaging characteristics similar to intracranial tuberculoma. There may be mild cord swelling in the region of the tuberculoma with adjacent edema. On T1-weighted images, lesions demonstrate hypointense to isointense signal. On T2-weighted images, lesions demonstrate a variable appearance. According to Gupta et al., smaller lesions tend to have T2 signal hyperintense to cord gray matter, while larger lesions tend to be hypointense to cord gray matter (Gupta et al. 1994). Post-contrast imaging typically shows rim or nodular enhancement. After treatment, follow up MR examination often shows initial swelling and later decrease in lesion size with clinical improvement of symptoms. 4.3.6.4.2  Sarcoidosis Sarcoidosis is a multisystem disease of unknown origin with a wide variety of clinical and radiologic manifestations. It is a non-caseating granulomatous disease that can involve the spine and spinal cord. The disease shows a predilection for adults less than 40 years of age, peaking in those 20–29 years old. Spinal cord involvement is comparatively rare. When present, it typically occurs in the early stage of the disease and responds rapidly to steroid treatment. Imaging Findings When the cord is involved, MRI demonstrates involvement of the cervical and upper thoracic spinal regions. The involved segment usually demonstrates fusiform cord enlargement with the lesion located in the intramedullary compartment. Focal or diffuse hyperintensity on T2-weighted images and iso to hypointensity on T1-weighted images is typically seen due to associated edema. Contrast-enhanced images often demonstrate areas of enhancement that are predominantly in the per­ iphery of the spinal cord. Enhancement may be patchy and multifocal (Fig. 4.3.38). Junger et al. (1993) hypothesized that patients with spinal cord sarcoidosis progress in four phases: 1 A linear leptomeningeal pattern of enhancement is seen with early inflammation. 2 Secondary centripetal spread of the leptomeningeal inflammatory process occurs through the VirchowRobin spaces, showing parenchymal involvement with faint enhancement and diffuse swelling. 3 Inflammation decreases, and the enlarged spinal cord tends to return to normal size with focal or multifocal intramedullary lesions.

4.3  Intramedullary Diseases of the Spinal Cord

Fig. 4.3.38  Sarcoid. Sagittal T2-weighted (a) and proton density–weighted (b) images demonstrate ill-defined hyperintensity spanning a few segments in the upper cervical cord. After contrast administration (c), an area of solid enhancement is seen at the C4 level

4 The cord begins to reduce in size until it reaches a final stage of atrophy and no enhancement. Phases 2 and 3 are the most frequent at clinical presentation. Other rare findings such as calcifications, cyst formation, and extradural involvement have also been described. Following steroid therapy, significant reduction or resolution of enhancement and T2 prolongation is expected. However, improvement on MR imaging may lag behind clinical improvement. 4.3.6.5  Differential Diagnosis Because of the potentially treatable nature of infectious and inflammatory processes, early diagnosis is beneficial and a high index of suspicion should be maintained in the right clinical setting. Any myelopathic patient with risk factors (immunosuppression, prior trauma or surgery, systemic bacteremia, meningitis, or congenital abnormality such as dermal sinus) deserves consideration of the possibility of spinal cord infection when there are abnormal MRI findings of cord swelling, hyperintensity on T2-weighted images, intramedullary enhancement, or edema extending away from an enhancing area. The MR imaging findings for infectious myelitis are often nonspecific. The differential diagnosis is extensive and includes transverse myelitis, multiple sclerosis, acute disseminated encephalomyelitis (ADEM), cord tumor, infarction, and vascular malformation, among other etiologies. Some imaging and clinical information may give us clues to differentiate the above-listed entities from infectious myelopathy. Transverse myelitis often demonstrates centrally located T2 hyperintensity with peripheral enhancement, with an acute to subacute clinical onset. It is

usually characterized by loss of motor and sensory function and is preceded by a viral illness. This diagnosis is usually a diagnosis of exclusion. Multiple sclerosis often demonstrates multiple peripherally located T2 hyperintense lesions. It is typically less than two vertebral bodies in length. Acute disseminated encephalomyelitis (ADEM) typically is associated with a recent history of vaccination or immune insult and usually affects the brain and occasionally the spinal cord. Spinal cord tumors tend to have slower clinical progression and may resemble granulomatous lesions. In practice, it is often difficult to differentiate many of the entities mentioned in this section by radiological findings alone due to similar imaging characteristics. Therefore, it should be emphasized that imaging findings should be correlated with the patient’s history, physical examination and laboratory results in the process of patient work up. A trial of the appropriate antimicrobial therapy may be indicated in certain situations. 4.3.7  Intramedullary Traumatic Injury Spinal cord injury has long been recognized as one of the most devastating injuries one can suffer, as was eloquently stated circa 2,500 BCE: “When you examine a man with dislocation of a vertebrae of his neck, and you find him unable to move his arms and legs, then you have to say: a disease one cannot treat” (Dubendorf 1999). Approximately 10,000 new spinal cord injuries occur in the United States each year, with 82% of these injuries occurring in males, and 60% occurring between the ages of 16 and 30 years old (Dubendorf 1999). Spinal cord injury predominantly affects employed young adults, resulting in a tremendous financial loss to society with regards to overall lifetime productivity (Flanders and Croul 2001). The total lifetime cost for an incomplete spinal motor injury is US $364,491, while the total lifetime cost for a high cervical injury is US $1,713,267 (Flanders and Croul 2001). Motor vehicle accidents continue to be the number one cause of spinal cord injuries, followed by falls, violent attacks, and sports injuries (Dubendorf 1999). In this section, the clinical manifestations of spinal cord trauma, as well as the different systems for classifying spinal cord injury, are discussed. The pathophysiological response to spinal cord injury, with the subsequent MR imaging findings of both acute and chronic injury will be reviewed as well. Finally, special attention will be focused on spinal trauma with regards to the pediatric patient. 4.3.7.1  Clinical Manifestations Spinal injury may result in quadriplegia, which involves one of the cervical vertebral segments, leading to paralysis of all four limbs. Paraplegia occurs when there is

653

654

4  Spine and Spinal Canal

injury to the thoracic, lumbar, or sacral segments of the spine, resulting in dysfunction of the lower extremities (Flanders and Croul 2001). A complete injury occurs when there are no motor and sensory function three segments below the level of injury. The neurologic level of injury is considered the most caudal segment of the spinal cord with normal motor and sensory function on both sides (Flanders and Croul 2001). A particular type of hyperextension injury in the cervical spinal region can result in a specific injury pattern known as central cord syndrome. This is often seen in young patients with congenital central canal stenosis, or elderly patients with degenerative spondylosis of the cervical spine (Santoro et al. 2004). Central cord syndrome manifests as a disproportionate impairment of motor function involving the upper extremity as compared to the lower extremity. In addition, the impairment predominantly affects distal motor function as compared to proximal motor function (Santoro et al. 2004). The mechanism of injury is thought to be related to compression of the cord between an anterior disk bulge or osteophytic bar, with subsequent buckling of the ligamentum flavum and resulting posterior compression of the spinal cord (Santoro et al. 2004). Another specific injury pattern is the anterior cord syndrome, often associated with traumatic disk herniation, in which there is injury to the anterior two thirds of the spinal cord, resulting in deficits of the pain and temperature sensory tracts, with preservation of the touch and vibration sensory tracts. In contrast, posterior cord syndrome results in loss of the dorsal column, with subsequent loss of proprioception (Ross 2004) Brown-Sequard syndrome occurs when there is traumatic insult to half of the spinal cord, resulting in ipsilateral loss of touch and motor function, with contralateral loss of pain and temperature sensation. Brown-Sequard syndrome may be superimposed on central cord syndrome. Autonomic hyperreflexia arises in spinal cord injury above the level of T6, resulting in paroxysmal hypertension, with associated agonizing headaches (Ross 2004). 4.3.7.2  Classification Vertebral column and spinal cord injuries can be classified according to neurologic deficits, type of trauma, mechanism of injury, and anatomic location of injury (Takhtani and Melhem 2000). Frankel’s functional classification (Table 4.3.6) of spinal cord injury relates the injury to motor and sensory loss/recovery, with grade A representing complete motor and sensory loss, and grade E representing complete neurologic recovery (Takhtani and Melhem 2000). Classification by mechanism of injury is related to the forces exerted on the spinal cord, which include hyperflexion, hyperextension, vertical rotation, rotation with hyperflexion or hyperextension,

Table 4.3.6  Frankel’s functional classification of spinal cord injury (Takhtani and Melhem 2000) Grade

Neurologic deficit

A

Complete motor and sensory loss

B

Preserved sensory function

C

Preserved motor activity (nonfunctional)

D

Preserved motor activity (functional)

E

Complete neurologic recovery

lateral flexion, and indeterminate forces. The anatomic classification relates to the exact level of injury (Takhtani and Melhem 2000). 4.3.7.3  Pathophysiology Common underlying processes are noted in all etiologies of spinal cord injury, referred to as primary and secondary injury mechanisms. Primary injury refers to the primary damage to the spinal cord as a result of the blunt impact, which includes cord concussion, contusion, lacerations, transections, and intraparenchymal hemorrhage (Dubendorf 1999). The spinal cord may also be damaged by compression resulting from acceleration/deceleration injury, hyperflexion, hyperextension, axial loading, and severe rotation. The degree of spinal cord injury depends on the extent of mechanical damage from the primary event, as well as those processes that occur during the secondary phase of spinal cord injury (Dubendorf 1999). The secondary phase of neurologic injury consists of the interdependent chain of events that occur at the systemic and cellular level following the primary injury. These processes occur within minutes of the initial injury and last for days to weeks following injury (Dubendorf 1999). Immediately following spinal cord injury, ischemia occurs secondary to interruption of the microvascular circulation of the cord, particularly with regards to the gray matter. The causes of this ischemia include direct vascular injury, hypotension from neurogenic shock, loss of the normal autoregulation, vasospasm, and thrombosis. These microvascular circulation changes are noted at the level of the injury initially, and then subsequently expand to include a wider zone of injury within the spinal cord segment, ultimately spreading both rostrally and caudally beyond the initial level of injury. Decreased availability of oxygen eventually leads to tissue necrosis, anaerobic cellular metabolism, and lactic acidosis (Dubendorf 1999). In addition to the loss of normal impulse conduction in those fibers damaged from direct injury, there is loss of impulse conduction in the anatomically intact fibers

4.3  Intramedullary Diseases of the Spinal Cord

related to a shift in the intracellular/extracellular ion concentration, particularly calcium. Following injury to the spinal cord, calcium levels decrease extracellularly while increasing intracellularly, impairing the function of the cell and therefore catalyzing neuronal demise. Increased levels of intracellular calcium also stimulate protease and lipase activity, initiating lipid perioxidation of the cell membranes, allowing free fatty acids and free radical release from the damaged cell, resulting in further vasospasm and progressive ischemia. The overall inflammatory process continues for several days, and involves endothelial damage, release of inflammatory mediators, changes in vascular permeability, development of edema, migration of peripheral inflammatory cells, and activation of microglia. The final result of this inflammatory response is evident pathologically as petechial hemorrhage progressing to hemorrhagic necrosis, infarction, atrophy, glial scarring, and cyst formation. These pathologic findings have distinct imaging characteristics on MR imaging (Dubendorf 1999). 4.3.7.4  Technique MR evaluation of the spinal cord and the vertebral column are generally performed together using the same protocol. In addition to routine T1 and T2-weighted sequences, T2* GRE imaging can identify even subtle areas of parenchymal hemorrhage which may not be seen otherwise (Fig. 4.3.39). STIR imaging can also provide useful additional information, particularly regarding ligamentous injury. Contrast is usually not necessary. Recent technologic advances in MR scanning techniques have shown promise with regards to imaging the injured spinal cord. Research continues with regard to diffusion weighted imaging of the injured spinal cord, and in animal models changes were noted on diffusion weighted images despite normal conventional T2-weighted spinecho images (Saifuddin 2001). Early elevation on ADC images is predictive of early cyst and syringomyelia formation, and quantification of ADC abnormality may improve the precision with which MRI can predict neurologic recovery (Flander and Croul 2001). Functional MRI is a valuable tool for assessing neural activity across spinal cord lesions, the functional effects of spinal cord injury on the brain, and the efficacy of traditional and novel treatments (Krzyzak et al. 2005). Functional MRI with blood oxygen level dependent (BOLD) contrast provides a non-invasive method for assessing neuronal activity (Krzyzak et al. 2005). Utilizing fMRI and BOLD imaging, studies have demonstrated preservation of motor and some somatosensory cortical representations in the absence of overt movements or conscious sensations several years after spinal cord injury, which bodes well for potential therapies aimed at restoring spinal cord connections (Krzyzak et al. 2005).

Fig. 4.3.39  Lateral plain film of the cervical spine demonstrates an anterior subluxation of the C6 vertebra. Subsequent MRI evaluation demonstrates injury to the spinal cord in this region. Sagittal proton density (b) and T1-weighted (c) images demonstrate only subtle signal abnormality at the C6–C7 level. Sagittal T2-weighted imaging (d) shows narrowing of the spinal canal, cord hyperintensity, and mild focal cord expansion at this level. Sagittal gradient-echo recalled T2-weighted image (e) reveals a small focus of hypointensity (arrow), indicating a small focus of intramedullary hemorrhage which was not detectable on other sequences

4.3.7.5  Imaging Findings Plain films and CT continue to be instrumental in the evaluation of the acutely injured spine, particularly with regard to the identification of spinal instability and vertebral fractures (Takhtani and Melhem 2000). However, both plains films and CT scans are less than optimal for characterizing direct spinal cord injury and for identifying non-osseous causes of cord compression (Takhtani and Melhem 2000). MRI is an excellent means for visualizing both spinal cord and adjacent vertebral column/paraspinal soft tissue pathology. MRI can adequately identify cord hemorrhage, cord edema, cord swelling, cord compression, and nerve root damage. MRI is also excellent at identifying extra axial and vertebral column injury, such as epidural hematoma, ligamentous damage, traumatic disk hernia-

655

656

4  Spine and Spinal Canal

tion, vertebral body fracture and subluxation, and paraspinal muscular strains and hematoma formation, as is discussed in the chapter on extradural disease of the spine. 4.3.7.6  Acute Findings Spinal cord injury can be classified as concussive, contusive, or compressive. Concussive injuries to the spinal cord are purely functional and fully reversible derangements, attributed to transient deficiencies in spinal cord microcirculation (Takhtani and Melhem 2000) Cord contusions may range from simple edema or petechial hemorrhage to severe hemorrhage and liquefactive necrosis, the most severe manifestation of which is complete transection. Cord compression may be related to bone fragments, vertebral column subluxation or dislocation, disk herniation, spondylotic bar, or epidural hematoma formation. MRI is far superior compared to any other imaging modality for imaging spinal cord injury (Takhtani and Melhem 2000). Cord concussive injury, as was previously noted, is a purely functional and fully reversible derangement. This is attributed to deficiencies at the level of the microcirculation of the cord status post-injury. This may uncommonly be demonstrated by spinal cord edema, seen as hyperintense signal on T2-weighted images, which is often transient in nature (Takhtani and Melhem 2000). Various patterns of cord injury have been described which strongly relate to neurologic outcome. Kulkarni et al. developed a system involving three types of spinal cord injury (Table 4.3.7) (Kulkarni et al. 1988). The type 1 injury represents an intramedullary hematoma, and is depicted as a central large focus of hypointensity on T2-weighted images, with a thin surrounding rim of hyperintensity. The type 1 injury may demonstrate a heterogeneous appearance on T1-weighted images. Methemoglobin, seen both intracellularly and extracellularly in the subacute stage of bleeding depending on the time course, may be depicted as hyperintense signal on T1-weighted images (Kulkarni et al. 1988). The type 2 injury depicts cord swelling/contusion, in which there is increased cord caliber with hyperintense signal noted

on T2-weighted images (Fig. 4.2.40). There is no hypointense signal on T2-weighted images in the type 2 injury, and T1-weighted images may be normal or demonstrate hypointense signal. The type 3 injury is a mixed pattern, and often demonstrates a small central focus of hypointense signal with a thick surrounding rim of hyperintensity on T2-weighted images, and a normal appearance on T1-weighted images (Fig. 4.2.40) (Kulkarni et al. 1988). Of the three types of spinal cord contusion/hemorrhagic injuries described by Kulkarni et al., the type 1 pattern with its depiction of intramedullary hemorrhage carries the worst prognosis (Kulkarni et al. 1988). The majority of patients with the type 1 injury pattern have complete neurologic injury at presentation, and do not show subsequent improvement. The type 2 injury pattern, the most common injury pattern, is often associated with little if any neurologic deficits at presentation and

Fig. 4.3.40  Sagittal T2-weighted image (a) demonstrates contusion or concussive injury of the spinal cord at the C6 level. There is subtle, poorly marginated intramedullary hyperintensity at this level, adjacent to a traumatic anterior subluxation of the C6 vertebral body. This patient had a jumped and fractured C6–C7 facet on the right side. Axial gradient-echo image (b) through this level demonstrates ill defined intramedullary hyperintensity on the right side

Table 4.3.7  Kulkarni MRI signal patterns in acute spinal trauma (Kulkarni et al. 1988) MRI pattern

T1-weighted image

T2-weighted image, central

T2-weighted image, peripheral

I

Inhomogeneous

Large area of hypointensity (hemorrhage)

Thin rim of hyperintensity (extracellular methemoglobin)

II

Normal

Hyperintensity (presumably edema)

Hyperintense (presumably edema)

III

Normal

Isointense

Thick rim of hyperintensity

4.3  Intramedullary Diseases of the Spinal Cord

demonstrates the best prognosis for recovery. The type 3 pattern of spinal cord injury, or the mixed type pattern, demonstrates variable deficits in presentation, as well as variable prognosis for recovery (Kulkarni et al. 1988). MRI findings associated with poor neurologic function at initial presentation and at long-term follow-up include the presence and extent of intra-axial hematoma, the extent of cord edema, the presence and extent of cord compression, and the presence of an extra-axial hematoma resulting in spinal cord compression (Selden et al. 1999). The results of the initial, admitting neurologic examination remain the best single predictor of potential motor and sensory improvement following spinal cord injury (Selden et al. 1999). However, these additional MRI characteristics, both at initial presentation and at follow-up, provide significant additional prognostic information for the potential neurologic recovery (Selden et al. 1999). In the proper clinical setting, with regard to central spinal cord syndrome, MRI demonstrates spinal cord edema, which is seen as hyperintense signal on T2-weighted images, with the absence of hemorrhage. The severity of the clinical symptoms correlates with the extent and amount of edema and subsequent T2 hyperintensity within the spinal cord (Saifuddin 2001). As previously mentioned, MRI is superior for evaluating the causes of spinal cord compression, particularly when considering non-osseous causes of spinal compression. Spinal cord compressive injuries (Figs. 4.2.41, 4.2.42) may produce MRI findings similar to spinal cord contusion, with the notable addition of compressive lesions including bony fragments, spinal subluxations, disk herniations, epidural hematomas, and spondylotic bars (Takhtani and Melhem 2000). For spinal cord compression to be definitively diagnosed, a reduction of 50% or greater in the anterior–posterior dimension must be

Fig. 4.3.41  Spinal cord compressive injury. Sagittal T2-weighted (a), STIR (b), and proton density–weighted (c) images demonstrate traumatic cord compression with focal marked spinal canal stenosis and hyperintense signal in the adjacent spinal cord at the C3–C4 level. This patient also had a C3 vertebral body fracture

Fig. 4.3.42  Spinal cord compressive injury. Sagittal T2-weighted image (a) demonstrates displaced C1 and odontoid fractures compressing the upper cervical spinal cord. There is hyperintensity and abnormal contour of the cord in this region. Note the extension of abnormal hyperintense signal into the brainstem. Axial T2-weighted image (b) through the cervical spine also demonstrates traumatic spinal stenosis, with cord compression and abnormal hyperintense signal within the spinal cord

present (Takhtani and Melhem 2000). Acute compressive lesions often mandate immediate surgical intervention, necessitating prompt identification by the radiologist. Experimental and clinical evidence supports improvement and resolution of both acute and chronic neurologic deficits following surgical decompression, which indicates the importance of distinguishing spinal cord contusion and hemorrhage from direct cord compression (Takhtani and Melhem 2000). 4.3.7.7  Chronic Findings Patients with chronic spinal cord injury may present with challenging clinical features, including increasing myelopathy, ascending neurological level, increasing pain, and increasing muscle spasm. Posttraumatic spinal cord injury has multiple chronic manifestations, including myelomalacia, intramedullary cyst and syrinx formation, and spinal cord atrophy (Potter and Saifuddin 2003). Other manifestations of chronic injury include cord tethering and cord disruption. Patients may demonstrate progressive signs and symptoms related to chronic spinal cord injury, which may be related to progressive syrinx formation (43%), extended atrophy (26%), or progressive myelomalacia (21%) (Potter and Saifuddin 2003). Posttraumatic myelomacia occurs at the site of previous trauma, and usually manifests as increased signal on T2-weighted images which may or may not demonstrate post-contrast enhancement. The margins of these lesions may not be distinct, and the cord caliber may or may not be enlarged or atrophic (Potter and Saifuddin

657

658

4  Spine and Spinal Canal

2003). Posttraumatic myelomalacia precedes overt syrinx formation, and is considered by some to be a “pre-syrinx” state (Ross 2004). Small cysts may develop in areas of posttraumatic myelomalacia, demonstrated on MRI as well-defined foci of increased signal on T2-weighted images with corresponding decreased signal on T1-weighted images. These small cysts are thought to represent the residua of prior intramedullary hemorrhage, with subsequent liquefaction of necrotic cord tissue (Potter and Saifuddin 2003). Like cysts elsewhere in the central nervous system, there is no post-contrast enhancement. Cyst formation is commonly associated with cervical spinal injuries, and is often confined to the vertebral body level which demonstrates maximal bony protrusion into the spinal canal. These small cysts are often seen adjacent to focal areas of myelomalacia (Potter and Saifuddin 2003). With the passing of time, ranging from several months to many years after the initial trauma, an elongated syrinx may form at the site of prior trauma, extending both superiorly and inferiorly (Potter and Saifuddin 2003). The etiology of syrinx formation is not completely understood, and it has been suggested that some posttraumatic cysts may communicate with the CSF space, resulting in the extension of the cyst cavity due to subsequent CSF flow. Other theories suggest syrinx formation is related to the development of dural adhesions with resultant syrinx cavity formation (Potter and Saifuddin 2003). The “one-way valve” theory suggests that CSF is allowed into, but not out of the posttraumatic cystic cavity (Ross 2004). The “slosh-and-suck” theory suggests that increased epidural venous flow from certain activities that increase pressure around the spinal cord (i.e., coughing/sneezing/straining) leads to formation of an enlarging cystic cavity, as this increased pressure cannot be dissipated due to abnormal disruptions in CSF flow (Ross 2004). A posttraumatic syrinx will be depicted on MRI as an elongated focus of increased signal on T2-weighted images with corresponding decreased signal on T1weighted images (Fig. 4.2.43). Although the signal characteristics are usually similar to CSF, in some instances the T1-weighted images may be slightly hyperintense relative to CSF, due to increased protein content (Potter and Saifuddin 2003). The syrinx will demonstrate no significant post-contrast enhancement. In addition, the margins of the syrinx will be well-defined and sharply marginated. The syrinx is typically tapered at one or both ends, and often can appear septated (Potter and Saifuddin 2003). Usually, there is expansion of the cord with thinning of the cord tissue. Syrinx cavities may range in length from 2 to 20 vertebral body segments, and they may extend both rostrally and caudally with equal incidence. Flow voids may occasionally be seen within syrinxes on T2-weighted images, suggesting a high pressure syrinx, which predicts a good response to surgical drain-

Fig. 4.3.43  Syrinx as a chronic sequela of traumatic cord injury. T1-weighted (a) and T2-weighted (b) sagittal images demonstrate syringohydromyelia (arrows) in an area of remote traumatic injury. Note the tapered inferior margin of the syrinx and thinning of the surrounding cord parenchyma

age as opposed to low pressure syrinx, which does not demonstrate flow voids (Potter and Saifuddin 2003). Spinal cord atrophy is often seen with complete spinal cord injuries. The spinal cord is considered atrophic if it measures less than 7 mm in AP dimensions in the cervical spine, and less than 6 mm in AP dimensions in the thoracic spine (Potter and Saifuddin 2003). Atrophy is considered extended if it is at least two segments beyond the level of the associated vertebral body injury. Extended atrophy is the most common abnormality seen in patients imaged more than 20 years after injury, with a prevalence of 62% (Potter and Saifuddin 2003). Cord tethering occurs when the spinal cord is attached to the bony wall of the spinal canal, and is often associated with other manifestations of chronic spinal cord injury, including atrophy and cyst formation (Potter and Saifuddin 2003). Cord disruption is defined as complete absence of spinal cord tissue, typically at the site of injury or more distally. Cord disruption is associated with other changes of chronic spinal cord injury as well, including atrophy and myelomalacia (Potter and Saifuddin 2003). 4.3.7.8  Pediatric Spinal Cord Injury Imaging of the pediatric spine deserves special consideration, as the incidence and distribution of spinal injury differ between the pediatric population and adults (Flanders and Croul 2001). Pediatric spinal trauma accounts for 1–10% of all spinal cord injury, and most commonly occurs secondary to falls (56%), motor vehicle accidents (23%), and sports-related injuries (16%)

4.3  Intramedullary Diseases of the Spinal Cord

(Flanders and Croul 2001). Flynn et al. (2002) stated that plain films were unreliable for evaluating all but the worst soft tissue injuries in children, and that even CT scanning could miss soft tissue, ligamentous, and osseous injuries. Injuries to the upper cervical spine commonly occur in young children, secondary to hypermobility related to ligamentous laxity, open ossification centers, underdeveloped neck musculature, and the increased head-to–body weight ratio (Flynn et al. 2002). These normal anatomic and physiologic variations result in a greater risk of non-osseous injury to pediatric patients as compared to adults. The increased elasticity of the spinal support structures persists throughout the first decade of life, and most injuries at the cranial–cervical junction are seen in children younger than 8 years old (Flynn et al. 2002) Spinal cord injury without obvious radiographic abnormality, or SCIWORA, has been noted in 23–67% of pediatric spinal cord injuries according to some studies, and it is thought to be more common in younger children (Flanders and Croul 2001). The pathophysiology of SCIWORA in children is due to spinal cord stretching, while the same entity in adults is due to spinal cord compression with preexisting spinal stenosis or osteophyte formation (Flynn et al. 2002). In a study by Flynn et al., it was noted that 23.4% of pediatric trauma patients with normal plain films had abnormal findings on MRI (Flynn et al. 2002). In the same study, management of pediatric trauma patients was altered in 34 % of cases, and plain film diagnosis was confirmed in the other 66 % of cases (Flynn et al. 2002). 4.3.4.9  Subacute Combined Degeneration Last, we discuss a process that does not quite fit into our selected categories. Few metabolic conditions are encountered in imaging of the spinal cord. One rather characteristic entity is subacute combined degeneration as a result of vitamin B12 deficiency. The process is thought to cause demyelination and vacuolation in the dorsal and lateral columns of the spinal cord (Timms et al. 1993). Clinically, patients can presents with position sense disturbance, bilateral dysesthesias, spastic paraparesis, or tetraparesis (Hemmer et al. 1998). The lower extremities are usually more affected than are the upper extremities. The process often begins in the thoracic levels and may progress to involve other regions (Timms et al. 1993). Imaging findings on MR can be fairly characteristic when present. Signal abnormality within the dorsal columns is typical, and is usually best seen on T2-weighted images. On sagittal images, this is seen as an intermediate-to-long segment of continuous linear T2 hyperintensity in the posterior spinal cord. On axial images, focal T2 hyperintensity will be present within the dorsal columns (Fig. 4.2.44) Corresponding hypointensity may be

Fig. 4.3.44  Subacute Combined Degeneration. This patient was a 69-year-old female who presented with leg weakness and gait instability. Sagittal T2-weighted (a) and STIR images (b) demonstrate linear hyperintensity in the posterior spinal cord, which extends throughout most of the cervical spinal cord. Axial T2-weighted images (c) demonstrate that the signal abnormality is within both dorsal columns. An axial post-contrast T1-weighted image (d) shows the lesions to be hypointense and non-enhancing

seen on T1-weighted imaging. The lesions may demonstrate enhancement. Imaging findings may resolve after treatment with cobalamin (vitamin B12) monthly injections. The clinical response to treatment is better when the disease is treated earlier in its course (Ravina et al. 2000; Timms et al. 1993). References 1.

2.

3.

Aminoff MJ, Barnard RO, Valentine L (1974) The pathophysiology of spinal vascular malformations. J Neurol Sci 23:255–263 Andersson T, van Dijk JMC, Willinsky RA (2003) Venous manifestations of spinal arteriovenous fistulas. Neuroimaging Clin N Am 13:73–93 Anson JA, Spetzler RF (1992) Classification of spinal arteriovenous malformations and implications for treatment. Barrow Neurological Institute Quarterly 8:2–8

659

660

4  Spine and Spinal Canal 4.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18. 19.

20.

Baker KB, Moran CJ, Wippold FJ et al (2000) MR imaging of spinal hemangioblastoma. AJR Am J Roentgenol 174:377–382 Bemporad JA, Sze GS (2000) MR imaging of spinal cord vascular malformations with an emphasis on the cervical spine. Magn Reson Imaging Clin N Am 8:581–596 Berger JR, Sabet A (2002) Infectious myelopathies. Semin Neurol 22:133–142 Bot JCJ, Barkhof F, Polman, CH, Lyklama à Nijeholt GJ, de Groot V, Bergers E, Ader HJ, Castelijns JA (2004a) Spinal cord abnormalities in recently diagnosed MS patients: added value of spinal MRI examination. Neurology 62:226–233 Bot JCJ, Blezer EL, Kamphorst W, Lycklama à Nijeholt GJ, Ader HJ, Castelijns JA, Ig KN, Bergers E, Ravid R, Polman C, Barkhof F (2004b) The spinal cord in multiple sclerosis: relationship of high-spatial-resolution quantitative MR imaging findings to histopathologic results. Radiology 233:531–540 Bowen BC, DePrima S, Pattany PM, Marcillo A, Madsen P, Quencer RM (1996) MR angiography of normal intradural vessels of the thoracolumbar spine. AJNR Am J Neuroradiol 17:483–494 Browne TR, Adams RD, Roberson GH (1976) Hemangioblastomas of the spinal cord: review and report of 5 cases. Arch Neurol 33:435–441 Bydder G, Brown J, Niendorf HP et al (1985) Enhancement of cervical intraspinal tumors in MR imaging with intravenous gadolinium-DTPA. J Comut Assist Tomogr 9:847–851 Caldemeyer KS, Smith RR, Harris TM, Edwards MK (1994) MRI in acute disseminated encephalomyelitis. Neuroradiology 36:216–20 Choi KH, Lee KS et al (1996) Idiopathic transverse myelitis: MR characteristics. AJNR Am J Neuroradiol 17:1151–1160 Chong J, Di Rocco A, Tagliati M, Danisi F, Simpson DM, Atlas SW (1999) MR findings in AIDS-associated myelopathy. AJNR Am J Neuroradiology 20:1412–1416 Costigan DA, Winkelman MD (1985) Intramedullary spinal cord metastasis: a clinicopathological study of 13 cases. J Neurosurg 62:227–233 Criscuolo GR, Oldfield EH, Doppman JL (1989) Reversible acute and subacute myelopathy with dural arteriovenous fistulas. J Neurosurg 70:354–359 Crum B, Mokri B, Fulgham J (2000) Spinal manifestations of vertebral artery dissection. Neurology 55:304–306 Dubendorf P (1999) Spinal cord injury pathophysiology. Crit Care Nurs Q 22:31–35 Facon D, Ozanne A et al (2005) MR diffusion tensor imaging and fiber tracking in spinal cord compression. AJNR Am J Neuroradiol 26:1587–1594 Feydy A, Carlier R et al (1997) Brain and spinal cord MR imaging in a case of acute disseminated encephalomyelitis. Eur Radiol 7:415–417

21. Filippi M, Rocca MA, Moiola L, Martinelli V, Ghezzi A, Capra R, Salvi F, Comi G (1999) MRI and magnetization transfer imaging changes in the brain and cervical cord of patients with Devic’s neuromyelitis optica. Neurology 53:1705–1710 22. Flanders AE, Croul SE (2001) Spinal Trauma. In: Atlas SW (ed) Magnetic resonance imaging of the brain and spine, 3rd edn. Lippincott Williams & Wilkins, Philadelphia, pp 1769–1824 23. Flynn JM, Closkey RF, Mahboubi S, Dormans JP (2002) Role of magnetic resonance imaging in the assessment of pediatric cervical spine injuries. J Pediatr Orthop 22:573–577 24. Fujikawa A, Tsuchiya K, Takeuchi S, Hachiya J (2004) Diffusion-weighted MR imaging in acute spinal cord ischemia. Eur Radiol 14:2076–2078 25. Garg RK (2003) Acute disseminated encephalomyelitis. Postgrad Med J 79:11–17 26. Gilbertson JR, Miller GM, Goldman MS, Marsh WR (1995) Spinal dural arteriovenous fistulas: MR and myelographic findings. AJNR Am J Neuroradiol 16:2049–2057 27. Gupta RK, Gupta S, Kumar S, Kohli A, Misra UK, Gujral RB (1994) MRI in intraspinal tuberculosis. Neuroradiology 36:39–43 28. Hemmer B, Glocker FX et al (1998) Subacute combined degeneration: clinical, electrophysiological, and magnetic resonance imaging findings. J Neurol Neurosurg Psychiatr 65:822–827 29. Hurst RW (2001) Vascular disorders of the spine and spinal cord. In: Atlas SW (ed) Magnetic resonance imaging of the brain and spine, 3rd edn. Lippincott Williams & Wilkins, Philadelphia, pp 1825–1854 30. Jellinger KA (1997) Spinal cord watershed. Neurology 48:1474 31. Jellinger K, Kothbauer P, Sunder-Plassman E et al (1979) Intramedullary spinal cord metastasis. Neurology 22:122–1231 32. Junger SS, Stern BJ, Levine SR, Sipos E, Marti-Masso JF (1993) Intramedullary spinal sarcoidosis: clinical and magnetic resonance imaging characteristics. Neurology 43:333–337 33. Khong P-L, Ho H-K et al (2002) Childhood acute disseminated encephalomyelitis: the role of brain and spinal cord MRI. Pediatr Radiol 32:59–56 34. Kim RC, Smith HR, Henbest ML, Choi BH (1984) Nonhemorrhagic venous infarction of the spinal cord. Ann Neurol 15:379–385 35. Komachi H, Tsuchiya K et al (1995) Radiation myelopathy: a clinicopathological study with special reference to correlation between MRI findings and neuropathology. Journal of the neurologic sciences 132:228–232 36. Koskiniemi M, Rantalaiho T (2001) Infections of the central nervous system of suspected viral origin: a collaborative study from Finland. J Neurovirol 7:400–408 37. Krings T, Mull M, Gilsbach JM, Thron A (2005) Spinal vascular malformations. Eur Radiol 15:267–278

4.3  Intramedullary Diseases of the Spinal Cord 38. Krzyzak AT, Jasinski AJ, Weglarz WP, Adamek D, Sagnowski P, Baj M (2005) Visualization of the extent of damage in a rat spinal cord injury model using MR microscopy of the water diffusion tensor. Acta Neurobiol Exp (Wars) 65:255–264 39. Kulkarni MV, Bondurant FJ, Rose SL, Narayana PA (1988) 1.5 Tesla magnetic resonance imaging of acute spinal trauma. Radiographics 8:1059–1082 40. Kurtzke JF, Heltberg A (2001) Multiple sclerosis in the Faroe Islands: an epitome. J Clin Epidemiol 54:1–22 41. Lasjaunias P, Berenstein A (1987) Spine and spinal cord arteries and veins. In; Lasjaunias P, Berenstein A (eds) Surgical neuroangiography, vol. 3. Springer, Berlin Hedilberg New York, pp 15–87 42. Losseff NA, Webb SL, O’Riordan JI, Page R, Wang L, Barker GJ, Tofts PS, McDonald WI, Miller DH, Thompson AJ (1996) Spinal cord atrophy and disability in multiple sclerosis. A new reproducible and sensitive MRI method with potential to monitor disease progression. Brain 119(Pt 3):701–708 43. Lowe G (2000) Magnetic resonance imaging of intramedullary spinal cord tumors. J Neurooncol 47:195–210 44. Lublin FD, Reingold SC (1996) Defining the course of multiple sclerosis: results of an international survey. Neurology 46:907–911 45. Lycklama à Nijeholt GJ, Thompson A, Filippi M, Miller D, Polman C, Fazekas F, Barkhof, F (2003) Spinal-cord MRI in multiple sclerosis. Lancet Neurol 2:555–62 46. Marchioni E, Ravaglia S et al (2005) Postinfectious inflammatory disorders: subgroups based on prospective followup. Neurology 65:1057–1065 47. Mascalchi M, Ferrito G, Quilici N et al (2001) Spinal vascular malformations: MR angiography after treatment. Radiology 219:346–353 48. Masson C, Pruvo JP, Meder JF, Cordonnier CT et al (2004) Spinal cord infarction: clinical and magnetic resonance imaging findings and short term outcome. J Neurol Neurosurg Psychiatry 75:1431–1435 49. McDonald WI, Compston A, Edan G, Goodkin D, Hartung H-P, Lublin, FD, McFarland HF, Paty DW, Polman CH, Reingold SC, Sandberg-Wollheim M, Sibley W, Thompson A, van den Noort S, Weinshenker BY, Wolinsky JS (2001) Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 50:121–127 50. Medonca AM (2001) Spinal infection and inflammatory disorders. In: Atlas SW (ed) Magnetic resonance imaging of the brain and spine, 3rd edn. Lippincott Williams & Wilkins, Philadelphia, pp 1855–1869 51. Mikulis DJ, Ogilvy CS, Mckee A, Davis KR, Ojeman RG (1992) Spinal cord infarction and fibrocartilagenous emboli. AJNR Am J Neuroradiol 13:155–160 52. Mishra R, Kaw R (2005) Foix-Alajouanine syndrome: an uncommon cause of myelopathy from an anatomic variant circulation. South Med J 98:567–569

53. Mouzaki A, Tselios T, Papathanassopoulos P, Matsoukas I, Chatzantoni K (2004) Immunotherapy for multiple sclerosis: insights into new clinical strategies. Clin Neurovasc Res 1:325–340 54. Niino M, Isu T, Tashiro K (1999) Nonhemorrhagic venous infarction of the spinal cord without spinal vascular malformation. J Neurol 246:852–854 55. Pallis CA, Louis S, Morgan RL (1961) Radiation myelopathy. Brain 84:460–479 56. Parmar H, Shah J, Patkar D, Varma R (2000) Intramedullary tuberculomas. MR findings in seven patients. Acta Radiol 41:572–577 57. Pattany PM, Saraf-Lavi E, Bowen BC (2003) MR angiography of the spine and spinal cord. Top Magn Reson Imaging 14:444–460 58. Potter K, Saifuddin A (2003) MRI of chronic spinal cord injury. Br J Radiol 76:347–352 59. Rampling R, Symonds P (1998) Radiation myelopathy. Curr Opin Neurol, 11:627–632 60. Ravina B, Loevner LA, Bank W (2000) MR findings in subacute combined degeneration of the spinal cord: a case of reversible cervical myelopathy. AJR Am J Roentgenol 174:863–865 61. Reimer R, Onofrio BM (1985) Astrocytomas of the spinal cord in children and adolescents. J Neurosurg 63:669–675 62. Rocca MA, Mastronardo G, Horsfield MA, Pereira C, Iannucci G, Colombo B, Moiola L, Comi G, Fillipi M (1999) Comparison of three MR sequences for the detection of cervical cord lesions in patients with multiple sclerosis. AJNR Am J Neuroradiol 20:1710–1716 63. Ross JS (2004) Diagnostic imaging: spine. AMIRSYS, Salt Lake City 64. Saifuddin A (2001) MRI of acute spinal trauma. Skeletal Radiology 30:237–246 65. Santoro A, Piccirilli M, Frati A et al (2004) Intramedullary spinal cord cavernous malformations: report of 10 new cases. Neurosurg Rev 27:93–98 66. Saraf-Lavi E, Bowen BC, Quencer RM, Sklar EML, Holz A, Falcone S, Latchaw RE, Duncan R and Wakhloo A (2002) Detection of spinal dural arteriovenous fistulae with MR imaging and contrast-enhanced MR angiography: sensitivity, specificity, and prediction of vertebral level. AJNR Am J Neuroradiol 23:858–867 67. Savvas A, Albuquerque-Jonathan G et al (2003) MRI findings in acute idiopathic transverse myelopathy. Pediatr Radiol 33: 624–629 68. Schultheiss TE, Stephens LC (1992) Radiation myelopathy. AJNR Am J Neuroradiol 13:1056–1058 69. Scotti G, Gerevini S (2001) Diagnosis of acute transverse myelopathy. The role of neuroradiological investigations and review of the literature. Neurol Sci 22:S69–S73 70. Selden N, Quint D, Patel N, d’Arcy H, Papadopoulos S (1999) Emergency magnetic resonance imaging of cervical spinal injuries: clinical correlation and prognosis. Neurosurgery Online 44:785–792

661

662

4  Spine and Spinal Canal 71. Singh S, Alexander M, Korah IP (1999) Acute disseminated encephalomyelitis: MR imaging features. AJR Am J Roentgenol 173:1101–1107 72. Singh S, Prabhakar S et al (2000) Acute disseminated encephalomyelitis and multiple sclerosis: magnetic resonance imaging differentiation. Australas Radiol 44:404–411 73. Suzuki K, Meguro K, Wada M, Nakai K, Nose T (1998) Anterior spinal artery syndrome associated with severe stenosis of the vertebral artery. AJNR Am J Neuroradiol 19:1353–1355 74. Sze G (2002) Neoplastic disease of the spine and spinal cord. In: Atlas S (ed) Magnetic resonance imaging of the brain and spine. Lippincott Williams & Wilkins, Philadelphia, pp 1715–1768 75. Sze G, Krol G, Zimmerman RD et al (1988) Intramedullary disease of the spine: diagnosis using gadolinium-DTPA enhanced MR imaging. Am J Neuroradiol 9:847–858 76. Takhtani D, Melhem ER (2000) MR imaging in cervical spinal trauma. Magn Reson Imaging Clin N Am 8:615–633 77. Tartaglia MC, Arnold DL (2006) The role of MRS and fMRI in multiple sclerosis. Adv Neurol 98:185–202 78. Tartaglino LM, Friedman DP, Flanders AE, Lublin FD, Knobler RL, Liem M (1995) Multiple sclerosis in the spinal cord: MR appearance and correlation with clinical parameters. Radiology 195:725–732 79. Tench CR, Morgan PS, Jaspan T, Auer DP, Constantnescu CS (2005) Spinal cord imaging in multiple sclerosis. J Neuroimging 15:94S–102S 80. Thron, AK (1988) Vascular anatomy of the spinal cord: neuroradiological investigations and clinical syndromes. Springer, Berlin Heidelberg New York 81. Timms SR, Cure JK, Kurent JE (1993) Subacute combined degeneration of the spinal cord: MR findings. AJNR 14:1224–1227

82. Wadia NH, Dastur DK (1969) Spinal meningitides with radiculo-myelopathy: clinical and radiological features. J Neurol Sci 8:239–260 83. Wang P, Shen W(1991) Magnetic resonance imaging in two patients with radiation myelopathy. J Formosan Med Assoc 90:583–585 84. Wang P, Shen W, Jan J (1992) MR imaging in radiation myelopathy. AJNR Am J Neuroradiol 13:1049–1055 85. Weidauer S, Nichtweiss M, Lanfermann H, Zanella FE (2002) Spinal cord infarction: MR imaging and clinical features in 16 cases. Neuroradiology 44:851–857 86. White ML (2000) Cervical spine: MR imaging techniques and anatomy. Magn Reson Imaging Clin N Am 8:453–469 87. Willinsky RA, Brugge K, Montanera W et al (1995) Posttreatment MR findings in spinal dural arteriovenous malformations. AJNR Am J Neuroradiol 16:2063–2071 88. Wingerchuk DM, Hogancamp WF, O’Brien PC, Weinshenker BG (1999) The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 53:1107–1114 89. Yoshioka K, Niinuma H, Ohira A, Nasu K, Kawakami T, Sasaki M, Kawazoe K (2003) MR angiography and CT angiography of the artery of Adamkiewicz: noninvasive preoperative assessment of thoracoabdominal aortic aneurysm. Radiographics 23:1215–1225 90. Yuh WT, Marsh EE 3rd, Wang AK, Russell JW, Chiang F, Koci TM, Ryals TJ (1992) MR imaging of spinal cord and vertebral body infarction. AJNR Am J Neuroradiol 13:145–154 91. Zimmerman RA, Bilaniuk LT (1988) Imaging of tumors of the spinal canal and cord. Rad Clin North Am 26:965–1007 92. Zweig G, Russell EJ (1990) Radiation myelopathy of the cervical spinal cord: MR findings. AJNR Am J Neuroradiol 11:1188–1190

Chapter 5

Thorax and Vasculature

5.1 

Lungs, Pleura, and Mediastinum  . . . . . .   666 G. Layer and H.U. Kauczor

5.1.1 

General Requirements for Imaging of Thoracic Organs  . . . . . . . .   666

5.1.2 

Basic MR Sequences for Imaging of the Chest  .. . . . . . . . . . . . . .   666

5

5.1.7.9  Asthma  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   682 5.1.7.10  Neoplasia  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   683 5.1.8 

Diseases of the Pleura  . . . . . . . . . . . . . . . . .   688

5.1.8.1  Pleural Effusion  .. . . . . . . . . . . . . . . . . . . . . .   688 5.1.8.2  Pleural Mesothelioma  . . . . . . . . . . . . . . . . .   689

5.1.2.1  Lung Parenchyma  .. . . . . . . . . . . . . . . . . . . .   667

5.1.9 

5.1.2.2  Mediastinum  . . . . . . . . . . . . . . . . . . . . . . . . .   668

5.1.9.1  Lymphoma  . . . . . . . . . . . . . . . . . . . . . . . . . . .   690

5.1.2.3  Chest Wall  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   668

5.1.9.2  Thymoma  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   692

5.1.2.4  Proposal for a Comprehensive Standard Protocol  .. . . . . . . . . . . . . . . . . . . .   668

5.1.9.3  Goiter  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   692

5.1.2.5  Extensions to the Standard Protocol  . . . .   670 5.1.2.6  Lung Imaging with Open Low-Field Scanners or Ultrahigh-Field Scanners  .. .   670 5.1.3 

MRI of Ventilation  . . . . . . . . . . . . . . . . . . . .   670

5.1.4 

Use of Contrast Agents  . . . . . . . . . . . . . . . .   671

5.1.5 

Signal Intensities and Contrast Behavior    672

5.1.6 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   672

Mediastinal Disease   .. . . . . . . . . . . . . . . . . .   690

5.1.9.4  Cysts and Inflammation  . . . . . . . . . . . . . . .   692 5.1.9.5  Neurogenic Tumors  . . . . . . . . . . . . . . . . . . .   693 5.1.10 

Differential Diagnosis  . . . . . . . . . . . . . . . . .   693

5.1.11 

Value of MRI with Regard to Other Imaging Modalities  .. . . . . . . . . .   694

5.1.12 

Diagnostic Procedure  . . . . . . . . . . . . . . . . .   696 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   696

5.2 

High-Risk Screening Breast MRI  . . . . . .   700 E.A. Morris

5.1.6.2  Mediastinum  . . . . . . . . . . . . . . . . . . . . . . . . .   672

5.2.1 

Importance of Early Detection  . . . . . . . . .   700

5.1.6.3  Pleura and Chest Wall  .. . . . . . . . . . . . . . . .   674

5.2.2 

Pathology of Breast Cancer: What Are We Looking for?  .. . . . . . . . . . . .   701

5.2.3 

Why Consider MRI?  . . . . . . . . . . . . . . . . . .   701

5.2.4 

Defining the High-Risk Population  . . . . .   701

5.1.6.1  Lungs  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   672

5.1.7 

Lung Diseases  . . . . . . . . . . . . . . . . . . . . . . . .   674

5.1.7.1  Congenital Pathologies  . . . . . . . . . . . . . . . .   674 5.1.7.2  Infiltrative Lung Diseases  . . . . . . . . . . . . . .   677 5.1.7.3  Atelectasis  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   677 5.1.7.4  Pneumonia  .. . . . . . . . . . . . . . . . . . . . . . . . . .   678 5.1.7.5  Pulmonary Edema  . . . . . . . . . . . . . . . . . . . .   678 5.1.7.6  Diffuse Interstitial Lung Disease  .. . . . . . .   679 5.1.7.7  Emphysema and COPD  . . . . . . . . . . . . . . .   679 5.1.7.8  Cystic Fibrosis  . . . . . . . . . . . . . . . . . . . . . . . .   681

5.2.4.1  Family History  .. . . . . . . . . . . . . . . . . . . . . . .   703 5.2.4.2  Genetic Testing  . . . . . . . . . . . . . . . . . . . . . . .   704 5.2.4.3  Clinical History  .. . . . . . . . . . . . . . . . . . . . . .   704 5.2.5 

Overview of High-Risk MRI Screening Studies  . . . . . . . . . . . . . . . .   704

5.2.6 

Description of High-Risk Screening MRI Studies  . . . . . . . . . . . . . . . . . . . . . . . . . .   704

664

5  Thorax and Vasculature 5.2.7 

National Guidelines  . . . . . . . . . . . . . . . . . . .   706

5.3.2.16  Congenitally Corrected Transposition  . .   731

5.2.8 

Current Issues with Using MRI for Screening  . . . . . . . . . . . . . . . . . . . . . . . . .   706

5.3.2.17  Heterotaxia Syndrome  .. . . . . . . . . . . . . . . .   731

5.2.9 

Increased Call-Backs and Biopsies  .. . . . .   707

5.2.10 

Inconsistency of DCIS Detection  .. . . . . .   707

5.2.11 

MRI Interpretation   . . . . . . . . . . . . . . . . . . .   707

5.2.12 

MRI Technique  . . . . . . . . . . . . . . . . . . . . . . .   708

5.2.13 

Research Needed  .. . . . . . . . . . . . . . . . . . . . .   708

5.2.14 

Summary  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   709 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   709

5.3 

Heart  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   711

5.3.1 

Acquisition Techniques and Protocols  . .   711 B.J. Wintersperger

5.3.1.1  Basic Considerations  . . . . . . . . . . . . . . . . . .   711 5.3.1.2  MR Hardware and Software for Cardiac Imaging  .. . . . . . . . . . . . . . . . . .   711 5.3.1.3  Techniques in Cardiac MRI  .. . . . . . . . . . .   711

5.3.2.18  Asplenia Syndrome (Right Isomerism)     732 5.3.2.19  Polysplenia Syndrome (Left Isomerism)    732 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   732 5.3.3 

5.3.3.1  Introduction: Background Information   and Role of MRI  . . . . . . . . . . . . . . . . . . . . . .   734 5.3.3.2  Dilated Cardiomyopathy  .. . . . . . . . . . . . . .   735 5.3.3.3  Hypertrophic Cardiomyopathy  .. . . . . . . .   736 5.3.3.4  Restrictive Cardiomyopathy  .. . . . . . . . . . .   739 5.3.3.5  Arrhythmogenic Right Ventricular Dysplasia  .. . . . . . . . . . . . . . . . .   740 5.3.3.6  Future Considerations and Conclusions    742 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   742 5.3.4 

References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   717 5.3.2 

Congenital Heart Disease: Cardiac Anomalies and Malformations     718 T.R.C. Johnson

5.3.2.1  Situs, Atrial, and Ventricular Morphology  .. . . . . . . . . .   719 5.3.2.2  Shunts   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   719 5.3.2.3  Atrial Septal Defect  . . . . . . . . . . . . . . . . . . .   719 5.3.2.4  Ventricular Septal Defect  . . . . . . . . . . . . . .   721 5.3.2.5  Endocardial Cushion Defects  . . . . . . . . . .   721 5.3.2.6  Patent Ductus Arteriosus  . . . . . . . . . . . . . .   721 5.3.2.7  Atrioventricular Malformations  . . . . . . . .   721 5.3.2.8  Univentricular Heart   .. . . . . . . . . . . . . . . . .   724 5.3.2.9  Tricuspid Atresia  .. . . . . . . . . . . . . . . . . . . . .   724 5.3.2.10  Ebstein’s Anomaly  .. . . . . . . . . . . . . . . . . . . .   724 5.3.2.11  Truncus Arteriosus  .. . . . . . . . . . . . . . . . . . .   724 5.3.2.12  Vascular Anomalies   .. . . . . . . . . . . . . . . . . .   726 5.3.2.13  Pulmonary Artery Anomalies  .. . . . . . . . .   727 5.3.2.14  Anomalies of the Aorta and Supra-Aortic Vessels  . . . . . . . . . . . . . .   729 5.3.2.15  Complete Transposition of the Great Arteries  .. . . . . . . . . . . . . . . . . .   730

Primary Cardiomyopathies  . . . . . . . . . . . .   734 K. Nikolaou

Secondary Cardiomyopathies and Specific Heart Muscle Diseases  .. . . .   744 A. Huber

5.3.4.1  Amyloidosis  .. . . . . . . . . . . . . . . . . . . . . . . . .   744 5.3.4.2  Sarcoidosis  . . . . . . . . . . . . . . . . . . . . . . . . . . .   746 5.3.4.3  Iron-Overload Cardiomyopathy  .. . . . . . .   748 5.3.4.4  Endomyocardial Disease (Löffler’s Endocarditis   and Endomyocardial Fibrosis)  .. . . . . . . . .   749 5.3.4.5  Metabolic Storage Disease  . . . . . . . . . . . . .   749 5.3.4.6  Friedreich’s Ataxia  .. . . . . . . . . . . . . . . . . . . .   750 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   750 5.3.5 

Valvular Heart Diseases  .. . . . . . . . . . . . . . .   751 A. Huber

5.3.5.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . . .   751 5.3.5.2  Conventional Imaging Modalities  . . . . . .   751 5.3.5.3  MRI of Cardiac Valves  .. . . . . . . . . . . . . . . .   752 References   . . . . . . . . . . . . . . . . . . . . . . . . . . .   760 5.3.6 

Pericardial Diseases  . . . . . . . . . . . . . . . . . . .   761 K. Bauner

5.3.6.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   761 5.3.6.2  Imaging of the Pericardium  .. . . . . . . . . . .   762 5.3.6.3  Pericardial Diseases  . . . . . . . . . . . . . . . . . . .   762 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   766

5  Thorax and Vasculature 5.3.7 

Ischemic Heart Disease  .. . . . . . . . . . . . . . .   766 K. Nikolaou

5.3.7.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   766 5.3.7.2  Coronary MR Angiography and Bypass MR Angiography  . . . . . . . . . .   767 5.3.7.3  Myocardial Function  . . . . . . . . . . . . . . . . . .   769 5.3.7.4  Myocardial Viability: Delayed-Enhancement MRI  . . . . . . . . . . .   771 5.3.7.5  Myocardial Perfusion  .. . . . . . . . . . . . . . . . .   773 5.3.7.6  Future Perspectives and Conclusions  . . .   774 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   775 5.3.8 

Cardiac Tumors  .. . . . . . . . . . . . . . . . . . . . . .   778 B.J. Wintersperger

5.3.8.1  Basic Consideration in MRI  .. . . . . . . . . . .   778 5.3.8.2  Epidemiology of Cardiac Masses  . . . . . . .   779 5.3.8.3  Benign Cardiac Tumors  . . . . . . . . . . . . . . .   779 5.3.8.4  Malignant Cardiac Tumors  .. . . . . . . . . . . .   782 5.3.8.5  Cardiac Pseudotumors or Tumor-Like Lesions  . . . . . . . . . . . . . . . .   785 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   786 5.4 

MR Angiography  .. . . . . . . . . . . . . . . . . . . .   788

5.4.1 

MRA Techniques and Acquisition Techniques  .. . . . . . . . . . .   788 H.J. Michaely and S.O. Schönberg

5.4.3.3  Clinical Applications  . . . . . . . . . . . . . . . . . .   810 5.4.3.4  MRA of the Intracranial Vasculature  .. . .   813 5.4.4 

MRA of the Renal Arteries  .. . . . . . . . . . . .   817 S.O. Schönberg and H.J. Michaely

5.4.4.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   817 5.4.4.2  Anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   817 5.4.4.3  Renal Artery Disease  .. . . . . . . . . . . . . . . . .   818 5.4.4.4  Renal Vein Thrombosis   . . . . . . . . . . . . . . .   824 5.4.4.5  Transplantation  .. . . . . . . . . . . . . . . . . . . . . .   824 5.4.4.6  Challenges for Renal MRA for Replacement of DSA  .. . . . . . . . . . . . . .   827 5.4.5 

Diseases of the Aorta  .. . . . . . . . . . . . . . . . .   829 D. Theisen

5.4.5.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   829 5.4.5.2  Congenital Anomalies  .. . . . . . . . . . . . . . . .   829 5.4.5.3  Acquired Diseases  .. . . . . . . . . . . . . . . . . . . .   830 5.4.5.4  Aortic Trauma  .. . . . . . . . . . . . . . . . . . . . . . .   836 5.4.5.5  Aortic Stent-Graft Imaging  .. . . . . . . . . . . .   836 5.4.6 

Peripheral MRA  . . . . . . . . . . . . . . . . . . . . . .   837 H.J. Michaely and H. Kramer

5.4.6.1  Peripheral Arteries  .. . . . . . . . . . . . . . . . . . .   837 5.4.6.2  Arteriosclerosis  . . . . . . . . . . . . . . . . . . . . . . .   837 5.4.6.3  Atherosclerosis  . . . . . . . . . . . . . . . . . . . . . . .   837

5.4.1.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   788

5.4.6.4  Thoracic Outlet Syndrome  .. . . . . . . . . . . .   841

5.4.1.2  MRA Techniques  . . . . . . . . . . . . . . . . . . . . .   788

5.4.6.5  Raynaud’s Syndrome  . . . . . . . . . . . . . . . . . .   841

5.4.1.3  Contrast Agents for MRA  .. . . . . . . . . . . . .   793

5.4.6.6  Peripheral Veins and Venous Thrombosis  . . . . . . . . . . . . . . .   842

5.4.1.4  General Technical Considerations  . . . . . .   794 5.4.1.5  Clinical Applications  . . . . . . . . . . . . . . . . . .   802 5.4.2 

Pulmonary MRA  .. . . . . . . . . . . . . . . . . . . . .   804 C. Fink

5.4.2.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   804 5.4.2.2  Technical Considerations  .. . . . . . . . . . . . .   804 5.4.2.3  Clinical Applications  . . . . . . . . . . . . . . . . . .   805 5.4.3 

MRA of the Supra-Aortic and Intracranial Vasculature  . . . . . . . . . . .   809 C. Fink, U. Attenberger, H.J. Michaely, and S.O. Schönberg

5.4.3.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . . .   809 5.4.3.2  Imaging Technique  .. . . . . . . . . . . . . . . . . . .   809

5.4.6.7  Vascular Malformations  . . . . . . . . . . . . . . .   842 5.4.6.8  Vasculitis and Inflammatory Diseases  .. .   843 5.4.6.9  Hereditary Disorders with Peripheral Arterial Involvement  . . .   845 5.4.6.10  Popliteal Artery Entrapment  . . . . . . . . . . .   845 5.4.6.11  Imaging of Hemodialysis Shunts   .. . . . . .   845 5.4.7 

Whole-Body MRA  . . . . . . . . . . . . . . . . . . . .   846 H. Kramer and H. Schlemmer

5.4.7.1  Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   846 5.4.7.2  Angiographic Modalities  . . . . . . . . . . . . . .   846 5.4.7.3  Whole-Body MRA  . . . . . . . . . . . . . . . . . . . .   846 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   851

665

666

5  Thorax and Vasculature

5.1  Lungs, Pleura, and Mediastinum G. Layer and H.U. Kauczor 5.1.1  General Requirements for Imaging of Thoracic Organs While MRI is an imaging method of first choice in almost all parts of the human body its acceptance for thoracic imaging is rather low even today. One of the reasons is of course that CT has developed dramatically. Multidetector spiral technique (MDCT) allows for isotropic voxels and reconstruction in any user-defined section; the method is robust and fast. The major problem in imaging thoracic organs by MRI is the continuous motion of all compo­ nents induced by heart pulsation and breathing. Both are most prominent in the lower and anterior sections of the chest. Technical challenges to overcome these effects are one major reason why MRI of the chest was for a long time limited to the posterior chest wall and the thoracic outlet. Both locations are relatively static and could be examined with classical T1- and T2-weighted spin-echo and fast spin-echo techniques. The easiest way to overcome respiratory motion is to use breath-hold techniques that ideally cover the whole thorax within a single 20-s breath hold (Alsop et al. 1995). Longer acquisitions can be split into a few clusters that are then acquired within several breath holds. However, splitting the acquisition introduces additional artifacts, especially if the breath holds are not reproducible (Biederer et al. 2003). Cardiac pulsation may be overcome with single shot techniques such as T2 HASTE or ultra fast turbo spin-echo UTSE using very short echo times (Lutterbey et al. 1996). The disadvantage of the breathhold techniques is their usually lower spatial resolution and lower signal-to-noise ratio. For the chest wall and the mediastinum, clear T1 and T2 contrast can be mandatory, in particular for the characterization of mediastinal masses. Since conventional spin-echo and fast spin-echo imaging requires acquisition times of a couple of minutes, dedicated techniques are needed for the direct compensation of cardiac and respiratory motion. The most straightforward method to directly compensate for respiratory motion is a simple pneumatic belt or a compressible cushion placed at the upper abdomen or the chest of the patient. The ideal position of the device depends on the individual breathing pattern of the patient at rest. The pressure changes due to compression and decompression of the device are directly registered by the MR scanner. The trigger signal is usually set to expiration, because this is the longest and most reproducible phase of the respiratory cycle (Biederer et al. 2002a; Plathow et al. 2005). Imaging at end expiration might be favorable for imaging lung parenchymal disease, since signal intensity increases with deflation. However, appropriate instructions of the patient remain

the key to high image quality without respiratory motion artifacts. More sophisticated techniques are navigator sequences for monitoring the movement of the diaphragm with continuous real time image acquisition from a small volume placed at the top of the diaphragm. The images are evaluated electronically and trigger settings can be applied interactively. As a disadvantage, respiratory motion cannot be monitored during acquisition of the diagnostic images when the navigator has to be suppressed. It is therefore recommended to additionally apply a respiratory belt, if available, for uninterrupted control of the breathing maneuvers. Similarly, navigator techniques can also be used to adjust several slice blocks of a multi-breathhold acquisition in case of a variable depth of inspiration. Control for cardiac pulsation artifacts is achieved with either ECG monitoring or peripheral pulse oximetry. However, any triggering or gating procedure prolongs acquisition time and compensation for motion artifacts usually remains incomplete (Biederer et al. 2002a). Simultaneous double-triggering or gating for respiration and cardiac pulsation is usually not favorable, although recommended by some authors (Leutner et al. 1999). Fast sequences such as single-shot turbo spin-echo, fast lowangle shot gradient-echo, and T1-weighted 3D gradientecho such as volume interpolated breath-hold examination (VIBE) are usually quite robust to cardiac motion. Their image quality largely depends on the breath-hold capability and compliance of the patient more than on compensation for cardiac motion. Particular problems are encountered, if MR is applied for imaging of lung parenchyma. MR signal originates from protons within water molecules or organic substances. The lungs contain only 800 g of tissue and blood that are distributed over a volume of 4–6 l. Thus, proton density and signal intensity are extremely low compared to other parts of the human body. Even more, local field inhomogeneities due to susceptibility artifacts at tissue–air or liquid–air interfaces of the alveoli result in extremely short T2*. Therefore, the usual appearance of the lung on conventional MR images is that of a “black hole” inside the thorax. However, this may be favorable, since any pathology with its higher proton density and therefore higher signal appears with a strong inherent contrast against the black surrounding lung tissue. However, subtle changes of lung signal due to small lesions or fine reticulations will be missed. 5.1.2  Basic MR Sequences for Imaging of the Chest MRI of the lung parenchyma requires fast sequences, preferably for breath-hold imaging with reasonably high spatial resolution and short TE, to receive as much lung signal as possible within the short interval before signal decay (Fig. 5.1.1).

5.1  Lungs, Pleura, and Mediastinum

Fig. 5.1.1  Basic MR sequence techniques for, e.g., lung cancer imaging. Illustrated is the case of a lung cancer in the left lower lobe with mediastinal lymphadenopathy and rib metastasis. a Axial 3D GRE (VIBE), b axial contrast-enhanced 3D GRE (VIBE), c coronal T2 TSE, d axial T2 HASTE, e axial T2 TIRM

5.1.2.1  Lung Parenchyma Ex vivo experiments and in vivo experience have shown that solid lung pathology can be well detected with fast T1-weighted gradient-echo sequences (T1 GRE). For lung nodules larger than 4–5 mm, 3D gradient-echo sequences reach the detection rates of conventional helical CT with single row detector technique. Only recently, it was shown that 3D sequences are superior to 2D techniques (Biederer et al. 2002a, 2003). Modern scanners with parallel imaging technique are capable of acquiring 3D data sets of the whole chest with voxel sizes as small as 1.6 × 1.6 × 4 mm or isotropic voxels of 2 × 2 × 2 mm

within one breath hold. Parallel imaging techniques use arrangements of multiple coils to acquire additional information along the phase encoding direction. k-Space data are partially replaced by spatial information, which is given by differences in the signal intensities of a certain location depending on its distance to the individual coils. The result is a substantial improvement in image acquisition speed, usually two- or threefold. For the lung, acceleration factors of 2–3 are reasonable; more than 3 cannot be recommended because of the unavoidable signal loss. If available, GRAPPA algorithms appear to produce a slightly better signal-to-noise ratio than does SENSE.

667

668

5  Thorax and Vasculature

T1 GRE images do not cover infiltrative processes very well. Experimental results show that the signal intensity of fluid inside lung tissue on T1 GRE is much too low to be of diagnostic use (Biederer et al. 2002a, b). Therefore it is commonly accepted practice to use primarily T2-weighted sequences for the evaluation of pulmonary infiltrates. For fast T2-weighted imaging, we recommend the use of single shot techniques with partial-Fourier acquisition (e.g., HASTE) or ultra short TE (UTSE). As an alternative to breath-hold T2 FSE, fast T2-weighted spin-echo sequences with respiratory triggering have produced reasonable results (Leutner et al. 2001; Both et al. 2005). A great advantage of MRI compared to CT is the potential to demonstrate functional aspects with timeresolved image series. For a basic protocol, axial steady state gradient-echo imaging (trueFISP) can be acquired with slice overlap during free breathing. The result is a series covering the whole chest. It yields basic information on pulmonary motion during the respiratory cycle as well as on the size, shape, and patency of the central pulmonary vessels. Thus, it can also be used as a fast technique to exclude massive acute pulmonary embolism (Kluge et al. 2006). An additional coronal series can be placed on the highest elevation of the diaphragmatic dome to be acquired with a temporal resolution of 3–10 images per second, depending on the performance of the MR scanner. 5.1.2.2  Mediastinum The mediastinum contains the heart, large vessels, trachea, esophagus, and neural structures as well as lymphatic tissues and the thoracic duct. The heart and great vessels are addressed in Sect. 5.3. Typical indications for cross-sectional imaging of the mediastinum are masses originating from the present structures. The size and position of a tumor can be assessed with MRI as well as with CT. Both modalities contribute to the characterization of tumors, e.g., with the detection of fat or calcifications inside a teratoma. In this case, CT has an even superior sensitivity for small calcifications. Usually, CT is acquired with the administration of contrast media to obtain a sufficient contrast between vessel and soft tissues while nonenhanced MRI may be sufficient in many cases due to its inherent and excellent soft tissue contrast. However, for the assessment of indeterminate lesions, the application of contrast media is also recommended for MRI. Non-enhanced MRI can benefit from black-blood techniques which reduce flow artifacts. This especially holds true for fast T2-weighted imaging where darkblood preparation allows for identification of vessel walls and better differentiation vessels and lymph nodes, but is used at the expense of signal intensity. ECG-triggered sequences provide excellent detail of structures close to the

heart. Navigator or respiratory belt-triggered sequences do not enhance image quality so much as justify the significantly longer acquisition time. For scanning mediastinal masses a quick method is the black-blood-prepared HASTE. Identification and classification of mediastinal processes can be made possible by using unsaturated and fat-saturated non-enhanced 3D gradient-echo sequences and repeating the latter after contrast administration. With an in-plane resolution of 1.6 mm, even very small lymph nodes can be detected. It is imperative, however, to apply fat saturation to contrast-enhanced images. Otherwise enhanced lymph nodes will “drown” within the signal of the surrounding fatty tissue. 5.1.2.3  Chest Wall The evaluation of the chest wall is of interest in cases of infiltration of peripheral bronchogenic carcinoma, pleural mesothelioma, or metastases. Other indications include inflammatory processes with empyema or abscess. There are only minor technical problems in imaging the relatively static chest wall with sufficient proton density and therefore adequate signal. Motion artifacts may be minimized with exact positioning of the coil at the location of the pathology. Tumors of the anterior chest wall may be favorably examined in prone position and with regard for the proper phase encoding direction. Standard T1-weighted and T2-weighted TSE sequences with respiratory gating are adequate in most cases. Contrast application is required in all tumor cases. To delineate the tumor-fat border, subtraction or fat saturation is essential. A T2-weighted fat-saturated sequence, e.g., STIR, is especially helpful for demarcating increased water content due to soft tissue infiltration or edema. 5.1.2.4  Proposal for a Comprehensive Standard Protocol The following recommendations for a thoracic imaging protocol refer to common MR sequence components of current standard installations. The necessary hardware is widely available. The recommended protocols of Table 5.1.1 are based on parallel acquisition techniques, but multi-breath-hold acquisitions can be used instead. The preparation for the examination includes instruction of the patient for the breathing maneuvers, the application of a respiratory belt and the selection of a phased array body coil for thoracic imaging. ECG is not required on a routine base but should be used if cardiac imaging sequences are planned, e.g., in case of tumor infiltration into the pericardium or large vessels of the mediastinum. The imaging protocol starts with a

5.1  Lungs, Pleura, and Mediastinum

gradient-echo localizer in inspiration. The first sequences are acquired in breath hold, usually starting with the coronal T2 HASTE followed by the transverse T1 3D GRE (VIBE). Then the first set of coronal SS GRE sequences is acquired in free breathing, giving the patient some time to recover from the breath-hold maneuvers. This is followed by the STIR image series, which is acquired with

multiple breath holds. The basic protocol is concluded with the single slice dynamic SS GRE series for diaphragmatic function. For this series, the patient is instructed to breathe deeply. After this, the standard protocol may be extended by additional series (e.g., after contrast administration) depending on the indication. Total room time up to this point will be approximately 15–20 min.

Table 5.1.1  Proposal for a sequence protocol for thoracic MRI of the lung, mediastinum or chest wall Thoracic MRI protocol

 CE

 Perfusion

 MRA

Sequence

2D FLASH localizer

T2 HASTE

T1 3D GRE

T2 TIRM

T1 3D GRE CM

4D MRA

3D FLASH angio

Respiratory phase

Inspiration/ breath hold

Inspiration/ breath hold

Inspiration/ breath hold

Inspiration/ breath hold

Inspiration/ breath hold

Inspiration

Inspiration

Slice orientation

Coronal/ transverse/ sagittal

Coronal

Transverse

Transverse

Transverse

Coronal

Coronal

Preparation

Anterior– posterior Right–left

Right–left

Anterior– posterior

Anterior– posterior

Anterior– posterior

Right–left

Right–left

FOV (mm) (FOV phase %)

500 (100)

450 (100)

400 (87.5)

400 (75)

400 (87.5)

500 (100)

500 (83.3)

Base resolution

256

256

256

320

256

256

384

Phase resolution (%)

75

100

100

75

100

54

90

Slice thickness (mm)

10

8

4

6

4

5

1.6

Phase partial Fourier

6/8

4/8

off

off

off

6/8

6/8

Pixel size (mm)

2.6 × 2.0

1.8 × 1.8

1.6 × 1.6

1.7 × 1.3

1.6 × 1.6

3.6 × 2.0

1.2 × 1.0

Distance factor (%)

50

0

20

10

20

20

20

TR (ms)

8.9

600

3.15

3500

3.15

1.64

2.75

TE (ms)

4.38

31

1.38

106

1.38

0.64

1.12

Flip-angle (°)

30

180

8

150

8

40

25

Band width (Hz/pixel)

180

610

500

252

500

1220

384

iPAT (no. of ref. lines)

0

2 (14)

2 (35)

2 (66)

2 (35)

2 (24)

2 (24)

Large FOV (dist. corr.)

Off

On

Off

Off

Off

Off

Off

669

670

5  Thorax and Vasculature

5.1.2.5  Extensions to the Standard Protocol For routine purposes it might appear sufficient to conclude the study if the non-enhanced scans show completely normal findings. Nevertheless, application of intravenous contrast material markedly improves the diagnostic yield of 3D GRE imaging of the lung by the clearer depiction of vessels, hilar structures, and pleural enhancement. Parenchymal disease and solid pathologies are also enhanced. Thus, a study to exclude pulmonary malignancies or for staging purposes should usually comprise a contrast-enhanced series, preferably with a fat-saturated 3D GRE sequence. Contrast enhancement is also necessary in case of pleural processes (empyema, abscess, metastatic spread of carcinoma, mesothelioma) or for the further evaluation of solid masses as well as for functional imaging or angiography (see Table 5.1.1 and chpater 5.4.2). Since the sequences listed so far only contain a non-fat-saturated 3D GRE sequence, it might be helpful to acquire an additional fat-saturated scan before intravenous contrast administration to allow for a direct comparison of contrast uptake. Further options to extend the standard protocol are T1- and T2-weighted SE or FSE sequences with respiratory triggering (or gating). Traditionally, T1-weighted images are recommended for the detection of lymph nodes and tumor infiltration into the chest wall, but only the T2-weighted sequences contribute to the evaluation of lung parenchyma pathology and provide equal information about the chest wall and mediastinum. We therefore did not include T1-SE series into a lean protocol, since all aspects are already covered by other sequences.

have been found to be particularly useful. Also known as SS FFE, TrueFISP or balanced steady-state acquisition with rewound gradient echo (BASG), they can be applied as 2D or 3D multislice acquisitions or in a single thickslice technique. 5.1.3  MRI of Ventilation The visualization of lung ventilation is feasible by different MRI techniques. It can be achieved while using conventional proton MRI by so-called oxygen-enhanced imaging as well as dedicated non-proton MRI techniques using hyperpolarized noble gases (Fig. 5.1.2).

5.1.2.6  Lung Imaging with Open Low-Field Scanners or Ultrahigh-Field Scanners The above-described problems of lung imaging with MR suggest that a lower field strength, e.g., of 0.5 T, should produce fewer susceptibility artifacts than does 1.5 T and achieve an increase in lung signal intensity (Heussel et al. 2002). At the same time, the problems suggest that lung imaging at 3 T might be unfavorable (Lutterbey et al. 2005). However, practice shows that the gradient systems of the low-field systems are usually weaker, so that the principal advantages at lower field strength have not been realized so far. At the same time the first images obtained at 3 T are of similar quality compared to those acquired at 1.5 T. Nevertheless, at present no advantage can be seen in using ultrahigh-field MR for scanning lungs while lowfield scanners are economic and yield the advantages of open systems regarding patient compliance, in particular for children. Principally, MR at 0.5 T can be performed with T1 GRE and T2 FSE sequences as well, but steadystate gradient-echo sequences with strong T1/T2-contrast producing high signal of solid and liquid pathology also

Fig. 5.1.2a,b  Helium MRI. a Healthy nonsmoking volunteer. The ventilated alveolar space shows fairly homogeneous high signal intensity with good delineation of the signal-free pulmonary stroma and vessels. b Smoker with chronic obstructive pulmonary disease. Inhomogeneous appearance of the ventilated alveolar space with large ventilation defects in both upper fields

5.1  Lungs, Pleura, and Mediastinum

Oxygen-enhanced proton MRI is a rather straightforward method with reasonable technical requirements and cost. The methodology relies on the paramagnetic properties of oxygen due to its two free electrons (Edelman et al. 1996). Thus, the difference between two different oxygen concentrations provides the contrast. Thus, ventilation imaging relies on two different acquisitions with slice selective inversions and refocusing pulses, one with the subject breathing room air, and one with the subject breathing 100% oxygen. The observed increase of the signal intensity is due to a shortening of the T1-time, probably mostly originating from molecular oxygen (Loffler et al. 2000; Muller et al. 2001). It can actually originate from the alveolar space, the interstitial space, and the blood. There are hints that dissolved molecular oxygen within the pulmonary veins is the major contributor to the signal increase. Newer observations also indicate that breathing pure oxygen will lead to a vasodilatation of the pulmonary vasculature, thus increasing the signal intensity within the lung. As such oxygen-enhanced MRI does not only provide ventilation imaging but reflects the complex combination of ventilation, diffusion, and perfusion. Consequently, the interpretation of the imaging findings from oxygen-enhanced MRI can be difficult since the lack of a signal increase can be related to any problem along the way of oxygen from inhalation to ventilation, perfusion, diffusion, and perfusion. Respiratory gating in end expiration and cardiac triggering gating in diastole improve the signal-to-noise ratio (Molinari et al. 2006). Parallel imaging enables more complete coverage of the lung (Dietrich et al. 2005). The effects of oxygen enhancement are dependent on the magnetic field and will actually increase at higher field strength. The use of hyperpolarized noble gases for ventilation imaging using MRI was first described in 1994 (Albert et al. 1994). Helium-3 and xenon-129 can be hyperpolarized by dedicated polarized light (laser). Hyperpolarization then describes a state of polarization approximately five orders of magnitude above the Boltzmann factor for thermal polarization, which is the same for protons or helium-3 and xenon-129 in their original states. Different techniques for hyperpolarization have been developed especially for helium-3 where direct and indirect optical pumping can lead to a polarization rate of more than 80% within a few hours. In recent years, the performance of optical pumping of xenon-129 has also seen enormous progress. For both gases dedicated techniques have been developed to preserve the polarization for some time in order to have reasonable polarization rates available at the time of administration as an inhaled “contrast agent”. It has to be noted that ventilation MRI using hyperpolarized noble gases is still quite experimental and expensive, as the world’s reserves of helium-3, which is a byproduct of tritium decay, are very limited. Also, the purification of xenon-129 from abundant xenon in the atmosphere is expensive. In addition, the necessary polarizers require a

significant investment as well as the availability of broadband MR imaging capabilities for measurements on the helium or xenon Larmor frequencies. Most of the work in this field has been done using hyperpolarized helium-3. This is mainly due to the higher polarization rates, which can be achieved as well as a more favorable gyromagnetic ratio. In addition, helium-3 is inert and not soluble. Thus, it is not absorbed but remains within the alveolar space and is exhaled within the next few breaths, making it a selective “contrast agent” for ventilation with no inherent side effects. In addition, the anoxic or hypoxic gas mixture to be inhaled makes helium-3 preferable for single breath-hold acquisitions. In contrast, xenon is highly lipophilic and absorbed into the blood and then distributed through the body. Thus, the image reflects contributions from xenon atoms within different compartments, the alveolar space, the interstitium, and the vasculature. As the environments vary, the individual contributions can be separated by spectral analysis using MR spectroscopy or chemical shifting imaging. At the same time, xenon is a narcotic gas which accumulates within fatty tissue, especially the brain. These potential side effects have to be taken into consideration. 5.1.4  Use of Contrast Agents The acquisition of contrast-enhanced images is an important part of MRI of the chest. In clinical routine the intravenous administration of paramagnetic gadolinium chelates is well established. Although non-enhanced sequences provide good soft tissue and vascular contrast, there are a significant number of indications where contrast-enhanced scans add substantial value in the detection, characterization, and differentiation of the diagnosis of a lung disease. In general, contrast-enhanced MR angiography of the pulmonary arteries or investigation of pulmonary perfusion is more robust and provides better image quality than do non-enhanced techniques. The analysis of the contrast enhancement patterns of nodules and masses is an important feature in the radiological diagnosis. The contrast enhancement of lymph nodes in the mediastinum can be substantial and may be a contribution for the diagnosis (Laissy et al. 1994; Hasegawa et al. 2003) especially when fat-suppressed sequences are used. The intravenous application of contrast agents is also helpful in the differentiation between residual tumor, e.g., in mediastinal lymphoma, scar tissue as well as for the separation of central tumors and post-obstruction atelectasis. Specific contrast agents, such as ultrasmall particles of iran oxide (USPIO) have not yet gained clinical acceptance, especially as they are not yet approved for clinical use. These contrast agents accumulate within the macrophages of normal lymph nodes resulting in a signal drop

671

672

5  Thorax and Vasculature

on T2-weighted sequences several hours to a few days after intravenous administration. As neoplastic foci within lymph nodes do not host macrophages, malignant lymph nodes will not show this significant decrease of signal intensity. First clinical studies have demonstrated very high sensitivity for the use of USPIOs in the differentiation between benign and malignant lymph nodes Specificity, however, was quite low (Nguyen et al. 1999; Pannu et al. 2000). Iron oxides can also be used as blood pool contrast agents for MR angiography of the pulmonary arteries (Ahlstrom et al. 1999). 5.1.5  Signal Intensities and Contrast Behavior A general challenge in the assessment of signal intensities and contrast in MRI of the chest is the use of GREweighted sequences which will result in variable repetition times due to ECG gating. Subsequently the images from different patients and even different investigations from the same patient will exhibit different signal intensities and contrast properties. The contrast is always dependent on the selected sequence and further determined by the heart rate. T2-weighted sequences with long repetition times and very fast sequences exhibited the best reproducibility. For the determination of relaxation times substantial methodological challenges for measurements in vivo have to be taken into account. This is especially valid for the chest with its inherent motion due to respiration and cardiac cycle. Also the interpretation of in vitro measurements is not straightforward, since significant properties of the tissue in vivo might be obscured. Within the lung parenchyma tumors and inflammatory changes exhibited prolonged T1-relaxation times with regard to the normal tissue. Adenocarcinoma might represent an exception since it can show relatively short T1 relaxation times. In general, T1 and T2 relaxation times of tissue will correlate with the ratio of water within the tissue. Thus, they will be prolonged in pathological, especially neoplastic and inflammatory, tissue. In vitro measurements have generated evidence that T2-relaxation times can be used to differentiate between vital and necrotic tumor tissue as well as atelectasis. However, the measurements of relaxation times cannot be applied to definitively say whether a pulmonary lesion is benign or malignant, because the overlap between these two different tissues is large. It is also true for the assessment of mediastinal lymph nodes. Although, normal and pathological lymph nodes show prolonged T1 relaxation times with regard to the mediastinal fat leaving a good delineation on a T1-weighted image such differences are not specific although malignant lymph nodes might show slightly longer T1 and T2 relaxation times in comparison to benign lymph nodes, e.g., in sarcoid disease.

In clinical routine, the varying signal intensities seen with different relaxation times are very helpful for differentiating between fibrosis and tumor as well as cysts and solid processes and in diagnosing and monitoring hemorrhage and hematoma. Additionally, differentiation between obstructive and non-obstructive atelectasis is feasible. Table 5.1.2 summarizes the signal characteristics of different structures. 5.1.6  Normal Anatomy 5.1.6.1  Lungs The close proximity of airspace and soft tissue at the alveolar level generates multiple air–soft-tissue interfaces that together with the low proton density of the aerated lung result in susceptibility artifacts too pronounced for meaningful imaging of detailed anatomical structures of lung parenchyma. The fine structure of the lung parenchyma cannot be depicted sufficiently well with MRI and the equivalent of a “lung window” has not yet been achieved. Elaborate MR techniques (Bader et al. 2002, Biederer et al. 2002b) to overcome these problems have until now not been implemented in clinical practice and are as yet no match for CT. Current strategies to overcome the already described problems include either short echo times in GRE sequences or short echo spacing in TSE sequences. Advantages of the T1 GRE are the small number of artifacts due to respiratory motion, better visualization of gross lung anatomy, and short acquisition time. Lung vascularization and segmentation are made apparent by the bright flow signal of the vessels. T2 TSE sequences have better spatial resolution and therefore can outline not only the main and the lobar bronchi but even peripheral smaller airways and smaller pulmonary vessels within the lung parenchyma. 5.1.6.2  Mediastinum Many schemes have been devised by for subdividing the mediastinum. It is important to realize that every subdivision of the mediastinum is arbitrary and that a consensus about numbers and borders of mediastinal compartments has not yet been reached. Anatomically, a separation of the mediastinum into at least the anterior, middle, and posterior compartments seems meaningful. The anterior mediastinum is bounded posteriorly by a line along the anterior trachea and the posterior pericardium. The middle mediastinum is bounded posteriorly by the anterior margins of the thoracic spine. All mediastinal compartments lie inferior to the thoracic inlet and superior to the diaphragm.

5.1  Lungs, Pleura, and Mediastinum Table 5.1.2  Enhancement patterns of intrathoracic conditions on different pulse sequences and their differential diagnostic relevance T1 weighted

T2 weighted

Comment

Asymptomatic cysts

↓

↑

Homogeneous

Cysts with a high protein content

→/↑

↑

Homogeneous

Abscess

↓/→

↑

Inhomogeneous

Seroma

↓

↑

Similar to asymptomatic cyst

Lipoma

↑

↑

Obstructive atelectasis

↓/→

↑

Non-obstructive atelectasis

↓/→

↑

Lung cancer

↓/→

→/↑

Typically inhomogeneous

Acute bleeding

↑

↑

Homogeneous

Chronic bleeding

↓

→/↑

Peripheral ring of low SI, Susceptibility artifacts

Hematoma

→/↑

→/↑/↓

Susceptibility artifacts on T2*w GRE images

5.1.6.2.1  Thyroid On MRI, the normal thyroid gland reveals relatively low signal intensity on T1-weighted and intermediate signal intensity on T2-weighted images. For measurements of the diameters and the exclusion of cysts and/or masses thin section SE or TSE images are adequate. 5.1.6.2.2  Parathyroid Normal parathyroid glands are not visualized by MRI. Ectopic enlarged parathyroid glands are typically hypoor isointense to muscle tissue in T1-weighted, and hyperintense on T2-weighted. They are best detected by fat saturated T2-weighted images or in T1-weighted after fat saturation and gadolinium injection. 5.1.6.2.3  Thymus The thymus is a bilobed organ situated in the upper part of the anterior mediastinum between right and left mediastinal pleura, the sternum and the aortic arch. Thymus size varies with age. To differentiate thymus hyperplasia from the normal gland, the most useful measurement is thickness of the lobes. The maximum normal thickness is 15–20 mm in adolescents and up to 5 mm in adults. The left thymic lobe is always identified on MRI, whereas the smaller right lobe may not be visible. The MRI signal

of the gland is homogeneous and age-dependent. The average T1 relaxation times decrease with age and range from 1,000 ms in infants to 400 ms in adults (DeGeer et al. 1986). T2 relaxation times hardly change with age and vary from 40 to 60 ms. In adolescents but not in adults, the T1 signal therefore reliably differentiates the thymus from surrounding lymph nodes. Differentiation in adults is facilitated by employing short SE sequences which yield lower signals in lymph nodes than in the thymus. 5.1.6.2.4  Lymph Nodes The anterior mediastinum contains multiple lymph node chains that follow the thoracic vessels such as the internal mammary nodes, the prevascular nodes, the pericaval and periaortic nodes and the nodes accompanying the brachiocephalic artery, the brachiocephalic, subclavian, and internal jugular veins. In the middle mediastinum the main lymph nodes are the paratracheal, tracheobronchial and subcarinal groups. Main lymph nodes in the posterior mediastinum are the paraesophageal, para-aortic and intercostal groups. Lymph nodes as small as 0.5 cm may be detected on MRI if they are surrounded by sufficient adipose tissue (Fig. 5.1.3). Lymph nodes demonstrate intermediate signal intensity on T1-weighted images, contrasting with the high signal from the adipocytes. Vessels may be differentiated by their low-to-void signal. Mediastinal lymph nodes are usually oval-shaped, as studies of pathology specimens show. Their large di-

673

674

5  Thorax and Vasculature

able thickness. On MRI the esophagus shows intermediate signal intensity; its lumen is invisible without dilatation by gas or water. On sagittal sections, the esophagus courses through the mediastinum as a narrow oblong structure behind the heart and anterior to the descending aorta. Identification of the esophagus is enhanced by the highly intense paraesophageal fat signal. The paraesophageal fat may, however, be slim or absent in which case the differentiation of the esophagus from the heart and vessels may become impossible. 5.1.6.2.7  Great Vessels

Fig. 5.1.3  Two paracardiac mediastinal lymph nodes in a patient with lung cancer. Zoomed transverse T1-weighted image; SE (1 RR/22); slice thickness = 8 mm. The low-signal-intensity lymph nodes (white arrow) are clearly seen in the pericardial fatty tissue despite their small size (0.8 cm). Unlike CT, MRI requires no contrast administration

ameters are variable; the small nodal diameter (which is usually encountered on axial images) varies from 8 to 12 mm, depending on which mediastinal group the nodes originate from. 5.1.6.2.5  Trachea The trachea courses caudad through the middle mediastinum and bifurcates at the level of the fourth thoracic vertebra into the two main bronchi that form an angle of 55–70°. Angles wider than 90° are pathological. The right main bronchus is longer and runs more vertically than the left. Trachea and bronchi are readily identified as signal-free structures, and their course may be best appreciated on sagittal or parasagittal sections. Oblique coronal images best demonstrate the configuration of the tracheal bifurcation. 5.1.6.2.6  Esophagus The esophagus traverses the chest in the posterior mediastinum to the left and anterior to the thoracic spine and enters the abdomen through the esophageal hiatus. In its upper parts the esophagus and the trachea are enveloped in a common fascia within which the esophageal tube is additionally surrounded by adipose tissue of vari-

On T1-weighted images the great vessels show a poor or absent signal and are well outlined against the surrounding mediastinal fat. Routinely, all cardiac chambers, the aorta, the pulmonary arteries, the venae cavae and the azygos and hemiazygos veins are outlined on MRI. 5.1.6.3  Pleura and Chest Wall The visceral and parietal layer of the pleura cannot be differentiated by MRI (Fig. 5.1.4). Only in cases of pleural effusion or tumor the separation of both layers is obvious. Also the chest wall is anatomically not optimally demonstrated by MRI. The ribs have an individual double angulated position in the body. This means that the bone marrow and also the cortical bone are not demonstrated in an optimal way by MRI. On the other hand, the excellent soft tissue contrast is very helpful for distinguishing the different muscle layers, the subcutis, and cutis of the chest wall. In addition, the vertebral bodies and articulations, the spinal canal and the myelon are delineated very well (Fig. 5.1.5). 5.1.7  Lung Diseases 5.1.7.1  Congenital Pathologies Ultrasonography is the primary imaging modality for the evaluation of fetal or maternal anomalies. This method is safe, relatively inexpensive, easily accessible, and allows real-time imaging. Continuous technical improvements in ultrasonography in the last 10–15 years have led to improved diagnostic accuracy for fetal malformations. In cases of complex anomalies, however, MRI provides additional information and has evolved as a valuable diagnostic method for evaluating fetal pathology. Examinations should be avoided in the first trimenon, although a teratogenic effect of diagnostic MRI is not known. For thoracic fetal imaging, especially congenital cystadenomatoid malformation, diaphragmatic hernia, pulmonary arteriovenous malformations, pulmonary sequestration,

5.1  Lungs, Pleura, and Mediastinum

Fig. 5.1.4a,b  Peripheral adenocarcinoma with infiltration of the pleura. a T2 TIRM shows inhomogeneous high signal intensity with broad contact to the pleura. b ce-VIBE shows strong inhomogeneous enhancement and infiltration of the pleura

Fig. 5.1.5a,b  Spinal infiltration of a chondrosarcoma of the chest wall with infiltration of the neuroforamen. a.T2 TIRM shows inhomogeneous high signal intensity and obvious infiltration of the pleura and the neuroforamen. b ce-VIBE shows inhomogeneous perfusion of the tumor

and anomalous pulmonary venous return are of diagnostic interest. But also oligohydramnions, due to premature rupture of membranes, is an important risk factor for compromised fetal lung growth. In these situations MR volumetry can be used to measure the size of the fetal lung quite accurately. Together with the evaluation of lung signal intensities on T2-weighted sequences, fetuses with pulmonary hypoplasia can be readily detected. The fetal lung tissue has conspicuous signal decrease in these cases.

liferation of hamartomatous tissue in terminal bronchioles with suppression of alveolar tissue and an increase in cystic components. The prevalence is estimated of about 1:10.000. In cases of indistinguishable sonographic findings, fetal MRI is the modality of choice for proving the diagnosis and preliminary appraisal of intensive care therapy and postnatal extracorporeal membrane oxygenation. Furthermore, fetal MRI often facilitates assessment and planning of intrauterine surgical procedures. Diagnostic criteria are their expansive character, often with mediastinal shift, deep ipsilateral diaphragm, and atelectatic lung tissue with diminished signal in T2 weighting. According to the type of the CCAM (Stocker I to III), in the majority of cases cystic lesions mostly affect a single lobe. The differentiation of CCAM from sequestrations is difficult, especially if the left lower lobe is involved.

5.1.7.1.1  Congenital Cystadenomatoid Malformation Congenital cystadenomatoid malformation (CCAM) is the result of a dysfunction of the fetal lung development with unclear etiology. It is characterized by a pro-

675

676

5  Thorax and Vasculature

5.1.7.1.2  Congenital Diaphragmatic Hernia

5.1.7.1.4  Pulmonary Sequestrations

Congenital diaphragmatic hernia (CDH) has a prevalence of about 1:5,000. It is located on the left side in about 85%, on the right in 13% of cases. Bilateral involvement is rare. Despite therapeutic advances the mortality is still high and accounts for about 50%. The main reason is the association with pulmonary hypoplasia. The diagnosis of congenital diaphragmatic hernia (CDH) can be difficult by ultrasonography, particularly due to the similarity of liver and lung parenchyma. Besides morphological aspects, e.g., herniation of abdominal structures into the chest, small amounts of compressed lung can be visualized on MRI. T2-weighted sequences allow for a differentiation of fetal lung tissue with higher signal intensity as muscle, but lower signal compared with amnion fluid. The herniation of bowel, liver, stomach or spleen, and the mediastinal shift is easily depicted (Fig. 5.1.6). The feasibility of using volumetric measurements on MRI may be visualized. Volumetric management on MRI may help to predict high-risk fetuses, to accelerate decisions about pre- and postnatal management, and improve outcome.

Pulmonary sequestrations consist of non-functioning lung tissue, which is not in continuity with the tracheobronchial tree. A majority of them are located in the posterior lower lobes and are more often left-sided. Intralobar sequestrations are characterized by normal surrounding with visceral and parietal pleura and a venous drainage by lung veins, whereas extralobar sequestrations have a separated pleural cover and drain into the vena cava or azygos vein. The atypical systemic arteries arise in both cases from the descending or abdominal aorta in most cases. They may be visualized by MRI, including CEMRA, with a high degree of confidence, thus avoiding conventional angiography in most cases. Clinically striking is often an episode of pneumonia. Sequestrations contain mucus, inflammatory, granulomatous, and fibrotic tissue, and they exhibit high signal intensity on T2-weighted images and intermediate-to-high signal intensity on non-enhanced T1-weighted images. After contrast administration the lesions demonstrate strong enhancement (Hang et al. 1996).

5.1.7.1.3  Pulmonary Arteriovenous Malformations

5.1.7.1.5  Anomalous Pulmonary Venous Return

They consist of a racemose convolute of vascular structures and may appear as pulmonary nodules but can be differentiated by the visualization of feeding arteries and draining veins (Fink et al. 2003). Flow within the malformation can be highly variable. Patent vessels will show flow voids, whereas slow flow will exhibit low signal and thrombosed vessels, and hemorrhage will be made apparent by their high signal intensity. After contrast administration malformations are strongly enhancing, while high-caliber feeding arteries and draining veins are easily depicted at CEMRA. CEMRA and pulmonary perfusion MRI are well suited to be used for the management of pulmonary arteriovenous malformations including planning of embolotherapy and follow-up (Ohno et al. 2002a).

MRI and MRA can be successfully used for accurate identification of pulmonary venous confluence and total or partial anomalous pulmonary venous return, such as is seen in Scimitar syndrome. Additional information of concomitant bronchial and visceral abnormalities can be obtained. Flow measurements will allow for quantitation of shunt volumes. The limitations of echocardiography and invasiveness of angiography make CEMRA the modality of choice for the assessment of anomalies of the pulmonary veins (Fig. 5.1.7).

Fig. 5.1.6  a–c “Classical” left-sided congenital hernia without liver shift in coronal plane. Displacement of the stomach, small intestine, and colon into the thoracic cavity. Mediastinal shift to the right

5.1  Lungs, Pleura, and Mediastinum

Fig. 5.1.7  Contrast-enhanced MRA shows the anomalous drainage of the right upper pulmonary vein and its connection to the superior vena cava (arrow)

5.1.7.2  Infiltrative Lung Diseases MRI still plays a minor role in the detection and diagnosis of inflammatory and destructive lung diseases when compared to projection radiography and computer tomography. As already mentioned the challenges of MRI of the chest, such as limited spatial resolution, low proton density, susceptibility, and motion artifacts limit the visualization of the normal lung parenchyma. Especially the signal-to-noise ratio is low since the amount of protons is low. However, many inflammatory and diffuse lung diseases are associated with an increase of fluid, cells, and tissue within the lung. Thus, they provide their own protons to be detected by MRI, so-called disease-related contrast. Several studies have demonstrated that the typical radiological findings of lung fibrosis, such as a reticular or cystic pattern including ground glass opacities indicating alveolitis, can also be translated to MRI. With the use of T1- and T2-weighted sequences, all but the very subtle findings of diffuse interstitial lung disease can be demonstrated by MRI. By using T1-weighted sequences pre- and post-contrast as well as T2 or T2* sequences the different tissue components and the activity of inflammation can be differentiated and evaluated semiquantitatively. This is a major advantage with regard to CT. Thus, MRI can be used to differentiate between an active alveolar process and fibrotic scars. This will provide important indications for the anti-inflammatory treatment. Research activities are ongoing to evaluate the potential clinical role of MRI in these diseases, especially compared to CT but also to nuclear medicine. 5.1.7.3  Atelectasis Atelectasis is defined as an area of lung which does not contain air and has lost volume. The amount of tissue and blood provides enough signal intensity for MR visualization (Fig. 5.1.8). Obstruction atelectasis and compression atelectasis differ depending on the underlying cause, and MRI can demonstrate their different compositions (Herold et al. 1991). Thus, atelectasis caused by central obstruction, e.g., due to lung cancer, will exhibit

Fig. 5.1.8a,b  Subtotal atelectasis of the right upper lobe shows the typical volume loss. a T2 TIRM with high signal intensity, b ce-VIBE with strong enhancement of the atelectasis

a significantly higher signal intensity on T2-weighted sequences than non-obstructive, i.e., compression atelectasis. This difference in signal intensity is probably explained by the larger amount of mucus accumulating in cases of bronchial obstruction. Mucus typically contains a lot of fluid with protons and longer T2 times, leading to increased signal intensity. In contrast compression atelectasis mainly consists of compressed lung tissue and blood and only a limited amount of mucus leading to shorter T2 times and lower signal intensity. In oncology, the delineation between a central tumor and peripheral atelectasis is important for accurate staging and surgical or radiotherapy planning. Besides T2-weighted imaging different enhancement patterns of tumor and atelectasis on post-contrast T1-weighted imaging provide important useful, though not definitive, information for differentiation which cannot be obtained from CT. Atelectasis is characterized by high early enhancement with washout, whereas bronchial carcinoma shows a more prolongated enhancement pattern with lower slope.

677

678

5  Thorax and Vasculature

5.1.7.4  Pneumonia Pneumonias are infectious infiltrates of the lung tissue mostly representing an increase of inflammatory tissue and fluid within the interstitium and/or alveolar space. Depending on the underlying microorganism different, rather typical patterns of infiltration can be observed on MRI as have already been described using the gold standard of CT (Puderbach 2006). Typical bacterial pneumonia is represented by a lobar or segmental infiltration or consolidation with high signal intensity. It can be detected on T1-weighted imaging as an isointense and especially on T2-weighted imaging as a hyperintense lesion. It also shows rather strong enhancement after contrast administration (Fig. 5.1.9). Accompanying atelectasis or effusions are likewise easy to identify. Pneumonia can also occur as typical bronchial pneumonia represented by bronchocentric infiltrates leaving out the subpleural space. They also have rather high signal intensity and a typical distribution pattern. Atypical bacterial pneumonia is associated with a ground glass pattern on CT. This translates into an area of slightly increased signal intensity, especially on T2-weighted sequences, which can be detected by MRI. The acquisition of high-quality T2-weighted images of the lung might be a major challenge. Acquisition can be done as a breath-hold T2-weighted HASTE sequence. The major advantage is the single-shot nature of this sequence. The images are almost free from respiratory motion artifacts. The respiratory-gated T2-weighted ultrashort turbo spin-echo (UTSE) sequences can be regarded as the reference standard because they will provide very high signal-to-noise ratios and lead to easy detection of pulmonary infiltrates (Lutterbey et al. 1998). The use of a T1-weighted VIBE sequence post-contrast will also nicely demonstrate the increased perfusion and vascular permeability within the infiltrate. MRI has been shown to be especially useful in the detection and characterization of fungal pneumonia (Blum et al. 1994; Herold et al. 1989). In the early phase, angioinvasive aspergillosis is characterized by an infiltrate with high signal intensity and strong enhancement after contrast administration. In a later phase abscess formation is frequent. MRI has shown much higher sensitivity than CT in the detection of such infiltrates with abscess formation than CT. Characteristic is the hyperintense consolidation with the hypointense ring on T2-weighted imaging, the so-called reverse-target sign. Recent work has shown that MRI can be used for early detection of pulmonary infections in a high-risk population, such as immunocompromised patients with neutropenic fever (Fig. 5.1.10). In this population MRI performs as well as CT, and it might be preferable due to the lack of radiation exposure (Leutner et al. 2000; Eibel et al. 2006).

Fig. 5.1.9a–c  Typical bacterial bronchopneumonia shows consolidation of the right lower lobe. a T2 HASTE with high signal intensity, b T1 VIBE, c ce-VIBE shows strong enhancement of the infiltrate

5.1.7.5  Pulmonary Edema Experimental studies have shown that MRI can depict and allow for quantitative assessment of lung water. Pulmonary edema can be detected on non-enhanced images and total lung water calculated from proton density obtained from a multi-spin-echo sequence. After assessment of pulmonary intravascular volume, e.g., by 3D TOF angio-

5.1  Lungs, Pleura, and Mediastinum

grams, extravascular lung water can be calculated. After induction of diffuse alveolar damage in pigs a substantial increase of the signal-to-noise ratio caused by an increase of extravascular lung water has been shown (Caruthers et al. 1998). By using the rate of signal-to-noise ratio change, the mismatch of transcapillary filtration flow and lymph clearance as well as the filtration coefficient can be estimated. Another technique uses a macromolecular contrast agent that very nicely demonstrates the sequelae of diffuse alveolar damage with increased pulmonary permeability and extravasation of the contrast agent. There was a good correlation with the dye-dilution technique serving as the reference technique (r = 0.89) (Bock et al. 1997). Further in vitro work has confirmed these observations with an excellent correlation between MR measurements and gravimetric analysis of lungs. Two distinct regions (inflated and deflated) could be separated (Estilaei et al. 1999). However, the translation into humans and clinical routine use, e.g., to the benefit of patients from the intensive care unit, has not been achieved yet. 5.1.7.6  Diffuse Interstitial Lung Disease Several studies have demonstrated that the typical radiological findings of lung fibrosis, such as a reticular or cystic pattern including ground glass opacities indicating alveolitis, can also be translated to MRI. With the use of T1- and T2-weighted sequences all but the very subtle findings of diffuse interstitial lung disease can be demonstrated by MRI. By the use of T2-weighted sequences and post-contrast T1-weighted acquisitions, the inflammatory activity of the disease can also be evaluated semiquantitatively, which is a major advantage relative to CT (Fig. 5.1.11). Thus, with MRI, it is possible to differentiate between an active alveolar process and fibrotic scars, providing important indications for the anti-inflammatory treatment. 5.1.7.7  Emphysema and COPD

Fig. 5.1.10a–c  CT and MRI of consolidating pneumonia in an immunocompromised patient with acute myeloid leukemia. While CT using the soft tissue (a) and lung window (b) settings only shows the dense infiltrate without definitive evidence of abscess, T2 ultrashort turbo spin-echo (UTSE) MRI (c) demonstrates both central abscess and the typical low-signal-intensity ring associated with invasive aspergillosis

Attempts to use MRI for the depiction of pulmonary structure in COPD will always be compared to the use of CT. Due to physics, MRI will never outperform CT when it comes to visualization of bronchial dilation, wall thickening, or emphysematous destruction. Emphysematous destruction can hardly be diagnosed by a loss of signal in the periphery since there already is not much signal to be detected from a normal lung. It is much easier to detect consecutive hyperinflation by the size or the volume of the thorax as well as the reduced blood volume in the severely affected areas by perfusion MRI (Fig. 5.1.12). Gas exchange in the lungs is maintained by a balance between ventilation and perfusion. In patients with

679

680

5  Thorax and Vasculature Fig. 5.1.11a–d  Alveolar proteinosis: extensive involvement of both lungs with typical crazy paving pattern on CT (a). T2 TSE (b) shows high signal intensity infiltration, while T2 HASTE (c) only shows intermediate signal intensity. Contrast-enhanced T1 VIBE (d) shows strong enhancement of the infiltrations

COPD and emphysema, ventilation is impaired due to airway obstruction and parenchymal destruction. The structural loss of lung parenchyma leads to a significant increase of closed volume and a decreased ventilation of those regions, with subsequent hypoxic vasoconstriction including a local reduction in pulmonary blood flow. The loss of pulmonary arteries is related to the severity of parenchymal destruction and the mechanical compression of pulmonary arteries in case of hyperinflation. MR perfusion allows for a high diagnostic accuracy in detecting perfusion abnormalities (Sergiacomi et al. 2003; Fink et al. 2004; Morino et al. 2006). Furthermore, MR perfusion ratios correlate well with radionuclide perfusion scintigraphy ratios (Ohno et al. 2004a; Molinari et al. 2006). The perfusion abnormalities in COPD differ clearly from those caused by vascular obstruction. While in embolic obstruction wedge-shaped perfusion defects occur, a generally low degree of contrast enhancement is found in COPD with emphysema. Furthermore the peak signal intensity is reduced. These features allow for easy visual differentiation. MRI has been used for imaging the right ventricle for quite a long time, and a loose correlation between increased right ventricular mass and the severity of emphysema was found (Boxt 1996). Assessment of right ventricular function can be done either by flow measurements in the pulmonary trunk or by short-axis cine acquisition of the right ventricle. In clinically stable, normoxic COPD patients the position of the heart is rotated and shifted to a more vertical position in the thoracic cavity due to hyperinflation of the lungs, enlarging the

retrosternal space. The right ventricular muscular mass is increased while the right ventricular ejection fraction is unchanged (Vonk-Noordegraaf et al. 1997). The concentric right ventricular hypertrophy is the earliest sign of right ventricular pressure overload in patients with COPD. This structural adaptation of the heart does not alter right and left ventricular systolic function (VonkNoordegraaf et al. 2005). As sufficient gas exchange depends on matched perfusion and ventilation, assessment of regional ventilation in human lungs is important for the diagnosis and evaluation of pulmonary emphysema. Helium-3 was used in several studies showing a high sensitivity to airflow obstruction and impairment of ventilation in emphysematous patients (Altes et al. 2001; Ley et al. 2004). Helium-3 proved also to be a valuable tool for quantification of ventilation and correlated well with parenchymal destruction in emphysema (Zaporozhan et al. 2004; Swift et al. 2005). Oxygen-enhanced MR ventilation imaging in COPD showed regional changes in ventilation, and the maximum mean relative enhancement ratio was excellently correlated with the diffusion capacity for carbon monoxide. The mean slope of relative enhancement was strongly correlated with the forced expiratory volume in 1 s, and the maximum mean relative enhancement had a good correlation with the high-resolution CT emphysema score (Ohno et al. 2002b). Hyperinflation severely reduces the mechanical advantage of the diaphragm with respect to the rib cage to respiratory motion, while it commonly increases the con-

5.1  Lungs, Pleura, and Mediastinum

al. 2005). For data acquisition time resolved techniques are used based on FLASH or trueFISP sequences, which allow for a temporal resolution of 100 ms per frame. In contrast to normal subjects with regular, synchronous diaphragm and chest wall motion, patients with emphysema frequently show reduced, irregular, or asynchronous motion, with a significant decrease in the maximum amplitude and the length of apposition of the diaphragm (Suga et al. 1999). Parts of the diaphragm move paradoxically, relative to the change in lung area. The paradoxical diaphragmatic motion correlated with hyperinflation, although severe hyperinflation tended to restrict both normal and paradoxical diaphragmatic motion (Iwasawa et al. 2002). After lung volume reduction surgery, patients showed improvements in diaphragm and chest wall configuration and mobility (Suga et al. 1999). 5.1.7.8  Cystic Fibrosis

Fig. 5.1.12a–c  Lung cancer of the left upper lobe with direct infiltration of the mediastinum, left-sided pleural effusion and mediastinal subcarinal lymphadenopathy shown by T2 HASTE (a) and T2 TSE (b). Note hyperinflation of the right lung indicating emphysema. MR perfusion (c) shows central patchy perfusion defects on the right due to hypoxic vasoconstriction in emphysema as well as a large perfusion defect on the left presumably due to the tumor

tribution of the rib cage itself. This complex interaction between chest wall and diaphragm motion can be visualized by dynamic cine MRI (Cluzel et al. 2000; Plathow et

Cystic Fibrosis (CF) is an autosomal recessive disorder caused by mutations of a gene, located on the long arm of chromosome 7. It codes for the CFTR (cystic fibrosis transmembrane-regulator protein), which functions as an anion channel. The impaired CFTR function causes aberrations of volume and ion composition of airway surface fluid, leading to viscous secretions with the consequence of bacterial colonization, chronic lung infection, airway obstruction, and consecutive destruction of the lung parenchyma. Despite improved understanding of the underlying pathophysiology and introduction of new therapies, CF is still the most frequent life-shortening inherited disease in the white population. Advantages in knowledge and medical care resulted in a dramatic increase of the life span for these patients. The median of survival in CF patients in Germany has risen to more than 35 years. CF affects most body systems. But the majority of morbidity and mortality in CF patients is due to their lung disease. As life expectancy is limited by pulmonary complications, repeated imaging of the chest is required. The standard radiological tools for imaging of the chest are chestX-ray and CT. Both modalities, especially CT, cause a high cumulative dose of radiation within several years. Thus, a radiation-free imaging method for therapy monitoring and follow-up is highly desirable. MRI of the chest was proposed as a potential imaging alternative in patients with CF in 1987 (Fiel et al. 1987). But at that time the technique of MRI was not capable of producing results comparable to those of CT or an adequate clinical contribution. The introduction of parallel imaging into clinical practice has made fast image acquisition possible, enabling a substantial improvement in MRI of cystic fibrosis. The major changes in CF are bronchial wall thickening, mucus plugging, bronchiectasis, air fluid level, consolidation, and segmental/lobar

681

682

5  Thorax and Vasculature

destruction (Puderbach et al. 2006). In CF patients the bronchial wall of small airways is visualized in the lung periphery due to bronchial wall thickening. The different signal intensities of the bronchial walls on T2weighted images or on T1-weighted images after contrast administration are due to different amounts of fluid most likely representing inflammatory activity. Mucus plugging is very well visualized by MRI, demonstrating a high T2weighted signal because of the tremendous proton density of the fluid. In central mucus plugging there is high T2-weighted signal filling the bronchus within its course partly or completely. Peripheral mucus plugging shows a grape-like appearance of small T2-weighted high-signalintensity areas almost like the “tree-in-bud” phenomena in small-airway inflammation on CT. As mucus plugging does not show contrast enhancement it can be differentiated from airway inflammation. This is a great advantage as these two entities cannot be distinguished by CT. Depending on the stage of disease CF patients have a high risk of hemoptysis. The localization of the origin of bleeding can be crucial for the outcome of the patient. With conventional non-invasive imaging modalities the specification of visualized material in the bronchial lumen is not possible. On CT mucus and blood in the bronchial lumen cause a density increase and cannot be distinguished. On MRI the two can clearly be distinguished, as mucus shows high T2-weighted and low T1-weighted signal intensity, whereas fresh blood shows low signal intensity on both. The MRI appearance of bronchiectasis is dependent on bronchial level, bronchial diameter, wall thickness, wall signal, and the signal of the bronchial lumen. Central bronchiectasis is well visualized independent from wall thickening and wall signal because of the anatomically thicker wall of the central bronchi even without pathological changes. Peripheral bronchi starting at the third to fourth generation are poorly visualized by MRI. CT is able to show bronchi in a healthy lung down to the eighth generation. Peripheral bronchi are demonstrated by MRI only when they are pathological in case of delineation from the surrounding tissue by bronchial wall thickening (Fig. 5.1.13a) and/or a high wall signal and/or mucus filling of the dilated lumen. Airway obstruction will lead to hypoxia and subsequent vasoconstriction in the peripheral segments, which can be nicely demonstrated by MR perfusion (Fig. 5.1.13b). Air fluid levels occur in saccular or varicose bronchiectasis. They can be visualized by MRI because of the extremely high fluid proton density on T2-weighted images. The delineation of a bronchus with a partial mucus plug or a severely thickened wall with high T2-weighted signal can be difficult on HASTE sequences. Combining the signal characteristics of the HASTE and VIBE sequences (high T2-weighted signal, low T1-weighted signal; wall delineation by contrast media enhancement) the air fluid level can be differentiated.

Fig. 5.1.13a,b  Cystic fibrosis. T2 HASTE (a) shows moderate bronchiectasis in both lungs partly associated with perfusion defects as demonstrated by MR perfusion (b)

Consolidations in CF are mainly caused by alveolar filling with inflammatory fluid, which causes a high T2weighted signal. Moreover the anatomical segmental or lobar distribution is appreciated. Comparable to CT, MRI is able to visualize the positive bronchopneumogram as T2-weighted hypointense areas following the course of the bronchi within the consolidation. 5.1.7.9  Asthma Asthma is a disease characterized by chronic inflammation and reversible obstruction of the small airways resulting in impaired pulmonary ventilation. Morphological proton MRI has not been used for imaging asthma patients as the airway changes expected are too subtle to be depicted. There are some indications that perfusion MRI will show small peripheral defects after bronchial obstruction and hypoxic vasoconstriction, but this has not been confirmed. Helium-3 MRI has been used extensively in asthma. Due to its sensitivity, ventilation

5.1  Lungs, Pleura, and Mediastinum

defects were even depicted in asymptomatic patients with normal lung function. These ventilation defects were more numerous and larger in the symptomatic asthmatics who had abnormal spirometry (Altes et al. 2001). In a group of patients with moderate-to-severe persistent asthma, there were more defects of 3 cm or more in diameter than in the group with mild-intermittent and mild-persistent disease (p = 0.021) (De Lange 2006). The number of defects is also inversely related to the percent predicted FEV1 as well as forced expiratory flows (Samee et al. 2003). Ventilation defects studied over time demonstrated no change in appearance over 30–60 min, however over a longer period Helium-3 MRI shows defects which resolved and appeared. After methacholine challenge the number of defects increased. Similarly, imaging of the lungs after exercise (n = 6) showed increased ventilation defects in parallel with decreases in FEV1. The administration of inhaled bronchodilators leads to the resolution of ventilation defects. Thus, Helium-3 MRI is well suited to monitor the natural course of the disease as well as the effects of therapy. 5.1.7.10  Neoplasia 5.1.7.10.1  Benign Tumors Benign lung tumors (see “Overview,” below) are rare entities in comparison with lung cancer and account for an only minimal percentage of all thoracic neoplasms. Distinguishing features of benign tumors are absent or slow growth and unchanged morphology over years of followup. Systematic MR descriptions of benign lesions are rare as these tumors most often are found incidentally and are either followed by imaging methods other than MRI or excised immediately. Among the benign tumors the hamartomas are most frequent, representing 55% of cases. They are benign lesions composed to variable degrees of muscle, cartilage, adipose, connective, and epithelial tissue. Hamartomas usually are well circumscribed, often slightly lobulated, and in 30% contain chondroid (popcorn-like) calcifications. In characterizing hamartomas, MRI is inferior to CT or radiography because of its inferior spatial resolution and because it does not depict calcifications well. Hamartomas have been described as lobulated tumors of medium signal intensity on T1-weighted SE and high signal intensity on T2-weighted SE; intralesional septae are enhanced after intravenous contrast administration (Sakai et al. 1994). In individual cases MRI may aid in distinguishing hamartomas from bronchogenic carcinoma or lung metastases by detecting adipose tissue (SPIR, STIR, inphase/opposed-phase sequences) and by showing a comparatively low enhancement in a lesion.

The historic term “bronchial adenoma” is regarded as a misnomer. Historically, it described various entities such as carcinoid tumors (> 80%) and cylindromas (> 10%) that are now considered low-grade carcinomas. The term cylindroma is not recommended since it comprises mostly adenoidcystic (i.e., of mucous gland origin) and rarely mucoepidermoidal (i.e., of columnar epithelial carcinomas). MR characteristics of these rare malignant tumors have not yet been described. Likewise, truly benign tumors other than hamartomas have not been studied systematically with MRI to date. Overview of Benign Lung Tumors (According to Müller 1992) • Benign epithelial tumors • Papillary tumors - Squamous papillomas - Transitional cell papillomas - Micropapillomatosis • Adenomas - Pleomorphic adenomas (mixed tumors) - Monomorphic adenomas (cystadenoma, oncocytoma, clear cell tumors) • Carcinoids (carcinoid tumors, apudoma) • Benign mesenchymal tumors • Chondromas • Osteomas - Osteoplastic tracheobronchopathy - Dendriform pulmonary ossification • Lipomas • Myxomas • Fibromas • Leiomyofibromas • Angiogenic tumors (angiomas) • Pulmonary sequestration • Hemangiopericytoma Other rare lung tumors • Neurogenic tumors • Granular cell tumors (myoblastomas) • Paragangliomas (chemodectoma, glomus tumors) Tumor-like conditions • Hamartomas - Tuberous sclerosis - Lymphangioleiomyomatosis (lymphangiomatosis) - Fetal adenoma • Plasma cell granulomas (inflammatory pseudotumors) • Pseudolymphomas (lymphoproliferative lesions) • Amyloid tumors • Endometriosis • Sclerosing angiomas • Intravascular and sclerosing bronchioloalveolar tumors (IVSBAT) • Histiocytosis X

683

684

5  Thorax and Vasculature

5.1.7.10.2  Bronchogenic Carcinoma Lung cancer or bronchogenic carcinoma is the leading cause of cancer mortality in adults, and incidence rates still continue to rise especially in women. Diagnostic imaging serves a triple role in patients suspected to have lung cancer. Initially, it will substantiate the diagnosis and characterize the lesion in question. In addition, if tumor is likely, imaging serves as part of the staging process and later in the course of the disease will aid in therapy monitoring. For patient management, both histology and location of tumors are relevant. There are four main histological types of lung cancer of epithelial origin accounting for the majority of cancers of the lung (Table 5.1.3): squamous cell carcinoma, small cell carcinoma, adenocarcinoma, and large cell carcinoma. Mixed histological patterns of cancer are common and in some cases two different histologies may be found in one single radiological tumor manifestation. Most common is squamous cell carcinoma, which is predominantly located centrally and shows intrabronchial nodular growth with abundant necroses. A hallmark of small cell carcinoma is the early hematogenous spread, which is the reason that in this cancer type staging does not follow the TNM classification, but rather distinguishes between limited and extensive disease. Adenocarcinomas more often occur in women than in men; they are most often found peripherally and hardly show necrosis. Clinically it is important to distinguish between central mediastinal and peripheral intrapulmonary sites of tumor manifestation. Approximately 75% of lung cancers arise centrally, originating from segmental or subsegmental bifurcations of the bronchial tree. By contrast, the less frequent peripheral bronchogenic carcinoma stems from the mucosa of airways distal to the subsegmental bronchi. An uncommon location is the lung apex, where

tumors are referred to as superior sulcus or Pancoast tumors. The clinical presentation is distinct from that of the other tumor sites and includes shoulder and arm pain and the Horner syndrome (ptosis, miosis, anhidrosis). With the exception of small cell tumors already mentioned, all other carcinomas, i.e., non-small-cell lung cancers of the various types, are staged following the TNM system (Table 5.1.4). T Stage T staging takes into account not only tumor size and invasion of surrounding tissues but also the secondary effects of tumor growth on neighboring structures such as atelectasis or obstructive pneumonitis. On T1-weighted MRI images, lung cancers demonstrate intermediate signal intensity, comparable to muscle, while the T2-weighted signal is higher. If tumors are of sufficient size they may show inhomogeneities that suggest necroses and/or calcifications. After intravenous paramagnetic contrast agents tumors regularly demonstrate inhomogeneous and rather strong enhancement. The most important features for determining the T stage are tumor diameter and tumor extension towards the main bronchi, towards the mediastinal vessels and, if the tumor is located in the periphery, towards the pleura and chest wall. MRI reliably depicts tumors larger than 4–5 mm, i.e., all tumor stages. Imaging in the coronal and sagittal planes is often helpful in diagnosing tumor extension into adjacent structures (Figs. 5.1.1, 5.1.4). Perfusion MRI adds functional information (Fig. 5.1.12). For apically located Pancoast tumors MRI was recommended for a long time as the imaging method of choice in the staging of chest wall invasion (Fig. 5.1.14) (Heelan et al. 1989; McLoud et al. 1989). Spiral CT, with the possibility of multiplanar reconstructions in isotropic voxels, has changed the situation.

Table 5.1.3  5-Year survival rates (percentages) and characteristic findings in lung cancer Adenocarcinoma

Large cell carcinoma

Squamous carcinoma

Small cell carcinoma

5-Year survival

12

13

25

1

Mediastinal lymph nodes

41

42

42

72

Distant metastases

36

35

23

47

Central involvement

18

32

40

78

Peripheral localization

71

59

27

29

Concomitant atelectasis

10

13

36

17

Concomitant pneumonia

15

23

15

22

2

 4

7

0

Liquefactions

5.1  Lungs, Pleura, and Mediastinum Table 5.1.4  TNM classification of lung cancer T: primary tumor T1

Tumor 3 cm or less in greatest dimension, surrounded by lung   or visceral pleura, without bronchoscopic evidence of invasion more   proximal than the lobar bronchus (i.e., not in the main bronchus)

T2

Tumor with any of the following features of size or extent: • More than 3 cm in greatest dimension • Involves main bronchus, 2 cm or more distal to the carina • Invades visceral pleura • Associated with atelectasis or obstructive pneumonitis that   extends to the hilar region but does not involve the entire lung

T3

Tumor of any size that directly invades any of the following: • Chest wall (including superior sulcus tumors) • Diaphragm • Mediastinal pleura • Parietal pericardium: or tumor in the main bronchus less than 2 cm distal   to the carina but without involvement of the carina; or associated atelectasis or obstructive pneumonitis of the entire lung

T4

Tumor of any size that invades any of the following: mediastinum, • Heart • Great vessels, • Trachea, esophagus • Vertebral body, carina • Separate tumor nodule(s) in the same lobe • Tumor with malignant pleural effusion

N: regional lymph nodes NX

Regional lymph nodes cannot be assessed

N0

No regional lymph node metastasis

N1

Metastasis in ipsilateral peribronchial and/or ipsilateral hilar lymph nodes   and intrapulmonary nodes, including involvement by direct extension

N2

Metastasis in contralateral mediastinal and/or subcarinal lymph nodes

M: distant metastasis MX

Distant metastasis cannot be assessed

M0

No distant metastasis

M1

No distant metastasis, includes separate tumor nodule(s)   in a different lobe (ipsilateral or contralateral)

Soft tissue invasion is of particular importance in centrally located T4 carcinomas. These large tumors are well outlined on MRI as are their frequent extensions into the mediastinum, the heart, the great vessels, or the esophagus. The T1 signal of these tumors provides good contrast between the black signal void of vessels and the bright mediastinal fat. Therefore, tumor extension into a mediastinal vessel or the aortic wall (Fig. 5.1.15) is reliably diagnosed, even without intravenous contrast. ECG-trig-

gered contrast-enhanced MRA improves the image quality and the detection of hilar and mediastinal invasion of bronchogenic carcinoma (Ohno et al. 2001). Because of its high soft-tissue contrast MRI is the method of choice for the detection of early chest wall infiltrations in peripherally located bronchial carcinomas (Padovani et al. 1993). Sensitivity and specificity is about 90%. Most reliable are T1-weighted fat-saturated images after contrast. The accuracy of diagnosis of invasion of

685

686

5  Thorax and Vasculature

Fig. 5.1.14a–c  Aggressive fibromatosis of the left cervicothoracic junction (Pancoast-type tumor). T2 TIRM (a) shows moderately high signal intensity, while pre-contrast T1 VIBE (b) only shows low signal intensity. Post-contrast T1 VIBE (c) shows

strong homogeneous enhancement. Note that the extension through the cervicothoracic junction is best illustrated by the sagittal plane

Fig. 5.1.15a,b  Lung cancer in the left hilum with histologically proven infiltration of the aortic wall. Spiral CT (a) shows broad contact of the tumor with the aortic wall, infiltration can neither

be excluded nor confirmed. ECG-triggered non-enhanced T1weighted image the same level (b) demonstrates consumption of perivascular fat and infiltration of the aortic wall

5.1  Lungs, Pleura, and Mediastinum

bones depends on the region, as infiltration of the ribs is diagnosed earlier on CT because it outlines cortical bone more accurately. Additionally, MRI may show double angulated structures such as the ribs in each imaging plane with section artifacts. Thus, only the advanced stages of osseous invasion of the ribs when the tumor has penetrated through the cortical bone and contrasts with the high T1-weighted signal of the bone marrow are well detected by MRI. Likewise, intratumoral calcifications are reliably detected on conventional radiographs or CT but are usually overlooked on MRI for their low signal intensity which approaches that of air. In contrast MRI is a reliable tool for the demonstration of spinal tumor infiltration and especially for the evaluation of the myelon. Here recognition of details is much better than with any other imaging modality. A standard protocol of MRI to determine the T stage of bronchial carcinoma should include a 3D T1-weighted fat saturated sequence (e.g., VIBE) before and after GdDTPA, an ECG-triggered T2-weighted single-shot TSE (HASTE) sequence and, if applicable, respiratory gated T2-weighted TSE and dynamic True-FISP sequences for the evaluation of respiratory mobility. N Stage In about half the patients with lung cancer, mediastinal lymph node metastases are found at the time of diagnosis. Lymph node involvement is an important prognostic indicator. Nodal spread is classified as N1, N2, and N3. The latter class includes the supraclavicular and scalene nodes, which were formerly considered distant metastases and are amenable to radiation therapy. Accurate nodal staging is related to surgical indication and therefore important for patient management and prognosis. On MRI the common mediastinal lymph node metastases are best depicted by T1-weighted sequences if they are surrounded by mediastinal fat. In these cases the signal void of great vessels and large airways as well as the bright mediastinal fat nicely contrast with the soft tissue of the intermediate-signal-intensity lymph node. T2-weighted sequences do not yield further information because the signal-to-noise ratio is less favorable, and the signal intensities of lymph nodes and the mediastinal fat are too similar to those seen on T1-weighted images. Paramagnetic intravenous contrast agents offer no advantage because they reduce the contrast between lymph nodes and fat. For both techniques the combination with a robust fat saturation is extremely important. Using HASTE or T1-weighted FS GRE, oone may get high signal contrast of the nodes in low-signal-intensity surroundings. Intravenous gadolinium may also increase the specificity of MRI in mediastinal lymph node staging, especially if used in conjunction with dynamic T1weighted SE sequences (Laissy et al. 1994). One study with large numbers concluded that STIR TSE MR images enable differentiation of lymph nodes with metastasis

from those without metastasis (Ohno et al. 2004b). Intravenous iron oxide (USPIO) in future may aid in differentiating metastatic from non-neoplastic lymph node enlargement as tumor-harboring lymph nodes reportedly accumulate less iron oxide than do reactive nodes (Nguyen et al. 1999; Will et al. 2006). Overall, MRI without specific contrast media stages lymph node metastasis in lung cancer with an accuracy comparable to that of contrast-enhanced CT. For both techniques a mean sensitivity and specificity of about 70% is reported. Values depend on the size threshold. The best discrimination can be obtained by a value of 13 mm. Nevertheless, some reactive lymph nodes are greater and some lymph node metastases are smaller than 13 mm. Signal intensities and contrast enhancement are not able to discriminate metastatic lymph nodes as reliably as PET seems to. Although new papers may appear with sophisticated analyses of the signal intensities and contrast behavior of benign and malignant nodules, at present, MRI analysis gives only crude hints. M Stage Distant metastasis is frequently associated with lung cancer and can involve many organs. MRI of the adrenal glands, a frequent site of metastasis, is of special interest. MR studies using chemical shift imaging can help differentiate adrenal adenomas from metastases by demonstrating the lipid within the benign adrenal mass. Sensitivities and even specificities up to 100% were reported for this easy technique. MRI is also the imaging modality of choice for liver metastases and bone marrow infiltrations as the earliest form of skeletal metastases. Recently whole-body MRI has been applied in tumor patients for early and complete staging. 5.1.7.10.3  Lung Metastases A solitary pulmonary nodule is defined as an approximately round lesion less than 3 cm in diameter that is completely surrounded by pulmonary parenchyma without other pulmonary abnormalities (Ost et al. 2003). It is found on about 0.15% of all chest radiographs. MDCT is able to detect nodules as small as 1–2 mm, although it is well known that radiologist overlook up to 30% of small nodules on MDCT. On the other hand, nodules smaller 7 mm are of course a very common finding in healthy individuals as well. About 30% of benign nodules are resected unnecessarily. The diagnostic challenge therefore includes not only detection but also differentiation of malignant from benign lung nodules. Lung metastases are common and have been found in 20–30% of all patients with malignant tumors. They arise hematogenically and are multiple as a rule. Miliary spreading and also lymphangiosis carcinomatosa occur rarely but are possible. The most common primary tu-

687

688

5  Thorax and Vasculature

mors are seminoma, melanoma, sarcoma, and hypernephroma. Nowadays MRI is able to detect about 95% of all pulmonary nodules larger than 3 mm compared with MDCT (Fig. 5.1.16). Techniques are T2-weighted HASTE, contrast-enhanced 3D GRE, or triggered T2-weighted TSE (Yamashita et al. 1999). For the HASTE sequence Schröder et al. (2005) found a sensitivity of 92.1% for nodules 3 mm or larger in diameter and a sensitivity of 97.6% for nodules larger than 5 mm. Schäfer et al. (2005) detected 84% of all nodules and 96% of nodules larger than 4 mm with a 3D GRE technique and 81% of all nodules and 96% of nodules larger than 4 mm with a HASTE sequence. Lutterbey (1998) successfully used a triggered T2-weighted TSE sequence with a sensitivity of 98% for nodules larger than 3 mm. The use of intravenous contrast medium shows no additional benefit for nodule detection, but cell carcinoma promising results for nodule characterization. Several studies (Schaefer et al. 2004, 2006; Kim et al. 2004; Kono et al. 2004) have shown that the enhancement pattern of malignant tumors is faster and higher than that of benign nodules. Absence of enhancement is a strong predictor that a lesion is benign. The parameters to analyze are the maximum enhancement ratio, the slope or time to maximum enhancement ratio and the washout ratio. Typically a malignant lesion should show a slope of more than 15% per minute and a maximum enhancement ratio of about 100%. Disadvantages of MRI are the limited spatial resolution for lesions smaller than 7 mm and the inability to detect small calcifications, which are also a predictor for benign character. 5.1.8  Diseases of the Pleura 5.1.8.1  Pleural Effusion Two types of pleural effusion can be differentiated. The transudate is a clear or serous fluid with a specific gravity below 1.016 and a protein concentration below 3 g/dl. The most common pleural transudates result from congestive heart failure or from severe hypoproteinemia as is frequently seen in liver cirrhosis. Pleural fluid with a specific weight above 1.01 and a protein content above 3.0 g/dl is termed an exudate. Exudates may result from cancers or metastasis in lung or pleura, or from an unspecific or viral pleuritis. A suppurative exudate forming a pyothorax may accompany bacterial pneumonias including tuberculosis or may occur postoperatively. Sanguineous effusions (hemothorax) occur after bleeding into the pleural space due to trauma, surgery, or aortic rupture. Chylothorax refers to a pleural collection of lymphatic fluid that may result from obstruction of major lymph ducts by a mediastinal tumor or from lymphatic fistulas. Chylothorax after trauma or surgery is also common.

Fig. 5.1.16a–c  Multiple small pulmonary metastases (<1 cm) in a patient with renal cell carcinoma. All pulmonary nodules demonstrated by CT (a,b) are also visualized as high-signal-intensity lesions on the corresponding T2 ultrashort turbo spinecho MR image (c)

Pleural effusions on MRI present as semilunar structures between the chest wall and lung. Even small fluid volumes can be detected within the pleural space. Sagittal and coronal sequences may provide valuable additional

5.1  Lungs, Pleura, and Mediastinum

information as the costophrenic and costomediastinal angles are better perceived. Their signal characteristics vary with the composition of the effusion. Transudates are of low T1-weighted signal intensity but high T2-weighted signal intensity. The protein-rich exudates, including hemothorax and chylothorax, demonstrate an intermediate or high T1-weighted signal but share a high T2-weighted signal with the transudates. 5.1.8.2  Pleural Mesothelioma Many malignant pleural mesotheliomas (MPM) have in the past been misclassified as pleural effusions from bronchogenic cancer and therefore, accurate epidemiological data are lacking (Aisner et al. 1995). MPM has been termed a signal tumor of asbestos exposure as the occurrence of MPM is linked dose dependently to exposure. The incidence of MPM continues to rise, and new diagnoses of MPM are expected to peak between 2010 and 2030 (Peto et al. 1999). MPM currently has an incidence rate of 2 to 4 per 100,000 (Neumann et al. 2001). The disease is three times more common in males than in females. MPM is a serosal (epithelium lining the coelom) tumor and derives from pluripotent serosal cells that may differentiate into epithelium or mesenchyma, giving rise to the two principal histopathological types of mesothelioma. The WHO classifies three microscopic subtypes of MPM: the epithelioid, the biphasic, and the sarcomatoid mesothelioma. The most frequent is epithelioid MPM in 50–60% of cases followed by the biphasic type in 20–40%. Sarcomatoid tumors are found in approximately 10% (Mackay et al. 1991). Histological diagnosis and classification requires ample biopsy material which is usually obtained by minimally invasive techniques such as videoassisted thoracic surgery (VATS). The first staging system for MPM used in the past was proposed 1976 by Butchart (Butchart et al. 1976). However, using the Butchart system it could not be proven that patients in stage I had a more favorable prognosis than those in stage II. Therefore, Boutin et al. (1995) introduced a modified staging system that incorporated the still-accepted version of the staging system of the International Mesothelioma Interest Group. The current TNM classification of MPM was developed to incorporate MPM into the TNM syllabus. Moreover, in its current form the TNM system accommodates the progress made in surgical treatment and in multimodal therapy concepts. Transferring the studies of Pisani et al. (1988) into the actual TNM system, most patients are not diagnosed in stage I, i.e., a stage amenable to modern therapies, but in stages II or III where therapeutic options are still unsatisfactory. MPM is often diagnosed in the later stages because there are no early symptoms. Clinical symptoms such as nonspecific thoracic pain and dyspnea usually occur late

in the course of the disease. Pain may be explained by invasion of nerves whereas dyspnea results from the pleural effusions that are usually voluminous at the time of diagnosis. Tumor invasion into the adjacent organs may result in dysphagia if the esophagus is involved and may cause superior vena cava syndrome if the vein is compressed or invaded. Distant metastases are rarely found in MPM and thus the clinical presentation is mainly determined by thoracic tumor extent. Tumor stage and the histologic subtype are the most important prognostic factors in MPM, and for the former, pretherapeutic imaging is essential. Imaging for the purpose of staging should be scaled to the extent of disease. If a tumor is advanced and will only be treated palliatively, it may well suffice to evaluate the patient once by CT and subsequently follow the patient by chest radiographies and ultrasound. In these patients cross-sectional imaging beyond the initial staging seems warranted only if complications arise. In contrast, for younger patients and those operated on or otherwise treated with curative intent, meticulous evaluation during both staging and follow-up are required, and MRI may be indicated. Typical of unilateral MPM is a contraction of the affected hemithorax because the tumor encases the lung and the lobes as it grows along the pleural fissures, forming lobulated thickenings that even involve the mediastinal pleura (Fig. 5.1.17). A pleural tumor is a nodular pleural thickening of more than 3 mm (Metintas et al. 2002), and nodular thickening of more than 1 cm is highly suggestive of a malignant process. The tumors found in MPM typically extend more than 8 cm craniocaudally and more than 5 cm in the axial plane. Tumor calcification is rarely observed. In about 50% of cases enlarged mediastinal lymph nodes greater than 1.5 cm are detected. With respect to T1 stage, current cross-sectional imaging methods are limited in that they cannot differentiate between visceral and parietal pleura. Therefore, in stage I it is impossible to reliably exclude disease of the visceral pleura which would translate into stage Ib, whereas in stage Ia only the parietal pleura is affected. As most patients are diagnosed in the higher tumor stages it is of greater importance to reliably diagnose invasion of lung parenchyma or the diaphragm (T2), involvement of mediastinal or chest wall fat planes, or extension into the pericardium (T3) (Fig. 5.1.18) or into or even through mediastinal organs, chest wall or the diaphragm (T4). On MRI MPM demonstrates muscle-like intensities on T1-weighted images and on T2-weighted sequences often yields high signal intensities that may be considerably inhomogeneous depending on the proportion of fluid within the lesions. Contrast agents will not aid in the differential diagnosis of pleural lesions and are not required in MRI or in CT (Knuuttila et al. 2001). It is recommended that data be acquired in all imaging planes. Axial images will show tumor extent along the intercostal pleura, will demonstrate the typical ipsilateral

689

690

5  Thorax and Vasculature

Fig. 5.1.18  Pleural mesothelioma on the right side. ECG-triggered T1-weighted image obtained shows an isointense tumor with pericardial infiltration clearly depicted by tumor extension into the anterior subpericardial fatty tissue (black arrow)

Fig. 5.1.17a,b  Pleural mesothelioma. T2 TIRM (a) shows multiple high signal intensity pleural tumor nodules on the left with involvement of the fissure as well as pleural effusion on the right, mediastinal lymphadenopathy with intermediate signal intensity on tumor nodule in the left chest wall. Contrastenhanced T1 VIBE (b) shows rim enhancement of the pleural nodules while lymph nodes and chest wall nodule show rather enhancement

shift of the mediastinum and ipsilateral contraction of the hemithorax, and will depict mediastinal invasion. Sagittal sequences will visualize involvement of the diaphragm, and coronal images will best show the costophrenic angles and lateral or apical parts of the chest wall. If surgery is intended, both histopathological examination and precise knowledge about the tumor extension is crucial for planning a procedure. Patz et al. (1992) have found excellent sensitivities (90–100%) but poor specificities (< 50%) of both CT and MRI in 41 patients with MPM. Layer et al. (1999) found higher specificities but lower sensitivities. Two factors may explain the differing findings: Criteria for tumor invasion differed between the two studies, and the study by Layer et al. included a greater number of advanced tumors. Diagnostic accuracy, which is characterized by both sensitivity and specificity, was comparable between both studies.

The radiological modalities employed in follow-up of MPM vary in accordance with the therapy chosen. In the rare cases of curative therapy, radiological follow-up has to satisfy the highest standards to detect complications or relapses and therefore serial MRI or CT examinations are justified. For the majority of patients who are treated palliatively, chest radiographs and ultrasound will suffice to follow the progression of disease and to diagnose most complications. If major complications such as the superior vena cava syndrome occur, a more detailed local restaging by cross-sectional imaging is needed. Routine whole-body imaging or extensive restaging including bone scintigraphy or abdominal ultrasound seems unwarranted in patients under palliative treatment, as distant metastases are rare in MPM. 5.1.9  Mediastinal Disease 5.1.9.1  Lymphoma The most common cause of a mediastinal mass is lymph node enlargement, either from nodal metastasis of bronchogenic carcinoma or lymphoma (either Hodgkin’s disease [HD] or non-Hodgkin’s lymphoma [NHL]). Masses can occur in all parts of the mediastinum. HD has a biphasic age distribution with a first peak in the mid twenties and a second beyond 50 years. It is less common than NHL which has a much more complex histological differentiation and occurs in all age groups. In both forms of lymphoma an intrathoracic manifestation with enlarged

5.1  Lungs, Pleura, and Mediastinum

nodes is common in more than 50% of cases. NHL is more prone to extranodular involvement, while HD often affects the thymus. Whereas an asymmetric manifestation in hilar disease is common, mediastinal affection is characterized by a chimney aspect. Such an aspect is quite atypical for mediastinal lymph node enlargement of other etiologies. About 50% of all bronchogenic carcinomas already are combined with enlarged positive mediastinal nodes at the point of diagnosis. Lastly, a different important cause of mediastinal masses is sarcoidosis. At present, anatomic MR imaging of lymph nodes is primarily a problem-solving tool for cases with inconclusive CT results (Boiselle et al. 1998). Non-enhanced T1 GRE and T2 HASTE are limited in the detection of metastatic mediastinal lymph nodes. Principally, spectral fat saturation does enhance the visibility of lymph nodes on T2-weighted images such as those acquired with HASTE, but particularly in the mediastinum it is often inhomogeneous and incomplete. Therefore, short inversion recovery sequences (STIR or TIRM) are recommended for the detection and follow-up of mediastinal lymph nodes in multiple breath-hold acquisition technique (Fig. 5.1.19). Fat saturated post-contrast T1 GRE acquisitions (Fig. 5.1.19) are very sensitive. Hodgkin’s disease and non-Hodgkin’s lymphoma display a similar appearance in MRI: They are isointense towards muscle tissue and relatively homogeneous on T1-weighted images. On T2-weighted images they are typically bright because of increased water content due to high cell density, comparable to edema and inflammation. Post-contrast, enhancement is pronounced. Because of the high risk of recurrence follow-up examinations have to be performed regularly. While bulky disease is easily visible on any MR image, a respiratory triggered T2- or T1-weighted coronal image can display the lymphoma masses in very great detail. Internal nodular texture can be discerned and the relationship to vessels is well demonstrated. Another important aspect of staging is the identification of local infiltration. An ECG-triggered FSSP gradient-echo cine study will help in case a local enhancement cannot be identified. In this circumstance heart motion will be linked to tumor motion and local contraction will be limited. In the same manner ECG-triggered, fat-saturated FLASH 2D sequences can be used to identify local infiltration of vessel walls. Residual “scar” lymphoma tissue is characterized by lower signal intensity than active tumor tissue (Rahmouni et al. 2001). The diagnostic dilemma is that there is no qualitative or quantitative cut-off point for differentiating tissues in order to optimize follow-up or early therapy response evaluation. The accuracy of MR performed at 1 and at 3 months after the end of therapy is not satisfying. This represents a clinical problem because the most important clinical decisions have to be made just in this early post-treatment phase (Di Cesare et al. 2004). For differential diagnosis of enlarged mediastinal lymph nodes MRI cannot deliver a substantial benefit. A

Fig. 5.1.19a–c  Hodgkin’s disease of the cervicothoracic junction with extensive nodular and pulmonary involvement shows intermediate signal on T1-weighted (a), high signal on the TIRM sequence (b) and strong enhancement on post-contrast VIBE (c)

691

692

5  Thorax and Vasculature

specific pattern for histology is not shown by signal intensities in any sequence, contrast enhancement behavior, relaxation time measurements (Glazer et al. 1988), spectroscopy, or any other technique. 5.1.9.2  Thymoma In case of the anterior mediastinum a very common tumor is the thymoma (Maher and Shepard 2005). It appears as a well-defined rounded or lobulated mass that is mostly homogeneous and in the typical position of the thymus. An irregular border and compression of vessels is suggestive of tumor infiltration. Thymomas typically show isointense signal on T1-weighted images and are bright on T2-weighted images. Post-contrast, enhancement is low (Kushihashi et al. 1996). They might contain cystic features. They have to be differentiated from large and poorly defined, infiltrative masses, which also might contain cysts, hemorrhage, and necrosis. These features can occur in thymic carcinoma and mediastinal germ cell tumors (e.g., teratoma). The differentiation between malignant thymic infiltration, benign thymoma, and thymic rebound after chemotherapy or radiation is generally not possible in MRI. 5.1.9.3  Goiter An intrathoracic goiter is the most frequent space occupying lesion in the anterior mediastinum, accounting for 10% of all mediastinal masses. On radiographs the mediastinum above the aortic arch is widened in an arcuate sweep with the widest diameter at the thoracic inlet. The trachea is often compressed and may be displaced. Ultrasound will only visualize the superior but not the intrathoracic parts of an enlarged thyroid. Intrathoracic extension of a goiter may be estimated on radioisotope studies and is well outlined on CT. MRI adequately depicts the form, shape, and location of thyroid masses in the mediastinum. Sagittal sequences are especially helpful as they demonstrate the craniocaudal extent of the gland well (Fig. 5.1.20). The differentiation of thyroid tissue from the trachea and the great vessels in the mediastinum and the neck on MRI is superior to other imaging modalities (Belardinelli et al. 1995) and does not require intravenous contrast. This is of particular advantage if hyperthyroid goiter or thyroid malignancy are suspected. Goiter is iso- to hyperintense on T1- and T2-weighted images. As a drawback, MRI will not adequately demonstrate intrathyroidal calcifications that frequently exist in multinodular goiters. In general, signal intensities of intrathoracic thyroid tissue are not different from those of other mediastinal lesions. However, due to the multitude of possible regressive changes such as necroses, cysts, calcium deposits and hemor-

Fig. 5.1.20  Large intrathoracic goiter with displacement of the trachea and supra-aortic vessels. Slight inhomogeneities indicate regressive changes

rhage, MRI will show considerable signal inhomogeneities within the gland tissue. A differentiation between benign and malignant changes is impossible by CT or MRI and differential diagnoses include malignant lymphomas, thymomas, atypically located teratomas, and also neurogenic tumors. Routine use of CT or MRI is not indicated in the evaluation of a thyroid nodule, but each is useful in selected circumstances. Either CT or MRI can accurately determine substernal extension and invasion of surrounding structures, such as esophagus, larynx, or trachea, and should be used only if such extension or invasion is suspected. CT delivers an iodine load to the patient that can delay thyroid scanning or radionuclide therapy and also cause a subclinical hyperthyroidism to enter thyroid crisis. Thus, MRI is preferred. 5.1.9.4  Cysts and Inflammation Cystic lesions with round, T2 hyper- and T1 hypointense formations can be found in any part of the mediastinum. A circular increase in intensity with a T2-weighted high signal is suggestive of cystic lesions or abscess formation. In case of a mediastinitis there is a diffuse enhancement of mediastinal tissues that have previously been isointense. Bronchogenic cysts are the most common bronchopulmonal malformation of the adult. They develop from a disordered growth of the tracheobronchial mucosa, and a third of them manifest in a mediastinal rather than an

5.1  Lungs, Pleura, and Mediastinum

5.1.10  Differential Diagnosis

Fig. 5.1.21  Bronchogenic cyst shows relatively high signal intensity of the cyst on the T1-weighted image due to its highly mucinous contents. Good visualization of the relationship of the cyst to the tracheal bifurcation and the right main bronchus

intrapulmonary location. Only infection or abnormal growth will lead to clinical discomfort and concerted diagnostic approaches. In other cases these cysts are found incidentally. The typical MR features include direct contact with the tracheobronchial system. Signal characteristics in T1-weighted imaging depend on the substance of the cyst (Murayama et al. 1995). Most of them contain protein-rich mucous content, leading to high signal intensity not only on T2- but also on T1-weighted imaging (McAdams et al. 2000) (Fig. 5.1.21). 5.1.9.5  Neurogenic Tumors Neurogenic tumors are the most common lesions in the dorsal mediastinum. They originate from peripheral nerves (e.g., neurofibroma), sympathetic (e.g., neuroblastoma), and parasympathetic ganglia. These tumors are heterogeneous with high signal intensity on T2weighted imaging and isointense but sometime whorled structures on T1-weighted imaging with rapid enhancement after contrast administration (Tanaka et al. 2005). Schwannoma and neurofibroma may manifest as a dumb bell–shaped paraspinal mass, one part within the spinal canal, another expanding through the neural foramen. MRI is the best-suited imaging method for the evaluation of these tumors because no other technique is able to depict the anatomical details of the relationship of the tumors to the spinal canal and the myelon in such a detailed manner.

MRI of the chest provides a wide range of diagnostic information to characterize pathological processes. As with CT, detection of structure enables the recognition of typical patterns based on shape, distribution, and signal intensity. In addition, further information can be obtained from the characterization of the different tissue compounds by T1- and T2-weighted imaging as well as fat-suppressed sequences and post-contrast acquisitions. Further characterization is feasible by MR angiography with high spatial or temporal resolution as well as perfusion or ventilation imaging. Very novel developments looking at respiratory mechanics are also of great benefit in special clinical indications. Vascular malformations are especially well depicted with MRI. Vascular malformations and pulmonary sequestrations are clear indications for MRI although no systematic studies are available. MRI is capable of demonstrating arterial feeders and draining veins of sequestrations. MRI is also very well suited for demonstrating anomalous pulmonary venous return, including the Scimitar syndrome. MRI can easily differentiate these entities from other diagnoses, such as diaphragmatic hernia, empyema, as well as benign or malignant pleural tumors. Very novel developments looking at respiratory mechanics are also of great benefit in special clinical indications. The major advantage is that all important aspects of the disease can be addressed with different MRI techniques. In children and adults, cystic fibrosis is a major indication for MRI. MRI is superior to chest radiography and almost as good as CT. Thus, it is very well suited for monitoring patients at annual follow-ups. MRI is also very helpful for differentiating cystic lesions within the mediastinum. The signal intensity of bronchogenic cysts is highly dependent on their contents. In general, an increasing amount of protein and blood will lead to higher signal intensity on the T1-weighted image, associated with very high signal intensity on T2-weighted images. Frontal oblique planes allow demonstrating the course of the tracheobronchial tree and exact relationship with the cystic lesion. However, even in this situation it might still be a challenge to tell whether the lesion actually originates from the trachea, bronchial tree, or esophagus, especially if there is not much mediastinal fat separating these structures. The importance of MRI for the differential diagnosis of pneumonia in young patients is already evident and will slowly find its way into clinical routine. The combination of visualization of an infiltrate, early detection of abscess formation as well as the subsequent effects on perfusion and ventilation generates new dimensions that cannot be obtained by CT. This amount of functional information compensates also for the lower spatial resolution of MRI when compared to CT in the diagnosis of interstitial lung disease. MRI also offers major advantages in making the

693

694

5  Thorax and Vasculature

differentiation between active alveolitic processes and fibrotic scars in interstitial lung disease. As the pattern of ground class attenuation is very sensitive but at the same time non-specific, further characterization of the underlying process is warranted. From CT, it is well known that ground class attenuation can represent a purely interstitial or alveolitic process or combination of both. In the future larger studies will have to generate the evidence that MRI is capable of solving this diagnostic dilemma. The potential of MRI to separate atelectasis from tumor and differentiate different types of atelectasis might markedly facilitate the final diagnosis of the underlying disease. By doing so, MRI will also substantially support the planning of surgery or radiotherapy of lung cancer. Clear differentiation between benign and malignant lung tumors is not yet possible with MRI. The use of different MRI techniques, such as measurement of T1- and T2-relaxation times, has not been successful. The detailed analysis of signal intensity curves after contrast administration only provides rough suggestions about the histology of the lesion. However, MRI has clear advantages in the determination of the exact extent of a tumor, especially when it comes to the infiltration of mediastinum, pleura, chest wall, or apex. Coronal and sagittal planes using T1-weighted fat-saturated sequences post-contrast and T2-weighted sequences are especially helpful to demonstrate the infiltration of the affected structures. The exact definition of the extent of the infiltration towards the brachial plexus in Pancoast tumors can easily be done by MRI. MRI is recommended in patients with T4 tumors with central infiltration, for evaluating the exact extent with regard to the pulmonary arteries, veins, thrombus formation, and tracheobronchial tree. The potential of MRI with regard to the detection of recurrent tumors after surgery or radiotherapy is still unclear. In the detection and characterization of pulmonary nodules (<3 cm) the role of MRI is also not yet determined. It is well known that MRI has a low sensitivity for the detection of calcifications, which are a major indicator for a benign lesion. However, it is important to detect a malignant or suspicious nodule as a positive finding. In this situation the simple non-visualization of a small calcified lesion that is considered benign can also be regarded as helpful. Malignant and inflammatory processes will have higher signal intensity on T2- and post-contrast T1-weighted imaging, so they are easily picked up. In nodules with sizes between 0.5 and 2.0 cm the time curve of contrast-enhancement on MRI might be useful to characterize the lesion. If the lesions are bigger, the assessment of perfusion will have a lower value with regard to their characterization. MRI is also very helpful for differentiating cystic lesions within the mediastinum. The signal intensity of bronchogenic cysts is highly dependent on their contents. In general, an increasing amount of protein and blood

will lead to higher signal intensity on the T1-weighted image, whereas on T2-weighted images very high signal intensity is common. Frontal oblique planes allow demonstration of the course of a trachea bronchial tree and the exact relationship between the cystic lesion and the trachea bronchial system. However, even in this situation it might still be a challenge to tell whether the lesion actually originates from the trachea and the bronchial tree or from the esophagus, especially if there is not much mediastinal fat between these structures. The hallmark of pleural effusion on axial images is the collection of fluid in the dependent part of a semicircularly widened pleural space. Typically, the signal intensity of the fluid collection is homogeneous. Diagnosis of loculated pleural effusions can be difficult as their signal characteristics may not differ from those of solid pulmonary, mediastinal, or pleural lesions. Likewise, MRI cannot differentiate reliably between the various fluid entities in the pleural space such as transudate, exudate, blood, chyle, or pus. But MRI may provide clues to the composition of the pleural fluid—the higher the protein or blood content of the fluid, the higher will be the T1-weighted signal of the effusion. The T2 signal is high in all effusions, and thus T2-weighted images do not add to the differential diagnosis of pleural fluid. The differential diagnosis of pleural masses or pleural thickening is not reliable. Both MPM and metastatic pleural masses from lung cancer are isointense with muscle on T1-weighted images and appear hyperintense and inhomogeneous on T2-weighted sequences. Likewise, the signal characteristics of pleural thickening from various causes are nonspecific and are indistinguishable from the aforementioned pleural masses. To date there is no evidence that the rare benign tumors of the pleura (localized benign pleural mesothelioma, chondroma, hemangioendothelioma) which are included in the differential diagnosis of MPM demonstrate specific MRI signal characteristics. Although MRI provides better soft tissue contrast than does CT, neither contrast-enhanced MRI nor the advanced MR imaging methods described earlier are superior to CT in the differential diagnosis of pleural disease. 5.1.11  Value of MRI with Regard to Other Imaging Modalities Until recently, MRI has only played a secondary role in the imaging of lung disease. MRI is now been accepted as a first-line imaging modality for congenital pulmonary and also cardiovascular anomalies in the chest (Table 5.1.5). This is mainly due to its potential to provide a comprehensive assessment including structure (T1-weighted), fluid content (T2-weighted), inflammation (T1-weighted post-contrast), MR angiography, MR perfusion, and functional MRI of the heart without radiation exposure. As such it can easily be applied during follow-up. The role

5.1  Lungs, Pleura, and Mediastinum Table 5.1.5  Value of MRI with regard to other imaging modalities in diseases of the lungs, pleura, and mediastinum MRI

Ultrasound

CT

Congenital lung   Fetal  Non-fetal

++  +++  ++

++  +++  0

+  0  +++

Infiltrative lung disease

++

0

+++

Atelectasis

+++

+

++

Pneumonia

++

0

+++

Dif. interstitial lung disease

++

0

+++

COPD

+

+

+

Cystic fibrosis

++

0

+

Asthma

++

0

+

Benign lung tumors

+

0

+

Bronchogenic carcinoma  T3/T4

+  +++

+  +

+++  ++

Lung metastases

++

0

+++

Pleural effusion

++

+++

+

Pleural mesothelioma

+++

++

++

Lymphoma

+++

+

++

Thymoma

+

0

++

Goiter

++

+++

+

Bronchogenic cysts

+++

++

++

Neurogenic tumors

+++

0

+

of MRI for the assessment of interstitial, inflammatory, and obstructive disease is not yet determined. MRI may play a major role in primary and secondary lung tumors. Peripheral nodules with a diameter of more 5 mm can be detected with a high degree of confidence. Nodules smaller than 5 mm are detected significantly less frequently on MRI than on CT. CT also has advantages in the detection of nodules adjacent to the pleura or the diaphragm. On the other hand MRI might have advantages in the differentiation between small peripheral pulmonary arteries and pulmonary metastases. T2-weighted and STIR images are reported to be especially sensitive.

The exact localization of pulmonary nodules on MRI is challenging, as peripheral bronchi and the fissure are often much more difficult to delineate on MRI than on CT. For staging of lung cancer MRI has advantages and disadvantages when compared with CT. MRI can show the exact location of the tumor with regard to the central pulmonary vessels, even without the use of contrast media. MRI is especially well suited for showing the potential infiltration of the chest wall or the thoracic apex, with its high soft-tissue contrast. In general MRI is less available, will take more time and will cost more than CT. For nodules as well as stage T1 and T2 tumors both modalities are equivalent, and in clinical routine CT will be preferred. Some studies have looked into the infiltration of adjacent structures defining stage T3 and T4 tumors. For depiction of the infiltration of the parietal pleura and the soft-tissue structure of the chest wall MRI is slightly superior to CT. In the past, the ability to directly acquire coronal or sagittal plane images was a clear advantage of MRI. However, since the advent of MDCT with the acquisition of almost isotropic data sets and multiplanar reformats, the spatial resolution of CT is even better. MRI still has the advantage of its excellent soft-tissue contrast which was already evident in the past, before the use of novel T2-weighted sequences as well as fat-suppressed contrast-enhanced T1-weighted images. Thus, MRI can still be regarded as the first-line modality for the assessment of chest wall and mediastinal infiltration. MRI is also the preferred imaging modality in the examination of Pancoast tumors. In patients who are scheduled for resection MRI will provide excellent assessment of the infiltration of the apex as well as the subclavian artery and the brachial plexus. Coronal and sagittal planes together with fat-suppressed contrast-enhanced T1-weighted images are superior to the artifact-prone axial CT scans. The anatomic information which will be provided by MRI is essential for planning appropriate thoracic surgery. MRI is also the preferred imaging modality when it comes to the potential infiltration of the posterior mediastinum, spine, and spinal canal, where MRI is certainly superior to CT. In general MRI should be performed in patients with advanced, potential T4 disease to show how a successful operation can be performed or to prove irresectability. It is well known that MRI is less sensitive than CT is in the detection of calcifications, which are a major indicator of a benign lesion. However, in the clinical situation, it is important to detect malignant or suspicious nodules that are 0.5–2.0 cm in size. In this situation the simple non-visualization of very small lesions as well as calcified lesions that are considered benign can also be regarded as helpful. Malignant and inflammatory processes will have higher signal intensity and so will easily be picked up. They will also nicely show contrast enhancement in most cases. In nodules 0.5–2.0 cm in size the time curve of contrast enhancement on MRI might be useful to charac-

695

696

5  Thorax and Vasculature

terize the lesion. If the lesions are bigger, the assessment of perfusion will have a lower value with regard to their characterization. CT remains the established imaging modality in the preoperative evaluation of pleural masses but CT is of no value in differentiating between the various etiologies of pleural processes. CT is effective in determining the extent of pleural lesions and their possible infiltration into adjacent structures. Regarding local invasion, there are limitations for CT, as exemplified in the question of pericardial involvement where an intact fat layer does not reliably exclude tumor infiltration. MRI has equaled CT in demonstrating tumor extent, and has even gained an advantage over CT in depicting invasion of the mediastinum or the peritoneum, especially if the latest advances in imaging are employed, such as phase-array coil, parallel imaging, ECG triggering, or breath-hold techniques. Ultrasound of the thorax and abdomen supplements pretherapeutic imaging and planning in MPM. Evaluation of the costal and diaphragmal pleura in right-sided tumors is comparable on MRI and CT. In left-sided masses, an adequate acoustic window can often not be obtained with MRI and ultrasound examination of the left costodiaphragmal pleura may be impossible.

applied early in high-risk, immunosuppressed patients especially to detect abscess formation and angioinvasive aspergillosis as these will have immediate therapeutic impact. Increasing importance has been gained by fetal MRI. In all cases of indetermined fetal sonography MRI is the method of choice for the detection and prognostic characterization of processes like malformations or hernia. Evaluation of pleural pathology begins with the posterior–anterior and lateral chest radiograph and with ultrasound. These initial techniques will outline pleural effusions sufficiently well and will demonstrate obvious thoracic parenchymal abnormalities. If a more detailed investigation is required, contrast-enhanced CT will be employed. If the underlying pathology is MPM, MRI has a role in clarifying equivocal CT findings of local invasion, especially of the spinal canal and the chest wall. However MRI is not an absolute requisite in the preoperative evaluation of patients, especially since multislice CT with isotropic resolution and multiplanar reconstruction has become available for routine chest imaging. References 1.

5.1.12  Diagnostic Procedure Chest X-ray is still the first radiological examination performed in the assessment of potential chest disease. If the detected pathology should be assessed in more detail or if clinical suspicion persists, pathology should be excluded, CT is the method of choice. The exact localization of lesions is easily done by CT. CT will also demonstrate small peripheral airways and vessels as well as normal interlobular septa which is important to differentiate subtle pathological changes. With MDCT a new level of quality was reached regarding the visualization of lung structure almost without motion artifacts. MRI should be used for staging lung cancer if CT is inconclusive with regard to the infiltration of the chest wall. In particular, Pancoast tumors are a clear indication of MRI. In addition MRI is recommended in all T4 tumors and in all lung cancer patients who have contraindications for the administration of iodinated contrast agents. Dedicated MRI should be performed to improve planning of surgery in complex situations. MRI has demonstrated a comparable sensitivity to CT for the detection of pulmonary nodules with a diameter > 5 mm. It may be used for the detection of nodules and also metastases, but it is not sufficient when it comes to staging prior to pulmonary metastasectomy. However, MRI has advantages in the characterization of nodules when dynamic contrast enhancement techniques are used. MRI has gained increasing importance in the detection and diagnosis of pneumonic infiltrates. It should be

2. 3.

4.

5.

6.

7.

8.

Ahlstrom KH, Johansson LO, Rodenburg JB, Ragnarson AS, Akeson P, Borseth A (1999) Pulmonary MR angiography with ultrasmall superparamagnetic iron oxide particles as a blood pool agent and a navigator echo for respiratory gating: pilot study. Radiology 211:865–869 Aisner J (1995) Current approach to malignant mesothelioma of the pleura. Chest 107:332S–344S Albert MS, Cates GD, Driehuys B, Happer W, Saam B, Springer CS Jr, Wishnia A (1994) Biological magnetic resonance imaging using laser-polarized 129Xe. Nature 370:199–201 Alsop DC, Hatabu H, Bonnet M et al. (1995) Multi-slice, breathhold imaging of the lung with submillisecond echo times. Magn Reson Med 33:678–682 Altes TA, Powers PL, Knight-Scott J, Rakes G, Platts-Mills TA, de Lange EE, Alford BA, Mugler JP III, Brookeman JR (2001) Hyperpolarized 3He MR lung ventilation imaging in asthmatics: preliminary findings. J Magn Reson Imaging 13:378–384 Bader TR, Semelka RC, Pedro MS, Armao DM, Brown MA, Molina PL (2002) Magnetic Resonance Imaging of pulmonary parenchymal disease using a modified breathhold 3D gradient-echo technique: initial observations. J Magn Reson Imaging 15:31–38 Belardinelli L, Gialdi G, Ceroni L et al. (1995) Comparison between computed tomography and magnetic resonance data and pahthologic findings in substernal goiters. Int Surg 80:65–69 Biederer J, Busse I, Grimm J, Reuter M, Muhle C, Freitag S, Heller M (2002a) Sensitivity of MRI in detecting alveolar infiltrates: experimental studies. Fortschr Roentgenstr 174:1033–1039

5.1  Lungs, Pleura, and Mediastinum 9.

10.

11.

12.

13.

14.

15.

16.

17.

18. 19.

20. 21.

22.

23.

24.

Biederer J, Reuter M, Both M, Muhle C, Grimm J, Graessner J, Heller M (2002b) Analysis of artefacts and detail resolution of lung MRI with breath-hold T1-weighted gradient-echo and T2-weighted fast spin echo sequences with respiratory triggering. Eur Radiol 12:378–384 Blum U, Windfuhr M, Buitrago-Tellez C, Sigmund G, Herbst EW, Langer M (1994) Invasive pulmonary aspergillosis. MRI, CT, and plain radiographic findings and their contribution for early diagnosis. Chest 106:1156–1161 Böck JC, Pison U, Wlodarczyk et al. (1997) Kernspintomographische Quantifizierung des Lungenödemgrades mit Hilfe eines makromolekularen Kontastmittels. Fortschr Roentgenstr 167:509–515 Boiselle PM, Patz EF, Vining DJ et al. (1998) Imaging of mediastinal lymph nodes: CT, MR, and FDG PET. Radiographics 18:1061–1069 Both M, Schultze J, Reuter M, Bewig B, Hubner R, Bobis I, Noth R, Heller M, Biederer J (2003) Fast T1- and T2weighted pulmonary MR-imaging in patients with bronchial carcinoma. Eur J Radiol 53:478–488 Boxt LM (1996) MR imaging of pulmonary hypertension and right ventricular dysfunction. Magn Reson Imaging Clin N Am 4:307–325 Butchart EG, Ashcroft T, Barnsley WC et al. (1976) Pleuropneumonectomy in the management of diffuse malignant mesothelioma of the pleura. Experience with 29 patients. Thorax 31:15–24 Caruthers SD, Paschal CB, Pou NA et al. (1998) Regional measurements of pulmonary edema by using magnetic resonance imaging. J Appl Physiol 98:7587–8750 Cluzel P, Similowski T, Chartrand-Lefebvre C, Zelter M, Derenne JP, Grenier PA (2000) Diaphragm and chest wall: assessment of the inspiratory pump with MR imaging-preliminary observations. Radiology 215:574–583 De Geer G, Webb WR, Gamsu G (1986) Normal thymus: assessment with MR and CT. Radiology 158:313–317 De Geer G, Webb WR, Sollitto R, Golden J (1986) MR characteristics of benign lymph node enlargement in sarcoidosis and Castleman’s disease. Eur J Radiol 6:145–148 De Lange EE (2006) Science to practice: what is new about detecting emphysema? Radiology 239:619–620 Di Cesare E, Cerone G, Enrici RM, Tombolini V, Anselmo P, Masciocchi C (2004) MRI characterization of residual mediastinal masses in Hodgkin’s diesease: long-term follow-up. Magn Reson Imaging 22:31–38 Dietrich O, Losert C, Attenberger U, Fasol U, Peller M, Nikolaou K, Reiser MF, Schönberg SO (2005) Fast oxygen-enhanced multislice imaging of the lung using parallel aquisition techniques. Magn Reson Med 53:1317–1325 Edelman R, Hatabu H, Tadamura E et al. (1996) Noninvasive assessment of regional ventilation in the human lung using exygen-enhanced magnetic resonance imaging. Nat Med 11:1236–1239 Eibel R, Herzog P, Dietrich O, Rieger C, Ostermann H, Reiser M, Schönberg S (2006) Nachweis von pneumonischen Infiltraten mit der MRT. Radiologe 46:267–274

25. Estilaei M, MacKay A, Whittall K, Mayo J (1999) In vitro measurements of water content ant T2 relaxatoin times in lung using a clinical MRI scanner. J Magn Reson Imaging 9:699–703 26. Fiel SB, Friedman AC, Caroline DF, Radecki PD, Färber E, Grumbach K (1987) Magnetic resonance imaging in young adults with cystic fibrosis. Chest 91:181–184 27. Fink C, Eichhorn J, Kiessling F, Bock M, Delorme S (2003) Time-resolved multiphasic 3D MR angiography in the diagnosis of the pulmonary vascular system in children. Fortschr Roentgenstr 175:929–935 28. Fink C, Puderbach M, Bock M, Lodemann KP, Zuna I, Schmahl A. Delorme S, Kauczor HU (2004) Regional lung perfusion: assessment with partially parallel three-dimensional MR imaging. Radiology 231:175–184 29. Glazer GM, Orringer MB, Chenevert TL et al. (1988) Mediastinal lymph nodes: relaxation time/pathologic correlation and implications in staging of lung cancer with MR imaging. Radiology 168:429–431 30. Hang JD, Guo QY, Chen CX, Chen LY (1996) Imaging approach to the diagnosis of pulmonary sequestration. Acta Radiol 37:883–888 31. Hasegawa I, Eguchi K, Kohda E, Tanami Y, Mori T, Hatabu H, Kuribayashi S (2003) Pulmonary hilar lymph nodes in lung cancer; assessment with 3D-dynamic contrast-enhanced MR imaging. Eur Radiol 45:129–134 32. Heelan RT, Demas BE, Caravelli JF et al. (1989) Superior sulcus tumors: CT and MR imaging. Radiology 170:637–641 33. Herold CJ, Kramer J, Sertl K, Kahls P, Mallek R, Imhof H, Tscholakoff D (1989) Invasive pulmonray aspergillos: evaluation with MR imaging. Radiology 173:717–721 34. Herold CJ, Kuhlmann JE, Zerhouni EA (1991) Pulmonary atelectasis: signal patterns with MR imaging. Radiology 178:715–720 35. Heussel CP, Sandner A, Voigtländer T, Heike M, Deimling M, Kuth R, Rupprecht T, Schreiber WG, Kauczor HU (2002) Prospective feasibility study of chest X-ray vs. thoracic MRI in breath-hold technique at an open low-field scanner. Fortschr Roentgenstr 174:1582 36. International Mesothelioma Interest Group (1995) A proposed new international TNM staging system for malignant pleural mesothelioma. Chest 108:1122–1128 37. Iwasawa T, Kagei S, Gotoh T, Yoshiike Y, Matsushita K, Kurihara H, Saito K, Matsubara S (2002) Magnetic resonance analysis of abnormal diaphragmatic motion in patients with emphysema. Eur Respir J 19:225–231 38. Kim JH, Kim HJ, Lee KH, Lee HL (2004) Solitary pulmonary nodules: a comparative study evaluated with contrastenhanced dynamic MR imaging and CT. J Comput Assist Tomogr 28:766–775 39. Kluge A, Müller C, Strunk J, Lange U, Bachmann G (2006) Experience in 207 combined MRI examinations for acute pulmonary embolism and deep vein thrombosis. Am J Roentgenol 186:1986–1996 40. Knuuttila A, Kivisaari A et al. (2001) Evaluation of pleural disease using MR and CT. With special reference to malignant pleural mesothelioma. Acta Radiol 42:502–507

697

698

5  Thorax and Vasculature 41. Kono R, Fujimoto K, Terasaki H, Müller NL, Kato S, Sadohara J, Hayabuchi N, Takamori S (2007) Dynamic MRI of solitary pulmonary nodules: comparison of enhancement patterns of malignant and benign small peripheral lung lesions. Am J Roentgenol 188:26–36 42. Kushihashi T, Fujisawa H, Munechika H (1996) Magnetic resonance imaging of thymic epithelial tumor. Crit Rev Diagn Imag 37:191–259 43. Laissy JP, Gay-Depassier PH, Soyer PH, Dombret MC, Murciano G, Sautet A, Aubier M, Menu Y (1994) Enlarged mediastinal lymph nodes in bronchogenic carcinoma: assessment with dynamic contrast-enhanced MR imaging. Radiology 191:263–267 44. Layer G, Schmitteckert H, Steudel A et al. (1999) MRT, CT und Sonographie in der präoperativen Beurteilung der Primärtumorausdehnung beim malignen Pleuramesotheliom. Fortschr Roentgenstr 170:365–370 45. Leutner C, Gieseke J, Lutterbey G et al. (1999) MRT versus CT in der Diagnostik von Pneumonien; Evaluation einer T2-gewichteten ultrachnellen Turbo-Spin echo-Sequenz (UTSE) Fortschr Roentgenstr 170:449–456 46. Leutner C, Schild H (2001) MRT des Lungenparenchyms. Fortschr Roentgenstr 173:168–175 47. Ley S, Zaporozhan J, Morbach A, Eberle B, Gast KK, Heussel CP, Biedermann A, Mayer E, Schmiedeskamp J, Stepniak A, Schreiber WG, Kauczor HU (2004) Functional evaluation of emphysema using diffusion-weighted 3helium-magnetic resonance imaging, high-resolution computed tomography, and lung function tests. Invest Radiol 39:427–434 48. Loffler R, Muller CJ, Peller M, Penzkofer H, Deimling M, Schwaiblmair M, Scheidler J, Reiser M (2000) Optimization and evaluation of the signal intensitiy change in multisection oxygen-enhanced MR lung imaging. Magn Reson Med 43:860–866 49. Lutterbey G, Gieseke J, Sommer T, Keller E, Kuhl C, Schild H (1996) Ein neuer Ansatz in der Magnetresonanztomographie der Lunge mit einer ultrakurzen Turbo-Spin-echoSequenz (UTSE) Fortschr Roentgenstr 164 50. Lutterbey G, Gieseke J, von Falkenhausen M, Morakkabati N, Schild H (2005) Lung MRI at 3.0: a comparison of helical CT and high-field MRI in the detection of diffuse lung disease. Eur Radiol 15:324–328 51. Lutterbey G, Leutner C, Gieseke J et al. (1998) Detektion fokaler Lungenläsionen mit der Magnetresonanz-Tomographie mittels T2-gewichteter Ultrashort-Turbo-Spin-echoSequenz im Vergleich zur Spiral-Computer-Tomographie. Fortschr Roentgenstr 169:365–369 52. Mackay B, Lukeman JM, Ordonez NG (eds) (1991) Tumors of the pleura and chest wall. In: Tumors of the lung. Major problems in pathology series, vol. 24. Saunders, Philadelphia, pp 323–364 53. Maher MM, Shepard JA (2005) Imaging of thymoma. Semin Thorac Cardiovasc Surg 17:12–19 54. McAdams HP, Kirejczyk WM, Rosado-de-Christenson ML, Matsumoto S (2000) Bronchogenic cyst: Imaging features with clinical and histopathologic correlation. Radiology 217:441–446

55. McLoud TC, Filion RB, Edelman RR, Shepard JA (1989) MR imaging of superior sulcus carcinoma. J Comput Assist Tomogr 13:233–239 56. Metintas M, Ucgun I, Elbek O et al. (2002) Computer tomography features in malignant pleural mesothelioma and other commonly seen pleural diseases. Eur J Radiol 41: 1–9 57. Molinari F, Fink C, Risse F, Tuengerthal S, Bonomo L, Kauczor HU (2006a) Assessment of differential pulmonary blood flow using perfusion magnetic resonance imaging: comparison with radionuclide perfusion scintigraphy. Invest Radiol 41:624–630 58. Molinari F, Gaudino S, Fink C, Corbo GM, Valente S, Pirronti T, Bonomo L (2006b) Simultaneous cardia and respiratory synchornization in oxygen-enhanced magnetic resonance imaging of the lung using a pneumotachograph for respiratory monitoring. Invest Radiol 41:476–485 59. Morino S, Toba T, Araki M, Azuma T, Tsutsumi S, Tao H, Nakamura T, Nagayasu T, Tagawa T (2006) Noninvasive assessment of pulmonary emphysema using dynamic contrast-enhanced magnetic resonance imaging. Exp Lung Res 32:55–67 60. Mueller NL, Mayo JR, Zwirewich CV (1992) Value of MR Imaging in the evaluation of chronic infiltrative lung diseases. AJR Am J Radiol 158:1205–1209 61. Muller CJ, Loffler R, Deimling M, Peller M, Reiser M (2001) MR lung imaging at 0.2 T with T1-weighted true FISP: native and oxygen-enhanced. Magn Reson Imaging 14:164–168 62. Murayama S, Murakami J, Watanabe H et al. (1995) Signal intensity characteristics of mediastinal cystic masses on T1-weighted MRI. J Comput Assist Tomogr 19:188–191 63. Neumann V, Gunthe S, Müller KM, Fischer M (2001) Malignant mesothelioma—German Mesothelioma Register, 1987–1999. Int Arch Occup Environ Health 74:383–395 64. Ngueyen BC, Stanford W, Thompson BH, Rossi NP, Kernstine KH, Kern JA, Robinson RA, Amorosa JK, Mammone JF, Outwater EK (1999) Multicenter clinical trial of ultrasmall superparamagnetic iron oxide in the evaluation of mediastinal lymph nodes in patients with primary lung carcinoma. Magn Reson Imaging 10:468–473 65. Nguyen BC, Stanford W, Thompson BH, Rossi NP, Kernstine KH, Kern JA, Robinson RA, Amorosa JK, Mammone JF, Outwater EK (1999) Multicenter clinical trial of ultrasmall superparamagnetic iron oxide in the evaluation of mediastinal lymph nodes in patients with primary lung carcinoma. Magn Reson Imaging 10:468–473 66. Ohno Y, Adachi S, Motoyama A, Kusumoto M, Hatabu H, Sugimura K, Kono M (2001) Multiphase ECG-triggered 3D contrast-enhanced MR angiography: utility for evaluation of hilar and mediastinal invasion of bronchogenic carcinoma. J Magn Reson Imaging 13:215–224 67. Ohno Y, Hatabu H, Takenaka D, Adach S, Hirota S, Sugimura K (2002a) Contrast-enhanced MR perfusion imaging and MR angiography: utility for managnement of pulmonary arteriovenous malformations for embolotherapy. Eur J Radiol 41:136–146

5.1  Lungs, Pleura, and Mediastinum 68. Ohno Y, Hatabu H, Takenaka D, Van Cauteren M, Fujii M, Sugimura K (2002b) Dynamic oxygen-enhanced MRI reflects diffusing capacity of the lung. Magn Reson Med 47:1139–1144 69. Ohno Y, Hatabu H, Higashino T, Takenaka D, Watanabe H, Nishimura Y, Yoshimura M, Sugimura K (2004a) Dynamic perfusion MRI versus perfusion scintigraphy: prediction of postoperative lung function in patients with lung cancer. Am J Roentgenol 184:73–78 70. Ohno Y, Hatabu H, Takenaka D, Higashino T, Watanbe H, Ohbayashi C, Yoshimura M, Satouchi M, Nishimura Y, Sugimura K (2004b) Metastases in mediastinal and hilar lymph nodes in patients with non-small cell lung cancer: quantitative and qualitative assessment with STIR turbo spin-echo MR Imaging. Radiology 231:872–879 71. Ost D, Fein AM, Feinsilver SH (2003) Clinical practice. The solitary pulmonary nodule. N Engl J Med 348:2535–2542 72. Padovani B, Mouroux J, Seksik L et al. (1993) Chest wall invasion by bronchogenic carcinoma: evaluation with MR Imaging. Radiology 187:33–38 73. Pannu HK, Wang KP, Borman TL, Bluemke DA (2000) MR imaging of mediastinal lymph nodes: evaluation using a superparamagnetic contrast agent. Magn Reson Imaging 12:899–904 74. Patz EF, Schaffer K, Piwnica-Worms DR et al. (1992) Malignant pleural mesothelioma: value of CT and MR imaging in predicting resectability. Am J Roentgenol 159:961–966 75. Peto J, Decarli A, La Vecchia C, Levi F, Negri E (1999) The European mesothelioma epidemic. Br J Cancer 79:666–672 76. Pisani RJ, Colby TV, Williams DE (1988) Malignant mesothelioma of the pleura. Mayo Clin Proc 63:1234–1244 77. Plathow C, Schoenbinger M, Fink C, Ley S, Puderbach M, Eichinger M, Bock M, Meinzer HP, Kauczor HU (2005) Evaluation of lung volumetry using dynamic three-dimensional magnetic resonance imaging. Invest Radiol 40:173–179 78. Puderbach M, Eichinger M, Gahr J, Ley S, Tuengerthal S, Schmahl A, Fink C, Plathow C, Wiebel M, Muller FM, Kauczor HU (2007) Proton MRI appearance of cystic fibrosis: Comparison to CT. Eur Radiol 17:716–724 79. Rahmouni A, Divine M, Lepage E, Jazaerli N, Belhadj K, Gaulard P, Golli M, Reyes F, Vasile N (2001) Mediastinal lymphoma: quantitative changes in gadolinium enhancement at MR imaging after treatment. Radiology 219:621–628 80. Sakai F, Sone S, Kiyono K et al. (1994) MR of pulmonary hamartoma: pathologic correlation. J Thorac Imaging 9:51–55 81. Samee S, Altes T, Powers P, de Lange EE, Knight-Scott J, Rakes G, Mugler JP 3rd, Ciambotti JM, Alford BA, Brookeman JR, Platts-Mills TA (2003) Imaging the lungs in asthmatic patients by using hyperpolarized helium-3 magnetic resonance: assessment of response to methacholine and exercise challenge. Allergy Clin Immunol 111:1205–1211 82. Schäfer JF, Vollmar J, Schick F, Seemann MD, Kamm P, Erdtmann B, Claussen CD (2005) Detektion von Lungenrundherden mit der Magnetresonanztomographie in Atemanhaltetechnik im Vergleich zur Spiral-Computer. Fortschr Roentgenstr 177:41–49

83. Schaefer JF, Vollmar J, Schick F, Vonthein R, Seemann MD, Aebert H, Dierkesmann R, Friedel G, Claussen CD (2004) Solitary pulmonary nodules: dynamic chontrast-enhanced MR imaging-perfusion differences in malignant and benign lesions. Radiology 232:544–553 84. Schaefer JF, Schneider V, Vollmar J, Wehrmann M, Aebert H, Friedel G, Vonthein R, Schick F, Claussen CD (2006) Solitary pulmonary nodules: association between signal characteristics in dynamic contrast enhanced MRI and tumor angiogenesis. Lung Cancer 53:39–49 85. Schroeder T, Ruehm SG, Debatin JF, Ladd ME, Barkhausen J, Goehde SC (2005) Detection of pulmonary nodules using a 2D HASTE MR Sequence: Comparison with MDCT. Am J Roentgenol 185:979–984 86. Sergiacomi G, Sodani G, Fabiano S, Manenti G, Spinelli A, Konda D, Di Roma M, Schillaci O, Simonetti G (2003) MRI lung perfusion 2D dynamic breath-hold technique in patients with servere emphysema. In Vivo 17:319–324 87. Suga K, Tsukuda T, Awaya H, Takano K, Koike S, Matsunaga N, Sugi K, Esato K (1999) Impaired respiratory mechanics in pulmonary emphysema: evaluation with dynamic breathing MRI. Magn Reson Imaging 10:510–520 88. Swift AJ, Wild JM, Fichele S, Woodhouse N, Fleming S, Waterhouse J, Lawson RA, Paley MN, Van Beek EJ (2005) Emphysematous changes and normal variation in smokers and COPD patients using diffusion 3He MRI. Eur J Radiol 54:352–358 89. Tanaka O, Kiryu T, Hirose Y, Iwata H, Hoshi H (2005) Neurogenic tumors of the mediastinum and chest wall: MR imaging appearance. Thorac Imaging 20:316–320 90. Vonk-Noordegraaf A, Marcus JT, Roseboom B, Postmus PE, Faes TJ, de Vries PM (1997) The effect of right ventricular hypertrophy on left ventricular ejection fraction in pulmonary emphysema. Chest 112:640–645 91. Vonk-Noordegraaf A, van Wolferen SA, Marcus JT, Boonstra A, Postmus PE, Peeters JW, Peacock AJ (2005) Noninvasive assessment and monitoring of the pulmonary circulation. Eur Respir J 25:758–766 92. Will O, Purkayastha S, Chan C, Athanasiou T, Darzi AW, Gedroyc W, Tekkis PP (2006) Diagnostic precision of nanoparticle-enhanced MRI for lymph-node metastases: a meta-analysis. Lancet Oncol 7:52–60 93. Yamashita Y, Yokoyama T, Tomiguchi S et al. (1999) MR imaging of focal lung lesions: elimination of flow and motion artifact by breath-hold ECG-gated and black-blood techniques on T2-weighted turbo SE and STIR sequences. J Magn Reson Imaging 9:691–698 94. Zaporozhan J, Ley S, Gast KK, Schmiedeskamp J, Biedermann A, Eberle B, Kauczor HU (2004) Functional analysis in single-lung transplant recipients: a comparative study of high-resolution CT, 3He-MRI, and pulmonary function tests. Chest 125:173–181

699

700

5  Thorax and Vasculature

5.2  High-Risk Screening Breast MRI E.A. Morris In the United States breast cancer is the most frequently diagnosed cancer in women, accounting for 30% of all malignancies, and it is the second leading cause of cancer death among women, as approximately 30% of all women who develop breast cancer will die of the disease (National Cancer Institute 1998; American Cancer Society 2006; Greenlee et al. 2001). Having the ability to screen for preclinical breast cancer (cancer in the in situ or early invasive state) with mammography has resulted in a 30% decrease in mortality in the United States. Currently, breast cancer screening is performed by mammography and recommended annually for all women over the age of 40 years in the United States (Smith et al. 2005). Many randomized controlled trials have demonstrated the efficacy and importance of mammography in saving lives as well as improving treatment options for patients with this disease. However, mammography is not a perfect test and is limited in certain patient populations and in certain breast densities, particularly the heterogeneously dense and extremely dense parenchymal pattern (Fig. 5.2.1). It has been well recognized that mammography has an overall false-negative rate of up to 20%, which rises to nearly 50% in younger, premenopausal, average-risk patients with dense breasts (Kolb et al. 2002). These limitations have spurred interest in alternate modalities, above and beyond mammography, to detect early breast cancer, including breast magnetic resonance imaging (MRI). 5.2.1  Importance of Early Detection Mammography trials in Europe and North America have shown that breast cancer mortality is 25–30% lower in the mammographic screened group than in the control group after 5 years, suggesting that detecting cancers before they are clinically apparent reduces mortality (Lee 2002; Shapiro et al. 1982; Tabar et al. 1985). Although the study designs of these studies were different, a meta-analysis has confirmed this reduction of breast cancer mortality after 5–7 years (Kerlikowske et al. 1995). This is the evidence that has established mammography as a routine screening method to decrease cancer mortality. Mammography is the only test to date that has been shown to decrease mortality from breast cancer. Nevertheless, what we have learned from the mammography trials can be applied to newer modalities such as MRI. In addition to reducing mortality, mammography may permit less invasive and aggressive therapy by detecting cancer at an earlier stage. A recent update on the Swedish experience demonstrated that screening mammography

Fig. 5.2.1a,b  51-year-old woman with a family history and personal history of breast cancer had undergone routine highrisk screening by MRI for 5 years with no abnormalities. Most recent MRI (a) performed demonstrated a 5-mm mass in the previously conserved breast that was suspicious in morphology. b Lesion did not meet threshold for color coding and graph demonstrated plateau. Biopsy confirmed 5 mm invasive ductal carcinoma and focal DCIS

5.2  High-Risk Screening Breast MRI

was able to detect smaller, node-negative tumors, thus allowing patients to receive less aggressive surgical as well as adjuvant therapy (Tabar et al. 2000). The data from the screening trials are fairly compelling in showing that early breast cancer detection is important for saving lives and improving treatment options. Since the onset of mammographic screening, ductal carcinoma in situ (DCIS) detection has markedly increased from 5% to approximately 30% of all cancers detected. In situ disease in breast cancer means that the malignant cells are contained in the ducts of the breast and have not yet invaded into the breast tissue per se through the ductal basement membrane. Therefore, the malignant cells do not have access to the lymphatics or blood vessels, explaining the lack of nodal disease or distant metastases with DCIS. Prior to the introduction of mammography, DCIS was a rare presentation of breast cancer, as it seldom presents clinically. DCIS now representing a large proportion (30–50%) of the carcinomas diagnosed on mammography (Lee 2002). The fact that DCIS can be detected, usually as calcifications on the mammogram, has enormous positive implications for arresting the development of potentially invasive breast cancer. The ability of mammography to detect calcifications (a bioproduct of necrotic malignant ductal cells) and therefore preinvasive disease is undoubtedly an important factor in the resultant mortality reduction. An important point in the discussion of DCIS is that a significant percentage (up to 50%) of DCIS does not demonstrate calcifications on mammography and therefore may go undetected. Other modalities, such as MRI, may play an important role in the detection of this previously undiagnosed pre-invasive early stage cancer. Any new screening test to detect breast cancer will be measured against this important ability to detect DCIS.

most experts agree that if DCIS is detected, surgical treatment is warranted as it is currently not possible at this writing to predict which lesions are significant. 5.2.3  Why Consider MRI? Although there is fairly clear consensus that mammography saves lives in the general population, there is also frustration and anxiety about the possibility of a false-negative mammogram, particularly in the young, high-risk population where breast density may obscure a mass. Investigation into improvements in mammographic technique has and is currently being explored to improve sensitivity of detection. Digital mammography has demonstrated a slight improvement in cancer detection in young women with dense breasts and possibly should be used in this patient population (Pisano et al. 2005). Interest in MRI for breast cancer screening developed as breast MRI has a very high sensitivity for the detection of invasive carcinoma (Harms et al. 1993; Heywang et al. 1989; Kaiser and Zeitler 1989; Orel et al. 1995) (Figs. 5.2.2, 5.2.3, 5.2.4). As MRI relies on the enhancement of cancers after contrast injection, the issue of breast density is not as problematic as in mammography. In numerous screening studies in different patient populations and with different techniques, MRI can detect most but not all invasive cancers, with a higher sensitivity than mammography (Kriege et al. 2004; Kuhl et al. 2000; Leach et al. 2005; Lehman et al. 2005a; Morris et al. 2003; Podo et al. 2002; Stoutjesdijk et al. 2001; Tilanus-Linthorst et al. 2000; Warner et al. 2004). One issue for screening, however, is the ability to detect DCIS, and the published MRI screening studies are not as uniform in demonstrating the ability to detect DCIS.

5.2.2  Pathology of Breast Cancer: What Are We Looking for?

5.2.4  Defining the High-Risk Population

Compared with other types of cancer, the fact that there is a preinvasive form of breast cancer, which can be detected by imaging techniques, is relatively good news. The theory is that if DCIS can be detected and treated, the patient will not develop metastatic disease, because the in situ carcinoma has not crossed the basement membrane from the ducts to invade the breast tissue and therefore has not had access to lymphatics or blood vessels. This increase in detection of DCIS has not been without controversy, as the relationship of DCIS to invasive carcinoma remains uncertain. While most experts will agree that DCIS can progress to invasion, it has been observed that not all DCIS lesions will end up invading the breast tissue during the lifetime of the patient. In fact, in autopsy series, DCIS, presumably quiescent, can be present in 15% of women (Welch and Black 1997). However,

Although a woman’s average lifetime risk for developing breast cancer in the Unites States is 1 : 8, some women are at high risk. Risk factors for breast cancer include genetic mutations of BRCA1 or BRCA2 (King et al. 2003), family history of breast cancer, previous history of breast cancer, previous biopsy of atypia (ductal or lobular), lobular carcinoma in situ (LCIS), or prior mantle radiation for Hodgkin’s disease (Dershaw 2000). There are three methods to identify high-risk patients: genetic testing, family history assessment, and review of clinical history. It is well known that women in general overestimate their individual risk of breast cancer. Risk stratification for breast cancer is a very complicated process and likely beyond the expertise and knowledge of the average radiologist (National Cancer Institute 2006). There are a plethora of risk assessment

701

702

5  Thorax and Vasculature

Fig. 5.2.2a,b  64-year-old woman with history of left benign biopsy 6 years ago, yielding lobular carcinoma in situ and atypical duct hyperplasia. High-risk screening MRI examination demonstrated bilateral suspicious lesions in the left breast (a) and the right breast (b). Neither of these was iden-

tified on targeted ultrasound. The patient underwent bilateral MRI-guided intervention yielding invasive ductal carcinoma in both breasts (9 mm in left breast and 10 mm in right breast). Nodes were negative bilaterally. Patient underwent bilateral breast conservation

Fig. 5.2.3  42-year-old BRCA1-positive woman. High-risk screening MRI demonstrates irregular 7-mm mass in the lower breast that was biopsied under MRI intervention, yielding invasive ductal carcinoma with negative nodes

Fig. 5.2.4a,b  42-year-old with right breast clumped enhancement (a) in the inframammary fold representing DCIS. The lesion was occult on mammography (b)

5.2  High-Risk Screening Breast MRI

models that have attempted to easily assess risk—many are designed for specific patient groups and all have their strengths and weaknesses. Computer programs based on these models exist that can calculate lifetime risk and can be helpful to assess risk, but these may under- or overestimate the actual risk. A genetic counselor who has been trained in the vagaries of the assessment models is the best person to assess the level of risk. Counseling is an important part of the entire process, enabling test results to be interpreted in the context of the individual patient’s situation. Counseling can also be helpful for intensified surveillance, such as the addition of MRI, as the strengths and limitations of the additional screening tests can be discussed. Furthermore, the counselor can discuss with the patient the results of the genetic testing for either BRCA gene, if performed. 5.2.4.1  Family History Screening issues for high-risk young women who are documented gene carriers or have a family pedigree very suspicious for hereditary breast cancer are limited. For carriers, the average age at first diagnosis is younger than the general population who develops sporadic breast cancer. Recent reports suggest that 50% of carriers succumb to breast cancer before the age of 50 years. Also apparent is the fact that many patients who are successfully treated initially will develop a second breast cancer (up to 60% of patients). There may even be a younger incidence peak (late 40’s) for BRCA1 patients, much earlier than that for sporadic breast cancer. Breast cancers developing in mutation carriers tend to be of higher nuclear grade, exhibit more medullary differentiation and exhibit more receptor negativity than sporadic breast cancers, indicating their more aggressive nature (Brekelmans et al. 2001; Easton et al. 1995; Ford et al. 1998; Tilanus-Linthorst et al. 2002). Management options for high-risk women with genetic mutations have ranged from close mammographic surveillance to prophylactic mastectomy (Hartmann et al. 2001). Bilateral prophylactic mastectomy is a treatment that is usually recommended to the documented patient who is gene positive. The surgery is usually well tolerated by the high-risk patient but has potential drawbacks, including incomplete removal of all breast tissue, loss of nipples and therefore sensation, and psychological issues. An added concern about recommending this treatment option stems from the fact that not all high-risk patients are documented carriers, and it is difficult to recommend this potentially deforming surgery to a patient who is considered high-risk but does not have documentation (Meijers-Heijboer et al. 2001; Rebbeck et al. 1996, 2004). Bilateral salpingo-oophorectomy is also an option and has been shown to decrease the risk of breast cancer by 30%. Obviously this procedure reduces the chance of ovarian cancer as well and it is best used after the age of child-

bearing (Kauff et al. 2002). Chemo-preventive agents such as tamoxifen have become available that may reduce the likelihood of breast cancer in women at high risk. However, these interventions do not completely eliminate the risk of breast cancer death (Narod et al. 2000). Approximately 5% to 10% of all breast cancer is a result of hereditary susceptibility. The prevalence of BRCA mutations is estimated to be 1 : 500 and 1 : 1,000 in the general population, however, in women of Jewish ancestry, the prevalence may be much higher in women who inherit a mutated form of the breast cancer susceptibility gene such as BRCA1 or BRCA2 have a high risk of breast cancer (20% by age 40, 50% by age 50, and 87% by age 70) or ovarian cancer. BRCA1 and BRCA2 account for approximately 50% of the hereditary cancer cases; the rest are probably attributable to yet unidentified genes. The BRCA1 gene is thought to be responsible for breast/ ovarian syndrome. BRCA1 accounts for half of familial breast cancers and 5–8% of all breast cancers. BRCA1associated cancers tend to be highly proliferative, with a high grade, and tend to be receptor negative. The BRCA1 gene is located on chromosome 17 and is a tumor suppressor gene, possibly important in the growth and regulation of epithelial cells in the breast. BRCA2 is located on chromosome 13 and accounts for approximately 35% of familial breast cancer, less than BRCA1. It is thought that BRCA2 contributes fewer cases of early-onset breast cancer than does BRCA1. BRCA2 carriers are at risk of developing other cancers: prostate, bladder, pancreatic, and Hodgkin’s disease. Only 1–2% of women have a family history that is suggestive of an autosomal dominant gene despite the fact that the majority of women in the general population have at least one relative with breast cancer. In most of these the family history does not confer an increased risk (i.e., the cancer was sporadic) or it confers a low increase in risk due to a low-penetrance gene. Features that suggest that the breast cancer may be due to a high penetrance gene include two or more close relatives with breast or ovarian cancer, breast cancer occurring before age 50 in a close relative, a family history of both breast and ovarian cancer, one or more relatives with two cancers (breast and ovarian cancer or two independent breast cancers) and male relatives with breast cancer. Several models can help clinicians assess the likelihood of a BRCA mutation including the Gail, Claus, and Tyrer-Cuzick models. These models attempt to assess breast cancer risk based on family history, sometimes in combination with other factors such as prior breast biopsy or reproductive history; however, these models may yield different estimates of an individual woman’s risk. Two additional decision models have been developed to estimate the likelihood of a BRCA mutation, called the BRCAPRO and BOADICEA, and these models may fare better at predicting an individual’s risk (Berry et al. 2002).

703

704

5  Thorax and Vasculature

5.2.4.2  Genetic Testing Genetic testing for BRCA mutation is generally offered to adult members of families with a known BRCA mutation or to women with at least a 10% chance of carrying a mutation based on clinical history or one of the risk assessment models. If a woman from a family with a documented BRCA mutation tests negative, then it can be safely assumed that she has no elevated risk for breast cancer. However, in a family without a known mutation, failure to find a mutation does not decrease the risk estimate. 5.2.4.3  Clinical History Mammographic density has been shown to increase the risk of breast cancer. Most cancers develop in the breast in the dense areas of the mammogram, and women with dense breasts have a four to six times greater risk for developing breast cancer than those with fattier breasts. Also, compared to the general population, patients with a prior history of breast cancer are at increased risk for developing a new breast cancer (0.5–1% per year). Lobular carcinoma in situ (LCIS) and atypical lobular hyperplasia (ALH) are associated with an increased risk of breast cancer developing, ranging from 10–20% (approximately 0.5–1% per year). This translates to a 6–10 times greater relative risk of developing breast cancer compared with the general population. The invasive cancers are usually invasive lobular cancers, though they can be of ductal histology as well. Atypical ductal hyperplasia (ADH) is also a high-risk lesion leading to a four to five times greater risk of breast cancer. If the patient has a family history and a history of ADH the risk doubles. Treatment for Hodgkin’s disease (HD) between the years 1955 and 1995 confers an increased risk due to mantle radiation. Risk increases significantly 15–30 years after treatment. More recent treatment uses lower doses of radiation and limits the radiation field, so patients treated for HD more recently are not at elevated risk for developing breast cancer. 5.2.5  Overview of High-Risk MRI Screening Studies Several studies performed in the United States and in Europe have shown that MRI benefits high-risk patients by detecting breast cancer when it is mammographically occult. Most of these studies are composed of small patient groups; nevertheless, nearly all the studies have found that there is a 2–4% prevalence of cancer detected on MRI that is occult on mammography and occult to physical examination. A majority of these cancers detected with MRI are small (under 2 cm) and are associated with

negative axillary nodes. Negative nodal status and small cancer size are associated with the reduction of mortality seen with mammographic screening. Therefore, it is interesting to reason that if MRI is able to detect small node-negative cancers, including DCIS, there may be an overall mortality benefit for these patients. Currently, however, these end-points are surrogate end-points. No randomized controlled trial has been mounted to date. At this time, such a trial is likely not feasible and may arguably be unethical given the encouraging results published so far in these high-risk patient groups. Patient populations that are considered high risk in the published MRI screening trials have differed. Some trials have performed screening on patients with high suspicion for or proven BRCA1 and -2 mutations. Others have looked at those with a broader range of risk factors including those with a personal history of breast cancer, those with previous biopsy demonstrating LCIS or ADH, or strong family history. Others have used risk assessment models such as the Gail model to determine high risk and have set entry criteria at varying levels. It appears from the trials that the higher the risk, the more likely an occult cancer will be detected on MRI. Moreover, some trials demonstrated that risk factors can be additive. However, despite these variations in patient selection for MRI screening, there is surprising consistency of the data from multiple sites in different countries with different imaging techniques. It appears that screening MRI is able to detect occult cancer in 2–4% of patients. While MRI can detect additional cancers, the downside of MRI screening is the added risk (in addition to that from mammographic screening) of false-positive findings as well as call-backs for findings that are not entirely normal but that don’t warrant a biopsy. In general, when screening MRI is performed in high-risk populations, 15–25% of patients will require biopsy based on MRI findings and 25–65% of patients who undergo biopsy will have breast cancer. Most of the time these biopsies can be performed with percutaneous needle biopsy under ultrasound or MR guidance, though in some cases in which this is technically unfeasible, surgical biopsy may be required with preoperative MR needle localization. Specificity, an issue with MRI, may be improved in the future with MR spectroscopy (MRS). Initial results in a small trial have demonstrated that MRS interrogation of suspicious lesions undergoing biopsy resulted in one false positive but no false negatives. Importantly, by using MRS, the biopsy rate could have been reduced more that 50%. Though breast MRS is not currently available clinically, its potential to improve specificity is exciting. 5.2.6  Description of High-Risk Screening MRI Studies Several prospective non-randomized studies have demonstrated the utility of breast MRI. These have been per-

5.2  High-Risk Screening Breast MRI

formed in several countries (The Netherlands, United Kingdom, Canada, Germany, and United States) with different patient populations (risk status, age, etc.) and different MRI techniques (low spatial resolution/high temporal resolution versus high spatial resolution/low temporal resolution). All added MRI to mammography and some studies also included ultrasound to detect early breast cancer. All of these studies reported significantly higher detection rates with MRI compared to mammography (or ultrasound). Overall sensitivities range from 71 to 100% for MRI and 16 to 40% for mammography in these high-risk populations. Table 5.2.1 demonstrates sensitivities and specificities of these trials. Table 5.2.2 demonstrates the cancer yield from MRI alone, the number of biopsies performed and the positive predictive value (PPV) of the biopsies performed. The largest study to date was performed by Kriege et al. (2004) in The Netherlands where 1909 unaffected women were screened at six centers. The patients ranged from 25 to 70 years and needed an estimated risk of 15% to be included in the study (19% were BRCA proven carriers). After a median follow up of 3 years, 50 breast cancers were diagnosed, most of them (88%) invasive. Forty-three percent of the invasive cancers were 1 cm or less and a third had positive axillary lymph nodes. In this study mammography detected more DCIS than MRI, a result that may have been related to the low-spatial-resolution techniques used. Another multi-center trial (22 centers) in the United Kingdom, called the MARIBS trial by Leach et al. (2005),

screened 649 patients aged 35–49 who had a lifetime risk of at least 25% (19% proven BRCA carriers). After a median of 3 rounds, 35 cancers were diagnosed, again most of them (83%) invasive. Forty-five percent of the cancers were 1 cm or less and 14% had spread to the axilla. Two interval cancers were detected. In Canada a single-center trial by Warner et al. (2004) studied 236 documented BRCA carriers aged 25–65 years for up to 3 years. Twenty-two cancers (72% invasive) were detected, with 50% of the cancers being 1 cm or less and 13% being node positive. One interval cancer was detected. Kuhl et al. (2000) in Germany screened 529 women with a lifetime risk of 20% at a single center for a mean of 5 years. Forty-three cancers were detected (79% invasive) with positive nodes in 16%. One interval cancer was detected. Morris et al. in the United States reported on a singlecenter trial of 367 high-risk women (5% proven BRCA carriers) with a lifetime risk of greater than 15% for a single round of screening (Morris et al. 2003). Fourteen cancers were detected (43% invasive) and 14% had positive nodes. Lehman et al. in the United States reported on a multicenter trial (13 centers predominantly in the United States) of 390 women with a lifetime risk of more than 25% for one round of screening (Lehman et al. 2005b). Four cancers were found on MRI and only one on mammography. Podo et al. (2005) in Italy studied 105 documented BRCA women at nine sites for a single round of screening.

Table 5.2.1  Sensitivity and specificity breast MRI screening studies The Netherlands

United Kingdom

Canada

Germany

United States

Italy

No. of women

1,909

649

236

529

390

105

Age

25–70

35–49

25–65

> 30

> 25

> 25

No. of cancers

50

35

22

43

  4

  8

MRI

80

77

77

91

100

100

Mammogram

33

40

36

33

  25

  16

United States

–

–

33

33

 –

  16

MRI

90

81

95a

95a

  95a

  99a

Mammogram

95

93

99.8a

93a

  98a

  0

United States

–

–

96

80

 –

  0

No. of centers

 6

22

 1

 1

  13

  9

Sensitivity (%)

Specificity (%)

Based on biopsy rates

a

a

a

705

706

5  Thorax and Vasculature Table 5.2.2  Cancer yield from MRI alone, biopsies, and PPV of biopsies performed The Netherlands

United Kingdom

Canada

Germany

United States

Italy

No. of women

1,909

649

236

529

390

105

No. of cancers only on MRI (%)

22 (1.2)

19 (2.9)

7 (3)

19 (3.6)

3 (0.8)

7 (6.7)

Biopsies recommended based on MRI (%)

59 (2.9)

NS

37 (15.7)

78 (14.7)

23 (6.3)

9 (8.6)

PPV of biopsies performed (%)

57

25

46

50

17

89

NS not stated

Eight cancers were documented–all of them identified on MRI and one of them identified on mammography. 5.2.7  National Guidelines In 2006, the National Institute for Health and Clinical Excellence (NICE), the independent organization responsible for providing national guidance on the promotion of good health, and the National Collaborating Centre for Primary Care in the United Kingdom published an update of the NICE familial breast cancer screening guideline, recommending yearly MRI screening for some women between the ages of 20 and 49 if they have a high risk of breast cancer, including women who have one of the high-risk genes. This recommendation was based on the results from the MARIBS study that was mounted in the United Kingdom as well as the other multi-center trials performed in Europe. In the United States, the American Cancer Society (ACS) guideline for the early detection of breast cancer was last updated in 2003 and stated that women at increased risk may benefit from additional screening strategies beyond those offered to women of average risk. These strategies include earlier initiation of screening, shorter screening intervals, or the addition of other screening modalities (ultrasound or MRI) other than mammography or physical examination. However, the evidence at that time was insufficient to justify widespread recommendations for these additional modalities. Since that time, additional published studies have become available, which have led to new updated guidelines that will be published in 2007. These guidelines will likely recommend MRI screening on an annual basis for selected groups of high-risk patients. However, the decision to perform such additional screening should be carefully evaluated by the physician and patient with a clear understanding of the limitations of the additional tests as well as the benefits.

5.2.8  Current Issues with Using MRI for Screening There are multiple hurdles still to overcome to disseminate this important technology to achieve high level screening for the high-risk population. On a practical level, MRI is expensive, and that simple fact may restrict this test to only those who can afford it. Until insurers uniformly pay for MRI screening this may be a barrier to obtaining this exam for some women. The technique and the criteria for interpretation have been standardized in the United States in the form of the BI-RADS® (Breast Imaging Reporting and Data System) lexicon, supported by the American College of Radiology (ACR) (2003). The ACR has also supported formation of a voluntary accreditation process where facilities can obtain accreditation for performing breast MRI. It is expected that the criteria for accreditation will be released in 2007. As part of the accreditation process it will likely insist that centers perform intervention under MRI so that patients are treated in a timely and appropriate fashion. It is only reasonable for women to expect the same quality assurance programs incorporated into mammography practices to be part of other image-based screening. Quality assurance includes ability to localize or biopsy, correlation of biopsy results with imaging patterns, and adequate surgical referral. MRI may not be feasible in some women, such as those with pacemakers, aneurysm clips, or severe claustrophobia. Moreover, breast MRI may result in a benign biopsy: Among high-risk women who undergo breast MRI screening, 3–15% have a subsequent biopsy yielding benign results. Also, whether detection of cancer on MRI positively affects patient survival (i.e., affects mortality) is not yet known. Whether MRI would benefit the average-risk woman is completely unknown as no data exist. It is speculated that MR screening performed on women with average risk would have a much lower detection rate so that false positives would become overwhelming and the positive

5.2  High-Risk Screening Breast MRI

biopsy rate would be too low to justify its use. At this writing there is no indication to perform screening breast MRI in an average-risk patient. It should be emphasized that the data obtained in the studies is performed at tertiary referral centers that serve a high-risk population. Additionally, radiologists at these centers undoubtedly have more experience at breast MRI interpretation, biopsy and follow up. Rolling this technology out into the general radiology community will be challenging as there is a learning curve involved in the performance and the interpretation of breast MRI. Furthermore, radiologists who offer this service will need to be well-versed in all breast imaging techniques, be able to biopsy and correlate biopsy results and offer appropriate follow-up to their patients, in effect creating a whole new level of service. 5.2.9  Increased Call-Backs and Biopsies When considering a screening test, many important variables need to be satisfied, particularly a high sensitivity. As MRI has probably the highest sensitivity to the detection of invasive cancer of all tests at this time, interest in screening for breast cancer with MRI has evolved. Initial reports that breast MRI could detect early breast cancer in the screening setting were positive, but there was appropriate skepticism. The disadvantage of MRI screening is the possibility of a false positive diagnosis. The morbidity, anxiety, and cost associated with false-positive findings cannot be underestimated. Because the specificity of breast MRI is not extremely high (in fact, it is lower than that of mammography in all published screening studies), the chance of finding a false-positive finding far outweighs the possibility of cancer detection. Nevertheless, when the positive biopsy rate of MRI is examined in the literature, it is approximately 30%, meaning that one in three women recommended for a biopsy for an MRI finding will have cancer. For many, this is an acceptable rate and offers women with high risk a chance to undergo a screening test in lieu of more radical options such as tamoxifen or bilateral prophylactic mastectomies. Additionally, the cost and anxiety associated with short-term follow-up examinations need to be considered for all those incidental foci that are seen on MRI and are not clearly benign (Liberan et al. 2003; Warren et al. 2002). There are no guidelines on how soon to follow these patients, and no clear-cut way to manage these sometimes difficult findings. Recall rates of 12% in a screening population of high-risk women have been described; however, in our experience with a high-risk population, the recall rate was 25%—a significant percentage, and certainly more than the recommended mammographic follow-up rate. Call-back rates and biopsy rates are higher for MRI than for mammography, but the increased sensitivity of MRI also leads to a higher number of cancers detected.

Recall rates decrease on subsequent rounds of screening (in one study to less than 10%) reflecting improved diagnostic confidence when a comparison study is available as well as increased experience of the reader. What is interesting is that there is no evidence to suggest, despite higher recall rates and biopsies, a negative psychological impact on women at high-risk for breast cancer undergoing MRI screening. 5.2.10  Inconsistency of DCIS Detection What is not consistent in these trials is the ability to detect DCIS, and this may have to do with imaging protocols, patient populations, and interpretation criteria. As DCIS is the earliest form of breast carcinoma, a screening test that detects this process is critical. Detection of invasive carcinomas, after cells have had access to the breast lymphatics and blood vessels may be too late and may not impact on survival in a positive manner. Detection of DCIS is one of the advantages that MRI holds over other modalities such as ultrasound. Ultrasound usually does not detect mammographically occult DCIS unless it presents as a mass. MRI is able to detect ductal enhancement or “clumped enhancement,” and the DCIS lesion does not need to manifest as a mass. Detection of DCIS on mammography has been shown to decrease mortality from breast cancer. It is assumed that DCIS detection on MRI will also have an impact on mortality. The histologic grade and type of DCIS does not appear to differ from that detected mammographically, so it is entirely possible that the MRI-detected DCIS is just as important as the mammographically detected DCIS. 5.2.11  MRI Interpretation MRI of the breast uses intravenous contrast for the identification of areas of increased blood flow due to angiogenesis; therefore, areas on MRI that are suspicious for malignancy will show increased contrast uptake. However, increased uptake alone is not the sole criterion for malignancy, as both benign and malignant processes will exhibit increased contrast uptake. Basically, any lesion in the breast where there is increased blood flow (including an area of hormonal changes) will generate increased uptake. Interpretation of MRI therefore can be a challenge. An international group of experts in breast MRI have developed a lexicon for interpretation to assist the radiologist in identifying features that may be considered suspicious in contrast to those that are not considered suspicious. This is now published by the ACR and available. The lexicon incorporates both morphologic and kinetic features of lesions. The morphology describes how the lesion “looks,” and kinetic analysis evaluates how the contrast is taken up and washed out by the lesion. In general, if the

707

708

5  Thorax and Vasculature

lesion morphology is suspicious then the lesion should be biopsied regardless of the kinetic information. The kinetic information is useful in the event that a lesion is not clearly suspicious. In these cases the addition of the kinetic information can prompt biopsy over potential follow up. Therefore, both the morphology and kinetics can be used to facilitate decision-making and provide complementary information about a lesion. Morphologic features that are considered suspicious are spiculated margins and rim enhancement (Liberman 2002). Clumped linear enhancement is suspicious for DCIS particularly if in a segmental distribution. Although clumped enhancement can also be seen with benign histologies; the descriptor is sufficiently suspicious alone to warrant biopsy. Interestingly, kinetic analysis appears not to be reliable in DCIS. Irregular masses with heterogeneous internal enhancement can also be suspicious and warrant biopsy; however the positive biopsy rate on these lesions is somewhat less. Well-circumscribed masses with non-enhancing internal septations are considered benign and characteristic of fibroadenoma. Kinetic features that are suspicious for malignancy are rapid uptake and wash-out. Rapid uptake occurs in the first 2 min, which is why imaging of the breast should be rapid and be performed under 2 min. Washout of the contrast material occurs due to arteriovenous shunting, seen often in malignant lesions. The problem is that there is overlap in many of these findings. Benign lesions will have features of malignant lesions and vice versa. 5.2.12  MRI Technique The ability to detect breast cancer (both invasive and in situ disease) is directly related to high-quality imaging, particularly the signal-to-noise ratio as well as the spatial resolution of the MR image. In order to detect early breast cancer simultaneous imaging of both breasts with high spatial resolution is favored. High-spatial resolution should be performed with a breast coil on a high field magnet with thin slices and high matrix (approximately 1-mm in-plane resolution). These technical parameters are considered to be the minimal requirements to perform an adequate breast MRI study. The ability to biopsy under MRI guidance is crucial and should be a part of any MRI screening service. The ACR is currently developing a voluntary accreditation process for performing breast MRI and this guideline will likely be available in 2007. In the realm of screening, quality assurance will become important. Sites that are considering offering this service should partner with a high-risk clinic and/or clinician with experience in counseling women at high risk. Only sites with significant experience in interpretation and capability of biopsy should accept patients for screening breast MRI. Sites that perform this examina-

tion are strongly encouraged to audit their practices to track call-back rates and biopsies as one would in mammography. 5.2.13  Research Needed Using surrogate endpoints such as size and nodal status is not a perfect solution. In order to demonstrate an improved mortality benefit for those patients screened with MRI, randomized controlled trials are the gold standard. But, randomized controlled trials are expensive, timeconsuming, involve large populations, and from time to time are also imperfect. This brings up an additional point. We live in an era with new technology being introduced at a rapid rate. While it is important to be prudent and cautious, does every new technology (such as MRI or digital tomosynthesis) need to be proven with randomized controlled trials? Or can we use surrogate endpoints? If we know that the detection of cancer before it has had a chance to metastasize will positively impact on the patient’s survival with mammography, would not that also apply to the newer technologies? At the time of this writing there are no planned randomized controlled trial,s and it is even questionable if one at this time would be either feasible or ethical given the recommendations of societies based on the existing evidence. In the absence of randomized trials, recurrence and survival data will be derived from observational studies. The age at which women should start MRI screening is unknown and currently the decision is based on shared decision-making taking into consideration individual situations. In general high-risk screening is started between the ages of 25 and 30, based on data from observational studies that women with BRCA1 mutations have a 3% incidence of breast cancer at age 30 and that the incidence increases thereafter. Most investigators agree that women should be screened annually starting at age. Conversely, the age at which to stop intensive surveillance is also unknown. There are no data for women beyond 69 years. Screening intervals are also unknown. Most high-risk MRI screening studies show few interval cancers; however, BRCA tumors are known to develop more rapidly than sporadic cancers (Tilanus-Linthorst et al. 2005). At this time it is reasonable to screen yearly with MRI with further research into this area much needed. Some investigators recommend staggering the screening examinations with mammography so that the patient can be screened with either mammography or MRI every 6 months. Other experts, however, recommend MRI and mammography at the same time. There is no evidence to support either approach. It would be a serious mistake for physicians and women to assume that a breast MRI would obviate the need for a mammogram. Peerreviewed published studies have shown that some cancers that have been seen on mammography, particularly

5.2  High-Risk Screening Breast MRI

DCIS, have not been seen on MRI. Any attempt to replace mammographic screening with any other modality will sacrifice the advantage to the screened population of the detection of these early, highly treatable breast cancers. Therefore, responsible medical advice to women can only state that if MRI has a role to play in breast cancer screening, it should only be used in conjunction with conventional mammography. It is possible that older women can be screened less than younger women based on the observation that the doubling times of inheritable cancers decrease with age.

4.

5.2.14  Summary

8.

While it is certain that the terrain of the medical field is changing with the advent of more consumer-controlled medical care and more patient choice in determining the use of procedures, it is our responsibility as physicians to honestly present all the data, so that patients can make a well-informed decision. What we know so far about breast MRI screening is that it is very valuable in the high-risk population with a lifetime cancer risk >20%, and should be recommended annually in that population. Breast-imaging radiologists are well aware of the limitations of mammographic screening and have been deeply involved in the search for techniques to further decrease breast cancer mortality. There are still many unanswered questions about breast MRI screening of highrisk patients and as radiologists we must be cautious not to oversell this technology and expect it to be the panacea for breast cancer detection. Continued surveillance with mammography still appears to be important and it is entirely possible that in the future breast MRI may be supplanted by another modality. In this new millennium, we have a highly sensitive test to detect breast cancer and if it is applied to the appropriate patients, it will more than likely continue to improve breast cancer mortality.

9.

5. 6.

7.

10.

11.

12.

13.

14.

15.

16.

References 1.

2.

3.

American Cancer Society (2007) Breast cancer facts & figures 2005–2006. Available at http://www.cancer.org/docroot/stt/stt_0.asp. Cited 1 February 2006 American College of Radiology (2003) ACR BI-RADS. ACR breast imaging reporting and data system, breast imaging atlas, 4th edn. American College of Radiology, Reston, Va. Berry DA, Iversen ES, Jr., Gudbjartsson DF et al. (2002) BRCAPRO validation, sensitivity of genetic testing of BRCA1/BRCA2, and prevalence of other breast cancer susceptibility genes. J Clin Oncol 20:2701–2712

17.

18.

19.

Brekelmans CT, Seynaeve C, Bartels CC et al. (2001) Effectiveness of breast cancer surveillance in BRCA1/2 gene mutation carriers and women with high familial risk. J Clin Oncol 19:924–930 Dershaw DD (2000) Mammographic screening of the high-risk woman. Am J Surg 180:288–289 Easton DF, Ford D, Bishop DT (1995) Breast and ovarian cancer incidence in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Am J Hum Genet 56:265–271 Ford D, Easton DF, Stratton M et al. (1998) Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. The Breast Cancer Linkage Consortium. Am J Hum Genet 62:676–689 Greenlee RT, Hill-Harmon MB, Murray T, Thun M (2001) Cancer statistics 2001. Cancer J Clin 51:15–36 Harms SE, Flamig DP, Hesley KL et al. (1993) MR imaging of the breast with rotating delivery of excitation off resonance: clinical experience with pathologic correlation. Radiology 187:493–501 Hartmann LC, Sellers TA, Schaid DJ et al. (2001) Efficacy of bilateral prophylactic mastectomy in BRCA1 and BRCA2 gene mutation carriers. J Natl Cancer Inst 93:1633–1637 Heywang SH, Wolf A, Pruss E, Hilbertz T, Eiermann W, Permanetter W (1989) MR imaging of the breast with GdDTPA: use and limitations. Radiology 171:95–103 Kaiser WA, Zeitler E (1989) MR imaging of the breast: fast imaging sequences with and without Gd-DTPA. Preliminary observations. Radiology 170:681–686 Kauff ND, Satagopan JM, Robson ME et al. (2002) Riskreducing salpingo-oophorectomy in women with a BRCA1 or BRCA2 mutation. N Engl J Med 346:1609–1615 Kerlikowske K, Grady D, Rubin SM, Sandrock C, Ernster VL (1995) Efficacy of screening mammography. A metaanalysis. JAMA 273:149–154 King MC, Marks JH, Mandell JB (2003) Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA. Science 302:643–646 Kolb TM, Lichy J, Newhouse JH (2002) Comparison of the performance of screening mammography, physical examination, and breast United States and evaluation of factors that influence them: an analysis of 27,825 patient evaluation Radiology 225:165–175 Kriege M, Brekelmans CT, Boetes C et al. (2004) Efficacy of MRI and mammography for breast-cancer screening in women with a familial or genetic predisposition. N Engl J Med 351:427–437 Kuhl CK, Schmutzler RK, Leutner CC et al. (2000) Breast MR imaging screening in 192 women proved or suspected to be carriers of a breast cancer susceptibility gene: preliminary results. Radiology 215:267–279 Leach MO, Boggis CR, Dixon AK et al. (2005) Screening with magnetic resonance imaging and mammography of a UK population at high familial risk of breast cancer: a prospective multicentre cohort study (MARIBS). Lancet 365:1769–1778

709

710

5  Thorax and Vasculature 20. Lee CH (2002) Screening mammography: proven benefit, continued controversy. Radiol Clin North Am 40:395–407 21. Lehman CD, Blume JD, Thickman D et al. (2005a) Added cancer yield of MRI in screening the contralateral breast of women recently diagnosed with breast cancer: results from the International Breast Magnetic Resonance Consortium (IBMC) trial. J Surg Oncol 92:9–15; discussion 15–16 22. Lehman CD, Blume JD, Weatherall P, Thickman D, Hylton N, Warner E, Pisano E, Schnitt SJ, Gatsonis C, Schnall M, DeAngelis GA, Stomper P, Rosen EL, O’Loughlin M, Harms S, Bluemke DA (2005b) International Breast MRI Consortium Working Group. Screening women at high risk for breast cancer with mammography and magnetic resonance imaging. Cancer 103:1898–1905 23. Liberman L, Morris EA, Lee MJ et al. (2002) Breast lesions detected on MR imaging: features and positive predictive value. AJR Am J Roentgenol 179:171–178 24. Liberman L, Morris EA, Benton CL, Abramson AF, Dershaw DD (2003) Probably benign lesions at breast magnetic resonance imaging: preliminary experience in highrisk women. Cancer 98:377–388 25. Meijers-Heijboer H, van Geel B, van Putten WL et al. (2001) Breast cancer after prophylactic bilateral mastectomy in women with a BRCA1 or BRCA2 mutation. N Engl J Med 345:159–164 26. Morris EA, Liberman L, Ballon DJ et al. (2003) MRI of occult breast carcinoma in a high-risk population. AJR Am J Roentgenol 181:619–626 27. Narod SA, Brunet JS, Ghadirian P et al. (2000) Tamoxifen and risk of contralateral breast cancer in BRCA1 and BRCA2 mutation carriers: a case-control study. Hereditary Breast Cancer Clinical Study Group. Lancet 356:1876–1881 28. National Cancer Institute (1998) SEER cancer statistics review, 1973–1995. National Cancer Institute, Bethesda, Md. 29. National Cancer Institute (2006) Genetics of breast and ovarian cancer (PDQ). Available at: http://www.cancer. gov/cancertopics/pdq/genetics/breast-and-ovarian/healthprofessional. Cited 1 February 2006 30. Orel SG, Schnall MD, Powell CM et al. (1995) Staging of suspected breast cancer: effect of MR imaging and MRguided biopsy. Radiology 196:115–122 31. Pisano ED, Gatsonis C, Hendrick E, Yaffe M, Baum JK, Acharyya S, Conant EF, Fajardo LL, Bassett L, D’Orsi C, Jong R, Rebner M (2005) Digital Mammographic Imaging Screening Trial (DMIST) Investigators Group diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 353:1773–1783 32. Podo F, Sardanelli F, Canese R et al. (2002) The Italian multi-centre project on evaluation of MRI and other imaging modalities in early detection of breast cancer in subjects at high genetic risk. J Exp Clin Cancer Res 21:115–124

33. Rebbeck TR, Levin AM, Eisen A et al. (1999) Breast cancer risk after bilateral prophylactic oophorectomy in BRCA1 mutation carriers. J Natl Cancer Inst 91:1475–1479 34. Rebbeck TR, Friebel T, Lynch HT et al. (2004) Bilateral prophylactic mastectomy reduces breast cancer risk in BRCA1 and BRCA2 mutation carriers: the PROSE Study Group. J Clin Oncol 22:1055–1062 35. Shapiro S, Venet W, Strax P, Venet L, Roeser R (1982) Tento fourteen-year effect of screening on breast cancer mortality. J Natl Cancer Inst 69:349–355 36. Smith RA, Cokkinides V, Eyre HJ (2005) American Cancer Society guidelines for the early detection of cancer, 2001. Cancer J Clin 55:31–44; quiz 55–36 37. Stoutjesdijk MJ, Boetes C, Jager GJ et al. (2001) Magnetic resonance imaging and mammography in women with a hereditary risk of breast cancer. J Natl Cancer Inst 93:1095–1102 38. Tabar L, Fagerberg CJ, Gad A, Baldetorp L, Holmberg LH, Grontoft O et al. (1985) Reduction in mortality from breast cancer after mass screening with mammography. Randomised trial from the Breast Cancer Screening Working Group of the Swedish National Board of Health and Welfare. Lancet 1:829–832 39. Tabar L, Vitak B, Chen HH et al. (2000) The Swedish TwoCounty Trial twenty years later. Updated mortality results and new insights from long-term follow-up. Radiol Clin North Am 38:625–651 40. Tilanus-Linthorst MM, Obdeijn IM, Bartels KC, de Koning HJ, Oudkerk M (2000) First experiences in screening women at high risk for breast cancer with MR imaging. Breast Cancer Res Treat 63:53–60 41. Tilanus-Linthorst M, Verhoog L, Obdeijn IM et al. (2002) A BRCA1/2 mutation, high breast density and prominent pushing margins of a tumor independently contribute to a frequent false-negative mammography. Int J Cancer 102:91–95 42. Tilanus-Linthorst MM, Kriege M, Boetes C et al. (2005) Hereditary breast cancer growth rates and its impact on screening policy. Eur J Cancer 41:1610–1617 43. Warner E, Plewes DB, Hill KA et al. (2004) Surveillance of BRCA1 and BRCA2 mutation carriers with magnetic resonance imaging, ultrasound, mammography, and clinical breast examination. JAMA 292:1317–1325 44. Warren RM, Pointon L, Caines R, Hayes C, Thompson D, Leach MO (2002) What is the recall rate of breast MRI when used for screening asymptomatic women at high risk? Magn Reson Imaging 20:557–565 45. Welch HG, Black WC (1997) Using autopsy series to estimate the disease “reservoir” for ductal carcinoma in situ of the breast: how much more breast cancer can we find? Ann Intern Med 127:1023–1028

5.3  Heart

5.3  Heart 5.3.1  Acquisition Techniques and Protocols B.J. Wintersperger 5.3.1.1  Basic Considerations Magnetic resonance imaging (MRI) of the heart and great vessels has improved substantially over the past decade, and it is entering the mainstream of diagnostic imaging. As commercial cardiac MRI systems become widely available, demand will grow exponentially. The demand for cost-effective, noninvasive, and safer technology ensures that cardiac MRI will become a mainstay in cardiac imaging. MRI of the heart, like that of any other body part, is based on standard MR principles but requires the use of specifically tailored imaging sequences. Besides detailed knowledge of cardiac anatomy and function, knowledge about the use of these specific sequences and data acquisition techniques is mandatory. This section focuses on specific features of cardiac MR imaging techniques and their use in imaging protocols. Details of the basics of MR imaging techniques and strategies are to be reviewed in the chapter on basic MR and MR imaging principles (Chap. 2). 5.3.1.2  MR Hardware and Software for Cardiac Imaging For the use of dedicated imaging techniques, some basic requirements need to be met. Most techniques in cardiac MRI require ultrafast data sampling, necessitating gradient systems that allow for short repetition time and echo time. Signal-to-noise ratio (SNR) is strongly dependent not only on the external field strength (B0), but also on the overall data acquisition time. With the short data sampling times in cardiac MRI a B0 field of 1.5 T is preferred over a lower field strength, although basic cardiac MR techniques may also be applied at 1 T. The increasing use of even 3 Tesla systems for cardiac MRI shows the demand for high SNR to allow even faster imaging techniques (Wen 1997; Sommer 2005; Gutberlet 2006; Wintersperger 2006a). In addition to the external field strength and the gradient system, the signal receiving part of the hardware is extremely important. Modern chest and dedicated cardiac coils consist of coil arrays with 4–32 elements for optimization of the SNR. However, the optimal setting, even with modern multi-element arrays, can only be achieved with an adequate number of receiver channels, which should match the number of single coil elements. In addition to the optimization of the SNR these coil technologies are well suited for modern acceleration techniques such as parallel imaging (Reeder 2005; Wintersperger 2006b). As the vast majority of cardiac imag-

ing techniques require ECG triggering or gating, an adequate source (internal or external) for a sufficient trigger signal is mandatory. In addition to the MR hardware, the sequence technologies and post-processing capabilities available are of fundamental importance to the success of cardiac MR imaging. The necessary imaging techniques may vary and are strongly dependent on the reason for MR imaging referral, but most vendors offer comprehensive cardiac MR imaging packages that include a variety of sequence techniques as well as post-processing options. 5.3.1.3  Techniques in Cardiac MRI 5.3.1.3.1  Physiologic Control The position of the heart within the center of the chest in proximity to the lungs and diaphragm exposes the heart to respiratory motion in addition to its own cyclic pulsation. The latter especially makes MR imaging of the heart more complicated than MR imaging of other organs. Prospective ECG Triggering/Retrospective ECG Gating The use of data sampling techniques without respect to the heart’s physiologic motion and the overlying respiratory motion typically results in image blurring and a nondiagnostic image quality. The challenge in cardiac MR is to cope with the pulsatile motion of the heart and the superimposed respiratory motion. Although rapid imaging techniques enable data to be acquired within just a fraction of the cardiac cycle, synchronization of the cardiac motion is still mandatory. Synchronization can only be omitted if the image acquisition time is short enough compared to the cardiac cycle. Thus image quality is unaffected by cardiac and respiratory motion, as is the case in real-time imaging. The overall image quality and success of a cardiac MR exam is strongly related to the quality of the trigger. It has to be emphasized that every effort needs to be made to get a good and sufficient trigger signal. The ECG signal may be traced by the scanner’s built-in physiologic monitoring unit or by an external MR-compatible ECG monitor that can be connected to the MR scanner for trigger signal transfer. The ECG allows not only triggering of the data acquisition, but also tailoring of the data acquisition within the cardiac cycle. Hence, data may be sampled within systole or diastole based on the user’s preferences. The ECG best reflects the heart’s cyclic motion and thus is most commonly used for data sampling in cardiac MR. However, the ECG trace in the bore is affected by the magneto-hydro-dynamic (MHD) effect, which leads to an increase in T wave amplitude that may be misinterpreted as an R peak. The MHD effect is based on flowing blood and is further amplified with the transition to

711

712

5  Thorax and Vasculature

higher field strengths (e.g., 3 T) (Stuber 2002). Modern vector ECG systems (VCG) are recommended to deal with MHD effect problems (Fischer 1999). Additional filter algorithms also minimize the possible influences of ultrafast switching gradients. In prospective ECG triggering, the detection of an R peak triggers the sequence execution for data sampling. After the sampling certain data is completed, the search for the next trigger is enabled. However, to sample data in consecutive heart beats, data sampling needs to spare parts of end-diastole to enable trigger recognition. With retrospective ECG gating, data sampling is performed continuously with simultaneous ECG tracing. Acquired data can then be sorted within k-space according to time stamps. This technique specially allows coverage of the entire cardiac cycle in cardiac function analysis or flow measurements. Alternatively, in case of a non-diagnostic ECG trace even after repositioning of the electrodes and leads, pulse triggering can be applied. However, this trigger method leads to limitations because of the trigger delay and the wider signal peak. Recent developments in imaging techniques may allow the use of alternative methods such as real-time imaging or self-gating, though most of these techniques lead to inferior spatial resolution compared with standard ECG triggered or gated techniques. Elimination of Respiratory Artifacts Elimination of respiratory artifacts can be achieved using the following different approaches: • Data averaging • Breath-hold imaging • Respiratory gating • Real-time imaging While data averaging was used before the advent of ultrafast data sampling techniques, it may only be used now in an uncooperative patient population. Particularly in pediatric cardiac MRI the technique may still be useful when scanning is performed with sedation instead of anesthesia. The data sampling is performed during shallow breathing, and the image quality improves with an increasing number of acquisitions. As multiple data sets are to be acquired, the acquisition time is typically in the range of several minutes. With the advent of ultrafast imaging techniques, most data acquisitions are currently performed in a breathhold setting thus eliminating respiratory artifacts. Patients need to be adequately instructed before the exam, as a reproducible and constant breath-hold level ensures sufficient volume coverage with 2D techniques. Although the diaphragm position is more reproducible in expiration, patients’ comfort and compliance is higher using inspiratory breath-hold settings. Respiratory gating is typically applied in data acquisitions of the heart that cannot be fitted into breath-hold

settings or can alternatively be used in a non-compliant patient population. Especially 3D coronary MR angiography is often implemented with respiratory gating, as high-spatial-resolution data sets are required (Post 1996; Stuber 1999; Spuentrup 2002; Spuentrup 2003; Sommer 2005). The data acquisition is thus not only synchronized to the patient’s ECG but also to the patient’s diaphragm position. Today, the most common setting of respiratory gating is based on a diaphragm position navigating pulse that traces the diaphragm during the respiratory cycle and gives feedback to sequence data sampling for acquisition or ordering of image data. Detailed applications are vendor dependent and will therefore not be discussed in further detail. However, with the dependency of data acquisition on the proper time point in the cardiac cycle and the position of the diaphragm, the data sampling effectiveness is rather low, resulting in prolonged data acquisitions (~5–10 min). Real-time imaging with ultrafast sampling techniques allows for acquisition of an entire image data set within a short window of the respiratory cycle (and also the cardiac cycle) thus eliminating motion artifacts. Based on the rather short window of possible data sampling, spatial resolution of these techniques is usually limited. 5.3.1.3.2  Data Sampling and Sequence Techniques in Cardiac MRI A variety of techniques are available for data sampling in cardiac MRI that are also scanner and vendor dependent. However, the most common techniques are comparable between different manufacturers and may only vary depending on the hardware and software system provided. This section gives basic information required for the dedicated cardiac application of various imaging techniques and their clinical use. Basic sequence techniques that are also used in cardiac MR are covered in detail in the technical chapters of this book (Chap. 2). k-Space-Filling Techniques Segmented k-Space Filling Segmented sequence techniques are the underlying basic principles of modern breath-hold imaging techniques. While older techniques only acquired a single k-space line at a time so that total acquisition time in heart beats matched the number of necessary phase encoding steps, segmented data acquisitions sample multiple k-space lines every heart beat to fill a segment of k-space. This technique has been combined with fast spin-echo (FSE) techniques as well as with gradient-echo techniques. With FSE techniques the turbo factor (number of echoes per 90° pulse) determines the number of lines acquired per heart beat. The length of the overall data acquisition then equals the total number of phase-encoding steps (Ny) divided by the turbo factor or lines acquired per segment. The acceleration though cannot be chosen arbitrarily, as

5.3  Heart

the acquisition of multiple k-space lines reduces the temporal resolution of the technique and image blurring may arise based on motion during that time period. Turbo factors (TF) of 9–20 are commonly used; the TF is typically lower with higher heart rates as cardiac motion is exaggerated and motion artifacts become more likely. The same principle can be applied for gradiant echo (GE) imaging techniques although RF excitations are typically performed for every single k-space line (except for EPI hybrid sequence techniques). Segmented GE techniques are used in static and dynamic cardiac imaging. In dynamic cardiac cine imaging the number of k-space lines per segment is directly related to the temporal resolution of the acquisition. While “binning” of multiple TR’s together to a single time frame reduces the total acquisition time to a breath hold (Atkinson 1991), the temporal resolution needed is in the range of 40 to 50 ms (Setser 2000). With a higher number of lines per k-space segment the acquisition shortens, but temporal resolution worsens. The number of heart beats required for data acquisition is the total number of phase encoding steps (Ny) divided by the number of lines/segment and the temporal resolution matches the number of lines/segment times the sequence TR. Single-Shot k-Space Filling The goal in the single-shot technique is to acquire all needed k-space lines for an image during one cardiac cycle. In FSE this is realized by a single excitation followed by multiple 180° refocusing pulses to fill all k-space lines required to calculate an image. When using spoiled gradient-echo (SGE) or steady-state free precession (SSFP) sequences, each k-space line may require an individual excitation unless EPI techniques are in use. Additional methods such as partial-Fourier can be used to reduce the number of required k-space lines, resulting in a reduction in the total acquisition time. Single-shot techniques are used to acquire rapidly within a single cardiac cycle. Due to the short acquisition time it is possible to acquire the data in free breathing and to minimize image blurring due to arrhythmias. While in most cases acquisition is performed in diastole to minimize blurring, for dynamic measurement in first-pass perfusion the slices will be acquired throughout the entire cardiac cycle. The same slices will be repeated in the following heart beats so that the bolus arrival and passing of the contrast media can be monitored. Spin-Echo/Fast Spin-Echo Techniques Detailed soft tissue differentiation in cardiac MR is provided by T1-weighted and T2-weighted spin-echo (SE) or FSE techniques. FSE techniques have replaced the use of conventional SE in cardiac MR to a large extent. With FSE multiple k-space lines are acquired using multiple 180° refocusing pulses after a single 90° excitation pulse. The multiple echoes are also referred to as echo train, and the length of the echo train determines the acceleration

compared to conventional SE and is known as the turbo factor (TF). In combination with k-space segmentation, these sequence techniques allow acquisition of a single slice per breath hold, necessitating multiple consecutive breath-hold periods for volume coverage. However, soft-tissue appearance and differentiation is dependent on repetition time (TR), echo time (TE), or inversion time (TI). With ECG triggering variations may apply, as TR depends on the patient’s heart rate, and minimal TR matches the length of a single cardiac cycle (R-R peak interval). While for T1-weighted FSE, TR is often longer than needed (~1,000 ms at 60 bpm), in T2-weighted FSE a longer TR can be achieved by using every other or even every third heart beat as a trigger. Thus data acquisition for T2-weighted FSE may exceed the length of a comfortable breath-hold unless a higher turbo factor is applied. Alternatively single-shot FSE may be employed for T2-weighted imaging. This technique is based a single radiofrequency excitation (90° pulse) that is followed by multiple refocused echoes to fill k-space in a single cardiac cycle. To reduce the number of phase encodings necessary, typically partial- or half-Fourier techniques (HASTE) are used. Compared with single-slice FSE these techniques allow much faster coverage of the whole cardiac volume. However, this technique is used at the expense of SNR, temporal and spatial resolution. In all SE-based sequences the flowing blood appears hypointense as it is typically only exposed to a single RF pulse, either the 90° pulse or the 180° pulse. A complete signal void, though, is only present if all spins excited by the 90° pulse have left the image slice and have not experienced the 180° pulse. To avoid disturbances based on the complicated flow patterns in the cardiac cavities and to intensify the hypointense blood appearance, a black-blood preparation is usually employed with the use of spin-echo based techniques. This preparation consists of a double inversion preparation pulse scheme with a non-selective 180° inversion pulse followed by a slice-selective 180° pulse within the imaging slice (Fig. 5.3.1). Data acquisition then follows after a certain time delay in diastole (Fig. 5.3.1). After gadolinium administration this black-blood preparation may not be as effective as in plain imaging. Clinical Applications Cardiac MR is known to be superior to most other cardiac imaging technologies in assessment of cardiac anatomy and morphology. It can be applied in assessment of pre- or postsurgical congenital heart disease (CHD), when a cardiac mass is suspected as well as in inflammatory diseases (de Roos 2000; Rebergen 2000; Constantine 2004). Any morphologic imaging though is most often combined with a functional approach for further detailed insight into pathophysiology and functional aspects of anomalies. Morphologic imaging for evaluation of CHD is typically performed along standard orthogonal axes (transversal, coronal, sagittal). This allows a better orientation

713

714

5  Thorax and Vasculature

performed also (Figs. 5.3.2, 5.3.3). While in CHD imaging the use of contrast is typically not necessary (except when contrast MRA is used), contrast administration is mandatory in imaging of a cardiac mass or an inflammatory disease (Table 5.3.1) (see also Sects. 5.3.4, 5.3.6, 5.3.8).

Fig. 5.3.1  Black-blood preparation scheme in FSE. For a preselected slice (1), a non-selective 180° pulse is applied (2), immediately followed by a selective 180° pulse (3) to restore the magnetization in the selected slice. During systolic contraction (4) and diastolic filling (5) the blood within the slice is replaced and data acquisition then follows in mid- to end-diastole (6)

in complex cases as additional angulations may be confusing (Table 5.3.1). Also, sufficient coverage from the arch to the diaphragm is recommended. In assessment of cardiac mass or inflammatory disease (e.g., pericarditis) imaging along the individual cardiac axes is often

Fig. 5.3.2a–d  Planning of the individual cardiac axes. The plane perpendicular to an axial scout view (a) between the center of the mitral valve and the apex defines the vertical long axis (VLA). b Using the VLA plane as a reference a perpendicular plane again between the center of the mitral valve and the apex is called the horizontal long axis (HLA). c The planes parallel to the AV valves and perpendicular to the long-axis views are known as short-axis views. d This axis is used for volumetric evaluation of the ventricles

Gradient and Fast Gradient-Echo Techniques In gradient-echo imaging the 180° refocusing pulse is omitted, and the echo is generated by bipolar gradients. Thus, gradient based echo generation may speed up data acquisition due to short TR. However, spin dephasing based on local field inhomogeneities (T2*) may affect imaging. This may be beneficial in localization of small calcifications of metal residues; however in case of metal implants (e.g., artificial cardiac valves, sternal wires) image quality may be impaired or image parts obscured by susceptibility artifacts. Although there is a huge variety of a gradient-echo sequences, in cardiac MR, SGE and SSFP techniques are most commonly applied. These techniques are the backbones of state-of-the-art cardiac MRI especially with respect to the evaluation of cardiac dynamics (e.g., function, perfusion, flow). Even in assessment of cardiac morphology, GE techniques might add valuable information as the techniques show

Fig. 5.3.3a–c  The three-chamber view (or inflow–outflow view) can easily be planned based on a stack of short-axis views with a three-point planning tool. Selection of points in the center of the mitral valve orifice and the center of the aortic valve (a) together with the third point in the apex (b) defines the three-chamber view that allows simultaneous visualization of the LV inflow (mitral valve) and LV outflow (outflow tract + aortic valve)

5.3  Heart Table 5.3.1  Cardiac anatomy and morphology MR imaging Technique

• Segmented FSE/TSE with black blood preparation

Common application

• Congenital heart disease (CHD), cardiac mass evaluation, inflammatory disease

Typical imaging parameters

• ETL 9–16 (T1-weighted), ETL 16–25 (T2-weighted) • Matrix 256, slice 4–8 mm (thinner slices in children and infants)

Imaging planes

• Orthogonal planes for CHD • adapted planes for other diseases (including cardiac axes)

Critical factor

• Suppression of blood flow signal, imaging time

Additional features

• Fat suppression techniques, STIR imaging • Single-shot T2-weighted techniques available (HASTE)

Limitations

• TR depending on heart rate (minimum: single R–R interval)

Remarks

• Provides best soft-tissue differentiation • Contrast application mandatory for mass and inflammation • Combination with functional imaging techniques and MR angiography

less blood-flow-dependent artifacts, although soft tissue contrast assessment is worse than with FSE techniques. With the use of magnetization preparation pulses GE techniques can be applied in various ways to highlight pathology. Clinical Applications The clinical use of various gradient-echo techniques will be highlighted within the disease-specific parts of this cardiac MR chapter, but basic considerations regarding the most common uses of gradient-echo techniques will be focused on below. Cine Imaging MR imaging provides accurate and reproducible results in assessment of cardiac volumes and allows reliable evaluation of regional wall motion. SGE or SSFP techniques may be applied though the latter provide a higher contrast-to-noise ratio (CNR). Temporal resolution is considered the most crucial factor in cardiac cine imaging in use for functional analysis. At least 40 fps (50 ms per frame) are necessary to come close to Nyquist criteria, although at faster heart rates (>60–70 bpm) further improvement of temporal resolution is recommended (Weissler 1968; Setser 2000). To be able to minimize partial volume effects and to accurately assess also regional wall motion according to accepted ventricular models (American Heart Association [AHA]), imaging needs to performed according to the individual cardiac axes (Figs. 5.3.2, 5.3.3) (Cerqueira 2002). Beside k-space segmentation of the sequences, phase-sharing and parallel-imaging algorithms are commonly applied to maintain high temporal and spatial resolution without exceeding the length of a comfortable breath-hold. The combination of these techniques even allows real-time cine imaging

without the necessity of breath-holding and may allow further insights into the physiology and pathophysiology of various cardiac diseases. However, spatial resolution may be sacrificed. For clinical purposes a standard approach to functional assessment is recommended that may differ between adults, children, and infants (Table 5.3.2). While volumetric analysis is typically performed on a stack of short-axis slices, regional wall motion assessment also requires imaging in long-axis orientations. Besides quantification of volumes or evaluation of wall motion, cine MRI is also employed for valve assessment and assessment of morphology in imaging of shunting defects (e.g., ASD, VSD). Imaging and quantification of flow itself may be provided by the use of dedicated gradient-echo techniques, so-called phase contrast (PC) cine. Therefore dedicated techniques with additional flow sensitive gradients are applied that are further elucidated in the technical parts of this book (Chap. 2) as well as in the discussion of the clinical applications of cardiac MR imaging (Sect. 5.3.5) Myocardial Perfusion Imaging Myocardial perfusion imaging is still rapidly evolving and may replace other modalities for perfusion imaging in the future. First pass perfusion is by far the most commonly applied technique, sampling data with an ultrafast GE technique following a rapidly injected contrast agent bolus. The most commonly employed techniques are inversion recovery (IR) or saturation recovery (SR) prepared ultrafast SGE techniques (e.g., turboFLASH, turbo field echo) or echo planar (EPI) hybrid techniques (Wilke 1997; Nagel 2003). Data acquisition does not use k-space segmentation but individual images are acquired in a snapshot technique. Sufficient coverage of the contrast agent passage through the heart and myocardium

715

716

5  Thorax and Vasculature Table 5.3.2  Cardiac cine MR imaging Technique

• Segmented SSPF or SGE sequences

Common application

• • • •

Typical imaging parameters

• Matrix 192–256, slice 4–8 mm (8 mm for functional analysis in adults)

Imaging planes

• Along cardiac axes for functional analysis • adapted planes for valve, mass, and CHD imaging

Critical factor

• Temporal resolution should meet 50 ms or better • Temporal resolution affected by segmentation factor

Additional features

• Combination with tagging or flow techniques

Limitations

• Breath-hold imaging may influence cardiac dynamics • Possible better insights into pathology with real-time cine (at lower resolution) • SSFP may lead to SAR limitations at 3 T

Remarks

• • • •

Global and regional cardiac function Cardiac valve function Cardiac mass mobility CHD imaging (shunt assessment)

SSFP provides higher CNR than SGE Full evaluation: stack of short-axis slices; 2-, 3-, and 4-chamber view Coverage from the AV valves to apex necessary (short axis) SGE shows higher flow susceptibility than SSFP (beneficial for valve imaging)

allowing for an assessment of perfusion deficits necessitates data acquisition over 40 to 60 beats with repeated acquisition of each slice within every single heart beat. Thus, the number of slices is usually limited to three to five, depending on the length of the R-R interval (Table 5.3.3). This allows a moderate to good coverage of the left ventricle according to the AHA segmental model (Cerqueira 2002). Slices should be positioned primarily in short-axis orientations (at least 3 slices) with the chance of additional slices orientated along the long axis. Spatial resolution should be optimized to a maximal acquisition window of 140–160 ms to minimize image blurring. At 1.5 T also magnetization prepared SSFP sequences may be applied as these techniques show a superior SNR (Wang 2005). First-pass perfusion images may be affected by dark band artifacts that typically appear within the subendocardial region of the myocardium (Arai 2000). These artifacts have been reported to be based on T2* susceptibility effects, Gibbs ringing due to low spatial resolution and on cardiac motion. However, these artifacts are only transient and commonly disappear before peak myocardial enhancement thus allowing differentiation from real perfusion deficits (Arai 2000; Barkhausen 2004). With the use of parallel imaging algorithms, these artifacts may be less pronounced as the acquisition window can be shortened and spatial resolution improved (Di Bella 2005). As myocardial perfusion imaging shows limited SNR and CNR compared to other cardiac MR techniques, the acceleration factor should be limited (R = 2) in order to maintain sufficient signal for data analysis.

In clinical use of perfusion imaging, the assessment of reversible perfusion deficits is the most common task necessitating rest and stress imaging (Barkhausen 2004). In a clinical protocol setting, stress imaging is preferably performed prior to stress imaging and an interval (~10–15 min) prior to the rest perfusion allows for contrast agent clearance (Table 5.3.3). Pharmacologic stress testing is typically performed using either adenosine or dipyridamole. The technique though is not limited to the assessment of myocardial perfusion but may also be used in imaging of contrast dynamics for other purposes (e.g., a cardiac mass). Delayed Enhancement Imaging Delayed enhancement (DE) imaging has pushed clinical use of cardiac MR, especially with respect to the assessment of myocardial viability (see also sects. 5.3.3, 5.3.4, 5.3.7) (Barkhausen 2004; Constantine 2004; Hunold 2005). DE imaging generally highlights differences of contrast agent concentrations within the myocardium. To maximize CNR, heavily T1-weighted inversion recoveryprepared gradient-echo techniques have been proposed (Simonetti 2001). Magnetization prepared SGE as well as SSFP techniques are helpful. However, optimal contrast for delineation of areas with higher concentrations of a contrast agent (e.g., myocardial infarction, myocardial fibrosis) necessitates optimization of the TI. As TI changes are higher early after contrast injection (1–2 mmol/kg of body weight), imaging is recommended approximately 5–15 min after injection (when focusing on infarct de-

5.3  Heart Table 5.3.3  Cardiac MR perfusion imaging Technique

• Multi-slice SR fast SGE, SSFP, or EPI hybrid techniques

Common application

• Myocardial perfusion assessment (rest and stress imaging) • Cardiac mass evaluation • Assessment of shunt disease

Typical imaging parameters

• Matrix 128–192, slice 8–10 mm; 3–5 slices every heart beat • TI of 220–300 ms (depending on imaging time point and contrast volume)

Imaging planes

• Cardiac short axis (according to AHA segments) • Additional long-axis slices (in case of at least 3 short-axis slices)

Critical factor

• SNR • Acquisition window

Limitations

• Spatial resolution and myocardial coverage • Absolute quantification of perfusion

Remarks

• Typical artifacts need to be known (dark banding artifacts) • SSFP techniques offer higher SNR than SGE techniques

Table 5.3.4  Delayed enhancement MR imaging Technique

• IR SGE or SSFP techniques

Common application

• • • •

Typical imaging parameters

• Matrix 192–256, slice 6 mm – 8 mm

Imaging planes

• Cardiac short axis • Long-axis slices (2-, 3-, 4-chamber view)

Critical factor

• SNR • Myocardial coverage • Data acquisition window (maximum 140–160 ms)

Additional features

• Phase sensitive information possible (PSIR)

Limitations

• No adequate plain imaging for contrast uptake quantification

Remarks

• Different features of DE distribution depending on underlying disease • Highly sensitive technique with rather mediocre specificity (without specific DE)

Myocardial viability Cardiomyopathies Inflammatory disease Mass assessment

lineation) (Table 5.3.4). In pathologies other than myocardial infarction the timing recommendation may be different. Optimization of the inversion time may be accomplished either using a TI scout technique or manual adjustment with several test images at different TI values (Table 5.3.4). New imaging techniques also enable stable infarct delineation even at suboptimal TI with additional phase sensitive image information (Huber 2005). While for high-resolution imaging, segmented acquisition schemes are recommended, single-shot multi-slice approaches may also be employed at the cost of spatial resolution (Huber 2006). Recent studies also push the

use of this imaging technique in assessment of cardiac masses, pseudo-masses, and thrombi mostly with a fixed TI of ~300–350 ms (Sect. 5.3.8, Figs. 5.3.53, 5.3.54, 5.3.55, 5.3.56, 5.3.57). References 1.

Arai AE (2000) Magnetic resonance first-pass myocardial perfusion imaging. Top Magn Reson Imaging 11:383–398

717

718

5  Thorax and Vasculature 2.

3.

4. 5.

6. 7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18.

Atkinson DJ, Edelman RR (1991) Cineangiography of the heart in a single breath hold with a segmented turboFLASH sequence. Radiology 178:357–360 Barkhausen J et al. (2004) Imaging of myocardial perfusion with magnetic resonance. J Magn Reson Imaging 19:750–757 Barkhausen, J et al. (2004) MRI in coronary artery disease. Eur Radiol 14:2155–2162 Cerqueira MD et al. (2002) Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 105:539–542 Constantine G et al. (2004) Role of MRI in clinical cardiology. Lancet 363:2162–2171 Di Bella EV et al. (2005) On the dark rim artifact in dynamic contrast-enhanced MRI myocardial perfusion studies. Magn Reson Med 54:1295–1299 Fischer SE et al. (1999) Novel real-time R-wave detection algorithm based on the vectorcardiogram for accurate gated magnetic resonance acquisitions. Magn Reson Med 42:361–370 Gutberlet M et al. (2006) Comprehensive cardiac magnetic resonance imaging at 3.0 Tesla: feasibility and implications for clinical applications. Invest Radiol 41:154–167 Huber AM et al. (2006) Phase-sensitive inversion recovery (PSIR) single-shot TrueFISP for assessment of myocardial infarction at 3 Tesla. Invest Radiol 41:148–153 Huber AM et al. (2005) Value of phase-sensitive inversion recovery (PSIR) for detection of myocardial infarction. Radiology 237:854–860 Hunold P et al. (2005) Myocardial late enhancement in contrast-enhanced cardiac MRI: distinction between infarction scar and non-infarction-related disease. AJR Am J Roentgenol 184:1420–1426 Nagel E et al. (2003) Magnetic resonance perfusion measurements for the noninvasive detection of coronary artery disease. Circulation 108:432–437 Post JC et al. (1996) Three-dimensional respiratory-gated MR angiography of coronary arteries: comparison with conventional coronary angiography. AJR Am J Roentgenol 166:1399–1404 Rebergen SA, de Roos A (2000) Congenital heart disease. Evaluation of anatomy and function by MRI. Herz 25:365–383 Reeder SB et al. (2005) Practical approaches to the evaluation of signal-to-noise ratio performance with parallel imaging: application with cardiac imaging and a 32-channel cardiac coil. Magn Reson Med 54:748–754 Roos A de, Roest AA (2000) Evaluation of congenital heart disease by magnetic resonance imaging. Eur Radiol 10:2–6 Setser RM et al. (2000) Quantification of left ventricular function with magnetic resonance images acquired in real time. J Magn Reson Imaging 12:430–438

19. Simonetti OP et al. (2001) An improved MR imaging technique for the visualization of myocardial infarction. Radiology 218:215–223 20. Sommer T et al. (2005) Coronary MR angiography at 3.0 T versus that at 1.5 T: initial results in patients suspected of having coronary artery disease. Radiology 234:718–725 21. Spuentrup E et al. (2002) Navigator-gated free-breathing three-dimensional balanced fast field echo (TrueFISP) coronary magnetic resonance angiography. Invest Radiol 37:637–642 22. Spuentrup E et al. (2003) Navigator-gated coronary magnetic resonance angiography using steady-state-free-precession: comparison to standard T2-prepared gradientecho and spiral imaging. Invest Radiol 38:263–268 23. Stuber M et al. (1999) Breathhold three-dimensional coronary magnetic resonance angiography using real-time navi­ gator technology. J Cardiovasc Magn Reson 1:233–238 24. Stuber M et al. (2002) Preliminary report on in vivo coronary MRA at 3 Tesla in humans. Magn Reson Med 48:425–429 25. Wang Y et al. (2005) Myocardial first pass perfusion: steady-state free precession versus spoiled gradient-echo and segmented echo planar imaging. Magn Reson Med 54:1123–1129 26. Weissler AM et al. (1968) Systolic time intervals in heart failure in man. Circulation 37:149–159 27. Wen H et al. (1997) The intrinsic signal-to-noise ratio in human cardiac imaging at 1.5, 3, and 4 T. J Magn Reson 125:65–71 28. Wilke N et al. (1997) Myocardial perfusion reserve: assessment with multisection, quantitative, first-pass MR imaging. Radiology 204:373–384 29. Wintersperger BJ et al. (2006a) Cardiac steady-state free precession CINE magnetic resonance imaging at 3.0 Tesla: impact of parallel imaging acceleration on volumetric accuracy and signal parameters. Invest Radiol 41:141–147 30. Wintersperger BJ et al. (2006b) Cardiac CINE MR imaging with a 32-channel cardiac coil and parallel imaging: impact of acceleration factors on image quality and volumetric accuracy. J Magn Reson Imaging 23:222–227

5.3.2  Congenital Heart Disease: Cardiac Anomalies and Malformations T.R.C. Johnson Congenital heart disease is one of the most frequent indications for cardiac MRI in specialized centers. Current developments with steady-state and multi-slice sequences with parallel imaging techniques have contributed to an improved “one-stop-shop” coverage to assess anatomy and morphology of the heart, cardiac function, and vascular anatomy as well as flow measurements in one MR examination. Regarding this unique comprehensive evaluation, MRI has an important role in the primary investigation, follow-up and surgical planning of congenital

5.3  Heart

heart disease. Echocardiography serves as a good firstline modality that is easy and fast to perform, and MRI is often additionally applied for the evaluation of complex heart disorders, surgical planning, and for follow-up when the sonic window diminishes with age. 5.3.2.1  Situs, Atrial, and Ventricular Morphology The situs and the morphology of atria and ventricles are important to determine the ontogenetic origin of the structures in order to correctly classify aberrations from the normal anatomy (Fig. 5.3.4) (Yoo et al. 1999; Fogel et al. 1996; Choe et al. 1997). The cardiac situs is determined by the location of the atria, independent of the position of the ventricles. The normal situs with the morphologic right atrium on the right and the left atrium on the left is termed situs solitus. If the morphologic right atrium is situated on the left and vice versa, this condition is called cardiac situs inversus. If the cardiac situs cannot be determined because of the atrial morphology, then this represents a situs ambiguous. This term is also used if both atria have a right or left morphology as in isomerism syndromes. The right atrium is characterized by the crista terminalis, some trabeculae, and the right atrial appendage with its broad base and triangular shape. The left atrium on the other hand has very smooth walls and a long and narrow appendage. The two ventricles are also characterized by their different shape and trabeculae. The endocardial surface of the left ventricle appears smoother and does not have trabeculae on the septal wall. The left ventricle has an oval configuration and hence appears round on short-axis slices. In the wall of the right ventricle there are many trabeculae, and the papillary muscles of the tricuspid valve also originate from the interventricular septum. The shape is bent around the left ventricle with a semilunar appearance on short-axis slices. 5.3.2.2  Shunts Many congenital heart diseases can involve intracardiac or vascular shunts. However, there are three very frequent types of shunts, the atrial or ventricular septal defects and the patent ductus arteriosus, which account for a large share of congenital heart diseases and that do not necessarily involve other pathologies. 5.3.2.3  Atrial Septal Defect Atrial septal defect (ASD) is the second most common congenital heart disease and the most common one to become symptomatic in adults. There are two different types of ASD, the ostium primum and the more common ostium secundum defect, which are situated in the two

Fig. 5.3.4  Normal cardiac anatomy. AO aorta giving rise to brachiocephalic artery, left carotid and subclavian artery; PA main pulmonary artery dividing into right and left pulmonary artery; RA right atrium supplied with venous blood from superior and inferior vena cava; RV right ventricle with tricuspid and pulmonary valve; LA left atrium supplied with arterialized blood from the pulmonary veins; LV left ventricle with mitral and aortic valve

parts of the atrial septum, the septum primum, and secundum. The further is situated in the inferior part of the septum close to the atrioventricular plane with the endocardial cushion, while the latter is located further right and cranial. The oval foramen is situated between the two and commonly closes shortly after birth. The ostium primum defect accounts for 30% of ASD and represents an endocardial cushion defect. The ostium secundum defect usually causes a rather large gap in the center and the cranial part of the septum and accounts for 60% of ASD. Another type, the sinus venosus ASD is less common with some 10%, and is usually small and located at the connection to the superior vena cava. A diagnostic problem regarding the atrial septum is its thin membranous configuration and its motion during the cardiac cycle. Thin-slice steady-state free precession sequences should be used to yield a high signal and a good contrast to reliably delineate defects, ideally in parallel four-chamber views and short-axis slices (Fig. 5.3.5). Most indicative are flow voids in the adjacent lumen. Additionally, perfusion sequences in a four-chamber view can show a direct passage of contrast material from the right to the left atrium or, less likely, an inflow of non-enhanced blood. Also, phase-contrast flow measurements can be used to quantify the shunt volume (Beerbaum et al. 2001; Brenner et al. 1992).

719

720

5  Thorax and Vasculature

Fig. 5.3.5  a Schematic diagram of an atrial septal defect (*). b Sagittal SSFP image of a 14-year-old female patient with atrial septal defect (*) and partial anomalous pulmonary venous return (+). c Four-chamber view SSFP image of a 12-year-old patient showing a susceptibility artifact (*) due to an Amplatz device after interventional closure of an atrial septal defect

Fig. 5.3.6  a Schematic diagram of a ventricular septal defect (*). b Short-axis SSFP image of a 10-year-old female patient with complex heart disease showing a large ventricular septal defect (+) and a hypertrophy of the myocardium of the right ventricle (*)

5.3  Heart

5.3.2.4  Ventricular Septal Defect Ventricular septal defect (VSD) represents the most common congenital heart disease with a share of 25–30% (Fig. 5.3.6). Three types of VSD can be differentiated, i.e., a membranous, a muscular, and a supracristal type. The membranous type is the most common and is usually rather small and located in the basal membranous part of the ventricular septum just below the atrioventricular valves, often close to the aortic valve. The muscular VSDs are less frequent and can occur as single or many, very small gaps in the muscular part of the interventricular septum. Many VSDs close spontaneously, but more than half do not. Eisenmenger syndrome with reversal of the shunt direction can occur after a few years with large VSDs, then representing a contraindication for surgical closure. The supracristal type is mostly located just below the pulmonary valve and can be hard to delineate. Due to the variable configuration and size of the defects during the cardiac cycle, the identification in imaging can be difficult. Therefore, similar techniques should be applied as mentioned above for the atrial septal defects (Wang et al. 2003).

Fig. 5.3.7  Schematic diagram of a patent ductus arteriosus (*)

rather small and difficult to demonstrate in very young or premature babies, so high-resolution MRA and coronal reconstructions can be helpful to show the connection between the inferior aspect of the distal aortic arch and the left pulmonary artery close to its origin. 5.3.2.7  Atrioventricular Malformations 5.3.2.7.1  Double-Outlet Right Ventricle

Endocardial cushion defects, also termed atrioventricular (AV) canal, involve the primum part of the atrial septum, the atrioventricular septum, and the membranous part of the ventricular septum. They can also involve defects of the atrioventricular valves, i.e., the anterior leaflet of the mitral valve and the septal leaflet of the tricuspid valve. Endocardial cushion defects can be associated with complex congenital heart disorders and have an increased incidence in left isomerism and Down’s syndrome. MRI can be helpful to determine the presence of ventricular hypoplasia in these patients (Parsons et al. 1991).

In double-outlet right ventricle, aorta and main pulmonary artery both arise from the right ventricular outflow tract (Fig. 5.3.8). This requires a large VSD to provide a connection of systemic and pulmonary circulation, which often causes cyanosis. The position of the semilunar valves is altered with both the aortic and the pulmonary valve in one coronal plane, sometimes with the aorta a little more ventral (Parsons et al. 1991). The location of the VSD is important because a supracristal VSD is more likely to cause a severe cyanosis than is a subaortic VSD. If there is a septum, the aorta usually is in overriding position and receives more blood from the right ventricle. In about half of the cases there is an additional pulmonary stenosis.

5.3.2.6  Patent Ductus Arteriosus

5.3.2.7.2  Hypoplastic Left Heart Syndrome

Patent ductus arteriosus (PDA) Botalli represents a persistent sixth left aortic arch (Fig. 5.3.7). The ductus arteriosus is needed to bypass the non-inflated lungs of the fetus and physiologically closes immediately at birth or within one year. Therefore, the incidence of PDA is higher in premature babies. It is also an essential part of many complex congenital heart diseases. For example, is it required as a right to left shunt for systemic perfusion in hypoplastic left heart syndrome or coarctation. In dtransposition, tetralogy of Fallot, or pulmonary atresia it serves as a left-to-right shunt, providing lung perfusion. If applied early, prostaglandins can keep the ductus open in these children. Often, the PDA is sufficiently visualized on cine SSFP images (Chien et al. 1991), but it can be

Hypoplastic left heart syndrome is characterized by a hypoplasia or atresia of the left cardiac structures, i.e., the mitral valve, the left ventricle, the aortic valve, and the ascending aorta (Fig. 5.3.9). It represents the most severe congenital heart disease associated with congestive heart failure, cyanosis, and cardiogenic shock. In the case of an aortic atresia, there is a retrograde flow in the hypoplastic ascending aorta to supply the carotid and coronary arteries. Generally, the diagnosis is made or confirmed in utero or at birth by echocardiography to administer prostaglandin and eventually perform emergency Rashkind atrial septostomy as soon as possible. There also have been first reports on primary diagnosis in fetal MRI (Fogel et al. 2005; Hata et al. 1995).

5.3.2.5  Endocardial Cushion Defects

721

722

5  Thorax and Vasculature

Fig. 5.3.8  a Schematic diagram of a double-outlet right ventricle (DORV) with a ventricular septal defect (+) and the aorta in overriding position (*) receiving more blood from the right ventricle. b Coronal HASTE image of a 4-year-old girl with DORV showing right (#) and left (°) ventricle, with the aorta (*) in overriding position and pulmonary artery (+) arising off-plane

from the right ventricle. c Axial HASTE image (same patient as in b) showing aorta (*) and pulmonary artery (+) side by side d) Four chamber view in SSFP technique of a four year old male patient with DORV showing the dilated right ventricle (*) and a large apical septal defect

Fig. 5.3.9  a Schematic diagram of a hypoplastic left heart syndrome. The left ventricle (*) is small, and the mitral valve atretic. The left ventricular outflow tract is atretic. The patent ductus arteriosus (#) provides a retrograde perfusion of the aortic arch to supply the carotid and subclavian arteries. The patent foramen ovale or atrial septal defect (+) is required to allow the

blood returning from the lung to re-enter the circulation. b Four chamber view SSFP image of a 4-year-old male patient showing the hypoplastic left ventricle (*), the dilated and hypertrophied right ventricle (+). The left atrium (#) is massively dilated. There is a continuity of the hemiazygos vein (°). c see next page

5.3  Heart Fig. 5.3.9  (continued) c Short-axis SSFP image of the same patient showing the absence of left ventricular myocardium (*) just below the valvular plane and the right ventricular hypertrophy (+) with prominent trabeculae

Fig. 5.3.10  a Schematic diagram showing the situation in hypoplastic left heart syndrome after Norwood I operation: The reconstructed aortic arch (*) arising from the right ventricle supplies the head and upper extremities. The PDA is closed. a Blalock-Taussig shunt (+) between right subclavian and pulmonary artery provides lung perfusion. b Volume-rendered reconstruction of MRA in a 3-year-old patient with hypoplastic left heart syndrome confirming patency of the BlalockTaussig-Shunt (*). c Volume-rendered reconstruction of MRA

in a 4-year-old child with univentricular heart and Norwood II repair. The superior vena cava is connected to the pulmonary arteries (*). The dorsal part of the aortic arch is not visualized due to metal artifacts from a stent. d Schematic diagram showing the situation after Fontan repair: The Blalock-Taussig shunt has been ligated (*). The pulmonary perfusion is now supplied by the Glenn shunt (+) between the superior vena cava and the pulmonary arteries and the Fontan tunnel (#) from the inferior vena cava through the right atrium

723

724

5  Thorax and Vasculature

MRI is then often applied after successful Norwood I operation (Fig. 5.3.10) with construction of a neoaorta from the pulmonary artery, atrial septectomy, and Blalock-Taussig shunting between right subclavian artery and right pulmonary artery for pulmonary perfusion (Kondo et al. 1991; Ichida et al. 1992). In MRI, cardiac function of the univentricular heart can be evaluated for Fontan operation. Also, patency of the Blalock-Taussig shunt and, after Norwood II operation, of the Glenn shunt between the superior vena cava and the pulmonary artery can be assessed (Masui et al. 2000). After Fontan repair, patency and size of the lateral tunnel between the venae cavae through the right atrium can be observed as well as the patency of the Glenn shunt and the closure of the Blalock-Taussig shunt (Tuma et al. 1993; Weiss et al. 2000; de Roos and Roest 2000). 5.3.2.8  Univentricular Heart The term univentricular heart is an ill-defined group of pathologies in which the interventricular septum is not developed and there is only one single ventricle (Fig. 5.3.11). In MRI, the more frequent left ventricular type (accounting for 85% of cases) can be differentiated from the right ventricular type by the morphology of the wall. Marked trabeculae indicate a morphological right ventricle, while a smooth wall is a criterion for a left ventricle. If there is a rudimentary chamber, a ventral location indicates a rudimentary right and a posterior location a rudimentary left ventricle. The abnormal morphology usually entails changes of the regional wall motion, which can be observed with myocardial tagging (Kurotobi et al. 1998). Regarding the atrioventricular valves, there can be all variations with one common valve, one patent and one atretic valve or two separate atrioventricular valves, then also termed double-inlet ventricle (Beekman et al. 1996). 5.3.2.9  Tricuspid Atresia Tricuspid atresia represents a quite frequent cause for severe cyanosis in newborns (Sade and Fyfe 1990). Viability requires both an atrial septal defect and a ventricular septal defect to let the blood pass from the systemic veins into the heart and from the left into the right ventricle and the pulmonary artery (Fig. 5.3.12). The anterior atrioventricular ring is usually replaced by fat tissue which is apparent on T1-weighted images; the right atrium is enlarged and the right ventricle hypoplastic, while the left ventricle is often dilated due to the volume overload. The increased systemic venous pressure can cause venous collaterals in the mediastinum or pericardium. After repair operations, MRI can be very helpful to non-invasively monitor cardiac function and dilatation as well as shunt patency (see “Hypoplastic left heart syndrome” for details).

Fig. 5.3.11  SSFP four-chamber view of a univentricular heart with a common atrioventricular valve in a three week old child with severe cyanosis

5.3.2.10  Ebstein’s Anomaly In Ebstein’s anomaly, the septal and posterior leaflets of the tricuspid valve are displaced into the right ventricle to the apex, thus “atrializing” parts of the right ventricle (Fig. 5.3.13) (Kastler et al. 1990; Choi et al. 1994; Link et al. 1988). The usually patent foramen ovale with the right-to-left shunt and the insufficiency of the tricuspid valve contribute to a volume overload of the right heart and can result in cyanosis. A massive dilatation of the right atrium is observed frequently, while the remaining functional right ventricle can be very small. A limitation of MRI in comparison to echocardiography regarding this disorder is the fact that the individual leaflets of the valve can be difficult to visualize, although steady-state sequences have improved the delineation a little. 5.3.2.11  Truncus Arteriosus The truncus arteriosus or common arterial trunk represents one single arterial vessel arising from the heart that then gives rise to the aorta and the pulmonary artery as well as the coronary arteries. This entity accounts for 2% of congenital heart disease and is associated with 85% mortality in the first year of life (Sierra et al. 2004). In about 30% the truncus arteriosus is associated with a right aortic arch, and it also occurs frequently with DiGeorge syndrome (T-cell immunodeficiency with aplasia of the thymus and parathyroid glands due to a deletion on the long arm of chromosome 22) (Goldmuntz et al. 1998). A large ventricular septal defect, usually in outlet position, is vital and causes cyanosis. The postnatal drop in pulmonary vascular resistance results in an overload of the pulmonary circulation. MRI can be helpful to differentiate between four types (Razavi et al. 2004). In 50%, aorta and pulmonary artery arise from one common

5.3  Heart

Fig. 5.3.12  a Schematic diagram of tricuspid atresia (*): The atrial septal defect (#) is required to allow the systemic venous blood to pass to the left atrium to re-enter the circulation. There can be a ventricular septal defect (°) supplying the hypoplastic right ventricle (+) and pulmonary outflow tract. There is a Blalock-Taussig shunt shown that is often performed to improve pulmonary perfusion. b Axial HASTE image of a five-year-old child with tricuspid atresia showing the continuous wall (*) between right atrium and ventricle (a valve prosthesis is situated farther cranially in this patient) and the large ventricular septal defect (+). c Axial HASTE image of the same patient showing the position of the pulmonary (*) dorsal to the aortic valve (+)

Fig. 5.3.13  a Schematic diagram of Ebstein’s anomaly with the tricuspid valve (*) displaced to the apex. b Four-chamber view in SSFP technique of a 12-year-old child with Ebstein’s anomaly showing the displaced septal leaflet (*) dividing the right ventricle into an atrialized (#) and a true, functional (+) part

725

726

5  Thorax and Vasculature

Fig. 5.3.14  a Schematic diagram of truncus arteriosus type I. There is a common semilunar valve (*) giving rise to both aorta and pulmonary artery overriding a ventricular septal defect (+). b Coronal FLASH MRA image of a 2-week-old male patient, showing a truncus arteriosus type I (*)

valve (type I) (Fig. 5.3.14). In 25% of cases, both pulmonary arteries arise separately from the posterior aspect of the truncus (type II) and in another 10% laterally from the truncus (type III). In another type (IV) referred to as pseudo-truncus, there are no pulmonary arteries, and the lung is perfused by collaterals from the ascending aorta, so-called mid-aortic pulmonary arterial collaterals (MAPCAs). Often, the common valve also is dysplastic. The volume overload usually results in a marked dilatation of the left atrium. If repair operation is delayed, severe pulmonary hypertension results. Rastelli repair with a conduit from the right ventricle to the pulmonary arteries is therefore often performed at very young age (Kersting-Sommerhoff et al. 1990). However, the size of the conduit then often becomes a problem with growth and requires early replacement (Khambadkone et al. 2005).

cava, a hepatic vein, the portal vein, or the ductus venosus (Ito et al. 2000; Livolsi et al. 1001). In partial anomalous pulmonary venous return (PAPVR), not all or only one pulmonary vein is connected to the right atrium (Fig. 5.3.15). This is often an incidental, asymptomatic finding. Sometimes, PAPVR is associated with a hypoplasia of the right lung. Often, there is an infradiaphragmatic pulmonary vein, referred to as scimitar vein (a scimitar is a curved Turkish sword) draining into the inferior vena

5.3.2.12  Vascular Anomalies 5.3.2.12.1  Anomalous Pulmonary Venous Return In total anomalous pulmonary venous return (TAPVR), all pulmonary veins connect to the right instead of the left atrium (Masui et al. 1991). In this situation, a large atrial septal defect or a patent foramen ovale is vital as a right-to-left shunt but usually results in a severe cyanosis. There are three different typical morphologic variations. In the supracardiac type, there is a common vertical vein connecting to the left innominate vein. In the cardiac type the pulmonary veins directly connect to the right atrium or the coronary sinus, usually in a common orifice. In the infradiaphragmatic type, the pulmonary veins connect to a common vessel draining into the inferior vena

Fig. 5.3.15  Volume-rendered reconstruction of MR-angiography of a 7-year-old male patient with partial anomalous pulmonary venous return, showing the right upper lobe vein draining into the right atrium (*)

5.3  Heart

cava (Baran et al. 1996). MRA is a comprehensive tool to non-invasively and reliably detect aberrant pulmonary veins (Prasad et al. 2004). In the coexistence of an intracardiac shunt, carefully planned phase contrast flow measurements can help to non-invasively quantify the hemodynamic contribution of both shunts to assist treatment planning (Wang et al. 2003). 5.3.2.12.2  Cor Triatriatum If there is a connection of the pulmonary veins but remains stenotic, an accessory dorsal atrium results that is separated from the atrium by a fibromuscular membrane (Masui et al. 1991). Depending on the obstruction caused by this intra-atrial septum, a pulmonary venous congestion or edema can result similar to mitral stenosis.

Fig. 5.3.16  a Schematic diagram of tetralogy of Fallot. There is a hypoplastic right ventricular outflow tract (*) and a misaligned aortic valve (#) overriding the ventricular septal defect (+). The increased right ventricular pressure causes muscular hypertrophy (°). b Four-chamber view in SSFP technique of a 5-year-old boy with recurrent right ventricular outflow tract stenosis after Fallot repair: The right ventricle (*) is dilated and the jet (+) at the tricuspid valve indicated a regurgitation due to dilatation of

5.3.2.13  Pulmonary Artery Anomalies 5.3.2.13.1  Valvular Stenosis Valvular stenosis is the most common site of congenital right ventricular outflow tract and pulmonary artery obstruction (Powell et al. 2000; Sahn 2000). The elevated pressure results in right ventricular muscular hypertrophy without dilatation. The central pulmonary arteries, especially the left one, usually show a post-stenotic dilatation. Papillary muscle dysfunction is common in prolonged right ventricular hypertension, resulting in tricuspid regurgitation. However, right heart failure is uncommon in children and usually only becomes apparent in the fifth decade. Subvalvular stenoses are usually a result of a local infundibular hypertrophy, but can also be associated with a dysmorphic valve. On the other hand, supravalvular stenoses usually result from rubella infection in utero.

the valvular annulus. c Axial SSFP image (same patient as in b) showing the relation of the diameters of aorta (*) and pulmonary artery (+). d Phase-contrast flow measurement in the same patient showing a spin inversion (+) in the pulmonary artery compared with the aorta (*) due to a stenosis with flow acceleration. An additional measurement with a wider range of velocity encoding is necessary for better quantification

727

728

5  Thorax and Vasculature

5.3.2.13.2  Tetralogy of Fallot In tetralogy of Fallot, an abnormal position of the crista supraventricularis results in a malalignment of the septal structures, thus causing a right ventricular outflow tract obstruction (Fig. 5.3.16). The resulting four changes are a combination of right ventricular outflow tract stenosis, subaortic ventricular septal defect, overriding aorta, and right ventricular hypertrophy. On plain film radiograms, a boot shape is characteristic for this pathology, and it is often associated with a right aortic arch. In patients older than 12 months, tetralogy of Fallot is the most common cyanotic heart disease. The oxygenation of blood can vary depending on the balance between the obstruction and the shunting, thus ranging from the classic appearance with cyanosis to “pink Fallot” with increased pulmonary blood flow and predominant congestive heart failure. Also, a complete pulmonary atresia with multiple aorticopulmonary collateral arteries (MAPCAs) can be interpreted as an extreme variant of this entity. MRI can be very helpful in the postoperative assessment of the valves and shunts (Powell et al. 2000). If the Ductus arteriosus does not remain patent, the surgical creation of left-right shunts is frequently required. The BlalockTaussig shunt between subclavian and pulmonary artery and the Waterston Cooley anastomosis between pulmonary artery and ascending aorta or the Potts shunt between the descending aorta and the pulmonary artery are frequent primary palliative measures. Definitive repair surgery includes closure of the ventricular septal defect and reconstruction of the right ventricular outflow tract, often with a conduit and pulmonary valve prosthesis. A frequent problem that often requires MR examination is shrinkage of the conduit or small size in relation to the growing patient, then causing a new right ventricular outflow tract obstruction (Khambadkone et al. 2004). Fontan repair is only seldom required in Fallot patients.

5.3.2.13.3  Pulmonary Atresia and Agenesis of the Pulmonary Artery In congenital atresia of the pulmonary valve, the distal pulmonary artery remains hypoplastic. Pulmonary perfusion is maintained by collateralization via large intercostal arteries. If there is an intact ventricular septum, a massive tricuspid regurgitation results, allowing the blood to pass through an obligatory atrial septal defect (Fig. 5.3.17) (Powell et al. 2000). If there is an agenesis of one of the pulmonary arteries, the corresponding lung remains hypoplastic, and the mediastinum is shifted to this side. Other changes such as soft tissue, fibrotic changes, fat, and air-containing structures can be visualized. 5.3.2.13.4  Pulmonary Sling The left pulmonary artery can arise from the right rather than the main pulmonary artery and then pass to the left between trachea and esophagus, thus forming a sling around the trachea close to its bifurcation. Therefore, an indentation of the trachea from posterior can be seen, which frequently causes severe stridor and asymmetric lung inflation in early life, sometimes also recurrent pulmonary infections (Ohlemann et al. 1995). In MRI especially axial planes are helpful for preoperative clarification of the anatomy. As this pathology can be associated with intracardiac congenital heart disease, the presence of coexisting cardiac findings should be paid attention to. Also, there may be a complete tracheal ring which consists of circular cartilaginous material and can be differentiated from mere compression by its round configuration with small diameter.

Fig. 5.3.17  a Four-chamber view SSFP image of a 6-month-old male patient showing a hypoplastic right ventricle (*) with a continuous ventricular septum. b Short-axis SSFP image of the same patient. c MRA shows a Glenn shunt (*) providing pulmonary perfusion from the superior vena cava

5.3  Heart

5.3.2.14  Anomalies of the Aorta and Supra-Aortic Vessels 5.3.2.14.1  Coarctation of the Aorta If the aorta shows a focal narrowing with an obstruction to the blood flow from dorsal at the level of the ductus arteriosus, this is called coarctation. Two types of coarctation can be differentiated, the preductal and the postductal types. The postductal type is caused by the postnatal contraction of fibrous ductal tissue in the aortic wall and is often an incidental finding in adults with is a pre-stenotic dilatation involving the left subclavian artery (Fig. 5.3.18) and a post-stenotic dilatation of the descending aorta. The preductal type represents a tubular hypoplasia and involves a longer segment of the aortic arch. There may be collaterals, especially of the intercostal arteries via the internal mammary and the superior epigastric arteries with characteristic rib notching in older children. The obstruction can cause pressure overload and left ventricular hypertrophy. In older children and adults, hypertension and diminished femoral pulses with different blood pressures in upper and lower extremities can result. Hypertension can be increased by renal hypoperfusion. In preductal coarctation, a patent ductus arteriosus with right-to-left shunting can relieve left ventricular pressure overload, if there is a coexistent ventricular septal defect as left-to-right shunt. There is a strong association with persistent ductus arteriosus (about 66%), bicuspid aortic valve (50%), and ventricular septal defects (33%). In Turner’s syndrome, coarctation occurs in about 25%. Resection with end-to-end anastomosis, graft interposition, or patch angioplasty is a common surgical approach. Oblique parasagittal and perpendicular planes provide best assessment of the stenosis and allow the measure-

ment of diameters. MRA with contrast enhanced gradient-echo techniques can help to detect collateral vessels. With additional phase contrast flow measurements it is possible to assess the pressure gradient in the stenosis and to quantify collateral flow. 5.3.2.14.2  Double Aortic Arch If both the right and the left fourth aortic arch persist, a double aortic arch results (Fig. 5.3.19). The ipsilateral subclavian and carotid arteries arise from both arches. The left descending aorta is formed by both arches with the right arch usually running behind the esophagus. Hemodynamic sequelae are rare, but the airway compression with stridor caused by the two arches encircling and indenting the trachea is usually severe. As the esophagus is also enclosed in the ring, the children usually also suffer from dysphagia, and the stridor worsens with feeding. Axial MR images can clarify the anatomy and coronal images can additionally help to identify the dominant, larger of both arches for surgical planning. 5.3.2.14.3  Lusory Artery An aberrant right subclavian artery that arises separately from the descending aorta and takes an unusual course behind the esophagus to the left is called lusory artery (Latin: ludere = to play) (Fig. 5.3.20). The aberrant course can cause compression of the esophagus and dysphagia. Airway compression and stridor are less frequent. This finding is easily detected on axial images. As a symmetrical pathology, a right descending aorta with aberrant left subclavian artery is frequently associated with complex

Fig. 5.3.18  a Schematic diagram of postductal type aortic coarctation (*). b MIP reconstruction of MRA in a 28-year-old female patient with postductal coarctation of the aorta (*) and prominent collateralization via the intercostal arteries

729

730

5  Thorax and Vasculature

Fig. 5.3.19  MIP reconstruction of MRA of a 10-day-old female baby, showing a double aortic arch with the respective carotid and subclavian arteries arising from both arches

congenital heart disease, i.e., truncus arteriosus, d-transposition of the great arteries, and tetralogy of Fallot. The aberrant left subclavian artery (“left lusory artery”) arises from the right descending aorta and takes a retroesophageal course to the left. This can mimic the presence of a double aortic arch. 5.3.2.14.4  Transposition of the Great Arteries For the differentiation of transpositions, the correct identification of morphological right and left ventricles and atria as well as the corresponding atrioventricular valves is a prerequisite. The relative position of aortic and pulmonary valve is another important diagnostic criterion to detect cardiac rotation anomalies: Normally, the pulmonary valve is situated left anterior to the aortic valve. In d-transposition of the great arteries (TGA) it is located left posterior and in l-TGA right posterior to the aortic trunk. 5.3.2.15  Complete Transposition of the Great Arteries In the presence of ventriculoarterial discordance with atrioventricular concordance, a complete transposition (D-TGA) of the great arteries with separation of pulmonary and systemic circulation results (Fig. 5.3.21). Therefore, a shunt such as a patent foramen ovale, a ventricular septal defect, or a patent ductus arteriosus is vital. The aorta arises from the right ventricle, while the pulmo-

Fig. 5.3.20  a Coronal MIP image of supra-aortic MRA in a 58year-old male patient, showing the separate origin of the right subclavian artery. b Same dataset in maximum intensity projection from right anterior oblique view. c Coronal MIP image of thoracic MRA in a 3-week-old child with a right aortic arch and a “left lusory artery”: The left subclavian artery arises from a Kommerell’s diverticulum, which represents a remnant of the left arch (*)

nary artery arises from the left ventricle. Thus, there is a pressure overload in the right ventricle supplying the systemic circulation and a volume overload of the left ventricle supplying the pulmonary circulation. The limiting factor is the increased pressure in the right ventricle entailing muscular hypertrophy, shunt reversal, and congestive heart failure. Early palliative interventions such as interventional creation of an atrial septal defect by Rashkind maneuver or surgical septostomy by Blalock Hanlon operation, or the prevention of ductus arteriosus closure with prostaglandin are frequently necessary. The definitive cardio surgical repair is mostly done at

5.3  Heart

is not commonly performed anymore because the pressure overload of the right ventricle, which has shown to be the main problem in adult patients, is not corrected with these procedures. Other problems after atrial switch operation apart from right ventricular failure include atrial thrombosis and a high incidence of arrhythmias. However, an arterial switch operation sometimes cannot be performed in older patients because the left ventricle maintaining the pulmonary circulation is not able to take over systemic arterial pressure, and then atrial switch can be the best option. 5.3.2.16  Congenitally Corrected Transposition

Fig. 5.3.21  Schematic diagram of a complete transposition of the great arteries. The aorta (*) arises from the right ventricle, while the pulmonary artery (+) arises from the left ventricle. An atrial septal defect (#) and a patent ductus arteriosus (°) provide shunts between pulmonary and systemic circulation

an age of 6–9 months by the Jatene operation, in which pulmonary artery and aorta are switched cranial to their valves, and the coronary arteries usually have to be implanted into the neoaorta. Postoperatively, the dorsally placed neoaorta is embraced ventrally and laterally by the pulmonary arteries. This can result in stenoses of the proximal pulmonary arteries, which is the most common problem after switch operation apart from stenoses of the anastomoses (Kersting-Sommerhoff et al. 1990). These changes can be visualized reliably in axial and sagittal MR images. Atrial switch surgery with a pericardial baffle (Mustard) or reorientation of the atrial septum (Senning)

Fig. 5.3.22  a Mid-ventricular short-axis SSFP image of a 38year-old female patient with previously unknown congenitally corrected transposition of the great arteries showing the anatomical left ventricle (*) with prominent trabeculae character-

If there is atrioventricular and ventriculoarterial discordance, this inversion of the ventricles and atrioventricular valves results in a correct connection between the atria and the corresponding arteries (Fig. 5.3.22). However, only in 1% of congenitally corrected transpositions (l-TGA) there are no associated cardiac anomalies, justifying the term “congenitally corrected transposition.” Most commonly, this pathology is associated with a large ventricular septal defect and pulmonary artery stenosis, which determine the overall prognosis with 50% mortality in 15 years. Surgical treatment focuses on associated abnormalities, but some authors regard double-switch operation as an option to prevent late systemic right ventricular failure. In MRI, the evaluation and follow-up of right ventricular function is the most important issue in these patients. 5.3.2.17  Heterotaxia Syndrome The terms heterotaxia syndrome, situs ambiguus, or Ivemark syndrome describe a disturbance of the normal

istic for right ventricular morphology. The anatomical right ventricle (+) has a smooth wall. b Axial HASTE image of the same patient showing the position of the pulmonary (*) right posterior to the aortic valve (+)

731

732

5  Thorax and Vasculature

asymmetry of the thoracic and abdominal organs that occurs sporadically and accounts for about 1–3% of congenital heart disease with a high mortality of 80% in the first year of life. There are two major types, the double rightsided anatomy called asplenia syndrome (Fig. 5.3.23) or right isomerism, and the double left-sided anatomy referred to as polysplenia syndrome or left isomerism. In both, congenital heart disease is the rule, and the liver usually shows a midline position and a rather symmetrical configuration and the cardiac apex and stomach usually lie opposite to each other. Chest films or babygrams usually provide first hints to the disease by the position of the cardiac apex and stomach bubble and the presence or absence of minor fissures in both lungs. MRI with axial T1 and HASTE images of the chest and upper abdomen, cardiac imaging with cine SSFP short and long-axis images, and contrast-enhanced MRA can provide a complete overview to allow classification of the anatomical changes. 5.3.2.18  Asplenia Syndrome (Right Isomerism) Both lungs are trilobed and have symmetrical steep bronchi crossing cranial to the artery. There usually are bilateral superior venae cavae. Aorta and vena cava inferior are located on the same, more often than on left, side of the spine. There is no spleen on either side. Both atrial

appendages have a right-sided configuration. Severe cyanotic heart disease is very common in these children, often involving an atrioventricular septal defect with a common atrioventricular valve, a double-outlet right ventricle, transposition of the great arteries or pulmonary stenosis or atresia. Also, an anomalous pulmonary venous return is observed frequently, often with connections below the diaphragm. 5.3.2.19  Polysplenia Syndrome (Left Isomerism) Both lungs are bilobed and have horizontal bronchi crossing below the arteries. There usually are superior venae cavae on both sides. The atrial appendages both exhibit a left-sided configuration. The inferior vena cava below the liver is often absent. Then, there is azygos continuity with a dilated azygos vein, and the hepatic veins usually drain separately into the right or common atrium. Multiple spleens can be found in the upper abdomen. The heart diseases observed in these children is less severe. Most frequent are a common atrium and a ventricular septal defect. References 1.

2.

3.

4.

5.

6.

Fig. 5.3.23  Maximum-intensity projection of MR angiography of a 7-year-old boy with an asplenia syndrome with trilobed lungs, large atrioventricular septal defect (*) with a common atrioventricular valve, transverse liver (+), right-sided pancreas (°) and stomach, and atresia of the inferior vena cava (#)

7.

8.

Baran R, Kir A, Tor MM, Ozvaran K, Tunaci A (1996) Scimitar syndrome: confirmation of diagnosis by a noninvasive technique (MRI). Eur Radiol 6:92–94 Beekman RP, Beek FJ, Meijboom EJ, Wenink AC (1996( MRI appearance of a double inlet and double outlet right ventricle with supero-inferior ventricular relationship. Magn Reson Imaging 14:1107–1112 Beerbaum P, Korperich H, Barth P, Esdorn H, Gieseke J, Meyer H (2001) Noninvasive quantification of left-to-right shunt in pediatric patients: phase-contrast cine magnetic resonance imaging compared with invasive oximetry. Circulation 103:2476–2482 Brenner LD, Caputo GR, Mostbeck G et al. (1992) Quantification of left to right atrial shunts with velocity-encoded cine nuclear magnetic resonance imaging. J Am Coll Cardiol 20:1246–1250 Chien CT, Lin CS, Hsu YH, Lin MC, Chen KS, Wu DJ (1991) Potential diagnosis of hemodynamic abnormalities in patent ductus arteriosus by cine magnetic resonance imaging. Am Heart J 122:1065–1073 Choe YH, Kim YM, Han BK, Park KG, Lee HJ (1997) MR imaging in the morphologic diagnosis of congenital heart disease. Radiographics 17:403–422 Choi YH, Park JH, Choe YH, Yoo SJ (1994) MR imaging of Ebstein’s anomaly of the tricuspid valve. AJR Am J Roentgenol 163:539–543 Dohlemann C, Mantel K, Vogl TJ et al. (1995) Pulmonary sling: morphological findings. Pre- and postoperative course. Eur J Pediatr 154:2–14

5.3  Heart 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Fogel MA, Weinberg PM, Fellows KE, Hoffman EA (1995) A study in ventricular-ventricular interaction. Single right ventricles compared with systemic right ventricles in a dual-chamber circulation. Circulation 92:219–230 Fogel MA, Wilson RD, Flake A et al. (2005) Preliminary investigations into a new method of functional assessment of the fetal heart using a novel application of ‘real-time’ cardiac magnetic resonance imaging. Fetal Diagn Ther 20:475–480 Goldmuntz E, Clark BJ, Mitchell LE et al. (1998) Frequency of 22q11 deletions in patients with conotruncal defects. J Am Coll Cardiol 32:492–498 Hata K, Hata T, Manabe A, Kitao M (1995) Hypoplastic left heart syndrome: color Doppler sonographic and magnetic resonance imaging features in utero. Gynecol Obstet Invest 39:70–72 Ichida F, Hashimoto I, Miyazaki A et al. (1992) [Magnetic resonance imaging: evaluation of the Blalock-Taussig shunts and anatomy of the pulmonary artery]. J Cardiol 22:669–678 Ito N, Maie S, Furukawa Y et al. (2000) [Left partial anomalous pulmonary venous return to the innominate and hepatic veins]. Nippon Naika Gakkai Zasshi 89:1185–1187 Kastler B, Livolsi A, Zhu H, Roy E, Zollner G, Dietemann JL (1990) Potential role of MR imaging in the diagnostic management of Ebstein anomaly in a newborn. J Comput Assist Tomogr 14:825–827 Kersting-Sommerhoff BA, Seelos KC, Hardy C, Kondo C, Higgins SS, Higgins CB (1990) Evaluation of surgical procedures for cyanotic congenital heart disease by using MR imaging. AJR Am J Roentgenol 155:259–266 Khambadkone S, Bonhoeffer P (2004) Nonsurgical pulmonary valve replacement: why, when, and how? Catheter Cardiovasc Interv 62:401–408 Khambadkone S, Coats L, Taylor A et al. (2005) Percutaneous pulmonary valve implantation in humans: results in 59 consecutive patients. Circulation 112:1189–1197 Kondo C, Hardy C, Higgins SS, Young JN, Higgins CB (1991) Nuclear magnetic resonance imaging of the palliative operation for hypoplastic left heart syndrome. J Am Coll Cardiol 18:817–823 Kurotobi S, Sano T, Naito H et al. (1998) Regional ventricular systolic abnormalities caused by a rudimentary chamber in patients with univentricular hearts. Am J Cardiol 82:86–92 Link KM, Herrera MA, D’Souza VJ, Formanek AG (1988) MR imaging of Ebstein anomaly: results in four cases. AJR Am J Roentgenol 150:363–367 Livolsi A, Kastler B, Marcellin L, Casanova R, Bintner M, Haddad J (1991) MR diagnosis of subdiaphragmatic anomalous pulmonary venous drainage in a newborn. J Comput Assist Tomogr 15:1051–1053

23. Masui T, Seelos KC, Kersting-Sommerhoff BA, Higgins CB (1991) Abnormalities of the pulmonary veins: evaluation with MR imaging and comparison with cardiac angiography and echocardiography. Radiology 181:645–649 24. Masui T, Katayama M, Kobayashi S et al. (2000) Gadolinium-enhanced MR angiography in the evaluation of congenital cardiovascular disease pre- and postoperative states in infants and children. J Magn Reson Imaging 12:1034–1042 25. Parsons JM, Baker EJ, Anderson RH et al. (1990) Morphological evaluation of atrioventricular septal defects by magnetic resonance imaging. Br Heart J 64:138–145 26. Parsons JM, Baker EJ, Anderson RH et al. (1991) Doubleoutlet right ventricle: morphologic demonstration using nuclear magnetic resonance imaging. J Am Coll Cardiol 18:168–178 27. Powell AJ, Chung T, Landzberg MJ, Geva T (2000) Accuracy of MRI evaluation of pulmonary blood supply in patients with complex pulmonary stenosis or atresia. Int J Card Imaging 16:169–174 28. Prasad SK, Soukias N, Hornung T et al. (2004) Role of magnetic resonance angiography in the diagnosis of major aortopulmonary collateral arteries and partial anomalous pulmonary venous drainage. Circulation 109:207–214 29. Razavi R, Miquel M, Baker E (2004) Diagnosis of hemitruncus arteriosis by three-dimensional magnetic resonance angiography. Circulation 109:E15–E16 30. Roos A de, Roest AA (2000) Evaluation of congenital heart disease by magnetic resonance imaging. Eur Radiol 10:2–6 31. Sade RM, Fyfe DA (1990) Tricuspid atresia: current concepts in diagnosis and treatment. Pediatr Clin North Am 37:151–169 32. Sahn DJ (2000) Accuracy of MRI evaluation of pulmonary blood supply in patients with complex pulmonary stenosis or atresia. Int J Card Imaging 16:479–480 33. Sierra J, Beghetti M, Kalangos A (2004) Truncus arteriosus repair with double aortic homograft. J Card Surg 19:252–253 34. Tuma S, Lizler J, Fendrych P, Hruda J, Bartakova H (1993) [Magnetic resonance imaging of the postoperative status in children with congenital heart defects]. Cesk Pediatr 48:645–647 35. Wang ZJ, Reddy GP, Gotway MB, Yeh BM, Higgins CB (2003) Cardiovascular shunts: MR imaging evaluation. Radiographics 23(Spec no.):S181–S194 36. Weiss F, Habermann CR, Lilje C et al. (2002) [MRI in postoperative assessment of univentricular heart disease: correlation with echocardiography and angiography]. RoFo 174:1537–1543 37. Yoo SJ, Kim YM, Choe YH (1999) Magnetic resonance imaging of complex congenital heart disease. Int J Card Imaging 15:151–160

733

734

5  Thorax and Vasculature

5.3.3  Primary Cardiomyopathies K. Nikolaou 5.3.3.1  Introduction: Background Information and Role of MRI Cardiomyopathies are a group of diseases that are associated with myocardial dysfunction and can be classified either as primary or secondary cardiomyopathies. Generally, they are chronic and progressive conditions with impairment of cardiac function (Ehlert et al. 2001). Based on morphological and functional parameters, according to the World Health Organization (WHO) classification (Richardson et al. 1996) primary cardiomyopathies are divided into four groups: dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM), arrhythmogenic right ventricular cardiomyopathy (ARVC) and restrictive cardiomyopathy (RCM) (Table 5.3.5). The term “secondary cardiomyopathy” is used to specify diseases with the clinical indications of a cardiomyopathy, but can be attributed to a certain pathophysiological mechanism such as exposure to toxic substances, metabolic syndromes, or systemic diseases (see Sect. 5.3.4). Accordingly, the diagnosis “primary cardiomyopathy” requires exclusion of other etiological factors. MRI is well suited to the evaluation of myocardial disease (Miller and Riessen 2005). Cardiac MR (CMR) is non-invasive and does not expose the patient to ionizing radiation. CMR provides accurate and reproducible multiplanar, static as well as dynamic images of the heart with high-temporal and -spatial resolution in any desired plane without limitation by the acoustic window (Bellenger et al. 2000), and with excellent reproducibility (Grothues et al. 2002). The true geometry of the myocardium and the ventricles can be depicted, without the need of any geometrical assumptions.

Especially in patients with cardiomyopathy, the heart is frequently not ellipsoidal, but shows a distorted morphology due to ventricular dilatation or thickening. As mentioned, CMR can be used to evaluate both morphological and functional parameters. Systolic function can be appraised in various views such as the two-chamber view, four-chamber view, or on the short-axis views. Diastolic dysfunction is increasingly recognized as abnormal in many patients with cardiomyopathy, as demonstrated in Framingham data (Redfield et al. 2003). Recognition of diastolic dysfunction is important in view of the increased associated morbidity and mortality (Mandinov et al. 2000). Typical features of diastolic dysfunction assessable by CMR are abnormal peak filling rates, or dilatation of the left atrium. Myocardial tagging can be of additional help in the assessment of systolic and diastolic dysfunction (Zwanenburg et al. 2005). Furthermore, tissue alteration of the myocardium can be detected assessing regional contrast enhancement, T1 and T2 signal intensities, and chemical shift phenomena. Thereby, CMR sequences can derive more specific information about myocardial tissue that may provide the mechanistic basis for cardiomyopathies. T1-weighted, dark-blood spin-echo sequences visualize the myocardium with good contrast to adjacent structures such as epicardial fat and ventricular blood. Late enhancement imaging following gadolinium administration can be used to identify regions of fibrosis or scarring, as well as inflammatory and infiltrative processes. T2-weighted echo sequences may highlight myocardial edema or pericardial effusion. T2* sequences can be used to show siderosis (iron overload) in hemochromatosis and thalassemia. From a clinical point of view, an examination based on a single, efficient, and non-invasive MR study focusing on the clinically relevant features of cardiomyopathies is an objective and reproducible means for diagnosing and monitoring this disease.

Table 5.3.5  Classification of cardiomyopathies Dilated cardiomyopathy (DCM)

Dilatation and impaired function of the left ventricle or both ventricles.   The underlying cause is unknown in about half of the cases   (idiopathic DCM), but may be genetic, viral, metabolic, or toxic

Hypertrophic cardiomyopathy HCM (Hypertrophic obstructive cardiomyopathy: HOCM)

Left or right or both ventricles are hypertrophic, often asymmetrical,   septum is typically involved. May be obstructive if septal hypertrophy   involves the left ventricular outflow tract

Restrictive cardiomyopathy (RCM)

Rare form of cardiomyopathy. Restricted filling and reduced   diastolic dilatation of either or both ventricles, with normal   or near-normal systolic function. Might be primarily associated   with other diseases such as amyloidosis

Arrhythmogenic right ventricular dysplasia (ARVD)

Progressive fibrofatty replacement of the right-ventricular myocardium. Affects left ventricle in 15% of cases during life. Frequently causing arrhythmias, typically arising from the right ventricular outflow tract

5.3  Heart

5.3.3.2  Dilated Cardiomyopathy 5.3.3.2.1  Dilated Cardiomyopathy: Etiology, Histology, and Clinical Presentation Dilated cardiomyopathy (DCM) is the commonest form of primary cardiomyopathies (~60%). The incidence of primary DCM is around 5–8 cases per 100,000 population per year (Soler et al. 2003). It accounts for around 25% of cases of congestive heart failure in the United States (Takeda 2003). According to the WHO classification, its etiology is unclear in about half of the cases (idiopathic DCM) but may be genetic, viral, metabolic, or toxic (Takeda 2003; Dec and Fuster 1994). The most frequent cause of a dilated heart, however, is coronary artery disease (CAD), resulting in the so-called ischemic cardiomyopathy (see Sect. 5.3.7), which has to be separated from primary DCM. Macroscopically, the main feature of DCM is dilatation, usually of both ventricles, associated with impaired contractile function. There is an increase in the end-diastolic and end-systolic volume, resulting in a reduction in stroke volume and ejection fraction. Valvular abnormalities can be associated, including mitral and tricuspid regurgitation. Wall thickness can be normal so that the left ventricular (LV) mass index may be increased. However, in advanced stages, relative wall thinning becomes more frequent. Histologically, idiopathic DCM is characterized by progressive interstitial fibrosis, with a decrease in the number of contractile myocytes (Dec and Fuster 1994). Myocardial biopsy from affected regions typically shows widespread interstitial and perivascular fibrosis, particularly affecting the subendocardial layer. Regions of necrosis and cellular infiltration may be seen (de Leeuw et al. 2001). The clinical presentation usually involves heart failure, which is often progressive. Arrhythmias, thromboembolism, and sudden death

are common and may occur at any stage (White et al. 1998). Thromboembolism, mainly due to apical thrombi, is also observed frequently. 5.3.3.2.2  Dilated Cardiomyopathy: MR Protocol and Findings Requirements for a detailed assessment of DCM using any cardiac imaging modality are the following: evaluation of the size and volumes of the ventricular cavities (Fig. 5.3.24) assessment of the thickness of the ventricular walls, assessment of the ejection and shortening fractions, as well exclusion of other potentially reversible causes of left ventricular dysfunction. Magnetic resonance imaging can accurately reflect morphologic changes and quantify functional abnormalities in patients suffering from DCM, potentially monitoring the effects of pharmacological therapy or cardiac surgery (Doherty et al. 1992a; Parga et al. 2001). A typical CMR protocol in imaging DCM should include an ECG-gated T1-weighted multislice spin-echo sequence of the entire heart in axial planes to depict cardiac morphology. Functional imaging, e.g., along short-axis stacks, should depict functional and valvular abnormalities, and provide volumetric data, including stroke volume, ejection fraction, and LV mass. Myocardial tagging can further identify regional wall motion abnormalities and may provide a quantitative assessment of contractility (Table 5.3.6). Regional wall motion abnormalities or regional wall thinning might indicate ischemic cardiomyopathy as an underlying cause and should be further investigated. However, in idiopathic, non-ischemic DCM, the right ventricular function is typically affected. Also, the volume of the left atrium may be increased, carrying prognostic significance in DCM (Rossi et al. 2002). Valvular function can be assessed, e.g., using

Table 5.3.6  Dilatative cardiomyopathy: MR protocol and imaging findings Imaging protocol suggestions

Diagnostic criteria

Exclusion criteria

Myocardial function: Cine MR: Four-chamber view Two-chamber view Short axis

Decreased ejection fraction and/or decreased fractional shortening

Systemic hypertension

Function of cardiac valves: Phase-contrast imaging (flow maps)

Enlarged diastolic left ventricular   dimension

Coronary artery disease

Regional contractility: Myocardial tagging

Systemic disease known to cause DCM

Late enhancement: Detect myocardial fibrosis and exclude silent infarcts

Pericardial disease/congenital heart disease/cor pulmonale

735

736

5  Thorax and Vasculature

phase-contrast measurements. This is helpful to measure the extent of mitral regurgitation, where surgical repair may improve the outcome. As mentioned above, diastolic function should be assessed specifically. In studies of late gadolinium enhancement in DCM, fibrosis is absent in ~60% of the cases (McCrohon et al. 2003). In about a quarter of patients with DCM, there is patchy mid-wall hyperenhancement that has no correlation with coronary artery territories, i.e., late hyperenhancement in DCM is distinct from an ischemic pattern. The MR imaging findings of idiopathic DCM are frequently non-specific, and most secondary forms of DCM cannot be distinguished. However, it is possible to distinguish between cardiac dilatation due to ischemic cardiomyopathy, DCM due to cardiac hemochromatosis and idiopathic DCM. Although significant overlapping exists between the morphologic and functional changes of DCM secondary to myocardial iron overload and idiopathic DCM, the presence of focal or diffuse signal loss on spin-echo and gradient-echo images of the myocardium overloaded with iron are sufficient to confirm the diagnosis of cardiac hemochromatosis with MR imaging (Waxman et al. 1994). However, the MR signal intensity and enhancement characteristics of the myocardium are insufficient to distinguish between the dilated form of ischemic cardiomyopathy and nonischemic forms of DCM. Of course, selected imaging criteria, e.g., segmental wall thinning, are indicative of an old myocardial infarction, whereas uniformly reduced wall thickness of the myocardium and enlargement of the

Fig. 5.3.24a,b  55-year-old man with dilated cardiomyopathy. Cine-MR images in four-chamber view (a) and short-axis view (b) at end-diastole show significant dilatation of the LV cavity. Ejection fraction was <35% in this patient. RA right atrium,

left ventricular trabeculae are more typical of idiopathic DCM (Soler et al. 2003). Finally, in DCM, CMR can be used to monitor the efficacy of drug treatment (Strohm et al. 2001). This has been demonstrated in DCM for growth hormone, angiotensin-converting enzyme, or beta-blockers (Osterziel et al. 1998; Doherty et al. 1992b; Groenning et al. 2000). An important therapeutic option in patients with DCM is the use of resynchronization with pacemakers and the use of implantable cardioverter–defibrillators. At present, CMR is contraindicated in these patients, but dedicated MR-compatible devices are being developed. 5.3.3.3  Hypertrophic Cardiomyopathy 5.3.3.3.1  Hypertrophic Cardiomyopathy: Etiology, Histology, and Clinical Presentation Hypertrophic cardiomyopathy (HCM) is known to be inheritable (typically autosomal dominant), with variable phenotypic expression. The overall prevalence of HCM is about 1 in 500 of the general population (Prasad et al. 2004). The primary characteristic feature of HCM is myocardial hypertrophy of the LV in the absence of an obvious hypertrophic stimulus (e.g., hypertension or aortic valve stenosis) (Maron 2002). Typically, an asymmetrical involvement of the interventricular septum can be found.

LA left atrium, RV right ventricle, LV left ventricle (Courtesy of Dr. Armin Huber, Department of Clinical Radiology, LudwigMaximilians-University of Munich)

5.3  Heart

Regarding LV function, impaired diastolic function is a key feature of HCM, while systolic impairment or ventricular dilatation is only seen in late stages of the disease. In a subset of patients with septal hypertrophy close to the left ventricular outflow tract (LVOT), a pressure gradient along this septal bulging provokes increased resistance similar to an aortic valve stenosis, leading to hypertrophic obstructive cardiomyopathy (HOCM). Other functional features include systolic anterior motion of the mitral valve and mitral regurgitation (Fig. 5.3.25). Histologically, HCM is characterized by a pattern of myofibrillar disarray and fibrosis, due to a sarcomeric mutation, with a number of possible gene abnormalities (Arad et al. 2005). Also, the coronary arteries often show an intramural course, with a reduction in size of the lumen and thickening of the vessel wall (Devlin et al. 1999). The most relevant clinical symptoms in HCM patients are pectanginal complaints on exertion, dyspnea, fatigue, syncope, arrhythmia, and sudden death. In fact, HCM is the most common cardiac related cause of sudden death in active and apparently healthy adolescents and young

Fig. 5.3.25  62-year-old female patient with hypertrophic obstructive cardiomyopathy (HOCM). Four cine MR images in a three-chamber view, selected from various phases of the cardiac cycle, show the increasing obstruction of the left ventricular outflow tract during systole caused by a predominantly septal hypertrophy. Typical signal loss in the proximal aorta (arrow), indicative of the obstruction of the outflow tract, as well as a signal loss in the left atrium, indicative of a mitral valve insufficiency (arrowhead), can be observed. The left atrium (LA) is markedly enlarged. LV left ventricle, Ao aorta (Courtesy of Dr. Armin Huber, Department of Clinical Radiology, Ludwig-Maximilians-University of Munich)

adults (Liberthson 1996). HCM is often diagnosed on the grounds of unspecific cardiac symptoms, after exclusion of systemic or cardiac causes of myocardial hypertrophy. Typically, the primary diagnosis of HCM in patients with de novo symptoms described above is made by a combination of echocardiography with an abnormal ECG. 5.3.3.3.2  Hypertrophic Cardiomyopathy: MR Protocol and Findings In HCM, myocardial hypertrophy is characterized by heterogeneous involvement of the left ventricular wall, enlarged to very different degrees. The detection and quantification of myocardial thickness is the most relevant morphologic finding both to ascertain whether hypertrophy is present and in follow-up monitoring of HCM patients. Even in patients with good acoustic windows, echocardiography might be of limited value with regard to the assessment of the anterior and inferior wall and of the apex (Devlin et al. 1999). Cardiac MR (CMR) evaluates all myocardial segments with equal accuracy and can reveal unusual forms of hypertrophy that are difficult to assess with echocardiography, such as apical hypertrophies (Soler et al. 1997). Cine MR data in several views including two-chamber, four-chamber and shortaxis views provide for accurate measurement of the septum and the free wall of the left and right ventricle which is needed to ascertain the presence, distribution, and severity of the hypertrophic process (Table 5.3.7) (PonsLlaso et al. 1997). Isolated apical hypertrophy is best seen on two- and four-chamber views (Fig. 5.3.26). The overall LV mass may be significantly increased, and both atria may be dilated. Accurate measurement of myocardial mass in HCM patients is regarded as being an independent and powerful method for predicting morbidity and mortality in these patients (Scheffeold et al. 2005). On three-dimensional (3D) MR imaging, the Simpson rule is the best way to quantify myocardial mass (Soler et al. 1999). Using high-temporal-resolution steady-state freeprecession (SSFP) cine MR sequences, and appropriate analysis software, the peak filling rate and time to peak filling rate can be measured to assess diastolic function. Typically, a restrictive filling pattern is found due to reduced compliance. As described above, other relevant functional abnormalities in HCM include mitral abnormalities and LVOT obstruction. Mitral abnormalities are characterized by systolic anterior motion (SAM) of the anterior leaflet valve. CMR can aid in the identification of mitral regurgitation and also outflow tract obstruction by identifying a signal void in the jet flow from a narrowed outflow tract and secondly by using velocity mapping proximal and distal to the outflow tract (Didier et al. 2000). The turbulent jet that appears during systolic LVOT obstruction is easily detected and quantified with cine MR (White et al. 1996). Other functional changes in

737

738

5  Thorax and Vasculature Table 5.3.7  Hypertrophic cardiomyopathy: MR protocol and imaging findings Imaging protocol suggestions

Diagnostic criteria

Cine MR: global LV function Four-chamber view Two-chamber view Short axis

Primary imaging criterion: general or circumscript   (regional) left ventricular hypertrophy

Cine MR: Outflow tract: Gradient-echo cine MR of the LVOT in several planes

Septal hypertrophy of the LVOT causing   an LVOT stenosis: HOCM

Cine MR: mitral valve

Systolic anterior motion (SAM) of the anterior leaflet   of the mitral valve and mitral insufficiency

Regional contractility: Myocardial tagging on SA

Primarily impaired diastolic function, with preserved systolic function (until late stages of the disease)

Late enhancement: Detect myocardial fibrosis

Diffuse, patchy, mid-ventricular wall late enhancement   (not sub-endocardial), indicating increased diffuse fibrosis

LVOT left ventricular outflow tract, SA short-axis, HOCM hypertrophic obstructive cardiomyopathy

HCM patients, detected by the use of myocardial tagging, show abnormal regional contractility with reduced septal wall motion and impaired radial and circumferential strain patterns. Long-axis function is reduced, and ventricular torsion is usually increased (Ennis et al. 2003). The usefulness of these findings in routine clinical management and follow-up of the disease will, however, have to be explored in future studies. MR imaging is also useful for monitoring morphologic and functional changes during the natural course of the disease and in patient follow-up after pharmacological or surgical treatment (Schulz-Menger et al. 2000). Late gadolinium enhancement imaging may show areas of increased signal intensity, which can be patchy or diffuse. In a recent study on a medium-sized patient cohort with HCM, about 80% of patients showed late myocardial enhancement, with the

main distribution being in hypertrophic regions of the LV wall (Choudhury et al. 2002). Abundant connective tissue intermingled with myofibrillar bundles disarray is suggested as the reason of late enhancement, but no pathological confirmation has been obtained. The presence and extent of enhancement was associated inversely with regional contraction, and positively with regional hypertrophy. Other work has shown that HCM patients with gadolinium uptake are at an increased risk of cardiac events (Moon et al. 2003). In younger patients (<40 years), the extent of gadolinium uptake is linked to the risk of sudden cardiac death. In older patients (>40 years), the extent of gadolinium uptake is linked to the progression of heart failure. Differential diagnoses between HCM and infiltrative secondary cardiomyopathies with increased wall thickness, such as cardiac amyloidosis, should be

Fig. 5.3.26a,b  60-year-old man with asymmetrical, apical hypertrophic cardiomyopathy. Cine-MR images in a four-chamber and two-chamber view in diastole (a) and systole (b) show a markedly thickened left ventricular myocardium predominantly

of the apex, as compared with the basal segments. RV right ventricle, LV left ventricle (Courtesy of Dr. Armin Huber, Department of Clinical Radiology, Ludwig-Maximilians-University of Munich)

5.3  Heart Table 5.3.8  Differential diagnosis of restrictive cardiomyopathy (RCM) and constrictive pericarditis (CP) RCM

CP

Pericardium

Normal thickness

Thickened (>4mm), may be calcified

Atria

Dilated

Normal or dilated

Ventricles

Small or normal sized

Normal sized

Preserved systolic function Impaired diastolic function Myocardial mass and wall   thickness may be increased Septum Late enhancement

Normal ventricular wall thickness Paradox septal movement in diastole

Patchy myocardial uptake possible

observed. Although the wall thickness of amyloid heart disease may be suggestive of HCM, the LV ejection fraction and wall motion are typically normal or even higher than normal in HCM patients, whereas they decline in amyloidosis. The latter disorder leads most commonly to restrictive cardiomyopathy with dilated atria and thickening of the atrial walls and atrioventricular valves (see Sect. 5.3.4. for more information on secondary cardiomyopathies such as amyloidosis and sarcoidosis). Finally MR is a useful tool for non-invasive assessment of the morphologic and functional changes in right ventricular outflow obstruction, occasionally reported in connection with HCM (Katoh et al. 2001). 5.3.3.4  Restrictive Cardiomyopathy 5.3.3.4.1  Restrictive Cardiomyopathy: Etiology, Histology, and Clinical Presentation Restrictive cardiomyopathy (RCM) is a rare condition and the least common of primary cardiomyopathies. The key common features are impaired diastolic function with restricted ventricular filling and reduction in diastolic volume (Richardson et al. 1996). Also, dilated atria and dilatation of the inferior and superior vena cava are typical. RCM may be idiopathic or associated with other infiltrative diseases such as amyloidosis, hemochromatosis, or sarcoidosis (Kushwaha et al. 1997). Pathologically, RCM can be classified as obliterative and non-obliterative (or idiopathic). Obliterative RCM may result from an intracavitary thrombus filling the left ventricular apex

No late enhancement effects of the myocardium

and hampers the filling of the ventricles, e.g., in the end stage of eosinophilic syndromes. Non-obliterative, or idiopathic RCM, is characterized by progressive fibrosis of the myocardium but no thrombus formation (Hancock 2004). Restrictive cardiomyopathies cause symptoms and signs of left and/or right-side failure because they affect both ventricles. Some patients may have complete heart block due to fibrosis encasing the sinoatrial or atrioventricular nodes. The course of the disease varies depending on the pathology, treatment is often unsatisfactory, and even in advanced cases a heart transplant is necessary. The main diagnostic challenge is to differentiate this condition from constrictive pericarditis (CP) (Table 5.3.8). CP also presents with restrictive physiology but might be cured surgically (Schoenfeld et al. 1987). 5.3.3.4.2  Restrictive Cardiomyopathy: MR Protocol and Findings Using a combination of cines, T2-weighted spin-echo sequences, and post contrast T1-weighted images, CMR has good diagnostic accuracy for assessment of RCM. T2weighted images are particularly helpful with this form of cardiomyopathy because of the edema associated with inflammatory and granulomatous lesions. The features on MR images caused by impaired filling of the ventricles are common to both RCM and CP and include small or normal size of the ventricles and dilation of the atria as well as of the superior and inferior vena cava and hepatic veins, as described above. The normal pericardium, seen as a hypo-intense line less than 4 mm thick, is very clearly visible on spin-echo T1-weighted sequences. The diag-

739

740

5  Thorax and Vasculature

nostic accuracy of the existence of pericardial thicknesses of more than 4 mm on MR images has been shown to be >90% in constrictive pericarditis (Masui et al. 1992). In CP, thickening of the pericardium is most frequently localized on the free wall of the RV, but focal pericardial thickening may be observed. Phase-contrast cine MR imaging can also be used to quantify and monitor the restrictive filling pattern of the ventricles during therapy in RCM patients by measuring the diastolic flow across the mitral and tricuspid valves (Hartiala et al. 1993). 5.3.3.5  Arrhythmogenic Right Ventricular Dysplasia 5.3.3.5.1  Arrhythmogenic Right Ventricular Dysplasia: Etiology, Histology, and Clinical Presentation Arrhythmogenic right ventricular dysplasia (ARVD) must be considered as a part of the idiopathic cardiomyopathies, based on its nature of progressive heart muscle disease with unclear pathogenesis and etiology. ARVD is characterized histologically by fatty or fibro-fatty infiltration of the right ventricular myocardium (Tandri et al. 2004a; Basso et al. 1996). There is gradual replacement of myocytes by adipose and fibrous tissue, defining two morphologic predominant variants of ARVD: fatty and fibro-fatty ARVD (Corrado et al. 2000). The fatty form is characterized by almost complete replacement of the myocardium, typically without thinning of the ventricular wall, and it occurs exclusively in the right ventricle. The fibro-fatty variant is associated with significant thinning of the right ventricular wall, and the left ventricular myocardial wall may also be involved. Morphological alterations usually begin in the subepicardium and progress to the endocardium with subsequent thinning of the wall. Generally, for both forms, the regions most frequently involved are the RV inflow area, the apex, and the infundibulum. However, it is known from autopsy studies of hearts from individuals with no history of heart disorders, >50% of subjects had fat within their RV myocardial fibers, and the presence of intramyocardial fat increased with age. Consequently histological diagnosis of ARVD may be difficult in borderline cases (Fontaine et al. 1998). The exact prevalence of ARVD is not known, but is estimated to be around 1 in 5000 (Tandri et al. 2004b). The male-to-female ratio is 2.7 : 1 (Basso et al. 1996). A familial predilection has been identified with an autosomal dominant pattern with variable penetrance and incomplete expression. Several genetic disorders responsible for ARVD have been identified on chromosome 14 and recently on chromosome 3 (Rampazzo et al. 2003). Clinically, patients present with ventricular arrhythmias with left bundle branch block (LBBB) that may even lead to cardiac arrest and sudden cardiac death (Pinamonti et

al. 2000). ARVD is recognized as a major cause of sudden death in adult young men. The mechanism of sudden death is, in most cases, induction of ventricular tachycardia (VT), with degeneration into ventricular fibrillation. The islands of fibro-fatty tissue appear to generate macro re-entry electrical circuits and form the arrhythmogenic substrate for the onset of VT. Adrenergic stimulation such as physical exercise and catecholamine infusion appear to induce the arrhythmia (Thiene et al. 1988). The annual mortality rate of ARVD has been estimated as 3% without treatment and as 1% with pharmacological medical treatment. This is significantly reduced by the use of implantable cardioverter–defibrillators (ICDs). Therefore, an early and accurate diagnosis of this entity followed by appropriate therapy is of great importance. The definitive diagnosis of ARVC requires the histological finding of transmural fibro-fatty replacement of RV myocardium at necropsy, surgery, or endomyocardial biopsy. The latter is problematic, since the usual area sampled is the intraventricular septum, which is rarely affected. According to the 1994 Task Force Report on ARVD by McKenna et al. (McKenna et ala. 1994), the diagnosis of ARVD is based on the presence of structural, histological, electrocardiographical, and genetic factors, as well as presence of arrhythmia (Table 5.3.9). Detection of functional and structural alterations is performed with echocardiography, conventional angiography, MR imaging, or radionuclide angiography. To fulfill the appropriate criteria for ARVD, patients must have two major criteria, one major and two minor criteria, or four minor criteria. 5.3.3.5.2  Arrhythmogenic Right Ventricular Dysplasia: MR Protocol and Findings MRI offers several advantages in the diagnosis of ARVD and has helped to improve the diagnostic sensitivity and opportunity for early intervention (Sen-Chowdhry et al. 2005). The main diagnostic features of ARVD detected on MRI are quantitative measures of RV volumes, wall motion, and morphological abnormalities. When the left ventricle (LV) is affected (~15%), there is a mild decrease in LV function, although left-sided heart failure is unusual (Sen-Chowdhry et al. 2005). Cine CMR provides excellent contrast between the blood pool and the myocardial wall, and therefore can provide good quality information about regional RV wall motion and global RV function (Kayser et al. 2003). Impaired regional RV function is assessed in both the short-axis and transverse planes. Aneurysmal changes are also best appreciated with SSFP cines. Wall thinning and fibro-fatty replacement of myocardium is best visualized by T1-weighted spin-echo. Because fat infiltration of the RV has been described in a significant proportion of normal hearts in elderly patients, functional information is important in establishing the diagnosis of ARVD Kayser et al. 2003).

5.3  Heart Table 5.3.9  Diagnosis of ARVD: overview Factor

Major diagnostic criteria

Minor diagnostic criteria

RV dysfunction

Severely reduced RV ejection fraction

Mildly reduced RV ejection fraction,   regional RV hypokinesia

RV structural alterations

Severe global or segmental dilatation of the RV, localized RV aneurysms

Mild global or segmental   dilatation of RV

Tissue characterization

Fibrofatty replacement of RV myocardium at biopsy

Repolarization abnormalities Depolarization abnormalities

Inverted T waves (V2–V3) Epsilon waves or   QRS complex >110 ms in V1–V3

Arrhythmias Family History

Late potentials VT with LBBB Frequent VES

ARVD confirmed by surgery   or at necropsy

Family history of premature sudden   death due to suspected ARVD

RV right ventricle, VET ventricular extra-systole, LBBB left branch bundle block (Modified from McKenna et al. 1994)

Recapitulating, the typical criteria that can be demonstrated with MR imaging are the following (modified from Kayser et al. 2002) (see also: Table 5.3.10): 1 Fatty infiltration of the right ventricular myocardium with high signal intensity on T1-weighted images (Fig. 5.3.27) (major criterion) 2 Fibro-fatty replacement, which leads to diffuse thinning of the right ventricular myocardium (major criterion) 3 Aneurysms of the right ventricle and right ventricular outflow tract (major criterion) 4 Dilatation of the right ventricle and right ventricular out-flow tract (when severe, major criterion; when mild, minor criterion) 5 Regional contraction abnormalities (minor criterion)

6 Global systolic dysfunction (major criterion) and global diastolic dysfunction (minor criterion) In summary, MR imaging is useful for evaluating not only fatty replacement of the right ventricular myocardium, but also global and regional functional abnormalities of the right ventricle and right ventricular outflow tract. The demonstration of right ventricular abnormalities should be considered in the entire clinical context. Regarding potential limitations of CMR, The RV free wall is thin (normal thickness about 3 mm), resulting in limited ability to adequately quantify the RV thickness, in addition to the normal presence of epicardial and pericardial fat, which can cause some difficulty identifying intramyocardial fat (Kies et al. 2006).

Table 5.3.10  Diagnosis of ARVD: MR imaging features Major diagnostic criteria

Minor diagnostic criteria

Right myocardial high signal intensity on T1-weighted images

Mild RVOT dilatation

Diffuse myocardial thinning of the RV

Mild RV dilatation

Severe dilatation of the RV or the RVOT

Regional contraction abnormalities

Localized aneurysms of the RV and RVOT

Global diastolic dysfunction

Global systolic dysfunction

Prominent trabeculations

RV free-wall systolic bulging RV right ventricle, RVOT right ventricular outflow tract (Modified from Kayser et al. 2002)

741

742

5  Thorax and Vasculature 2.

3.

4.

5. Fig. 5.3.27  54-year-old male patient with arrhythmogenic right ventricular dysplasia (ARVD). Short-axis T1-weighted blackblood spin-echo images show extensive transmural fatty replacement of the right ventricular myocardium (RV) (arrows), which is a major criterion for the diagnosis of ARVD. LV left ventricle. (Courtesy of Dr. Armin Huber, Department of Clinical Radiology, Ludwig-Maximilians-University of Munich)

5.3.3.6  Future Considerations and Conclusions Concluding, primary cardiomyopathies are a complex group of heart diseases with characteristic morphologic and functional features. The detailed assessment of cardiac changes in patients suffering from primary cardiomyopathies can be comprehensively established by a single cardiac MR study. CMR is accurate and reproducible, assessing cardiac function, cardiac mass, morphology, fibrotic changes, and infiltrative processes. CMR can also assess ventricular remodeling, viability, and perfusion. Furthermore, CMR can be useful to monitor therapeutic response. In the future, CMR will focus on the role of myocardial velocity mapping, tagging, and ventricular torsion to appraise systolic and diastolic function in more detail. The increasing use of MR as part of the clinical routine for patients suspected to have primary cardiomyopathies will improve diagnosis and treatment, and may potentially enhance our understanding of these diseases. References 1.

Arad M, Penas-Lado M, Monserrat L, Maron BJ, Sherrid M, Ho CY, Barr S, Karim A, Olson TM, Kamisago M, Seidman JG, Seidman CE (2005) Gene mutations in apical hypertrophic cardiomyopathy. Circulation 112:2805–2811

6.

7.

8. 9.

10.

11.

12.

13.

Basso C, Thiene G, Corrado D, Angelini A, Nava A, Valente M (1996) Arrhythmogenic right ventricular cardiomyopathy. Dysplasia, dystrophy, or myocarditis? Circulation 94:983–991 Bellenger NG, Burgess MI, Ray SG, Lahiri A, Coats AJ, Cleland JG, Pennell DJ (2000) Comparison of left ventricular ejection fraction and volumes in heart failure by echocardiography, radionuclide ventriculography and cardiovascular magnetic resonance; are they interchangeable? Eur Heart J 21:1387–1396 Bluemke DA, Krupinski EA, Ovitt T, Gear K, Unger E, Axel L, Boxt LM, Casolo G, Ferrari VA, Funaki B, Globits S, Higgins CB, Julsrud P, Lipton M, Mawson J, Nygren A, Pennell DJ, Stillman A, White RD, Wichter T, Marcus F (2003) MR Imaging of arrhythmogenic right ventricular cardiomyopathy: morphologic findings and interobserver reliability. Cardiology 99:153–162 Brown CA, O’Connell JB (1995) Myocarditis and idiopathic dilated cardiomyopathy. Am J Med 99:309–314 Choudhury L, Mahrholdt H, Wagner A, Choi KM, Elliott MD, Klocke FJ, Bonow RO, Judd RM, Kim RJ (2002) Myocardial scarring in asymptomatic or mildly symptomatic patients with hypertrophic cardiomyopathy. J Am Coll Cardiol 40:2156–2164 Corrado D, Basso C, Thiene G (2000) Arrhythmogenic right ventricular cardiomyopathy: diagnosis, prognosis, and treatment. Heart 83:588–595 Dec GW, Fuster V (1994) Idiopathic dilated cardiomyopathy. N Engl J Med 331:1564–1575 Devlin AM, Moore NR, Ostman-Smith I (1999) A comparison of MRI and echocardiography in hypertrophic cardiomyopathy. Br J Radiol 72:258–264 Didier D, Ratib O, Lerch R, Friedli B (2000) Detection and quantification of valvular heart disease with dynamic cardiac MR imaging. Radiographics 20:1279–1299 Doherty NE 3rd, Seelos KC, Suzuki J, Caputo GR, O’Sul­ livan M, Sobol SM, Cavero P, Chatterjee K, Parmley WW, Higgins CB (1992a) Application of cine nuclear magnetic resonance imaging for sequential evaluation of response to angiotensin-converting enzyme inhibitor therapy in dilated cardiomyopathy. J Am Coll Cardiol 19:1294–1302 Doherty NE, Seelos KC, Suzuki J, Caputo GR, O’Sullivan M, Sobol SM, Cavero P, Chatterjee K, Parmlej WW, Higgins CB (1992b) Application of cine nuclear magnetic resonance imaging for sequential evaluation of response to angiotenin-converting enzyme inhibitor therapy in dilatated cardiomyopathy. J Am Coll Cardiol 19:1294–1302 Ehlert FA, Cannom DS, Renfroe EG, Greene HL, Ledingham R, Mitchell LB, Anderson JL, Halperin BD, Herre JM, Luceri RM, Marinchak RA, Steinberg JS (2001) Comparison of dilated cardiomyopathy and coronary artery disease in patients with life-threatening ventricular arrhythmias: Differences in presentation and outcome in the AVID registry. Am Heart J 142:816–822

5.3  Heart 14. Ennis DB, Epstein FH, Kellman P, Fananapazir L, McVeigh ER, Arai AE (2003) Assessment of regional systolic and diastolic dysfunction in familial hypertrophic cardiomyopathy using MR tagging. Magn Reson Med 50:638–642 15. Fontaine G, Fontaliran F, Frank R (1998) Arrhythmogenic right ventricular cardiomyopathies: clinical forms and main differential diagnoses. Circulation 97:1532–1535 16. Groenning BA, Nilsson JC, Sondergaard L, Fritz-Hansen T, Larsson HB, Hildebrandt PR (2000) Antiremodeling effects on the left ventricle during beta-blockade with metoprolol in the treatment of chronic heart failure. J Am Coll Cardiol 36:2072–2080 17. Grothues F, Smith GC, Moon JC, Bellenger NG, Collins P, Klein HU, Pennell DJ (2002) Comparison of interstudy reproducibility of cardiovascular magnetic resonance with two-dimensional echocardiography in normal subjects and in patients with heart failure or left ventricular hypertrophy. Am J Cardiol 90:29–34 18. Hancock EW (2004) A clearer view of effusive-constrictive pericarditis. N Engl J Med 350:435–437 19. Hartiala JJ, Mostbeck GH, Foster E, Fujita N, Dulce MC, Chazouilleres AF, Higgins CB (1993) Velocity-encoded cine MRI in the evaluation of left ventricular diastolic function: measurement of mitral valve and pulmonary vein flow velocities and flow volume across the mitral valve. Am Heart J 125:1054–1066 20. Katoh H, Murakami R, Shimada T (2001) Cine magnetic resonance imaging of isolated right ventricular outflow obstruction in hypertrophic cardiomyopathy. Clin Radiol 56:516–519 21. Kayser HW, van der Wall EE, Sivananthan MU, Plein S, Bloomer TN, De Roos A (2002) Diagnosis of arrhythmogenic right ventricular dysplasia: a review. Radiographics 22:639–648 22. Kayser HW, De Roos A, Schalij MJ, Bootsma M, Wellens HJ, van der Wall EE (2003) Usefulness of magnetic resonance imaging in diagnosis of arrhythmogenic right ventricular dysplasia and agreement with electrocardiographic criteria. Am J Cardiol 91:365–367 23. Kies P, Bootsma M, Bax J, Schalij MJ, van der Wall EE (2006) Arrhythmogenic right ventricular dysplasia/cardiomyopathy: screening, diagnosis, and treatment. Heart Rhythm 3:225–234 24. Kushwaha SS, Fallon JT, Fuster V (1997) Restrictive cardiomyopathy. N Engl J Med 336:267–276 25. Leeuw N de, Ruiter DJ, Balk AH, de Jonge N, Melchers WJ, Galama JM (2001) Histopathologic findings in explanted heart tissue from patients with end-stage idiopathic dilated cardiomyopathy. Transpl Int 14:299–306 26. Liberthson RR (1996) Sudden death from cardiac causes in children and young adults. N Engl J Med 334:1039–1044 27. Mandinov L, Eberli FR, Seiler C, Hess OM (2000) Diastolic heart failure. Cardiovasc Res 45:813–825 28. Maron BJ (2002 Hypertrophic cardiomyopathy: a systematic review. JAMA 287:1308–1320

29. Masui T, Finck S, Higgins CB (1992) Constrictive pericarditis and restrictive cardiomyopathy: evaluation with MR imaging. Radiology 182:369–373 30. McCrohon JA, Moon JC, Prasad SK, McKenna WJ, Lorenz CH, Coats AJ, Pennell DJ (2003) Differentiation of heart failure related to dilated cardiomyopathy and coronary artery disease using gadolinium-enhanced cardiovascular magnetic resonance. Circulation 108:54–59 31. McKenna WJ, Thiene G, Nava A, Fontaliran F, BlomstromLundqvist C, Fontaine G, Camerini F (1994) Diagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myocardial and Pericardial Disease of the European Society of Cardiology and of the Scientific Council on Cardiomyopathies of the International Society and Federation of Cardiology. Br Heart J 71:215–218 32. Miller S, Riessen R (2005) [MR imaging in cardiomyopathies]. Rofo 177:1497–1505 33. Moon JC, McKenna WJ, McCrohon JA, Elliott PM, Smith GC, Pennell DJ (2003) Toward clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance. J Am Coll Cardiol 41:1561–1567 34. Osterziel KJ, Strohm O, Schuler J, Friedrich M, Hanlein D, Willenbrock R, Anker SD, Poole-Wilson PA, Ranke MB, Dietz R (1998) Randomised, double-blind, placebo-controlled trial of human recombinant growth hormone in patients with chronic heart failure due to dilated cardiomyopathy. Lancet 351:1233–1237 35. Parga JR, Avila LF, Bacal F, Moreira LF, Stolf NG, Ramires JA, Bocchi EA (2001) Partial left ventriculectomy in severe idiopathic dilated cardiomyopathy: assessment of shortterm results and their impact on late survival by magnetic resonance imaging. J Magn Reson Imaging 13:781–786 36. Pinamonti B, Sinagra G, Camerini F (2000) Clinical relevance of right ventricular dysplasia/cardiomyopathy. Heart 83:9–11 37. Pons-Llado G, Carreras F, Borras X, Palmer J, Llauger J, Bayes DL (1997) Comparison of morphologic assessment of hypertrophic cardiomyopathy by magnetic resonance versus echocardiographic imaging. Am J Cardiol 79:1651–1656 38. Prasad SK, Assomull RG, Pennell DJ (2004) Recent developments in non-invasive cardiology. BMJ 329:1386–1389 39. Rampazzo A, Beffagna G, Nava A, Occhi G, Bauce B, Noiato M, Basso C, Frigo G, Thiene G, Towbin J, Danieli GA (2003) Arrhythmogenic right ventricular cardiomyopathy type 1 (ARVD1): confirmation of locus assignment and mutation screening of four candidate genes. Eur J Hum Genet 11:69–76 40. Redfield MM, Jacobsen SJ, Burnett JC Jr, Mahoney DW, Bailey KR, Rodeheffer RJ (2003) Burden of systolic and diastolic ventricular dysfunction in the community: appreciating the scope of the heart failure epidemic. JAMA 289:194–202

743

744

5  Thorax and Vasculature 41. Richardson P, McKenna W, Bristow M, Maisch B, Mautner B, O’Connell J, Olsen E, Thiene G, Goodwin J, Gyarfas I, Martin I, Nordet P (1996) Report of the 1995 World Health Organization/International Society and Federation of Cardiology task Force on the definition and classification of cardiomyopathies. Circulation 93:841–842 42. Rossi A, Cicoira M, Zanolla L, Sandrini R, Golia G, Zardini P, Enriquez-Sarano M (2002) Determinants and prognostic value of left atrial volume in patients with dilated cardiomyopathy. J Am Coll Cardiol 40:1425 43. Scheffold T, Binner P, Erdmann J, Schunkert H (2005) [Hypertrophic cardiomyopathy]. Herz 30:550–557 44. Schoenfeld MH, Supple EW, Dec GW Jr, Fallon JT, Palacios IF (1987) Restrictive cardiomyopathy versus constrictive pericarditis: role of endomyocardial biopsy in avoiding unnecessary thoracotomy. Circulation 75:1012–1017 45. Schulz-Menger J, Strohm O, Waigand J, Uhlich F, Dietz R, Friedrich MG (2000) The value of magnetic resonance imaging of the left ventricular outflow tract in patients with hypertrophic obstructive cardiomyopathy after septal artery embolization. Circulation 101:1764–1766 46. Sen-Chowdhry S, Prasad SK, McKenna WJ (2005) Arrhythmogenic right ventricular cardiomyopathy with fibrofatty atrophy, myocardial oedema, and aneurysmal dilation. Heart 91:784 47. Soler R, Rodriguez E, Rodriguez JA, Perez ML, Penas M (1997) Magnetic resonance imaging of apical hypertrophic cardiomyopathy. J Thorac Imaging 12:221–225 48. Soler R, Rodriguez E, Marini M (1999) Left ventricular mass in hypertrophic cardiomyopathy: assessment by three-dimensional and geometric MR methods. J Comput Assist Tomogr 23:577–582 49. Soler R, Rodriguez E, Remuinan C, Bello MJ, Diaz A (2003) Magnetic resonance imaging of primary cardiomyopathies. J Comput Assist Tomogr 27:724–734 50. Strohm O, Schulz-Menger J, Pilz B, Osterziel KJ, Dietz R, Friedrich MG (2001) Measurement of left ventricular dimensions and function in patients with dilated cardiomyopathy. J Magn Reson Imaging 13:367–371 51. Takeda N (2003) Cardiomyopathy: molecular and immunological aspects (review). Int J Mol Med 11:13–16 52. Tandri H, Bomma C, Calkins H, Bluemke DA (2004a) Magnetic resonance and computed tomography imaging of arrhythmogenic right ventricular dysplasia. J Magn Reson Imaging 19:848–858 53. Tandri H, Friedrich MG, Calkins H, Bluemke DA (2004b) MRI of arrhythmogenic right ventricular cardiomyopathy/ dysplasia. J Cardiovasc Magn Reson 6:557–563 54. Thiene G, Nava A, Corrado D, Rossi L, Pennelli N (1988) Right ventricular cardiomyopathy and sudden death in young people. N Engl J Med 318:129–133 55. Waxman S, Eustace S, Hartnell GG (1994) Myocardial involvement in primary hemochromatosis demonstrated by magnetic resonance imaging. Am Heart J 128:1047–1049

56. White RD, Obuchowski NA, Gunawardena S, Lipchik EO, Lever HM, Van Dyke CW, Lytle BW (1996) Left ventricular outflow tract obstruction in hypertrophic cardiomyopathy: presurgical and postsurgical evaluation by computed tomography magnetic resonance imaging. Am J Card Imaging 10:1–13 57. White RD, Trohman RG, Flamm SD, VanDyke CW, Optican RJ, Sterba R, Obuchowski NA, Carlson MD, Tchou PJ (1998) Right ventricular arrhythmia in the absence of arrhythmogenic dysplasia: MR imaging of myocardial abnormalities. Radiology 207:743–751 58. Zwanenburg JJ, Gotte MJ, Marcus JT, Kuijer JP, Knaapen P, Heethaar RM, van Rossum AC (2005) Propagation of onset and peak time of myocardial shortening in time of myocardial shortening in ischemic versus nonischemic cardiomyopathy: assessment by magnetic resonance imaging myocardial tagging. J Am Coll Cardiol 46:2215–2222

5.3.4  Secondary Cardiomyopathies and Specific Heart Muscle Diseases A. Huber 5.3.4.1  Amyloidosis Amyloidosis includes several diseases characterized by the deposition of twisted β-pleated sheet fibrils (amyloid) formed as a result of the misfolding of various proteins produced in different pathological states. The diverse types of amyloidosis are classified by the composition of the amyloid fibril. AL amyloidosis (previously known as primary amyloidosis) is characterized by light-chain immunoglobulin produced by monoclonal plasma cells (Lachmann et al. 2002). It is the most common form of amyloidosis in the United States and the most severe form. The accumulation of amyloid destroys the tissue architecture. In addition, there seems to be a toxic effect from the light chains, which causes severe organ damage with involvement of the kidneys, the heart, the liver, and the peripheral nerves (Brenner et al. 2004). Familial amyloidosis is an uncommon, autosomal dominant disease with a high degree of penetrance. Most cases result from the production of an unstable variant of the serum protein transthyretin (ATTR). Many different point mutations have been identified for ATTR amyloidosis. This type of amyloidosis results in many cases in neurological or cardiac dysfunction or both (Lachmann et al. 2002; Puille et al. 2002). One common mutant transthyretin is found in about 4% of the African American population (Jacobson et al. 1997). Its penetrance is unknown. This mutation causes cardiomyopathy. It is supposed to be one of the most common types of cardiomyopathy in older adult African American patients. Senile systemic amyloidosis is caused by a wild-type transthyretin. The heart is nearly exclusively affected. This type of amyloidosis may

5.3  Heart

become more important with increasing prevalence because of the higher average age of the population. Reactive systemic amyloidosis is the result of an overproduction of a non-immunoglobulin protein AA. The heart is usually not affected. In AL amyloidosis, cardiac involvement is common and the cause of death in the majority of cases (Gertz and Rajkumar 2002). Effective treatments for AL amyloidosis exist, but are only effective when the treatment is started before the heart is affected. Recently, the use of high-dose melphalan with autologous stem cell transplantation has been promising with regression of clinical manifestations of AL amyloidosis (Comenzo and Gertz 2002; Comenzo et al. 1998; Gertz et al. 2000; Moreau et al. 1998). Cardiac involvement means a poor prognosis in most forms of amyloidosis. Amyloidosis of the heart also predicts poor tolerance to high-dose chemotherapy and stem cell transplantation (Dubrey et al. 1997). Cardiac biopsy can reliably diagnose cardiac amyloidosis. However, it is an invasive procedure with a risk of complications. In patients with cardiac amyloidosis sampling errors hamper the ability of monitoring the clinical course of the disease. Thus, non-invasive diagnostic techniques are of interest for staging cardiac amyloidosis to assess prognosis of the patients and individual treatment options. Echocardiography is considered as the first non-invasive imaging modality to diagnose cardiac amyloidosis. Typical abnormalities seen on echocardiography include a small left ventricular size, biventricular and atrial septal thickening, atrial enlargement, and, in advanced disease, a restrictive left ventricular filling pattern (Figs. 5.3.28, 5.3.29). Imaging of the structural changes of the myocardium, however, is often difficult with echocardiography, especially differentiation of different causes of hypertrophy of the myocardium. A combination of typical signs of ECG and echocardiography allow for identification of patients with cardiac amyloidosis with a high accuracy in selected patient populations. However, in patients with higher age and common other diseases such as hypertension, left ventricular hypertrophy is unspecific to the disease, which causes the cardiomyopathy. Typical echocardiographic findings of cardiac amyloid such as hypertrophied ventricles and restrictive filling pattern are late signs that are associated with a short interval of survival of about 6 months. Therefore, echocardiography has limited impact in guiding treatment decisions. Thus, there is a need for a non-invasive technique to detect earlier changes in the disease. Maceira et al. compared cardiac MRI examination of patients with systemic amyloidosis and patients with hypertension (Maceira et al. 2005). The MRI data did show some interesting features in cardiac amyloid patients. MRI allowed for detection of global and subendocardial gadolinium enhancement of the myocardium in patients with cardiac amyloidosis. Myocardial enhancement was associated with increased ventricular mass and impaired left ventricular systolic function. Amyloidosis patients

had a fast gadolinium clearance from the blood pool, faster than the clearance in hypertensive patients without amyloidosis. This resulted in a reduced T1 contrast between the myocardium and blood in amyloid patients. Until now, the mechanism of the altered gadolinium kinetics in amyloidosis remains unclear. A faster gado-

Fig. 5.3.28a,b  45-year-old female patient with signs of global heart failure and histologically proven amyloidosis of the heart. CINE images in short-axis orientation, acquired during end-diastole (a) and end-systole (b). The systolic left ventricular function is normal with an ejection fraction of 77%. The left ventricular mass is elevated. The thickness of the left ventricular wall is increased (arrow)

745

746

5  Thorax and Vasculature

Fig. 5.3.29  55-year-old male patient with amyloidosis. Phasesensitive inversion recovery trueFISP images, acquired 10 min after application of a double-dose extravascular contrast agent in short-axis orientation, show circular subendocardial enhancement (arrow)

linium washout from blood and myocardium among amyloid patients may represent gadolinium distribution into the total body amyloid load. Gadolinium-enhanced MRI techniques have proven clinical value in characterizing myocardial infarction and fibrosis in other cardiac disorders. However, the histological basis of gadolinium delayed enhancement remains unclear. Current evidence suggests that it can be related to a combination of delayed washout kinetics and an increased volume of distribution of gadolinium in the interstitial space of abnormal myocardium. This mechanism seems to cause gadolinium enhancement in myocardial infarction. However, regions of delayed enhancement have been reported to correlate with the increased collagen content of myocardial fibrosis in hypertrophic cardiomyopathy. Similar to previous pathological studies of cardiac amyloidosis, histological assessment of this patient indicated little local tissue reaction caused by amyloid infiltration with minimal myocardial fibrosis. However, in these matched regions extensive delayed enhancement was demonstrated. Up to 40% of the subendocardial volume was infiltrated by amyloid in the study of Maceira et al. This finding suggested that interstitial expansion from amyloid infiltration without interstitial fibrosis could also cause delayed enhancement detectable by MRI. MRI has the potential to be a useful method for serial quantification of the extent of myocardial infiltration in patients with cardiac amyloidosis. Several findings have been described in cardiac amyloidosis, which can be identified by MRI. Each of the findings is not specific for amyloidosis; however, the presence of a combination of findings makes the diagnosis likely, especially if there is evidence of amyloidosis in an extra cardiac location, confirmed by biopsy. The morphologic assessment of the heart shows homogeneous hypertrophy of the ventricular and atrial myocardium. The ventricular cavities have normal or reduced volume. A severe concentric hypertrophy of both ventricles is suggestive of amyloidosis, when atrial hypertension or valvular disease is absent and the ventricles have a normal size. The increase in myocardial wall thickness in cardiac

amyloidosis can look like hypertrophic cardiomyopathy. Asymmetric septal hypertrophy is found in the minority of cases, in about 15–50% of the cases of cardiac amyloidosis. A systolic anterior motion of the mitral valve is an unspecific sign that can also be found in patients with hypertrophic obstructive cardiomyopathies can be present. An early systolic closure of the aortic valve may be observed. Other morphological changes include thickening of the papillary muscles and valve leaflets. The atrial volumes are usually enlarged because of the diastolic dysfunction due to amyloid deposition. Pleural and pericardial effusions are commonly seen. Cardiac amyloid deposits generally cause severe cardiac dysfunction with a poor prognosis. Congestive heart failure is predominantly based on diastolic dysfunction. Systolic dysfunction is observed very late in patients with amyloidosis. In the early phase ejection fraction is normal or high. Severe congestive heart failure in patients with cardiac amyloidosis can occur despite a normal or mildly reduced left ventricular ejection fraction. 5.3.4.2  Sarcoidosis Sarcoidosis is a granulomatous disorder of unknown etiology. Although disease affects the hilar lymph nodes and the lungs most commonly, many parts of the body can be affected. Cardiac involvement is present at autopsy studies in up to 30%. Even higher incidences up to 50% are observed in some races, e.g., Japanese people (Silverman et al. 1978; Virmani et al. 1980). The pathological investigations show patchy infiltration of the myocardium with three successive histological stages: first, edema; second, non-caseating epithelioid granulomatous infiltration; and third, fibrosis, leading to post-inflammatory scarring (Fig. 5.3.30) (Vignaux et al. 2002). The clinical signs depend on the extent of inflammation and on the location of the disease. Cardiac symptoms include benign arrhythmias, heart block, intractable heart failure, chest pain, and congestive heart failure as well as fatal ventricular fibrillations (Sharma 2003). In contrast to the high incidence of cardiac involvement at autopsy studies, only 5% of the patients with cardiac sarcoidosis are symptomatic. In most patients the disease has a subclinical course; however, there is a clear increased risk of sudden cardiac death because of ventricular arrhythmias or conduction block, accounting for 30–65% of deaths in patients with cardiac sarcoidosis. Cardiac involvement can occur before, during, or after involvement of the lungs or other regions in the body. If cardiac disease occurs in a patient with multi-systemic sarcoidosis, the probability of cardiac sarcoidosis is high. However, in patients with cardiac dysfunction as the first or sole manifestation of sarcoidosis, the diagnosis can be difficult. Early diagnosis and effective treatment, e.g., treatment with corticosteroids or immunosuppressive and immune modulator therapy,

5.3  Heart

Fig. 5.3.30  48-year-old man with cardiac sarcoidosis and without coronary artery disease. Magnitude reconstructed image of an IR trueFISP pulse sequence in short-axis orientation and phase-sensitive reconstructed image of a PSIR trueFISP pulse sequence in a four-chamber-view. The contrast enhancement in

the septum (arrows) is not transmural and not located in the subendocardial region. Thus, the distribution of the enhancement is not typical for myocardial infarction. The pathological enhancement is caused by a scar and fibrosis after cardiac involvement of sarcoidosis

or automatic implantable modulator therapy, in patients at risk are absolutely necessary to improve the long-term prognosis. The diagnosis of myocardial sarcoidosis can be difficult. Because of the more-focal-than-diffuse involvement of cardiac sarcoidosis, endomyocardial biopsy yields low sensitivity values. Most non-invasive tests have also a low sensitivity and specificity. The findings of ECG and echocardiography usually do not allow establishment of the diagnosis of sarcoidosis of the heart. The role of radionuclide techniques, scintigraphy with 201Tl and 67Ga, remains unclear and is limited by poor spatial resolution (Sharma 2003). Focal lesions of the myocardium can be identified by reduced 201Tl uptake; the size of these areas may decrease or disappear entirely during stress, in what is called a reverse distribution defect. In the presence of healthy coronary arteries, perfusion defects on 201Tl scintigraphy in patients with sarcoidosis support the diagnosis of cardiac involvement of sarcoidosis. A normal finding of 201Tl scintigraphy, however, cannot exclude cardiac sarcoidosis. Recently, 13NH3/18F-FDG-PET has been used for the identification of cardiac involvement and for the assessment of disease activity (Yamagishi et al. 2003). Myocardial abnormalities were seen as areas of decreased 13NH3 uptake (usually in the basal anteroseptal wall), or areas of increased 18F-FDG uptake (usually basal-mid anteroseptal and lateral). After corticoid therapy, the areas of decreased 13N-NH3 uptake remained unchanged while the areas of increased 18F-FDG uptake diminished in size or completely disappeared. Promising results were reported in several studies about MRI in the early diagnosis of cardiac sarcoidosis and follow-up examinations of the disease including as-

sessment of response to therapy (Vignaux et al. 2002; Matsuki and Matsuo 2000; Shimada et al. 2001). The imaging protocol in patients with suspected cardiac sarcoidosis should consist of a combination of T1-weighted sequences before and after administration of contrast agent and IR gradient-echo or SSFP pulse sequences as known from imaging-delayed enhancement in patients with myocardial infarction, T2-weighted TSE sequences, and STIR sequences and CINE SSFP pulse sequences. This protocol allows for assessment of morphological and functional abnormalities by T1-weighted TSE images and cine MRI, for the detection of edema by STIR images and T2-weighted TSE images and for the depiction of necrosis, inflammation, and scar tissue by contrast-enhanced IR gradient-echo techniques. Vignaux et al. (2002a, b) described three different MRI patterns of myocardial sarcoidosis. The first pattern is a nodular pattern with peripheral high and central low signal intensity on T2weighted images and on contrast-enhanced T1-weighted images. This pattern represents hyaline fibrotic tissue. The second pattern shows high signal intensity on T1weighted images with or without thickening of the myocardium. This pattern is patchy or focal and represents inflammation. The third pattern is a post-inflammatory pattern. Areas with focal increased signal intensity on T2weighted images are found, which do not show contrast enhancement. These regions can reveal thinning of the myocardium. In more than 50% of the cardiac-asymptomatic patients with involvement of multiple organs, cardiac abnormalities were found similar to those observed in patients with cardiac symptoms (Vignaux et al. 2002b). Frequent enhancement of the anteroseptal and

747

748

5  Thorax and Vasculature

anterolateral left ventricular (LV) wall after gadolinium administration was reported (Shimada et al. 2001). The location of the abnormalities by MRI can help for planning biopsy of the myocardium. Therefore sensitivity of biopsy may be increased if the regions with involvement of sarcoidosis can be discriminated from normal regions of the myocardium. In addition, regional and global ventricular function can be assessed. A regression of enhancement can be observed early after therapy, whereas the high signal intensity on T2-weighted images seem to remain for a longer time, i.e., more than 6 months (Vignaux et al. 2004). If no response to therapy can be observed, then the changes, visible by MRI, remain stable or show a progression. MRI seems to be an important imaging modality for the detection and identification of cardiac involvement in patients with sarcoidosis. However, the findings are not specific. Edema and enhancement patterns can be clearly differentiated from myocardial infarction, as the involved areas are located in the myocardium or are patchy and not strictly found in the subendocardium or completely transmural as in myocardial infarction. The appearance of changes of the myocardium in patients with myocarditis can be very similar to the changes found in patients with sarcoidosis, and enhancement patterns can be similar to those found in patients with cardiomyopathy. 5.3.4.3  Iron-Overload Cardiomyopathy Iron-overload cardiomyopathy is defined as the presence of systolic or diastolic cardiac dysfunction secondary to increased deposition of iron in the heart. The cardiac dysfunction is not dependent on other diseases, such as atherosclerosis, ischemia, or valvular disease, that may be simultaneously detected (Liu et al. 1994). Primary or secondary hemochromatosis can be the reason for the development of iron-overload cardiomyopathy. Primary hemochromatosis is an autosomal recessive disease affecting one in 220 persons and usually presents late in life. Secondary hemochromatosis is a common cause of mortality in the young adult population in the developed world. This is the result of the large population with hereditary anemias such as thalassemia or sickle cell anemia. These patients require chronic red cell transfusion starting at a young age. However, one of the complications of this transfusion therapy is iron overload. Chelation therapy can help to prevent iron-induced heart failure. The thalassemias are common monogenic disorders of hemoglobin synthesis. β-Thalassemias are the most important among the thalassemia syndromes and have become a worldwide clinical problem due to an increasing immigrant population. In β-thalassemia major, regular blood transfusions are necessary early in life. β-Thalassemia intermedia refers to a less severe phenotype, whereas βthalassemia/hemoglobin E disease encompasses a broad

phenotypic spectrum. Blood transfusions and increased gastrointestinal iron absorption result in iron overload and tissue damage. Among patients with β-thalassemia major, biventricular, dilated cardiomyopathy remains the leading cause of mortality. In some patients, a restrictive type of left ventricular cardiomyopathy (Richardson et al. 1996) or pulmonary hypertension is noted. The clinical course, although variable and occasionally fulminant, is more benign in recent than in older series. Myocarditis has been described as a cause of left-sided heart failure in younger patients. Pulmonary arterial hypertension is the principal cause of heart failure in β-thalassemia intermedia. Chelation therapy has improved prognosis in β-thalassemia major both by reducing the incidence of heart failure and by reversing cardiomyopathy. Estimation of the patient’s cardiac risk is mainly based on clinical criteria and serial echocardiography. A new cardiovascular magnetic resonance technique will probably fulfill the need for more precise risk stratification in βthalassemia syndromes. By increasing the proportion of patients on optimal chelation, survival in β-thalassemia major may be further improved. Recent advances in gene therapy are expected to result in the long-awaited cure of this disease. A key feature of the clinical scenario is that iron-induced cardiomyopathy is reversible if intensive chelation treatment is instituted in time before the onset of overt LV cardiac failure, which carries a poor prognosis. Thus, early detection of myocardial iron overload is important. The need for reliable assessments of myocardial iron loading has recently led to the use of a cardiovascular magnetic resonance (CMR) technique that measures the myocardial relaxation parameter T2*, which has a lower limit of normal of 20 ms (Anderson et al. 2001). Myocardial iron overload causes low values of T2* because iron interferes with local magnetic field homogeneity. This technique has been adapted for acquisition in a single breath hold (Westwood et al. 2003a), and has very good reproducibility (Anderson et al. 2001; Westwood et al. 2003a, b). Using this technique, it has been shown that global systolic functional parameters, such as the ejection fraction, are relatively late markers for the detection of iron loading (Anderson et al. 2001). LV diastolic function may be more sensitive as an early marker of myocardial iron overload, and a number of techniques employed in clinical practice have been used to assess diastolic function in thalassemia major, including radionuclide ventriculography (Kukuk et al. 1999), Doppler echocardiography of trans-mitral flow, and most recently, tissue Doppler imaging (Vogel et al. 2003). It has also recently become possible to measure diastolic function using CMR. Westwood et al. (2005) investigated whether direct assessment of myocardial iron loading using T2* CMR would be more sensitive for detecting early myocardial iron loading than indices of diastolic function in a cohort of thalassemia major patients. They concluded

5.3  Heart

that although diastolic dysfunction occurs in myocardial iron overload associated with TM, the large overlap with the normal ranges means that a single measurement has low sensitivity for diagnosing early myocardial iron loading. Myocardial T2* MRI is therefore preferable because it offers better definition of the categorical lower limit of normality. 5.3.4.4  Endomyocardial Disease (Löffler’s Endocarditis and Endomyocardial Fibrosis) Endomyocardial disease is a common form of secondary restrictive cardiomyopathy and includes Löffler’s endocarditis and endomyocardial fibrosis (Parillo 1990). Both diseases were previously thought to be two variants of the same disease. However, these entities have a different clinical presentation and have a geographically different distribution. Löffler’s endocarditis occurs in temperate countries, has a more aggressive and rapidly progressive course, and is related to hypereosinophilia. Endomyocardial fibrosis, however, is more common in equatorial Africa without hypereosinophilia. The restrictive pattern is caused by an intensive endocardial fibrotic wall thickening of the apex and subvalvular regions of one or both ventricles, resulting in an inflow obstruction of blood (Fig. 5.56). The underlying pathogenesis of Löffler’s endocarditis is related to hypereosinophilia with a toxic effect of eosinophils on the heart. Necrosis, endomyocarditis, and formation of intramyocardial thrombi are observed (Weller and Bubley 1994). Later, extensive fibrosis replaces the inflammatory process and causes thickening of the ventricular wall. Atrioventricular valve regurgitation, as a result of progressive scarring of the chordae tendianae, may occur. Atrial enlargement may be observed. The increased stiffness of the ventricular myocardium and reduced volume of the ventricular cavity by organized thrombus cause a pathological diastolic flow pattern in the AV valves. MRI is helpful in the diagnosis of endomyocardial disease and in monitoring the effects of medical treatment (Huong et al. 1997). A.31reas of myocardial fibrosis appear as areas of low signal intensity on enhanced images and as areas of high signal intensity with enhancement after contrast material application. Ventricular volumes and atrioventricular regurgitation can be accurately quantified by MRI. Delayed enhancement imaging allows for differentiation of mural thrombus from thickened myocardium (Fig. 5.3.31). 5.3.4.5  Metabolic Storage Disease Metabolic storage diseases with cardiac involvement are type I or II glycogenosis, Anderson-Fabry, Gaucher and Niemann-Pick disease, galactosialidosis, and mucopoly-

Fig. 5.3.31a,b  38-year-old male patient with hypereosinophilia and endomyocarditis. a CINE images in standard orientations (four-, two-, and three-chamber-views), acquired with an SSFP pulse sequence, demonstrate left ventricular hypertrophy and a mass in the apex of the left ventricular cavity (arrows). The mass is caused by a thrombus in the left ventricular cavity. b Contrast-enhanced images, acquired with an IR trueFISP after contrast material application reveal thrombotic material, which appears dark (arrows), in the apical left ventricular cavity and subendocardial enhancement (arrowheads), which appears bright. The areas with enhancement represent areas of endomyocarditis

saccharidosis. The cardiac involvement in these diseases is characterized by LV wall thickening, valvular involvement, and LV diastolic dysfunction, with often-normal LV systolic function. The LV wall thickening may even mimic HCM with or without obstruction of the LV outflow tract. In clinical practice, echocardiography is the first method used for the determination of LV wall thickness, systolic function, anatomical derangement, valvular dysfunction, and LV diastolic function (Huong et al. 1997). MRI is superior in accurate determination of systolic ventricular function, determination of the ventricular mass and tissue characterization. CE IR MRI promises to develop into an important tool for further differentiation of changes in the myocardium in hypertrophic cardiomyopathies (Moon et al. 2003a). Moon et al. (2003b) investigated contrast enhancement in patients with Anderson-Fabry disease, which seems to be commonly located in the basal inferolateral LV wall. The enhancement pattern usually shows no transmural extent or involvement of the subendocardial regions as known from myocardial infarction. Areas of signal intensity increase are rather patchy and located in the middle parts of the ventricular myocardium. This X-linked disorder of sphingolipid metabolism causes an idiopathic LV hypertrophy. While it is generally believed that the deposition of glycosphingolipid is within the myocytes, the abnormal enhancement also suggests the presence of myocardial fibrosis.

749

750

5  Thorax and Vasculature

5.3.4.6  Friedreich’s Ataxia Friedreich’s ataxia is a neurodegenerative disease. Typical symptoms are progressive limb and gait ataxia, areflexia, and pyramidal signs in the legs. This autosomal recessive disease of commonly associated with HCM (concentric or asymmetric). The production of frataxin, a protein of unknown function, seems to cause the disease. An oxidative stress or respiratory chain dysfunction is triggered by the production of frataxin. Idebenone, a potent freeradical scavenger, can be used for medical treatment of the disease. It protects the myocardium against this oxidative stress and has therefore a potential for controlling cardiac hypertrophy in Friedreich’s ataxia (Hausse et al. 2003). LV mass calculations in this study were performed with echocardiography. MRI, however, has the advantage of being a more accurate method than echocardiography. MRI is highly reproducible concerning wall mass determinations and is recommended for future studies for controlling the effects of new medications for treatment of ventricular hypertrophy.

9.

10. 11.

12.

13.

14.

15.

References

16.

1.

17.

2.

3.

4.

5.

6.

7. 8.

Anderson LJ, Holden S, Davies B et al. (2001) Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J 22:2171–2179 Brenner DA, Jain M, Pimentel DR, Wang B, Connors LH, Skinner M, Apstein CS, Liao R (2004) Human amyloidogenic light chains directly impair cardiomyocyte function through an increase in cellular oxidant stress. Circ Res 94:1008–1010 Comenzo RL, Gertz MA (2002) Autologous stem cell transplantation for primary systemic amyloidosis. Blood 99:4276–4282 Comenzo RL, Vosburgh E, Falk RH, Sanchorawala V, Reisinger J, Dubrey S, Dember LM, Berk JL, Akpek G, LaValley M, O’Hara C, Arkin CF, Wright DG, Skinner M (1998) Dose-intensive melphalan with blood stem-cell support for the treatment of AL (amyloid light-chain) amyloidosis: survival and responses in 25 patients. Blood 91:3662–3670 Dubrey SW, Cha K, Skinner M, LaValley M, Falk RH (1997) Familial and primary (AL) cardiac amyloidosis: echocardiographically similar diseases with distinctly different clinical outcomes. Heart 78:74–82 Gertz MA, Lacy MQ, Gastineau DA, Inwards DJ, Chen MG, Tefferi A, Kyle RA, Litzow MR (2000) Blood stem cell transplantation as therapy for primary systemic amyloidosis (AL). Bone Marrow Transplant 26:963–969 Gertz MA, Rajkumar SV (2002) Primary systemic amyloidosis. Curr Treat Options Oncol. 3:261–271 Gillmore JD, Davies J, Iqbal A, Madhoo S, Russell NH, Hawkins PN (2002) Allogeneic bone marrow transplantation for systemic AL amyloidosis. Br J Haematol 100:226 –228

18.

19.

20. 21.

22.

23. 24.

Hausse AO, Aggoun Y, Bonnet D et al. (2002) Idebenone and reduced cardiac hypertrophy in Friedreich’s Ataxia. Heart 87:346–349 Huong DL, Wechsler B, Papo T et al. (1997) Endomyocardial fibrosis in Behçet’s disease. Ann Rheum Dis 56:205–208 Jacobson DR, Pastore RD, Yaghoubian R, Kane I, Gallo G, Buck FS, Buxbaum JN (1997) Variant-sequence transthyretin (isoleucine 122) in late-onset cardiac amyloidosis in black Americans. N Engl J Med 336:466–473 Kucuk NO, Aras G, Sipahi T et al. (1999) Evaluation of cardiac functions in patients with thalassemia major. Ann Nucl Med 13:175–179 Lachmann HJ, Booth DR, Booth SE, Bybee A, Gilbertson JA, Gillmore JD, Pepys MB, Hawkins PN (2002) Misdiagnosis of hereditary amyloidosis as AL (primary) amyloidosis. N Engl J Med 346:1786–1817 Liu P, Olivieri N (1994) Iron overload cardiomyopathies: new insights into an old disease. Cardiovasc Drugs Ther 198:101–110 Maceira AM, Joshi J, Prasad SK, Moon JC, Perugini E, Harding I, Sheppard MN, Poole-Wilson PA, Hawkins PN, Pennell DJ (2005) Cardiovascular magnetic resonance in cardiac amyloidosis. Circulation 111:186–193 Matsuki M, Matsuo M (2000) MR findings of myocardial sarcoidosis. Clin Radiol 55:323–325 Moon JCC, McKenny WJ, Mc Crohon JA et al. (2003a) Toward clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance. J Am Coll Cardiol 41:1561–1567 Moon JCC, Mundy HR, Lee PJ, Mohiaddin RH, Pennell DJ (2003b) Myocardial fibrosis in glycogen storage disease type III. Circulation 107:47 Moreau P, Leblond V, Bourquelot P, Facon T, Huynh A, Caillot D, Hermine O, Attal M, Hamidou M, Nedellec G, Ferrant A, Audhuy B, Bataille R, Milpied N, Harousseau JL (1998) Prognostic factors for survival and response after high-dose therapy and autologous stem cell transplantation in systemic AL amyloidosis: a report on 21 patients. Br J Haematol 101:766–769 Parrillo JE (1990) Heart disease and the eosinopil. N Engl J Med 323:1560–1561 Puille M, Altland K, Linke RP, Steen-Muller MK, Kiett R, Steiner D, Bauer R (2002) 99mTc-DPD scintigraphy in transthyretin-related familial amyloidotic polyneuropathy. Eur J Nucl Med Mol Imaging 29:376–379 Richardson P, McKenna W, Bristow M et al. (1996) Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the Definition and Classification of Cardiomyopathies. Circulation 93:841–842 Sharma OP (2003) Diagnosis of cardiac sarcoidosis. An imperfect science, a hesitant art. Chest 123:18–19 Shimada T, Shimada K, Sakane T et al. (2001) Diagnosis of cardiac sarcoidosis and evaluation of the effects of steroid therapy by gadolinium-DTPA-enhanced magnetic resonance imaging. Am J Med 110:510–527

5.3  Heart 25. Silverman KJ, Hutchins GM, Bulkley BH (1978) Cardiac sarcoid: a clinicopathologic study of 84 unselected patients with systemic sarcoidosis. Circulation 58:1204–1211 26. Vignaux O, Dhote R, Duboc D et al. (2002) Detection of myocardial involvement in patients with sarcoidosis applying T2-weighted, contrast-enhanced, and cine magnetic resonance imaging : initial results of a prospective study. J Comput Assist Tomogr 26:762–767 27. Vignaux O, Dhote R, Blance P et al. (2004) Myocardial MRI in sarcoidosis: 3-year follow-up and evaluation of the effects of steroid therapy. J Cardiovasc Magn Reson 6(Abstract):44 28. Vignaux O, Dhote R, Blanche P et al. (2002) Clinical significance of myocardial magnetic resonance abnormalities in patients with sarcoidosis. A 1-year follow-up study. Chest 122:1895–1901 29. Virmani R, Bures JC, Roberts WC (1980) Cardiac sarcoidosis: a major cause of sudden death in young individuals. Chest 77:423–428 30. Vogel M, Anderson LJ, Holden S, Deanfield JE, Pennell DJ, Walker JM (2003) Tissue Doppler echocardiography in patients with thalassaemia detects early myocardial dysfunction related to myocardial iron overload. Eur Heart J 24:113–119 31. Weller PF, Bubley GJ (1994) The idopatic hypereosinopilic syndrome. Blood 83:2759–2779 32. Westwood MA, Anderson LJ, Firmin DN et al. (2003a) A single breath-hold multiecho T2* magnetic resonance technique for diagnosis of myocardial iron overload. J Magn Reson Imaging 18:33–39 33. Westwood MA, Anderson LJ, Firmin DN et al. (2003b) Interscanner reproducibility of cardiovascular magnetic resonance in the early diagnosis of myocardial iron overload. J Magn Reson Imaging 18:616–620 34. Westwood MA, Wonke B, Maceira AM et al. (2005) Left ventricular diastolic function compared with T2* cardiovascular magnetic resonance for early detection of myocardial iron overload in thalassemia major. J Magn Reson Imaging 22:229–232 35. Yamagishi H, Shirai N, Takagi M et al. (2003) Identification of cardiac sarcoidosis with 13N-NH3/18F-FDG PET. J Nucl Med 44:1030–1036

echocardiography and cardiac catheterization, MRI can be helpful to assess valve morphology, the degree of regurgitation or stenosis. In addition, MRI is known as the gold standard for the assessment of ventricular function. In cases of regurgitation of the mitral or the aortic valves, it can be very important to determine ejection fraction and ventricular volumes accurately for the indication of surgical valve replacement. In contrast to echocardiography, MRI allows for accurate assessment of the fraction of regurgitation, whereas grading of regurgitation in echocardiography is based on pressure gradients and visual criteria. In contrast to echocardiography, MRI has the unique advantage that it allows the investigator to choose the planes of the cine slices or flow measurements without any restrictions. However, echocardiography is limited to certain planes that can be visualized transthoracically. With development of stronger gradients, improved surface coils and introduction of parallel imaging, the spatial and temporal resolution of MRI could be improved. Furthermore, the acquisition time is shorter than years ago. However, the methods and principles of heart valve assessment have remained the same over the last 5 years. When investigating cardiac valves, four categories of information must be gathered: (1) the pathological findings of auscultation have to be clarified and the diseased valve must be identified; (2) The valvular anatomy including the number of valve leaflets, the thickness of the leaflets and the evidence or absence of endocarditis have to be assessed; (3) The degree of valvular stenosis or regurgitation has to be quantified; (4) The effect of valvular dysfunction on the ventricular function including size of the ventricles, ejection fraction and hypertrophy has to be determined. In the past, if all of this information could not be gathered from echocardiography, it was combined with X-ray angiography. MRI can now provide most of the desired information in a single investigation that is safe and noninvasive, without exposure to X-rays. In this section, an overview of currently available MRI techniques is presented, which can be used for the assessment of valvular heart disease.

5.3.5  Valvular Heart Diseases

5.3.5.2.1  Transthoracic Echocardiography

A. Huber 5.3.5.1  Introduction MRI is usually not considered to be the first imaging modality for the assessment of heart valves. In most cases transthoracic echocardiography or transesophageal echocardiography is sufficient. However, in cases with low image quality of echocardiography, for example in patients with adipositas or emphysema, or discrepant findings of

5.3.5.2  Conventional Imaging Modalities

Transthoracic echocardiography remains the most important and widely available investigation for the assessment of valvular heart disease (Carabello et al. 1997). The technique is noninvasive, safe, and accurate at localizing the disease valve. Quantification of valvular stenosis and valve area can be performed by different flow-dependent methods. However, quantification of valve regurgitation is less accurate and somewhat more dependent on the investigator and his subjective assessment of visual signs (Smith and Kie 1998). A semi-

751

752

5  Thorax and Vasculature

quantitative assessment can be achieved by measuring the regurgitant jet length and width, but there is poor correlation to X-ray angiographic measurements. Several echocardiography methods have been developed in order to quantify valvular regurgitation, but all of them have their limitations. The regurgitant fraction can be determined, but the procedure is time-consuming and has some limitations. The accuracy in determining the ventricular function is inferior to that of MRI, which is the established gold standard for the assessment of systolic ventricular function. Imaging of the proximal iso-velocity surface area has been proposed for the assessment of the degree of regurgitation by echocardiography. However, this technique relies on the assumption that the regurgitant jet orifice is both flat and circular, which is not the case for the majority of patients. Another limitation of transthoracic echocardiography is that the imaging planes are strongly dependent on good acoustic windows. 5.3.5.2.2  Transesophageal Echocardiography Transesophageal echocardiography is superior compared to transthoracic echocardiography in the assessment of atrial thrombus (Manning et al. 1993) and atrial septal wall defects. The technique can clarify valvular anatomy in patients with endocarditis or in patients with poor acoustic windows. With regard to the assessment of valvular regurgitation, transesophageal echocardiography has all the limitations of transthoracic echocardiography. A unique use of transesophageal echocardiography is intraoperative assessment of valvular function. 5.3.5.2.3  X-Ray Angiography X-ray angiography has been considered the gold standard for the assessment of valvular heart disease (Carroll 1993). Valvular stenosis can be quantified by calculating the transvalvular gradients and valve areas, using the Gorlin formula (Gorlin and Gorlin 1951). The grading system used for valvular regurgitation, however, can be regarded as very accurate or precise. The invasive examination is associated with a low risk of severe complications, up to 0.1%, like myocardial infarction, arterial embolization, thrombosis or dissection and even death. Today, noninvasive imaging modalities should be used before an invasive examination of the heart valves is performed. X-ray angiography should be used, if the pulmonary pressure or the coronary arteries have to be assessed or if there are discrepancies between the clinical symptoms and the results of noninvasive examinations.

5.3.5.3  MRI of Cardiac Valves 5.3.5.3.1  MR Acquisition Techniques for Cardiac-Valve Imaging Cine MR Imaging All images presented were acquired with a 1.5-T MRI system. An examination of the beating heart with MRI requires prospective ECG triggering or retrospective ECG gating, which allow image acquisition at various times in the same position during the cardiac cycle. cine MR imaging is a dynamic imaging modality, which was performed for a long time with gradient-echo pulse sequences. Recently, the gradient-echo readout technique was replaced by SSFP readout techniques, as the latter improve the contrast between blood and myocardium and reduce flow-dependent artifacts. Moreover, the contrast of cine SSFP techniques is not based on the signal of inflowing unsaturated blood, as it is for the cine gradient-echo techniques, but on the intrinsic high signal intensity of blood (Miller et al. 2002; Wintersperger et al. 2003). cine images should be acquired with a temporal resolution with a frame length below 50 ms. Typical values for the echo time and repetition time are 1.1–1.5 ms and 2.2–3 ms. Between 15 and 20 frames are usually necessary to cover the entire cardiac cycle. The images can be visualized in a cine loop, allowing a dynamic assessment. The last image in the series must be obtained as close as possible to the next R wave to encompass the greatest possible portion of the R-R interval and of the entire cardiac cycle. The cine techniques allow for accurate assessment of ventricular function based on the ejection fraction of the myocardial mass and the end-systolic and end-diastolic volumes. The global functional parameters can be normalized using the body surface area as well as age- and gender-related reference values. Abnormal flow patterns, which are observed in valvular disease, cause dephasing of the spins within a voxel and result in signal loss (flow void). These flow voids allow for direct visualization of valvular stenosis or regurgitation caused by high-velocity flow and turbulence (Didier et al. 1993). Its appearance depends on technical factors including display parameters (window width and level), flip angle, time of repetition (TR), and echo time (TE). With longer TE and TR values, small flow voids with lower velocity are demonstrated better with longer TE and TR values than with short ones. The gradient-echo readout techniques show flow voids in mild valvular disease with higher sensitivity than the SSFP readout techniques; however, for more severe valvular disease, the SSFP technique seems to be superior for a realistic visualization of the flow jet. These variables must be taken into account when evaluating flow anomalies. In general, the visual evaluation of the flow jets is inferior in grading of valvular stenosis and regurgitation, as the appearance of the flow jet is dependent on many facts and can be variable. Flow voids

5.3  Heart

with dark signal intensity are found in the large vessels or cardiac chambers behind the valves with stenosis. In valves with regurgitation the flow voids can be seen in the cardiac chamber, which is located proximal to the valve. Flow-Sensitive Imaging Flow-sensitive imaging techniques can be used to measure blood flow in vessels or in heart valves. Either velocity or volume per time unit can be determined. Flow-sensitive cine MR imaging techniques based on phase-contrast MR imaging or velocity-encoded (VEC) MR imaging are widely available. The method uses the phase differences of flowing spins compared to stationary spins along a magnetic gradient. There is a direct linear relationship to flow velocity. Two opposite gradient lobes are applied. Their shape and temporal distance define the maximum velocity that can be measured with a phase difference of –180° to +180°. An aliasing artifact can be observed if the real velocity in a vessel or valve is higher than the maximum, which can be encoded with the previously defined pulse sequence parameters. If an aliasing artifact can be identified, then the pulse sequence has to be repeated with parameters suitable for a higher maximum velocity. The phase-contrast technique allows quantification of blood velocity profiles at different times during the cardiac cycle (Mostbeck et al. 1992; Nayler et al. 1986; Bogren and Buonocore 1994). VEC MR imaging is based on the acquisition of two sets of images—one with and one without velocity encoding—that are usually acquired simultaneously. The subtraction of the two sets of images allows the calculation of a phase shift, which is proportional to flow velocity in the direction of the flow-compensation gradient. Magnitude images can be reconstructed to provide anatomic information, and phase images can be reconstructed to provide flow velocity information. The pixel signal intensity represents the phase shift within one pixel. Stationary tissue appears gray on this image, whereas flow in a positive direction along the flow-encoding axis appears bright, and flow in a negative direction appears dark. Therefore, antegrade and retrograde flow can be differentiated. Velocity can be encoded in planes that are perpendicular to the direction of flow by using section-selective direction (through-plane velocity measurement), and in planes that are parallel to the direction of flow by using phase-encoded or frequency encoded directions (in-plane velocity measurement), or, more recently, in 3D. However, VEC MR imaging has certain limitations and potential sources of error (Mohiaddin and Pennell 1998). Flow-related signal loss can be due to loss of coherence within a voxel, resulting in the inability to detect the phase of the flow signal above that of noise; to inappropriate selection of the velocity range, which in turn leads to poor detection of small vessels with slow flow; and to turbulence, which occurs in valvu-

lar stenosis and regurgitation. The latter can be overcome by using short TE sequences. VEC MR imaging can be used to calculate absolute velocity at any given time during the cardiac cycle at specified locations in the plane of data acquisition. Velocity can be measured for each pixel within a region of interest encircling all or part of the cross-sectional vessel area or across a valve annulus. The product of the entire area of a vessel or the orifice of a valve and the average flow velocity allow determination of the average flow per time unit. The integration of all flow volumes throughout the cardiac cycle within one valve yields the flow volume per heartbeat. These measurements can be used in the evaluation and quantitative assessment of valvular regurgitation and stenosis. Detection and Quantification of Valvular Regurgitation There are three methods for assessing the severity of valvular regurgitation: (1) visual assessment of the area of signal void, (2) calculation of the regurgitant fraction with ventricular volumetric measurements, and (3) quantification of the regurgitant blood flow with phasecontrast MR imaging in the position of the valve. Assessment of the Signal Void Cine gradient-echo MR imaging can be used to visualize and assess the severity of valvular regurgitation based on the area or volume of the signal void caused by the insufficient valve. Mitral regurgitation shows a signal void extending from the insufficient mitral valve into the left atrium in ventricular systole, whereas aortic regurgitation manifests as a signal void extending from the aortic valve into the left ventricle in ventricular diastole (Utz et al. 1988). Tricuspid regurgitation reveals a flow void in the right atrium and pulmonary regurgitation in the right ventricle. However, it has been demonstrated that the measurements of the dimensions of the signal void obtained with cine gradient-echo MR imaging are only semiquantitative indices of the severity of regurgitation (Duerinckx and Higgins 1994; Nishimura et al. 1989; Wagner et al. 1989). Among the different images acquired during an R–R interval, the image on which the signal void appears largest is used for the measurement. The signal void can vary significantly in size and shape depending on the observer and the image display settings. Thus, this method has limitations due to technical factors and physiopathological factors. Calculation of Regurgitant Fraction using Ventricular Volumes Ventricular volumes can be determined on a stack of short-axis views using cine MR imaging. The ventricular volumes can be determined exactly by the sum of the areas on the short axis slices during end-diastole and endsystole. In a normal heart, the right ventricular stroke volume is equal to the left ventricular stroke volume.

753

754

5  Thorax and Vasculature

The difference in stroke volume between a regurgitant ventricle (the left ventricle in mitral and aortic regurgitation) and a normal (right) ventricle is the regurgitant volume. The regurgitant fraction is calculated by dividing the regurgitant volume by the stroke volume of the regurgitant ventricle. The regurgitant fraction has been used to evaluate the severity of regurgitation. A regurgitant fraction of 15–20% represents mild regurgitation, a fraction of 20–40% represents moderate regurgitation, and a fraction greater than 40% represents severe regurgitation (Sechtem et al. 1988). One limitation of this method is that the formulas used to measure ventricular volumes yield more results in the elliptic left ventricle than in the right ventricle. Another limitation is that the method cannot be used in cases of multi-valvular disease. Right and left ventricular stroke volumes can be calculated as well by using VEC MR imaging to measure flow in the ascending aorta and pulmonary artery in a plane perpendicular to the vessel (Kondo et al. 1991).

Determination of the Transvalvular Pressure Gradient Modified Bernoulli Equation. VEC MR imaging allows evaluation of stenotic valves by directly measuring the increase in flow velocity in the jet of blood passing through the stenosis. The maximum of the velocity within the stenotic jet can be determined in planes perpendicular (through-plane velocity measurement) to the direction of flow. In-plane imaging can be considered as a method for visual assessment rather than for quantitative evaluation. In cases of very tight stenosis in which the jet is narrowed, in-plane measurement may be less reliable because of partial volume averaging and motion of the jet out of the imaging plane. The pressure gradient across the stenotic valve can be derived using the modified Bernoulli equation ∆P = 4 × (Vmax)2 (Wyttenbach et al. 1998), where ∆P is the pressure gradient across the stenosis and Vmax is the peak velocity (in meters per second) measured in the stenotic jet.

Quantification of Regurgitant Blood Flow with Phase-Contrast MR Imaging VEC MR imaging allows quantification of blood velocity profiles by generating a flow velocity map (Honda et al. 1993; Szolar et al. 1996). Retrograde flow can be identified, differentiated from antegrade flow, and quantified. For each time frame, the cross-sectional area is measured from a region of interest that is drawn manually around the vascular structure or heart valve on the magnitude image. This region of interest is then copied to the corresponding phase image. As the cross-sectional area of the heart valves and their orifices change during the cardiac cycle, a precise definition of the area must be done on the magnitude images. For the assessment of regurgitation the phase contrast measurements should be positioned at the valve annulus. For the assessment of stenosis these measurements should be positioned distal to the tips of the leaflets of the valve.

Continuity Equation. The valve area (AAo) can also be calculated with the continuity equation AAo = (AOT × VOT )/VAo (Wyttenbach et al. 1998) in planes perpendicular to the direction of flow. One slice is positioned in the plane of the aortic valve; a second slice is positioned perpendicular to the blood flow in the left ventricular outflow tract. AOT is the area of the outflow tract, VOT is the maximum velocity in the outflow tract, and VAo is the maximum velocity measured within the stenotic jet. A modified method also called a continuity equation is based on velocity time integrals during systole instead of the maximum velocity of the outflow tract. The following equation can also be used as continuity equation: AAo = AOT (VTIOT / VTIAo). VTIOT is the velocity time integral of the flow in the left ventricular outflow tract during systole. VTIAo is the velocity time integral of the flow in the orifice of the aortic valve during systole (Wyttenbach et al. 1998).

Detection and Quantification of Valvular Stenosis There are three methods for evaluating the severity of valvular stenosis: (1) assessment of the flow jet, (2) determination of the transvalvular pressure gradient by VEC MR imaging and (3) direct visualization of the valve area and orifice by cine MRI. Assessment of the Flow Jet Cine MR imaging can be used to determine the severity of valvular stenosis on the basis of the size and extent of the flow jet (Sechtem et al. 1987). In valvular stenosis, signal loss is caused by turbulent, high-velocity flow. Mitral stenosis manifests as a signal void extending from the mitral valve into the left ventricle during diastole, whereas aortic stenosis manifests as a signal void extending from the aortic valve into the aortic root during systole.

Planimetry by Cine MRI The valve area can be measured directly on the magnitude images or on the cine images, when the slice is positioned in the plane of the valve. Debl et al. (2005) compared planimetry of the aortic valve area determined by MRI in 33 patients to the results of catheterization using the Gorlin formula as well as to the results using planimetry with multiplanar transesophageal echocardiography. Image quality was adequate for determination of the aortic valve orifice in 82% by MRI, whereas image quality of transesophageal echocardiography for the assessment of the valve area was adequate in only 56% of patients because of calcification artifacts. The correlation between valve areas determined by MRI and by the invasive catheter-guided examination was 0.80 (p < 0.0001) and the correlation of areas determined by MRI and transesophageal echocardiography was 0.86 (p <0.0001). MRI overestimated the valve area by 15% compared with trans-

5.3  Heart

esophageal echocardiography and by 27% compared to the catheterization. Nevertheless, an MRI-derived aortic valve area below 1.3 cm2 indicated aortic stenosis with an area below 1 cm2, with a sensitivity of 96% and a specificity of 100%. However, overestimation of the valve area by MRI-based planimetry has not been confirmed in all studies (Debl et al. 2005; Caruthers et al. 2003; John et al. 2003; Kupfahl et al. 2004).

Aortic Stenosis Aortic stenosis is defined as a condition in which opening of the aortic valve in systole is restricted. In infants,

children, and young adults, the major causes of aortic stenosis are congenital malformation of the cusps, e.g., bicuspid aortic valve or annulus, and rheumatic disease. In patients over 60 years of age, the major causes are calcification of a bicuspid aortic valve and degeneration of the valve cusps or annulus. The optimal planes for identifying the signal void corresponding to the abnormal flow jet are the three-chamber view and a perpendicular slice to the three-chamber view parallel to the aortic root (Figs. 5.3.32, 5.3.33). The plane of the annulus of the aortic valve can be determined in these two planes. An aortic stenosis causes first left ventricular hypertrophy with thickening of the myocardium and later left ventricular dilatation. Table 5.3.11 reveals severity of aortic stenosis with respect to the determined aortic valve area (Bonow et al. 2005).

Fig 5.3.32  56-year-old female patient with regurgitation of the aortic valve. a Cine image in three-chamber orientation reveals a flow void. The flow void extends (arrow) from the aortic valve into the left ventricular outflow tract during diastole and hits the anterior leaflet of the mitral valve (arrowhead). b Slice for

through-plane flow measurement is orientated to the aortic valve perpendicular to a previously acquired slice in the threechamber view. c Velocity-encoded flow measurement allows for determination of the volume of regurgitation and the antegrade stroke volume

5.3.5.3.2  Clinical Applications

755

756

5  Thorax and Vasculature

Fig 5.3.33  73-year-old-patient with aortic valve stenosis. a cine SSFP image in three-chamber-view orientation, acquired during systole, reveals a flow void in the aortic root (arrowhead). b Cine SSFP image, acquired in an orientation that is perpendicular to the three-chamber view, reveals a flow void in the aortic root (arrow). c Magnitude-reconstructed images of a velocity-encoded gradient-echo sequence. The slice is positioned

oriented to the aortic valve. d Phase-contrast images of a velocity encoded gradient-echo sequence. The slice is orientated to the aortic valve. A region of interest is drawn in the central part of the orifice of the aortic valve. e Velocity–time curve reveals increased velocity maximum of 3.2 m/s. The modified Bernoulli equation ∆P = 4 × (Vmax)2 allows for calculation of the corresponding pressure gradient of 41 mmHg

5.3  Heart Table 5.3.11  Severity of aortic valve stenosis Severity of aortic valve stenosis

aortic valve area

Mean aortic valve pressure gradient

Mild

>1.5 cm2

<25 mmHg

Moderate

1–1.5 cm2

25–40 mmHg

Severe

<1 cm2

>40 mmHg

Aortic Regurgitation Aortic regurgitation may be due to abnormalities of the aortic cusps, a lesion of the annulus, or dilatation of the aortic root. The most frequent causes are rheumatic disease, aortoannular ectasia, endocarditis, and aortic root dilatation due to hypertension. The main hemodynamic consequence of aortic regurgitation is an increase in end-diastolic volume of the left ventricle. MRI allows for accurate determination of left ventricular volumes, diameter, ejection fraction, and fraction of regurgitation. However, evaluation with color Doppler flow mapping is semiquantitative (i.e., measures jet length and width) and may be affected by several technical limitations. Grading of regurgitation by invasive cardiac catheterization remains imprecise, and underestimation of the degree of regurgitation may occur in severe cases. Therefore, there remains a need for noninvasive, direct measurement of regurgitant volume (Fig. 5.3.32) and ventricular function, and MR imaging can fulfill this need (Wyttenbach et al. 1998). Left ventricular enlargement can also be demonstrated. Cine gradient-echo MR imaging can be used for quantitative evaluation of valvular regurgitation based on the area of the signal void corresponding to the regurgitant flow jet in diastole (Pflugfelder et al. 1989). Because the abnormal regurgitant flow jet is a 3D structure and may change shape and direction from one case to another, it is preferable to assess the area of the signal void in both planes. The degree of aortic regurgitation can best be quantified with VEC MR imaging. Combined Aortic Stenosis and Regurgitation Combined aortic stenosis and regurgitation is a condition encountered in the evolution of an abnormal bicuspid aortic valve. It is important to identify the dominant lesion, although in some cases the pathophysiological process may be balanced. MR imaging can demonstrate a bicuspid aortic valve (Fig. 5.3.34). Cine gradient-echo MR imaging can be used for quantitative evaluation of valvular stenosis and regurgitation as described earlier. However, VEC MR imaging is optimal for quantifying the degree of stenosis and regurgitation. Mitral Stenosis Mitral stenosis due to restricted opening of the mitral valve in diastole results in left ventricular inlet obstruc-

tion (Fig. 5.3.35) with a diastolic pressure gradient between the left atrium and the left ventricle. Rheumatic disease is the most common cause of mitral stenosis. Fusion of the commissures and thickening of the leaflets may result, and the chordae tendineae may be affected as well. Doppler echocardiography is the modality of choice in evaluating the degree of mitral stenosis and is usually sufficient for therapeutic treatment planning. MR imaging may be useful in cases in which findings at Doppler echocardiography are inconsistent with clinical data or are insufficient, particularly in those cases involving a limited acoustic window or complex flow patterns. These patterns are better demonstrated at MR imaging, which displays the entire abnormal flow jet in any plane or direction. In patients with atrial fibrillations dedicated MRI techniques should be used to avoid ECG artifacts. Cine SSFP imaging can be used as a real-time technique (Kupfahl et al. 2004), or with retrospective gating to cover RRcycles with varying lengths. The degree of mitral stenosis can be assessed on the basis of the size and extent of the abnormal flow jet in the left ventricle in diastole and associated anatomic findings (valve leaflet thickening and bulging, reduced valve motion, left atrial enlargement). The signal void corresponding to the abnormal flow jet is most clearly demonstrated on the four-chamber, threechamber, and two-chamber views. Cine MRI in the short axis orientation allows for direct assessment of the orifice of the mitral valve by planimetry. With the help of VEC MR imaging, maximum velocity within the stenotic jet can be calculated in planes either perpendicular or parallel to the direction of flow. The pressure gradient across the stenotic mitral valve can be calculated using the modified Bernoulli equation. Severity of mitral stenosis with respect to the determined aortic valve area is shown in Table 5.3.12 (Bonow et al. 2005). Mitral Regurgitation Mitral regurgitation can be due to abnormalities of the mitral annulus, mitral leaflets, chordae tendineae, or papillary muscles. The most frequent causes are rheumatic disease, in which the cusps are scarred and fibrotic; endocarditis, in which the cusps are destroyed; dysfunction due to infarction; dilated heart due to congestive heart failure; prolapse; and congenital abnormalities. The major hemodynamic consequence of mitral regurgitation

757

758

5  Thorax and Vasculature

is an increase in the total stroke volume of the left ventricle. The need for surgery is determined by the severity of symptoms and by whether the ejection fraction falls toward 60% or the left ventricular end systolic diameter approaches 45-mm echocardiography (Carabello and Crawford 1997). However, more precise measurements of the degree of regurgitation are also important. Measurements of the length and area of the mitral regurgitant jet obtained with Doppler echocardiography are semiquantitative. A direct, noninvasive, quantitative assessment of the regurgitant volume can be accurately obtained with MR imaging (Wyttenbach et al. 1998). Cine gradient-echo imaging can be used to evaluate the severity of

mitral regurgitation based on the area of the signal void in the left atrium. This signal void is best demonstrated on the four-chamber, three-chamber, and two-chamber views. A good correlation has been found between the ratio of maximum flow void area to the left atrial and left ventricular area and the severity of mitral regurgitation as estimated at pulsed and color Doppler echocardiography (Aurigemma et al. 1990; Nishimura et al. 1989). A good method for evaluating the severity of mitral regurgitation is to calculate the regurgitant fraction by measuring the right and left ventricular volumes on cine MRI. However, this calculation is valid only in patients with a single regurgitant valve (Duerinckx and Higgins 1994; Fujita et al.

Fig. 5.3.34a–d  27-year-old female patient. A bicuspid aortic valve causes a combined aortic valve stenosis and regurgitation. a The SSFP cine image orientated to the aortic valve demonstrates a bicuspid aortic valve. b The SSFP cine image, acquired during systole and orientated to the three-chamber view, demonstrates a flow void in the aortic root (arrow). c The SSFP cine

image, acquired during diastole and orientated to the threechamber view, demonstrates a flow void in the left ventricular outflow tract (arrow). d 34-year-old male patient. The SSFP cine image positioned in the plane of the aortic valve demonstrates a tricuspid aortic valve

5.3  Heart

1994). Mitral regurgitation can also be quantified by using VEC MR imaging to compare diastolic inflow across the mitral annulus with systolic outflow across the ascending aorta. These values are nearly identical in healthy individuals. Left ventricular inflow is increased in mitral regurgitation. The difference between the areas of the two superimposed curves corresponds to the volume of mitral regurgitation. However, evaluation of diastolic inflow across the mitral annulus remains difficult with VEC MR imaging due to motion of the mitral annulus during the cardiac cycle. With this method, patients with moderate to severe mitral regurgitation can be distinguished from healthy individuals and patients with mild regurgitation (Fujita et al. 1994; Hundley et al. 1995). Alternatively, the regurgitant volume can be calculated by measuring flow in the ascending aorta and pulmonary artery with VEC MR imaging in a plane perpendicular to the vessel as described earlier. Thus, the stroke volumes for the left and right ventricles can be calculated separately, with regurgitant volume being the difference between left ventricular stroke volume and right ventricular stroke volume (Caputo et al. 1991). However, the best way to quantify the volume of mitral regurgitation is probably

to combine ventricular volume measurements obtained with cine gradient-echo MR imaging with measurements of forward flow in the aorta obtained with VEC MR imaging. The left ventricular stroke volume index, which is the difference between the left ventricular end-diastolic volume index and the left ventricular end-systolic volume index, is calculated from cine gradient-echo images as described earlier. The left ventricular cardiac index is the product of the left ventricular stroke volume index and the heart rate. The regurgitant volume index is calculated by subtracting the forward cardiac index, determined at VEC MR imaging in a plane perpendicular to the proximal aorta from the left ventricular cardiac index. The regurgitant fraction is determined by dividing the regurgitant volume index by the left ventricular cardiac index. In a study by Hundley et al. (1995), these measurements correlated well with those obtained with catheterization and angiography, in which regurgitant flow was calculated by subtracting forward cardiac output, which was measured using the Fick principle or the indicator dilution method, from left ventricular output, which was determined with contrast material-enhanced ventriculography.

Table 5.3.12  Severity of mitral stenosis Severity of mitral stenosis

Mitral valve area

Exercise pulmonary artery wedge pressure

Rest pulmonary artery systolic pressure

Mild

>1.5 cm2

≤20 mmHg

<35 mmHg

Moderate

1–1.5 cm2

20–25 mmHg

35–50 mmHg

Severe

<1 cm2

>25 mmHg

>50 mmHg

Fig. 5.3.35a–c  67-year-old female patient with stenosis of the mitral valve. a Cine SSFP image in three-chamber view orientation. A flow void is arising from the mitral valve into the left ventricular cavity (arrow). The left atrium is enlarged (small arrow). b Cine SSFP image in four-chamber-view orientation. The posterolateral leaflet of the mitral valve reveals a doming (small arrow). The orifice is reduced. A flow void arises from the tips of

the leaflets of the mitral valve into the left ventricular cavity (arrow). Pleural effusions are present on both sides (arrowheads). c Cine SSFP image in short axis orientation. The reduced area of the orifice (arrow) can be determined by planimetry. d Phasecontrast image in short axis orientation reveals the flow (bright signal) in the reduced orifice of the mitral valve (arrow)

759

760

5  Thorax and Vasculature

Pulmonary Stenosis Pulmonary valvular stenosis is usually a congenital anomaly that is well tolerated for many years and consists of fusion of the leaflets at the commissures, which prevents the leaflets from opening completely in systole. In severe pulmonary valvular stenosis, patients present with symptoms of chronic right ventricular failure. Percutaneous valvuloplasty is one of the procedures used at present to treat patients with pulmonary stenosis. Doppler echocardiography is the modality of choice in evaluating the degree of pulmonary stenosis and is usually sufficient for treatment planning. MR imaging can be useful in cases in which Doppler echocardiographic findings are inconsistent with clinical data or are insufficient, particularly when the acoustic window is limited. The methods for the assessment of pulmonary stenosis are analogous to those described for the aortic valve and the left ventricle. Pulmonary Regurgitation Pulmonary regurgitation may be caused by (1) dilatation of the valve annulus secondary to pulmonary hypertension, (2) endocarditis, or (3) complications of surgical treatment for pulmonary stenosis, tetralogy of Fallot, or other conotruncal malformations. This is particularly true when reconstruction of the outflow tract does not include a valvular apparatus. The major hemodynamic consequence of pulmonary regurgitation is an increase in the end-diastolic volume of the right ventricle. Doppler echocardiography and angiography have limitations in evaluating right ventricular morphology and function. MR imaging, with its multiplanar and 3D imaging capabilities, more accurately depicts right ventricular dilatation and hypertrophy than does echocardiography. MR imaging also allows direct measurement of regurgitant volume in the right ventricle and evaluation of right ventricular function. However, the diagnostic accuracy in determination of right ventricular volumes is somewhat inferior compared with the accuracy in determination of the left ventricular volumes, for MRI, too. Tricuspid Regurgitation Tricuspid regurgitation may result from endocarditis, rheumatic disease, dilatation of the right ventricle and the tricuspid annulus secondary to mitral valvulopathy or pulmonary hypertension, dysfunction due to infarction, trauma, carcinoid syndrome, Marfan syndrome, or congenital abnormality, for example, Ebstein’s anomaly. The major hemodynamic consequence of tricuspid regurgitation is an increase in the total stroke volume of the right ventricle. The methods for assessment and quantification of tricuspid valve regurgitation are analogous to those described for the mitral valve. In addition, the pressure in the pulmonary artery can be derived from maximum velocity by changing the modified Bernoulli equation as follows: PPA = PRA + (4 × Vmax2), where PPA

is pulmonary artery pressure and PRA is right atrial pressure (estimated at 10 mmHg). Endocarditis and Prosthetic Valves Endocarditis is a severe complication of valve replacement with a mechanical prosthesis, and diagnosis must be made as quickly as possible because repeat surgery is usually mandatory. Transesophageal echocardiography is optimal for the assessment of valve morphology and is especially helpful in demonstrating vegetations in infectious endocarditis. However, MR imaging plays an important role in diagnosing paravalvular abscesses associated with infectious processes, which are difficult to visualize at echocardiography because of prosthesis-related artifacts that occur even with transesophageal technique. Patients with current prosthetic valves can safely undergo imaging with high-field-strength imagers (Caputo et al. 1991). Spin-echo imaging can demonstrate the exact location of the abscess and its relationship to the cavities and great vessels. Cine and VEC images allow for the assessment of valvular stenosis and regurgitation with the limitation of variable artifacts, caused by metal parts in valve prostheses (Caputo et al. 1991). References 1.

2.

3. 4.

5. 6. 7.

8.

9.

Aurigemma G, Reichek N, Schiebler M, Axel L (1990) Evaluation of mitral regurgitation by cine magnetic resonance imaging. Am J Cardiol 66:621–625 Bogren HG, Buonocore MH (1994) Blood flow measurements in the aorta and major arteries with MR velocity mapping. J Magn Reson Imaging 4:119–130 Bonow RO, Cheitlin MD, Crawfold MH, Douglas PS (2005) Task Force 3: valvular heart disease. JACC 45:1334–1340 Caputo GR, Kondo C, Masui T et al. (1991) Right and left lung perfusion: in vitro and in vivo validation with oblique-angle, velocity-encoded cine MR imaging. Radiology 180:693–698 Carabello BA, Crawford FA (1997a) Valvular heart disease. N Engl J Med 337:332–341 Carabello BA, Crawford FA (1997b) Valvular heart disease. N Engl J Med 307:1362–1367 Carroll JD (1993) Cardiac catheterisation and other imaging modalities in the evaluation of valvular heart disease. Curr Opin Cardiol 8:211–215 Caruthers, SD, Lin SJ, Brown P, Watkins MP, Williams TA, Lehr KA, AND, Wickline SA (2003) Practical value of cardiac magnetic resonance imaging for clinical quantification of aortic valve stenosis comparison with echocardiography. Circulation 108:2236–2243 Debl K, Djavidani B, Seitz J, Nitz W, Schmid FX, Muders F, Buchner S, Feuerbach. Riegger G, Luchner G (2005) Planimetry of aortic valve area in aortic stenosis by magnetic resonance imaging. Invest Radiol 40:631–636

5.3  Heart 10. Didier D, Ratib O, Friedli B et al. (1993) Cine gradientecho MR imaging in the evaluation of cardiovascular diseases. Radiographics 13:561–573 11. Didier D, Ratib O, Lerch R, Friedli B (2000) Detection and quantification of valvular heart disease with dynamic cardiac MR imaging. Radiographics 20:1279–1299 12. Duerinckx AJ, Higgins CB (1994) Valvular heart disease. Radiol Clin North Am 32:613–630 13. Fujita N, Chazouilleres AF, Hartiala JJ et al. (1994) Quantification of mitral regurgitation by velocityencoded cine magnetic resonance imaging. J Am Coll Cardiol 23:951–958 14. Gorlin R, Gorlin SG (1951) Hydraulic formula for calculation of the area of the stenotic mitral valve, other valves and central circulatory shunts. Am Heart J 1951, 41:1–29 15. Higgins CB, Sakuma H (1996) Heart disease: functional evaluation with MR imaging. Radiology 199:307–315 16. Honda N, Machida K, Hashimoto M et al. (1993) Aortic regurgitation: quantitation with MR imaging velocity mapping. Radiology 186:189–194 17. Hundley WG, Li HF, Willard JE et al. (1995) Magnetic resonance imaging assessment of the severity of mitral regurgitation: comparison with invasive techniques. Circulation 92:1151–1158 18. John AS, Dill T, Brandt RR, Rau M, Ricken W, Bachmann G, Hamm CW (2003) Magnetic resonance to assess the aortic valve area in aortic stenosis: how does it compare to current diagnostic standards? J Am Coll Cardiol 62:519–526 19. Kondo C, Caputo GR, Semelka R et al. (1991) Right and left ventricular stroke volume measurements with velocity encoded cine NMR imaging: in vitro and in vivo evaluation. AJR Am J Roentgenol 157:9–16 20. Kupfahl C, Honold M, Meinhardt G, Vogelsberg H, Wagner A, Mahrholdt H, Sechtem U (2004) Evaluation of aortic stenosis by cardiovascular magnetic resonance imaging: comparison with established routine clinical techniques. Heart 90:893–901 21. Manning WJ, Silverman DI, Gordon SP, Krumholz HM, Douglas PS (1993) Cardioversion from atrial fibrillation without prolonged anticoagulation with use of transesophageal echocardiography to exclude the presence of atrial thrombi. N Engl J Med 328:750–755 22. Miller S, Simonetti OP, Carr J, Kramer U, Finn JP (2002) MR Imaging of the heart with cine true fast imaging with steady-state precession: influence of spatial and temporal resolutions on left ventricular functional parameters. Radiology 223:263–269 23. Mohiaddin RH, Pennell DJ (1998) MR blood flow measurement: clinical application in the heart and circulation. Cardiol Clin 16:161–187 24. Mostbeck GH, Caputo GR, Higgins CB (1992) MR measurements of blood flow in the cardiovascular system. AJR Am J Roentgenol 159:453–461 25. Nayler GL, Firmin DN, Longmore DB (1986) Blood flow imaging by cine magnetic resonance. J Comput Assist Tomogr 10:715–722

26. Nishimura T, Yamada N, Itoh A, Miyatake K (1989) Cine MR imaging in mitral regurgitation: comparison with color Doppler flow imaging. AJR Am J Roentgenol 153:721–724 27. Pflugfelder PW, Landzberg JS, Cassidy MM et al. (1989) Comparison of cine MR imaging with Doppler echocardiography for the evaluation of aortic regurgitation. AJR Am J Roentgenol 152:729–735 28. Rebergen SA, van der Wall EE, Doornbos J et al. (1993) Magnetic resonance measurements of velocity and flow: technique, validation and cardiovascular applications. Am Heart J 126:1439–1456 29. Sechtem U, Pflugfelder PW, White RD et al. (1987) Cine MR imaging: potential for the evaluation of cardiovascular function. AJR Am J Roentgenol 148:239–246 30. Sechtem U, Pflugfelder PW, Cassidy MM et al. (1988) Mitral or aortic regurgitation: quantification of regurgitant volumes with cine MR imaging. Radiology 167:425–430 31. Smith MD, Kie GY (1998) Current echocardiographyDoppler approaches to the quantification of valvular regurgitation. Cardiology 6:168–181 32. Szolar DH, Sakuma H, Higgins CB (1996) Cardiovascular applications of magnetic resonance flow and velocity measurements. J Magn Reson Imaging 1:78–89 33. Utz JA, Herfkens RJ, Heinsimer JA, Shimakawa A, Glo­ ver G, Pelc N (1998) Valvular regurgitation: dynamic MR imaging. Radiology 168:91–94 34. Wagner S, Aufferman W, Buser P et al. (1989) Diagnostic accuracy and estimation of the severity of valvular regurgitation from the signal void on cine magnetic resonance images. Am Heart J 118:760–767 35. Wintersperger BJ, Nikoalou K, Dietrich O, Rieber J, Nittka M, Reiser MF, Schönberg SO (2003) Single breathhold real-time cine MR imaging: improved temporal resolution using a generalized autocalibrating partially parallel acquisition (GRAPPA) algorithm. Eur Radiol 13:1931–1936 36. Wyttenbach R, Bremerich J, Saeed M, Higgins CB (1998) Integrated MR imaging approach to valvular heart disease. Cardiol Clin 16:277–294

5.3.6  Pericardial Diseases K. Bauner 5.3.6.1  Introduction The pericardium is a two-layered membrane that envelops all four cardiac chambers and the origins of the great vessels. The outer fibrous parietal layer is attached to the sternum, the diaphragm and to costal cartilage. The serous layer is a thin mesothelial layer adjacent to the surface of the heart. The pericardial sac lies between variable amounts of epicardial and pericardial adipose tissue (Breen 2001). The parietal and the visceral layers are separated by a small amount of serous fluid (Fig.

761

762

5  Thorax and Vasculature

5.3.36), about 15–50 ml, which is mainly an ultrafiltrate of plasma (Edwards 2001). The pericardium limits the spread of infection and inflammation from the neighboring mediastinal structures, and is thought to prevent excessive cardiac dilatation as well as to reduce friction between the heart and adjacent structures. The presence of normal pericardium can also be detrimental in situations where fluid rapidly fills the limited interpericardial space and the ability to distend is exceeded, while slowly accumulating effusions cause the pericardium to stretch (Breen 2001). Many diseases can affect the pericardium, including infection, neoplasm, trauma, primary myocardial diseases, and congenital diseases. 5.3.6.2  Imaging of the Pericardium Two-dimensional echocardiography is the initial imaging modality applied in suspected pericardial disease. However, the restricted acoustic window limits the usefulness of this modality for imaging the entire pericardium. Loculated pericardial effusions, especially those in unusual locations, can be difficult to diagnose with echocardiography (Yousem 1987). Transthoracic echocardiography is not highly accurate for the detection of thickened pericardium associated with pericarditis while transesophageal echocardiography, which has shown more promise, is limited to the restricted field of view (Ling et al. 1997). MR imaging can provide comprehensive depiction of the pericardium, without the use of either iodinated contrast material or ionizing radiation and with a large field of

view. MR imaging allows characterization of pericardial effusion and pericardial masses with the use of a combination of T1-weighted, T2-weighted, and gradient-echo techniques, especially the balanced SSFP technique. The “normal” pericardium is seen as a very thin linear density surrounding the heart, but is often not visualized over much of the left ventricle (Bull et al. 1998). With MRI the pericardium appears as low-intensity signal between the high-intensity mediastinal and epicardial fat with a reported thickness up to 4 mm, on average ranging from 1 to 3 mm, being 1.2 mm in diastole and 1.7 mm in systole (Frank and Globits 1999). The measurement most likely includes the entire pericardial complex, physiologic fluid representing a significant component of the thickness measured. 5.3.6.3  Pericardial Diseases 5.3.6.3.1  Pericardial Defects Pericardial defects are uncommon. While the majority of cases are congenital, pericardial defects can be resulting from trauma or surgery. A wide spectrum of defects appears, ranging from small defects to the entire absence of the pericardium. The most common defect is the absence of the left side of the pericardium. Normally pericardium covers the aortopulmonary window, thus the absence allows interpositions of lung tissue between the aorta and the main segment of the pulmonary artery and occasionally, bulging of the left atrial appendage through

Fig. 5.3.36a,b  45-year-old male patient. Short-axis view, T1-weighted-TSE-MRI (a) and long-axis view, balanced SSFP (b). Note the thin pericardial layers (arrow) and the small amount of intrapericardial fluid (star). LV left ventricle, RV right ventricle

5.3  Heart

the defect. As a result of these abnormalities, the heart eventually rotates to the left (Wang et al. 2003). MRI may be helpful in diagnosing pericardial defects, but the occasional lack of visualization of the pericardium over the left ventricle and atrial appendage in normal individuals makes diagnosis difficult.

epicardial fat in GRE images. Conversely, hemorrhagic effusion is characterized by high signal intensity on T1weighted SE images and low intensity on SSFP cine images (Hancock 1990).

5.3.6.3.2  Pericardial Effusion

Inflammatory pericarditis can be caused by various infectious diseases (e.g. viral, bacterial, tuberculosis, fungal). Besides that, it can be a symptom of systemic diseases, as for example, connective-tissue diseases, it can be found in patients with uremia, or it can be a sequela of acute myocardial infarction or irradiation therapy in breast cancer patients or patients with lymphoma. However, often the underlying pathology is not found (idiopathic). The symptoms are mainly related to the severity of pericardial inflammation. In acute phase, inflammation of the pericardial layers is characterized by highly vascularized granulation tissue with fibrin deposits’ which show hyperenhancement in CE MRI (Fig. 5.3.39). Usually a variable amount of pericardial fluid is present, and the fibrin deposition may lead to a fibrinous adhesion of pericardial layers. Chronic inflammation is characterized by a progressive sclerosing pericarditis with fibroblasts, collagen, and a lesser amount of fibrin deposition. This may progress towards an end-stage, chronic fibrosing pericarditis. The main feature of the end stage is a stiff pericardium with constriction of the heart.

Pericardial effusion originates in the obstruction of venous or lymphatic drainage from the heart. Common causes of pericardial effusions include heart failure, renal insufficiency, infection, neoplasm, and injury. Although pericardial effusion usually has no or only limited impact on cardiac filling, acute accumulation of pericardial contents (e.g., fluid, blood, air) can lead to cardiac tamponade. Hemodynamic consequences reflect the volume of the effusion, rapidity of its accumulation, compliance of the pericardium and myocardium, cardiac compensatory mechanisms (contractility and heart rate), and total blood volume (Spodick 2001, 2003; Goldstein 1996) Thus, acutely developing effusions of even 300–400 ml can abruptly elevate interpericardial pressure and induce tamponade. However, given time, the pericardium can stretch and exhibit compliance. Accordingly, slow accumulation of fluid (even in amounts greater than 1 l associated with neoplasms) may be tolerated with little or no hemodynamic compromise. Hemodynamically significant effusions compress the cardiac chambers, particularly the thinner more compliant right heart chambers, throughout diastole (Figs. 5.3.37, 5.3.38). The appearance of pericardial fluid is different on SE and GRE cine MR images. Nonhemorrhagic fluid has low signal intensity on T1-weighted SE images and high intensity on SSFP images (Mulvagh et al. 1989). Transudate pericardial effusions are often even more intense than

Fig. 5.3.37a–c  56-year-old male patient with chronic pericardial effusion. On short-axis TSE-MRI (a), HASTE MRI (b) and balanced SSFP cine MRI (c) pericardial effusion (star) is clearly seen. In contrast to the low signal in T1-weighted TSE MRI and

5.3.6.3.3  Pericardial Inflammation

5.3.6.3.4  Constrictive Pericarditis The causes of constrictive pericarditis are numerous, including post-surgical, post-radiation, posttraumatic, and post-infectious presentations (Cameron et al. 1987). Often the cause is labeled idiopathic. A thickness exceeding

the inhomogeneous low signal intensity seen with the HASTE technique, the signal of pericardial fluid in cine MRI is highly suggestive of nonhemorrhagic fluid

763

764

5  Thorax and Vasculature

Fig. 5.3.38a,b  57-year-old female patient with loculated pericardial effusion. T2-weighted HASTE-MRI (a) and balanced-SSFP (b) long-axis in four-chamber view. Note the compression of the right ventricle

4 mm with calcifications, which may be missed in MRI, or without calcifications in a proper clinical setting is basically diagnostic of constriction. However, the presence of pericardial thickening by itself does not indicate constriction (Frank and Globits 1999). Additional findings seen with constriction include distorted contours of the ventricles, tubular-shaped ventricles, hepatic venous congestion, ascites, pleural effusions, and occasionally pericardial effusion. Often there is dilatation of the atria, coronary sinus, inferior vena cava, and the hepatic veins. Cine acquisitions show an abnormal motion of the interventricular septum in early diastole with sometimes

a prominent leftward convexity in the septum (Francone 2005). The pericardium can globally be thickened, but frequently only focal pericardinal thickening is identified. Compression of the right ventricle caused by focal thickening is more common than left-sided compressions. The absence of thickened pericardium argues against the diagnosis of constriction but does not rule it out. Clinically, it is difficult to differentiate between constrictive pericarditis and restrictive cardiomyopathy, both characterized by similar clinical symptoms and similar manifestations at cardiac catheterization and echocardiography. It is important, however, to distinguish

Fig. 5.3.39a–c  62-year-old male patient with chest pain and symptoms of diastolic heart failure. Four-chamber view. SSFP cine MRI long-axis (a) and short-axis (b) view. On contrast-enhanced-MRI, enhancement of pericardial layers (arrows) is seen with a pericardial effusion (star). LA left atrium, LV left ventricle

5.3  Heart

between constrictive pericarditis and restrictive cardiomyopathy because patients with constrictive pericarditis may benefit from pericardial stripping, whereas those with restrictive cardiomyopathy do not. The signal intensity of the thickened pericardium on ECG-gated MR images is variable and it may be difficult to differentiate a pericardial effusion from constrictive pericarditis. The purely fibrous or calcified pericardium in chronic pericardial disease has, as normal pericardium, low signal intensity on both T1- and T2-weighted images. In subacute forms of pericarditis, the thickened pericardium has moderate to high signal intensity on T1-weighted SE images. Enhancement of the thickened pericardium after the administration of gadolinium-based contrast material also suggests inflammation. The effusive-constrictive form of pericarditis involves both pericardial thickening and pericardial effusion (Fig. 5.3.40).

cation of gadolinium chelates. Occasionally cysts contain highly proteineous fluid, which may have high signal intensity on T1-weighted images. A discriminative feature is their common tendency to change in size or shape with respiration or body position.

The differential diagnosis of pericardial masses includes pericardial cyst, hematoma, and neoplasm.

Hematomas MR imaging is particularly useful for the diagnosis of pericardial hematomas, which have characteristic signal intensity on T1-weighted and T2-weighted images. High homogeneous signal intensity appears in acute hematoma, while subacute hematomas that are 1–4 weeks old show heterogeneous signal intensity with areas of high signal intensity on both T1- and T2-weighted images. On T1-weighted images or gradient-echo images, organized hematomas demonstrate a dark peripheral rim and low-signal-intensity internal foci that may present calcification, fibrosis, or hemosiderin deposition. High signal intensity areas on T1-weighted or T2-weighted imaging often correspond to hemorrhagic fluid. Coronary or ventricular pseudoaneurysm or neoplasms may resemble hematomas on MR images. However gadolinium-chelate application allows for differentiation of these entities, as hematomas do not enhance.

Pericardial Cysts Pericardial cysts are rare remnants of defective embryological development of the pericardium. They are clinically indistinguishable from pericardial diverticula and exist as unilocular, thin-walled structures that may either be attached intimately to the pericardium or are pedunculated. Most often they occur in the right cardiophrenic angle, but they can be located throughout the mediastinum (Breen 2001). Any cysts containing transudate fluid typically have low or intermediate signal intensity on T1weighted SE images and homogeneous high intensity on T2-weighted SE images. They do not enhance after appli-

Primary Pericardial Tumors Primary pericardial tumors are very rare, with mesotheliomas and sarcomas being the more common among them (Hancock 1990). Other malignant pericardial tumors seen are lymphoma and liposarcoma. Benign pericardial tumors include lipoma, teratoma, fibroma, and hemangioma. Primary pericardial mesothelioma may present as pericardial effusion eventually accompanied by nodules and plaques. Lymphoma, sarcoma, and liposarcoma typically manifest as large heterogeneous masses frequently associated with serosanguineous pericardial effusion. Biopsy and histopathological analysis are nec-

5.3.6.3.5  Pericardial Masses

Fig. 5.3.40a–c  40-year-old male patient with constrictive pericarditis and impairment of diastolic function. Thickened right (star) and left ventricular (arrow) pericardium with low signal intensity in T1-weighted TSE MRI (a) and SSFP cine MRI (b), high signal intensity in HASTE sequence (c). LA left atrium

765

766

5  Thorax and Vasculature 6.

7.

8. 9.

10.

11. Fig. 5.3.41  58-year-old male patient with small cell pericardial tumor. T2-weighted HASTE MRI short-axis view. Circular thickening of the pericardium (arrow) and intrapericardial fluid inclusions (star)

essary to achieve definitive diagnosis of most pericardial tumors. Metastatic Pericardial Disease Pericardial metastases are much more common than primary pericardial tumors; most of the lesions are seen at autopsy with an incidence of 10–12% of all patients with malignancies. Carcinomas most likely to metastasize to the pericardium are tumors of the breast and the lung (Fig. 5.3.41). Melanoma and lymphoma are two other common sources of pericardial metastases. Pericardial involvement by direct extension of pulmonary or mediastinal malignancies is commonly seen. References 1. 2. 3.

4.

5.

Breen JF (2001) Imaging of the pericardium. J Thorac Imaging 16:47–54 Bull RK, Edwards PD, Dixon AK (1998) CT dimensions of the normal pericardium. Br J Radiol 71:923–925 Cameron J, Oesterle SN, Baldwin JC, Hancock EW (1987) The etiologic spectrum of constrictive pericarditis. Am Heart J 113:354–360 Edwards ED (2001) Applied anatomy of the heart. In: Giulaini ER, Fuster V (eds) Cardiology: fundamentals and practice, 2nd edn. Mosby-Year Book, St. Louis, pp 47–51 Francone M, Dymarkowski S, Kalantzi M, Bogaert J (2005) Real-time cine MRI of ventricular septal motion: a novel approach to assess ventricular coupling. J Magn Reson Imaging 21:305–309

12.

13.

14.

Frank H, Globits S (1999) Magnetic resonance imaging evaluation of myocardial and pericardial disease. J Magn Reson Imaging 10:617–626 Goldstein JA (1996) Management of patients with pericardial diseases. Medical manage­ment of heart disease: the clinician’s consultant. Dekker, New York, pp 267–284 Hancock EW (1990) Neoplastic pericardial disease. Cardiol Clin 8:673–682 Ling LH, Oh JK, Tei C et al. (1997) Pericardial thickness measured with transesophageal echocardiography: feasibility and potential clinical usefulness. J Am Coll Cardiol 29:1317–1323 Mulvagh SL, Rokey R, Vick GW III, Johnston DL (1989) Usefulness of nuclear magnetic resonance imaging for evaluation of pericardial effusions, and comparison with two-dimensional echocardiography. Am J Cardiol 64:1002–1009 Spodick DH (2003) Acute cardiac tamponade. N Engl J Med 349:684–690 Spodick DH (2001) Pericardial diseases. In: Braunwald E, Zipes DP, Libby P (eds) Heart disease: a textbook of cardiovascular medicine, vol. 2, 6th edn. Saunders, Philadelphia, pp 1823–1876 Wang ZJ, Reddy GP, Gotway MB, Yeh BM, Hetts SW, Higgins CB (2003) CT and MR imaging of pericardial disease. Radiographics 23(Spec no.):S167–S180 Yousem D, Traill TT, Wheeler PS, Fishman EK (1987) Illustrative cases in pericardial effusion misdetection: correlation of echocardiography and CT. Cardiovasc Intervent Radiol 10:162–167

5.3.7  Ischemic Heart Disease K. Nikolaou 5.3.7.1  Introduction Cardiac MRI has matured into a multipurpose non-invasive imaging tool for the assessment of ischemic cardiomyopathy. The breadth of applications possible with cardiac MRI allows combined non-invasive assessment of myocardial perfusion, function, and myocardial viability—a task that usually requires use of myocardial scintigraphy and echocardiography. As such, cardiac MRI currently holds a strong position in the non-invasive work-up of patients with coronary artery disease (CAD). The distinct advantages of MRI over current conventional nuclear-based cardiac-imaging techniques, such as PET or myocardial scintigraphy, include its high spatial resolution and lack of exposure of the patient to ionizing radiation. Also, quantification of cardiac morphology and function by MRI is more accurate and image quality is more reproducible than in echocardiography, independent of the operator’s experience and skill level or the patient’s anatomy.

5.3  Heart

5.3.7.2  Coronary MR Angiography and Bypass MR Angiography 5.3.7.2.1  Coronary MR Angiography The small size and fast motion of the coronary arteries puts any non-invasive diagnostic imaging modality to the test. Both MRI and multislice computed tomography (MSCT) are being tested to determine whether they can depict the coronary arteries with a sufficiently high temporal and spatial resolution for a constant and adequate image quality (Nikolaou et al. 2003). Coronary MRI is going through constant developments, including new acquisition techniques such as parallel imaging (Park et al. 2005), higher field strengths (Sommer et al. 2005), intravascular contrast agents (Herborn et al. 2004), freebreathing and refined navigator techniques (Jahnke et al. 2004), whole-heart three-dimensional (3D) applications (Sakuma et al. 2005a), and others. On the other hand, MSCT is making considerable technical progress, with the innovation cycles between scanner generations becoming shorter and shorter. While the first 4-slice CT system was introduced in 1998, and 16-slice followed in 2002, already 2 years later, in 2004, the 64-slice generation was at hand (Nikolaou et al. 2004). And just recently, the first dual-source CT (DSCT) has been announced, considerably improving temporal resolution, being introduced to clinical practice in early 2006 (Achenbach et al. 2006). However, despite the ease and robustness of modern coronary MSCT angiography, magnetic resonance coronary angiography (MRCA) offers several advantages for coronary imaging. MR does not use ionizing radiation and does not necessarily require the injection of a contrast agent (e.g., using non-contrast-enhanced, timeof-flight techniques). Several MR techniques have been proposed for the detection of coronary stenosis with MR, and further development is ongoing. Still, the initial sensitivities and specificities reported in the first work on coronary MRA more than 12 years ago, ranging above 90% (Manning et al. 1993), could not be reproduced on a regular basis in the following years (Table 5.3.13). A recent meta-analysis (Danias et al. 2004), identifying all studies (MEDLINE and EMBASE) that evaluated CAD by both MRCA and conventional angiography in ≥10 subjects during the period 1991 to January 2004, analyzed diagnostic accuracy at the segment, vessel, and subject level. Overall, 39 studies were included, with data on 4,620 segments (993 subjects). Sensitivity and specificity for detection of CAD were 73 and 86%, respectively. Vessel-level analyses (16 studies, 2,041 vessels) showed a 75% sensitivity and 85% specificity. Subject-level analyses (13 studies, 607 subjects) showed an 88% sensitivity and 56% specificity. The authors of this meta-analysis concluded that in evaluable segments of the native coronary arteries, CMRA has moderately high sensitivity for detecting significant proximal stenoses and may have value for exclu-

sion of significant multi-vessel CAD in selected subjects considered for diagnostic catheterization. In contrast to coronary MSCT angiography, enabling coverage of the whole heart in just a few seconds, so far, using MRCA, full coverage of the coronary arteries within a reasonable amount of time is difficult to achieve. However, a recent study has reported on a whole-heart 3D acquisition technique (Sakuma et al. 2005a) using a steady-state free precession sequence with free breathing. In this study, MR angiography was successfully completed in 34 of 39 patients (87%); the average imaging time was 13.8 ± 3.8min, and the sensitivity and specificity of MR angiography for detecting significant stenosis were 82 and 91%, respectively. Other future developments in the area of coronary MR angiography include higher field strengths (3 T) (Stuber et al. 2002) and improved contrast techniques, such as balanced steady-state free-precession techniques (So et al. 2005), radial-imaging techniques (Bi et al. 2005), and improved navigator gating, enabling free breathing of the subject during examination (Fig. 5.3.42) (Park et al. 2005). Finally, new intravascular contrast agents may provide the long-awaited boost for reliable MRCA. Initial studies in subjects using such substances have shown promising results (Klein et al. 2000).

Fig. 5.3.42  Healthy male volunteer. Double-oblique coronal multiplanar reformatted image of the left anterior descending coronary artery (arrow, LAD) including a diagonal branch (arrowhead), obtained in a healthy adult subject, with T2-weighted navigator-gated free-breathing three-dimensional steady-state free precession coronary MR angiography by using radial kspace sampling. Ao aorta, LV left ventricle. (Courtesy of PD Dr. Elmar Spuentrup Department of Diagnostic Radiology, University Hospital, Aachen Technical University. Modified from Spuentrup et al. 2004, with permission)

767

768

5  Thorax and Vasculature Table 5.3.13  Diagnostic accuracy of coronary artery magnetic resonance angiography (study populations >20) Author

Patient

Technique

Assessable %  (no. of segments)

Sensitivity %  (no. of segments)

Specificity %  (no. of segments)

2D breath hold Manning et al. 1993

39

GE

98 (147/150)

90 (47/52)

92 (87/95)

Pennell et al. 1996

39

GE

NA

85 (47/55)

NA

Post et al. 1997

35

GE

89 (125/140)

63 (22/35)

89 (80/90)

van Geuns et al. 2000

38

GE

69 (187/272)

68 (21/31)

97 (151/156)

Regenfus et al. 2000

50

GE

77 (268/350)

86 (48/56)

91 (193/212)

Jahnke et al. 2004

40

SSFP

45 (143/320)

63 (12/19)

82 (102/124)

Kessler et al. 1997

73

GE

52 (236/455)

65 (28/43)

88 (169/193)

Sandstede et al. 1999

30

SE

77 (92/120)

81 (30/37)

89 (49/55)

van Geuns   et al. 1999

32

GE

74 (151/203)

50 (13/26)

91 (114/125)

Nikolaou et al. 2001

40

GE

74 (207/280)

71 (84/118)

56 (91/162)

Sardanelli et al. 2000

42

GE

86 (234/273)

82 (55/67)

89 (149/167)

109

GE

86 (374/434)

83 (78/94)

73 (204/280)

Wittlinger et al. 2002

25

SE

85 (102/120)

75 (18/24)

100 (78/78)

Regenfus et al. 2002

32

GE

69 (155/224)

60 (15/25)

88 (115/130)

Bogaert et al. 2003

21

TFE

72 (134/186)

56 (15/27)

83 (89/107)

Ikonen et al. 2003

69

GE

84 (233/276)

75 (64/85)

62 (92/148)

Jahnke et al. 2004

40

SSFP

79 (254/320)

72 (26/36)

92 (200/218)

3D breath hold

3D navigator

Kim et al. 2001

GE gradient echo, SSFP steady-state free precession, SE spin echo, NA not available

In conclusion, in spite of these recent technical improvements, MRCA will not soon equal the versatility or quantitative accuracy of catheter-based imaging techniques, but it may allow for non-invasive detection and exclusion of coronary obstructions, especially on a perpatient basis. Still, MRCA currently seems unsuited for the assessment of disease progression in patients with typical angina or unequivocally demonstrated myocardial ischemia on exercise testing or following percutaneous coronary intervention. Patients with a high pretest probability of having significant CAD are still best served by conventional coronary angiography. Based on these data and limitations, the current use of non-invasive MR angiography should focus on the exclusion of significant coronary disease in patients without a history of significant CAD and a low to intermediate pretest probability

(Schuijf et al. 2005). MRA might be especially promising if coronary angiography and stress perfusion or stress functional imaging of the myocardium could be combined (Foo et al. 2005). It has also been shown that MRCA can reliably be used in the detection of coronary artery anomalies (Casolo et al. 2005). The primary goal is the correct selection of suitable patient groups, to potentially reduce the number of unnecessary invasive catheter-based coronary angiographies in the future. 5.3.7.2.2  MR Angiography of Coronary Bypass Grafts Although percutaneous coronary intervention (PCI) including placement of coronary artery stents has been

5.3  Heart

Fig. 5.3.43a–c  65-year-old male patient with known CAD after bypass surgery, referred for bypass MR angiography to assess bypass graft patency. Contrast-enhanced three-dimensional MR angiography shows two venous bypass grafts (a arrows). One sequential bypass-graft is inserted on a diagonal branch of the left anterior descending coronary artery and on a marginal branch of the left circumflex coronary artery (b arrows). The second ve-

nous bypass graft is inserted on the distal right coronary artery (c arrows). Both bypass grafts are patent. Ao aorta, LV left ventricle, PA pulmonary artery, RV right ventricle (Courtesy of PD Dr. Karl-Friedrich Kreitner, Department of Radiology, Johannes Gutenberg-University of Mainz. Modified from Kreitner et al. 2004, with permission)

increasingly performed in recent years, coronary artery bypass graft surgery is still among the most common procedures performed in patients with CAD. However, early graft occlusion within the first months after surgery occurs in up to 10% of patients, and atherosclerotic processes often lead to late stenoses or occlusions in the majority of grafts (Fitzgibbon et al. 1996). In comparison with the native coronary arteries, bypass grafts are relatively easy to image because of their reduced overall motion and their larger lumen (Kreitner et al. 2004). Furthermore, their predictable and less convoluted course has allowed imaging of bypass grafts even with conventional magnetic resonance techniques. With schematic knowledge of the origin and touchdown site of each graft, conventional free-breathing ECG-gated two-dimensional spin-echo van (Geuns et al. 2000) and two-dimensional gradient-echo (Galjee et al. 1996) magnetic resonance in the transverse plane have been utilized to reliably assess bypass graft patency. Patency is generally determined by visualizing a patent graft lumen in at least two contiguous transverse levels along its expected course. If a patent graft lumen is not seen at any level, the graft is considered occluded. Contrast-enhanced CMR and three-dimensional non-contrast approaches offer slight improvements (Fig. 5.3.43; Table 5.3.14) (Molinari et al. 2000; Vrachliotis et al. 1997; Wintersperger et al. 1998).

The accuracy of ECG-gated SSFP sequences appears to be similar to that of spin-echo and gradient-echo approaches (Bunce et al. 2003). Data suggest that the use of phase velocity mapping or flow reserve for assessment of coronary artery bypass graft flow may be superior to graft imaging (Langerak et al. 2003). The limitations of coronary MRI bypass graft assessment include difficulties related to local signal loss/artifact due to implanted metallic objects (hemostatic clips, sternal wires). However, data suggest that with improving acquisition techniques, MRA of coronary artery bypass grafts might become a clinically useful option in patients with suspected bypass occlusion, competing with the promising results of the latest multislice CT studies (Anders et al. 2006). 5.3.7.3  Myocardial Function 5.3.7.3.1  Rest Cine MRI of Cardiac Function An essential feature in the assessment of ischemic myocardium includes the measurement of global and regional myocardial contractile function. Myocardium may become dysfunctional as a result of either acute reversible ischemic insults (myocardial stunning) or chronic gradual decrease in blood supply (hibernating myocardium).

769

770

5  Thorax and Vasculature Table 5.3.14  Sensitivity, specificity, and accuracy of coronary magnetic resonance imaging for assessment of coronary artery bypass graft patency Investigator

Technique

Grafts

Patency

Sensitivity

Specificity

Accuracy

Rubinstein et al. [54]

2D-SE

47

62%

90%

72%

83%

Galjee et al. [33K]

2D-SE

98

74%

98%

85%

89%

White et al. [32]

2D GRE

28

50%

93%

86%

89%

Galjee et al. [33K]

2D GRE

98

74%

98%

88%

96%

Molinari et al. [36]

3D GRE

51

76.5%

91%

97%

96%

Engelmann et al. [35]

CE-3D GRE

96 SVG

66%

92%

85%

89%

37 IMA

100%

100%

100%

Wintersperger et al. [38]

CE-3D GRE

39

87%

97%

100%

97%

Vrachliotis et al. [37]

CE-3D GRE

45

67%

93%

97%

95%

Langerak et al. [40]

3D TFE

166

71%

96%

92%

Myocardial wall thinning at rest is a reliable feature of scarred tissue resulting from extensive myocardial injury. The entire process of scarring usually takes about 4 months to develop from the time of acute injury (Baer et al. 1996). Baer et al. (1995) used a cut-off value of 5.5 mm as the lower limit of viable mean end-diastolic myocardial wall thickness. Besides this information on wall thickness and motility defects after myocardial infarction, however, cine MR imaging at rest is not suitable for the assessment of patients with suspicion of CAD, as the sensitivity for the detection of significant coronary artery stenoses is limited (Poon et al. 2002). 5.3.7.3.2  Stress Cine MRI of Cardiac Function In addition to cine rest MRI, cine stress MRI can evaluate additional important parameters. In recent years, pharmacological stress testing has evolved as an alternative to physical exercise for the detection of inducible myocardial ischemia or to assess myocardial viability. The diagnostic performance of various stress tests may vary considerably, depending on the imaging modality (echocardiography, MRI, nuclear techniques) or stress agent used. Routinely used pharmacological stress agents are adenosine or dipyridamole and the synthetic β-adrenergic agent dobutamine. There is controversy regarding the “optimal” pharmacological stress agent; according to guidelines, adenosine should be used mainly for myocardial perfusion measurements, whereas dobutamine is advised for the detection of stress-inducible wall motion abnormalities (IWMAs) (Paetsch et al. 2004). Cardiac MR is favorable regarding both functional

n.a.

measurements: It allows assessment of even subtle wall motion disturbances resulting from the consistently high endocardial border definition, and the measurement of myocardial perfusion can be integrated into the same examination, with the high spatial resolution of the scans facilitating the determination of the transmural extent of a regional perfusion deficit. Low- and high-dose dobutamine stress imaging (DSMR) of cardiac function can be useful. Low-dose dobutamine stimulation with infusion of 5 to 10 µg/kg per minute of dobutamine combined with cine cardiovascular MRI is a reliable means of differentiating viable from nonviable myocardium. It was shown that wall thickening during low-dose dobutamine of more than 2 mm at the first week after myocardial infarction (MI) related to a higher proportion of segments with normal WM after 6 months (Bodi et al. 2005). On the other hand, similar to dobutamine stress echocardiography (DSE), high-dose dobutamine stress MRI evaluates the contractile reserve in patients with suspected ischemic heart disease. Here doses of dobutamine of up to 40 µg/kg/min are used to diagnose the presence of coronary stenoses or cardiac ischemia. A recent study on the role of MRI in the comprehensive assessment of CAD comparing different stressing agents has confirmed that DSMR is superior to adenosine stress for the induction of IWMAs in patients with significant coronary artery disease (Paetsch et al. 2004). Compared with selective, invasive coronary angiography, Schalla et al. (2002) reported a sensitivity of 88% and a specificity of 83% for detecting significant coronary artery disease with high-dose dobutamine MRI. Recently, similarly good results have been reported in patients after revascularization. Wall et al. (2004) showed

5.3  Heart

that high-dose stress cardiovascular MR imaging can be used for follow-up of patients after coronary revascularization procedures. Diagnostic accuracy was similar to stress cardiovascular MR imaging data for patients suspected of having CAD and compared favorably with that of other established non-invasive techniques. Thus, high-dose DSMR is the method of choice for current state-of-the-art treatment regimens to detect ischemia in patients with suspected or known coronary artery disease but no history of prior myocardial infarction. DSMR has been shown to provide superior inter-study reproducibility in the assessment of clinically relevant changes in left ventricular (LV) dimensions and function as compared to echocardiography (Grothues et al. 2002. Cine cardiac MR tagging may further improve the diagnostic accuracy (Reeder et al. 2001; Paetsch et al. 2005).

infarction. Imaging modalities like ultrasound or nuclear-based imaging have one major limitation in the assessment of cardiac viability, in that none can adequately assess the transmural extent of viability in a damaged

5.3.7.3.3  Myocardial Tagging Myocardial tagging as a non-invasive method for the assessment of regional myocardial tissue function was introduced in 1988 (Zerhouni et al. 1988). Non-invasive alternatives such as echocardiography or conventional cine MR imaging suffer from the absence of reliably traceable landmarks. With tagging MR imaging, the myocardium is labeled by a spatial modulation of magnetization (Axel and Dougherty 1989) that leads to stripes or grids (called “tags”) that appear fixed to the myocardium (Fig. 5.3.44). The deformation of these tags in the cardiac cycle allows an assessment of the regional myocardial tissue function, including left ventricular rotation, radial contraction, and circumferential shortening (Sandstede et al. 2002; Maier et al. 1992). Using dedicated software, the quantification of left ventricular wall motion using tagged MR images allows characterization and follow-up of changes of left ventricular wall motion in various diseases of the heart (Johnson et al. 2004). 5.3.7.4  Myocardial Viability: Delayed-Enhancement MRI In patients with coronary artery disease and left ventricular dysfunction, the distinction between viable and nonviable myocardium is crucial for the prediction of functional recovery after surgical or interventional revascularization. Myocardial infarction is caused by an occlusion of a coronary artery. If duration of coronary occlusion is extended beyond 20 min, a wave front of necrosis marches from subendocardium over mid-myocardium to subepicardium over time (Kloner and Jennings 2001). Thus, the extent of MI is not only related to the size of the myocardium that is supplied by the occluded coronary artery, but also to the duration of occlusion that influences the development of transmural or non-transmural

Fig. 5.3.44a,b  55-year-old male patient with ischemic cardiomyopathy. Several tagged MR images obtained in left ventricular short-axis view obtained over the cardiac cycle (a). An enlarged tagged MR image obtained at same level as a in mid-systole depicts a reduced systolic contraction and rotational movement (b). Analysis of rotation and contraction calculated by subtraction of angle or radius, respectively, is performed at end-diastole and end-systole. This way, a quantitative analysis of regional wall motion is feasible (Courtesy of Dr. Thorsten Johnson, Department of Clinical Radiology, Ludwig-Maximilians-University of Munich, Modified from Johnson et al. 2004, with permission)

771

772

5  Thorax and Vasculature

myocardium (Poon et al. 2002). Delayed-enhancement magnetic resonance imaging (DE MRI) is a rather novel imaging technique that is rapidly becoming regarded as the most accurate myocardial viability study (Kim et al. 1999; Huber et al. 2005). Before a discussion of DE MRI or any other diagnostic modality, a definition of myocardial viability is necessary. Simply stated, the most precise definition of myocardial viability is the presence of living myocytes. Likewise, the most precise definition of nonviability is the presence of necrotic myocytes, or the replacement of myocytes by fibrosis or scar tissue. In principle, all ischemic events that occur before cell death are reversible, and the further a diagnostic technique deviates from directly assessing whether myocytes are present or absent, the more imprecise the technique becomes in determining myocardial viability. Here, DE MRI offers a distinct advantage over indirect assessment of viability, e.g., assessing residual function in an infarct-related myocardial territory, as DE MRI enables direct depiction of non-viable myocytes, i.e., of myocardial necrosis. Infarcted myocardium exhibits delayed hyper-enhancement after administration of gadolinium-based contrast material and can be imaged by using inversion-recovery (IR) techniques typically 10–30 min after administration of contrast material (Simonetti et al. 2001). In DE MRI, retention of contrast agent in the myocardium indicates the presence of nonviable myocardium regardless of the age of the infarct, i.e., delayed hyper-enhancement is observed in both acute and chronic MI. Analysing the underlying mechanisms, recent studies have postulated that in the acute setting, irreversible ischemic myocyte injury results in a loss of the cell membrane integrity, resulting in an edema of the infarct zone as well as an enlarged distribution space for the MR contrast agent within the necrotic myocytes (Sandstede 2003). This acute phase of myocardial injury is followed by subsequent leukocyte infiltration within the first 4 days and penetration of blood capillaries and connective tissue from the periphery. Up until the third week, muscle fibers are removed and then replaced by collagen fibers leading to an enlarged extracellular space within the following weeks. Thus, in a chronic infarction, the contrast agent has a larger distribution space due to increased interstitial space in fibrotic scar formation after MI. As described above, subendocardial involvement is always observed in an area subtended by the coronary artery. The degree of transmural extension depends on the severity of infarction (i.e., the duration of coronary artery occlusion). However, it is very important to observe that delayed contrast enhancement is not specific for myocardial necrosis, but can also be observed in other myocardial pathologies. A distinction from other causes of fibrosis or myocyte necrosis resulting in delayed contrast enhancement, such as dilated or hypertrophic cardiomyopathy or myocarditis, is possible assessing distinct distribution patterns of the contrast agent: while myocardial ischemia will always involve the

subendocardium, the distribution of late enhancement in other pathologies is typically patchy and/or in the midwall (Barkhausen et al. 2004a). In animal studies, MR images acquired 30 min after administration of contrast material showed hyperenhancement of myocardial tissue, with excellent spatial correlation between the hyperenhancing areas and the necrotic areas identified at histopathological examination after staining of the entire left ventricle with triphenyltetrazolium chloride. It was also shown that transient ischemia with reperfusion caused no hyperenhancement (Kim et al. 1999). Due to its high spatial resolution, DE MRI can be used to differentiate transmural and non-transmural myocardial infarction. Kim et al. (2000) correlated the percent of transmural extent of late enhancement with the recovery of systolic wall thickening after surgical or percutaneous revascularization. The smaller the area of late enhancement, the higher was the probability for mechanical improvement. Complete transmural infarctions could be distinguished from non-transmural infarctions by steps smaller than 25% of the thickness of the left ventricular myocardium. The contractility of dysfunctional but viable myocardium (i.e., no hyper-enhancement) can potentially improve after revascularization, even if the transmural extent of hyperenhancement is 50%. In recent years, a number of studies have dealt with the optimal imaging technique for DE MRI. It is known that regions of myocardial infarction exhibit higher signal intensity than do regions of normal myocardium on un-enhanced T2-weighted MR images (Higgins et al. 1983) and contrast-enhanced T1-weighted MR images (Oshinski et al. 2001). Since the initial investigations, many studies have been performed with a variety of pulse sequences to differentiate between infarcted and normal myocardium and to distinguish between reversible and irreversible ischemic injury. Simonetti et al. (2001) compared 10 different pulse sequences acquired with unenhanced and contrast-enhanced administration and found the greatest differences in regional myocardial signal intensity for the breath-hold segmented IR turbo FLASH (fast low-angle shot) sequence. In the literature, this is the pulse sequence that is the most investigated and established (Klein et al. 2002). In contrast-enhanced IR imaging, the inversion time (TI) hast to be adapted individually for each examination to null the signal of the normal myocardium, thus ensuring optimal contrast between infarcted and viable myocardium. There are different methods available to optimize the TI value. Images can be acquired with different TI values, and the best TI value can be chosen by trial and error. This can be time consuming. An alternative approach is to use a TI-optimizing sequence that acquires images with different TI values and reduced spatial resolution (Moran et al. 2002). Recently, to avoid this optimization procedure, Huber et al. (2005) have investigated the use of phase-sensitive IR (PSIR), demonstrating the advantages of the technique

5.3  Heart

for visualization of myocardial infarction without having to optimize the inversion time. In an additional study by Huber et al. (2006) it was shown, that by implementing this PSIR technique at 3 Tesla (as compared to 1.5 T), a superior contrast-to-noise (CNR) ratio can be achieved. The authors concluded that by combining PSIR imaging at 3 T with parallel imaging techniques, both high contrast and high spatial resolution could be advantageous to detect small subendocardial infarctions or contrast enhancement in other pathologies such as sarcoidosis, various cardiomyopathies (Fig. 5.3.45) and myocarditis. 5.3.7.5  Myocardial Perfusion Standard MR contrast-agents such as gadolinium chelate (e.g., Gd-DTPA) can be useful as contrast and perfusion agents at the same time (Fig. 5.3.46). The functional significance of any coronary artery lesion can be evaluated based on the characteristics of blood flow (and, therefore, myocardial enhancement) at rest and during stress testing using vasodilator agents like adenosine (Winters­ perger et al. 1999). Non-ischemic myocardium exhibits gradual signal enhancement and washout with the pas-

Fig. 5.3.45  A 67-year-old male patient with two myocardial infarctions in the perfusion territory of the right coronary artery and the circumflex artery (arrows). Several short-axis images, acquired with a phase-sensitive single-shot trueFISP sequence at 3 T, reveal hyperenhanced myocardium in the inferolateral and inferoseptal segments corresponding to myocardial infarctions. LV left ventricle, RV right ventricle (Courtesy of Dr. Armin Huber, Department of Clinical Radiology, Ludwig-Maximilians-University of Munich, modified from Huber et al. 2006, with permission)

sage of the T1-enhancing contrast agent (Wintersperger et al. 1999). Administration of adenosine accentuates the baseline perfusion defect in ischemic myocardium and helps differentiate ischemic from non-ischemic myocardium. Analogous to the detection of contractile reserve with low-dose dobutamine administration, vasodilatorinduced perfusion defects detect myocardial perfusion reserve. A valuable feature of cardiac MRI may be its ability to differentiate between different myocardial layers because of its high spatial resolution. This could allow differentiation between subendocardial and transmural perfusion defects. Subendocardial ischemia is believed to be the first indication of compromise of myocardial blood flow. Subendocardial perfusion defects also tend to be reversible and are typically present in segments that show normal or only mildly reduced contractility at rest. Fullthickness perfusion deficits, on the other hand, tend to be fixed and associated with severe myocardial dysfunction at rest (Sensky et al. 2000). This level of diagnostic detail is usually difficult to achieve in clinical practice, because established imaging modalities such as 201Tl scintigraphy and PET do not have sufficient spatial resolution to distinguish between the epicardial and subendocardial layers. MRI may be able to detect subendocardial perfusion defects in coronary territories supplied by stenotic but still-patent vessels. Additionally, the technique is attractive because it does not expose patients to ionizing radiation and because the acquired data can be analyzed by using quantitative or semiquantitative methods (Hsu et al. 2006). Until recently, however, the use of this technique was limited to the acquisition of not more than one image section per heartbeat, owing to the relatively slow data acquisition with conventional pulse sequences (Manning et

Fig. 5.3.46a,b  56-year-old woman with a significant stenosis of the right coronary artery (RCA). Representative perfusion MR images are shown for rest (a) and stress (b) examinations. No perfusion deficit can be seen in rest examination; however, after injection of 140 (µg/kg body weight)/min of adenosine, a marked perfusion deficit is visible in the inferoseptal and inferior part of the left ventricular wall (Courtesy of Dr. Bernd J. Wintersperger, Department of Clinical Radiology, LudwigMaximilians-University of Munich)

773

774

5  Thorax and Vasculature

al. 1991). The resulting incomplete myocardial coverage or low sampling rates may have limited the technical and diagnostic performance of myocardial perfusion MR imaging in early studies (Thiele et al. 2003). In recent years, parallel imaging has matured to a widespread application especially in cardiovascular MRI, as acquisition times can be reduced significantly using this approach. This way, multi-section MR perfusion imaging becomes feasible, acquiring data from several slices through the myocardium during the first pass of the contrast agent (Plein et al. 2005). Several recent studies have reported on high diagnostic accuracies of first-pass stress perfusion MRI for the detection of significant coronary artery stenoses as compared to SPECT or catheter angiography (Fenchel et al. 2005; Sakuma et al. 2005b). Future developments increasing the signal-to-noise ratio (SNR) of MR perfusion sequences, enabling greater myocardial coverage and faster image acquisition, as well as refined post-processing tools for a quantitative assessment of myocardial perfusion will be aimed at increasing the clinical applicability of stress perfusion MRI in the cardiologic routine workup of patients with CAD (Barkhausen et al. 2004b). 5.3.7.6  Future Perspectives and Conclusions 5.3.7.6.1  One-Stop-Shop of Myocardial Viability, Function, and Perfusion Comprehensive assessment of sequelae of coronary artery disease includes outlining of remodeled anatomy and function of the ventricles, assessment of myocardial perfusion at rest and after stress, detection of myocardial viability, and definition of the coronary anatomy and valvular pathology (Foo et al. 2005; Constantine et al. 2004). This plethora of information allows informed decision making in patients’ management. Cardiac MRI has been proposed as a non-invasive modality that could provide a one-stop evaluation of coronary artery disease. A number of integrated protocols have been proposed for comprehensive assessment of coronary artery disease (Table 5.3.15). Sensky et al. (2000) developed and tested a protocol that evaluates myocardial perfusion, function, and hibernation in the same setting over a period of about 50 min. This imaging protocol takes advantage of the short half-life of adenosine (as a vasodilating agent in perfusion imaging) and dobutamine (as a positive inotropic agent in cine MR functional imaging) and allows for the infusion of both agents in a relatively short interval to complete the function and perfusion analysis of the patient. This combined approach provides all the data required for the assessment of global and regional wall motion and the presence or absence of viable myocardium. Plein et al. (2002) have proposed a cardiac MRI protocol integrating multiphase gradient-echo cine magnetic resonance for function, first-pass myocardial perfu-

Table 5.3.15  Time sequence for a combined perfusion and function stress MR imaging protocol Time (min)

Infusion

MRI sequences

0

–

–

3

Adenosine

Stress perfusion

6

Adenosine

Stress perfusion

9

Adenosine

Stress perfusion

12

–

Out of scanner

15

–

Out of scanner

18

–

Cine rest

21

–

Cine rest

24

–

Perfusion rest

27

Dobutamine

Cine 5

30

Dobutamine

Cine 5

33

Dobutamine

Cine 5

36

Dobutamine

Cine 10

Courtesy of Prof. Dr. Michael Poon, Associate Clinical Professor, Medicine, Mount Sinai School of Medicine, New York, and Director of Cardiology, Cabrini Medical Center, New York. Modified from Poon et al. 2002, with permission

sion imaging at rest and during adenosine stress, delayed contrast enhancement for assessment of myocardial viability, and 3D respiratory navigator-gated coronary MR angiography. The entire protocol has been undertaken in about 1 h. A similar integrated MRI protocol has been tried in risk assessment of patients presenting with chest pain at the emergency department with acute coronary syndrome, with a sensitivity and specificity above 90% for the correct detection of significant CAD (Chiu et al. 2003). In a study comparing MRI with FDG-PET in the assessment of cardiac viability, Lauerma et al. (2000) reported that the combination of dobutamine stress, firstpass perfusion, and late contrast enhancement was the best approach for detecting cardiac viability in patients with multi-vessel coronary artery disease. 5.3.7.6.2  Conclusions The clinical utility of cardiac MRI in the diagnosis of CAD is increasing rapidly, due to the advances in the design of magnets and coils, gating techniques, pulse sequences, and post-processing software. Advances in fast cine car-

5.3  Heart

diovascular MRI sequences permit rapid acquisition of cine images in one quick breath hold. Development of user-friendly software for real-time imaging allows continuous monitoring of changes in cardiac motion allows for during inotropic stimulation. New real-time imaging techniques such as radial imaging may further shorten the image acquisition time and may eliminate the need for breath hold and perhaps the need for ECG gating. Currently the major limitations of CMRI include high initial equipment and installation costs, a small number of trained specialists in operating and interpreting the images, lack of MRI scanners equipped with specialized accessories for cardiac scanning, and the relatively lengthy time for image acquisition. However, cardiac dysfunction due to myocardial ischemia is a common clinical problem that is going to affect more and more people every day, as the average age of the population continues to increase. The ability to assess function, perfusion, and viability at one setting in less than 1 h has made cardiac MRI an appealing and cost-effective modality for the initial evaluation of ischemic heart disease. As the general acceptance and third-party reimbursement rate for performing cardiac MRI continues to increase, this imaging tool may finally become a part of everyday clinical cardiology practice.

6. 7.

8.

9.

10.

11.

12.

References 1.

2.

3.

4.

5.

Achenbach S, Ropers D, Kuettner A, Flohr T, Ohnesorge B, Bruder H, Theessen H, Karakaya M, Daniel WG, Bautz W, Kalender WA, Anders K (2006) Contrast-enhanced coronary artery visualization by dual-source computed tomography-Initial experience. Eur J Radiol 57:331–335 Anders K, Baum U, Schmid M, Ropers D, Schmid A, Pohle K, Daniel WG, Bautz W, Achenbach S (2006) Coronary artery bypass graft (CABG) patency: Assessment with highresolution submillimeter 16-slice multidetector-row computed tomography (MDCT) versus coronary angiography. Eur J Radiol 57:336–444 Axel L, Dougherty L (1989) Heart wall motion: improved method of spatial modulation of magnetization for MR imaging. Radiology 172:349–350 Baer FM, Voth E, Schneider CA, Theissen P, Schicha H, Sechtem U (1995) Comparison of low-dose dobutaminegradient-echo magnetic resonance imaging and positron emission tomography with 18Ffluorodeoxyglucose in patients with chronic coronary artery disease. A functional and morphological approach to the detection of residual myocardial viability. Circulation 91:1006–1015 Baer FM, Voth E, Deutsch HJ, Schneider CA, Horst M, de Vivie ER, Schicha H, Erdmann E, Sechtem U (1996) Predictive value of low dose dobutamine transesophageal echocardiography and fluorine-18 fluorodeoxyglucose positron emission tomography for recovery of regional left ventricular function after successful revascularization. J Am Coll Cardiol 28:60–69

13.

14. 15.

16.

17.

18.

Barkhausen J, Hunold P, Waltering KU (2004a) MRI in coronary artery disease. Eur Radiol 14:2155–2162 Barkhausen J, Hunold P, Jochims M, Debatin JF (2004b) Imaging of myocardial perfusion with magnetic resonance. J Magn Reson Imaging 19:750–757 Bi X, Park J, Larson AC, Zhang Q, Simonetti O, Li D (2005) Contrast-enhanced 4D radial coronary artery imaging at 3.0 T within a single breath-hold. Magn Reson Med 54:470–475 Bodi V, Sanchis J, Lopez-Lereu MP, Losada A, Nunez J, Pellicer M, Bertomeu V, Chorro FJ, Llacer A (2005) Usefulness of a comprehensive cardiovascular magnetic resonance imaging assessment for predicting recovery of left ventricular wall motion in the setting of myocardial stunning. J Am Coll Cardiol 46:1747–1752 Bogaert J, Kuzo R, Dymarkowski S, Beckers R, Piessens J, Rademakers FE (2003) Coronary artery imaging with realtime navigator three-dimensional turbo-field-echo MR coronary angiography: initial experience. Radiology 226:707–716 Bunce NH, Lorenz CH, John AS, Lesser JR, Mohiaddin RH, Pennell DJ (2003) Coronary artery bypass graft patency: assessment with true fast imaging with steady-state precession versus gadolinium-enhanced MR angiography. Radiology 227:440–446 Casolo G, Del Meglio J, Rega L, Manta R, Margheri M, Villari N, Gensini G (2005) Detection and assessment of coronary artery anomalies by three-dimensional magnetic resonance coronary angiography. Int J Cardiol 103:317–322 Chiu CW, So NM, Lam WW, Chan KY, Sanderson JE (2003) Combined first-pass perfusion and viability study at MR imaging in patients with non-ST segment-elevation acute coronary syndromes: feasibility study. Radiology 226:717–722 Constantine G, Shan K, Flamm SD, Sivananthan MU (2004) Role of MRI in clinical cardiology. Lancet 363:2162-2171 Danias PG, Roussakis A, Ioannidis JP (2004) Diagnostic performance of coronary magnetic resonance angiography as compared against conventional X-ray angiography: a meta-analysis. J Am Coll Cardiol 44:1867–1876 Engelmann MG, Knez A, von Smekal A, Wintersperger BJ, Huehns TY, Hofling B, Reiser MF, Steinbeck G (2000) Non-invasive coronary bypass graft imaging after multivessel revascularisation. Int J Cardiol 76:65–74 Fenchel M, Helber U, Kramer U, Stauder NI, Fra­ now A, Claussen CD, Miller S (2005) Detection of regional myocardial perfusion deficit using rest and stress perfusion MRI: a feasibility study. AJR Am J Roentgenol 185:627–635 Fitzgibbon GM, Kafka HP, Leach AJ, Keon WJ, Hooper GD, Burton JR (1996) Coronary bypass graft fate and patient outcome: angiographic follow-up of 5,065 grafts related to survival and reoperation in 1,388 patients during 25 years. J Am Coll Cardiol 28:616–626

775

776

5  Thorax and Vasculature 19. Foo TK, Ho VB, Saranathan M, Cheng LQ, Sakuma H, Kraitchman DL, Wu KC, Bluemke DA (2005) Feasibility of integrating high-spatial-resolution 3D breath-hold coronary MR angiography with myocardial perfusion and viability examinations. Radiology 235:1025–1030 20. Galjee MA, van Rossum AC, Doesburg T, Van Eenige MJ, Visser CA (1996) Value of magnetic resonance imaging in assessing patency and function of coronary artery bypass grafts. An angiographically controlled study. Circulation 93:660–666 21. Geuns RJ van, de Bruin HG, Rensing BJ, Wielopolski PA, Hulshoff MD, van Ooijen PM, Oudkerk M (1999) Magnetic resonance imaging of the coronary arteries: clinical results from three dimensional evaluation of a respiratory gated technique. Heart 82:515–519 22. Geuns RJ van, Wielopolski PA, de Bruin HG, Rensing BJ, Hulshoff M, van Ooijen PM, de Feyter PJ, Oudkerk M (2000) MR coronary angiography with breath-hold targeted volumes: preliminary clinical results. Radiology 217:270–277 23. Grothues F, Smith GC, Moon JC, Bellenger NG, Collins P, Klein HU, Pennell DJ (2002) Comparison of interstudy reproducibility of cardiovascular magnetic resonance with two-dimensional echocardiography in normal subjects and in patients with heart failure or left ventricular hypertrophy. Am J Cardiol 90:29–34 24. Herborn CU, Schmidt M, Bruder O, Nagel E, Shamsi K, Barkhausen J (2004) MR Coronary angiography with SH L 643 A: initial experience in patients with coronary artery disease. Radiology 233:567–573 25. Higgins CB, Herfkens R, Lipton MJ, Sievers R, Sheldon P, Kaufman L, Crooks LE (1983) Nuclear magnetic resonance imaging of acute myocardial infarction in dogs: alterations in magnetic relaxation times. Am J Cardiol 52:184–188 26. Hsu LY, Rhoads KL, Holly JE, Kellman P, Aletras AH, Arai AE (2006) Quantitative myocardial perfusion analysis with a dual-bolus contrast-enhanced first-pass MRI technique in humans. J Magn Reson Imaging 23:315–322 27. Huber AM, Schönberg SO, Hayes C, Spannagl B, Engelmann MG, Franz WM, Reiser MF (2005) Phase-sensitive inversion-recovery MR imaging in the detection of myocardial infarction. Radiology 237:854–860 28. Huber A, Bauner K, Wintersperger BJ, Reeder SB, Stadie F, Mueller E, Schmidt M, Winnik E, Reiser MF, Schönberg SO (2006) Phase-sensitive inversion recovery (PSIR) single-shot TrueFISP for assessment of myocardial infarction at 3 Tesla. Invest Radiol 41:148–153 29. Ikonen AE, Manninen HI, Vainio P, Hirvonen TP, Vanninen RL, Matsi PJ, Soimakallio S, Hartikainen JE (2003) Three-dimensional respiratory-gated coronary MR angiography with reference to X-ray coronary angiography. Acta Radiol 44:583–589 30. Jahnke C, Paetsch I, Schnackenburg B, Bornstedt A, Gebker R, Fleck E, Nagel E (2004) Coronary MR angiography with steady-state free precession: individually adapted breath-hold technique versus free-breathing technique. Radiology 232:669–676

31. Johnson T, Hahn D, Sandstede J (2004) [Quantitative analysis of left ventricular wall motion with MRI tagging]. Radiologe 44:158–163 32. Kessler W, Achenbach S, Moshage W, Zink D, Kroeker R, Nitz W, Laub G, Bachmann K (1997) Usefulness of respiratory gated magnetic resonance coronary angiography in assessing narrowings > or = 50% in diameter in native coronary arteries and in aortocoronary bypass conduits. Am J Cardiol 80:989–993 33. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J, Finn JP, Klocke FJ, Judd RM (1999) Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 100:1992–2002 34. Kim RJ, Wu E, Rafael A, Chen EL, Parker MA, Simonetti O, Klocke FJ, Bonow RO, Judd RM (2000) The use of contrastenhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 343:1445–1453 35. Kim WY, Danias PG, Stuber M, Flamm SD, Plein S, Nagel E, Langerak SE, Weber OM, Pedersen EM, Schmidt M, Botnar RM, Manning WJ (2001) Coronary magnetic resonance angiography for the detection of coronary stenoses. N Engl J Med 345:1863–1869 36. Klein C, Nagel E, Schnackenburg B, Bornstedt A, Schalla S, Hoffmann V, Lehning A, Fleck E (2000) The intravascular contrast agent Clariscan (NC 100150 injection) for 3D MR coronary angiography in patients with coronary artery disease. MAGMA 11:65–67 37. Klein C, Nekolla SG, Bengel FM, Momose M, Sammer A, Haas F, Schnackenburg B, Delius W, Mudra H, Wolfram D, Schwaiger M (2002) Assessment of myocardial viability with contrast-enhanced magnetic resonance imaging: comparison with positron emission tomography. Circulation 105:162–167 38. Kloner RA, Jennings RB (2001) Consequences of brief ischemia: stunning, preconditioning, and their clinical implications: part 2. Circulation 104:3158–3167 39. Kreitner KF, Ehrhard K, Kunz RP, Abegunewardene N, Oberholzer K, Horstick G, Hake U, Mildenberger P, Thelen M (2004) [Non-invasive assessment of coronary artery bypass grafts—an update]. RoFo 176:1079–1088 40. Langerak SE, Vliegen HW, Jukema JW, Kunz P, Zwinderman AH, Lamb HJ, van der Wall EE, De Roos A (2003) Value of magnetic resonance imaging for the noninvasive detection of stenosis in coronary artery bypass grafts and recipient coronary arteries. Circulation 107:1502–1508 41. Lauerma K, Niemi P, Hanninen H, Janatuinen T, VoipioPulkki LM, Knuuti J, Toivonen L, Makela T, Makijarvi MA, Aronen HJ (2000) Multimodality MR imaging assessment of myocardial viability: combination of first-pass and late contrast enhancement to wall motion dynamics and comparison with FDG PET-initial experience. Radiology 217:729–736

5.3  Heart 42. Maier SE, Fischer SE, McKinnon GC, Hess OM, Krayenbuehl HP, Boesiger P (1992) Evaluation of left ventricular segmental wall motion in hypertrophic cardiomyopathy with myocardial tagging. Circulation 86:1919–1928 43. Manning WJ, Atkinson DJ, Grossman W, Paulin S, Edelman RR (1991) First-pass nuclear magnetic resonance imaging studies using gadolinium-DTPA in patients with coronary artery disease. J Am Coll Cardiol 18:959–965 44. Manning WJ, Li W, Edelman RR (1993) A preliminary report comparing magnetic resonance coronary angiography with conventional angiography [Comments]. N Engl J Med 328:828–832 45. Molinari G, Sardanelli F, Zandrino F, Balbi M, Masperone MA (2000) Value of navigator echo magnetic resonance angiography in detecting occlusion/patency of arterial and venous, single and sequential coronary bypass grafts. Int J Card Imaging 16:149–160 46. Moran GR, Thornhill RE, Sykes J, Prato FS (2002) Myocardial viability imaging using Gd-DTPA: physiological modeling of infarcted myocardium, and impact on injection strategy and imaging time. Magn Reson Med 48:791–800 47. Nikolaou K, Flohr T, Knez A, Rist C, Wintersperger BJ, Johnson T, Reiser MF, Becker CR (2004) Advances in cardiac CT imaging: 64-slice scanner. Int J Cardiovasc Imaging 20:535–540 48. Nikolaou K, Huber A, Knez A, Scheidler J, Petsch R, Reiser M (2001) Navigator echo-based respiratory gating for 3DMR coronary angiography: Reduction of scan time using a slice-interpolation technique. J Comput Assist Tomogr 25:378–387 49. Nikolaou K, Poon M, Sirol M, Becker CR, Fayad ZA (2003) Complementary results of computed tomography and magnetic resonance imaging of the heart and coronary arteries: a review and future outlook. Cardiol Clin 21:639–655 50. Oshinski JN, Yang Z, Jones JR, Mata JF, French BA (2001) Imaging time after Gd-DTPA injection is critical in using delayed enhancement to determine infarct size accurately with magnetic resonance imaging. Circulation 104:2838–2842 51. Paetsch I, Jahnke C, Wahl A, Gebker R, Neuss M, Fleck E, Nagel E (2004) Comparison of dobutamine stress magnetic resonance, adenosine stress magnetic resonance, and adenosine stress magnetic resonance perfusion. Circulation 110:835–842 52. Paetsch I, Foll D, Kaluza A, Luechinger R, Stuber M, Bornstedt A, Wahl A, Fleck E, Nagel E (2005) Magnetic resonance stress tagging in ischemic heart disease. Am J Physiol 288:H2708–H2714 53. Park J, Larson AC, Zhang Q, Simonetti O, Li D (2005) High-resolution steady-state free precession coronary magnetic resonance angiography within a breath-hold: Parallel imaging with extended cardiac data acquisition. Magn Reson Med 54:1100–1106 54. Pennell DJ, Bogren HG, Keegan J, Firmin DN, Underwood SR (1996) Assessment of coronary artery stenosis by magnetic resonance imaging. Heart 75:127–133

55. Plein S, Ridgway JP, Jones TR, Bloomer TN, Sivananthan MU (2002) Coronary artery disease: assessment with a comprehensive MR imaging protocol—initial results. Radiology 225:300–307 56. Plein S, Radjenovic A, Ridgway JP, Barmby D, Greenwood JP, Ball SG, Sivananthan MU (2005) Coronary artery disease: myocardial perfusion MR imaging with sensitivity encoding versus conventional angiography. Radiology 235:423–430 57. Poon M, Fuster V, Fayad Z (2002) Cardiac magnetic resonance imaging: a “one-stop-shop” evaluation of myocardial dysfunction. Curr Opin Cardiol 17:663–670 58. Post JC, van Rossum AC, Hofman MBM, de Cock CC, Valk J, Visser CA (1997) Clinical utility of two-dimensional magnetic resonance angiography in detecting coronary artery disease. Eur Heart J 18:426–433 59. Reeder SB, Du YP, Lima JA, Bluemke DA (2001) Advanced cardiac MR imaging of ischemic heart disease. Radiographics 21:1047–1074 60. Regenfus M, Ropers D, Achenbach S, Kessler W, Laub G, Daniel WG, Moshage W (2000) Noninvasive detection of coronary artery stenosis using contrast-enhanced three-dimensional breath-hold magnetic resonance coronary angiography. J Am Coll Cardiol 36:44–50 61. Regenfus M, Ropers D, Achenbach S, Schlundt C, Kessler W, Laub G, Moshage W, Daniel WG (2002) Comparison of contrast-enhanced breath-hold and free-breathing respiratory-gated imaging in three-dimensional magnetic resonance coronary angiography. Am J Cardiol 90:725–730 62. Rubinstein RI, Askenase AD, Thickman D, Feldman MS, Agarwal JB, Helfant RH (1987) Magnetic resonance imaging to evaluate patency of aortocoronary bypass grafts. Circulation 76:786–791 63. Sakuma H, Ichikawa Y, Suzawa N, Hirano T, Makino K, Koyama N, Van Cauteren M, Takeda K (2005a) Assessment of coronary arteries with total study time of less than 30 min by using whole-heart coronary MR angiography. Radiology 237:316–321 64. Sakuma H, Suzawa N, Ichikawa Y, Makino K, Hirano T, Kitagawa K, Takeda K (2005b) Diagnostic accuracy of stress first-pass contrast-enhanced myocardial perfusion MRI compared with stress myocardial perfusion scintigraphy. AJR Am J Roentgenol 185:95–102 65. Sandstede J, Pabst T, Beer M, Geis N, Kenn W, Neubauer S, Hahn D (1999) Three-dimensional MR coronary angiography using the navigator technique compared with conventional coronary angiography. Am J Roentgenol 172:135–139 66. Sandstede J, Machann H, Machann W, Beer M, Johnson T, Harre K, Pabst T, Kenn W, Hahn D (2002) [Interindividual-variability of the analysis of regional myocardial wall function after myocardial infarction and revascularization]. RoFo 174:1147–1153 67. Sandstede JJ (2003) Assessment of myocardial viability by MR imaging. Eur Radiol 13:52–61

777

778

5  Thorax and Vasculature 68. Sardanelli F, Molinari G, Zandrino F, Balbi M (2000) Three-dimensional, navigator-echo MR coronary angiography in detecting stenoses of the major epicardial vessels, with conventional coronary angiography as the standard of reference. Radiology 214:808–814 69. Schalla S, Klein C, Paetsch I, Lehmkuhl H, Bornstedt A, Schnackenburg B, Fleck E, Nagel E (2002) Real-time MR image acquisition during high-dose dobutamine hydrochloride stress for detecting left ventricular wall-motion abnormalities in patients with coronary arterial disease. Radiology 224:845–851 70. Schuijf JD, Shaw LJ, Wijns W, Lamb HJ, Poldermans D, De Roos A, van der Wall EE, Bax JJ (2005) Cardiac imaging in coronary artery disease: differing modalities. Heart 91:1110–1117 71. Sensky PR, Jivan A, Hudson NM, Keal RP, Morgan B, Tranter JL, de Bono D, Samani NJ, Cherryman GR (2000) Coronary artery disease: combined stress MR imaging protocol-one-stop evaluation of myocardial perfusion and function. Radiology 215:608–614 72. Simonetti OP, Kim RJ, Fieno DS, Hillenbrand HB, Wu E, Bundy JM, Finn JP, Judd RM (2001) An Improved MR Imaging Technique for the Visualization of Myocardial Infarction. Radiology 218:215–223 73. So NM, Lam WW, Li D, Chan AK, Sanderson JE, Metreweli C (2005) Magnetic resonance angiography of coronary arteries with a 3-dimensional magnetization-prepared true fast imaging with steady-state precession sequence compared with conventional coronary angiography. Am Heart J 150:530–535 74. Sommer T, Hackenbroch M, Hofer U, Schmiedel A, Willinek WA, Flacke S, Gieseke J, Traber F, Fimmers R, Litt H, Schild H (2005) Coronary MR angiography at 3.0 T versus that at 1.5 T: initial results in patients suspected of having coronary artery disease. Radiology 234:718–725 75. Spuentrup E, Katoh M, Buecker A, Manning WJ, Schaeffter T, Nguyen TH, Kuhl HP, Stuber M, Botnar RM, Gunther RW (2004) Free-breathing 3D steady-state free precession coronary MR angiography with radial k-space sampling: comparison with Cartesian k-space sampling and Cartesian gradient-echo coronary MR angiography—pilot study. Radiology 231:581–586 76. Stuber M, Botnar RM, Fischer SE, Lamerichs R, Smink J, Harvey P, Manning WJ (2002) Preliminary report on in vivo coronary MRA at 3 Tesla in humans. Magn Reson Med 48:425–429 77. Thiele H, Plein S, Ridgway JP, Breeuwer M, Higgins D, Schuler G, Sivananthan M (2003) Effects of missing dynamic images on myocardial perfusion reserve index calculation: comparison between an every heartbeat and an alternate heartbeat acquisition. J Cardiovasc Magn Reson 5:343–352 78. Vrachliotis TG, Bis KG, Aliabadi D, Shetty AN, Safian R, Simonetti O (1997) Contrast-enhanced breath-hold MR angiography for evaluating patency of coronary artery bypass grafts. AJR Am J Roentgenol 168:1073–1080

79. Wahl A, Paetsch I, Roethemeyer S, Klein C, Fleck E, Nagel E (2004) High-dose dobutamine-atropine stress cardiovascular MR imaging after coronary revascularization in patients with wall motion abnormalities at rest. Radiology 233:210–216 80. White RD, Pflugfelder PW, Lipton MJ, Higgins CB (1988) Coronary artery bypass grafts: evaluation of patency with cine MR imaging. AJR Am J Roentgenol 150:1271–1274 81. Wintersperger BJ, Engelmann MG, von Smekal A, Knez A, Penzkofer HV, Hofling B, Laub G, Reiser MF (1998) Patency of coronary bypass grafts: assessment with breathhold contrast- enhanced MR angiography—value of a nonelectrocardiographically triggered technique. Radiology 208:345–351 82. Wintersperger BJ, Penzkofer HV, Knez A, Weber J, Reiser MF (1999) Multislice MR perfusion imaging and regional myocardial function analysis: complimentary findings in chronic myocardial ischemia. Int J Card Imaging 15:425–434 83. Wintersperger BJ, Penzkofer HV, Knez A, Huber A, Kerner M, Meininger M, Knesewitsch P, Scheidler J, Haberl R, Reiser M (2000) [Myocardial perfusion at rest and during stress. MR signal characteristics of persistent and reversible myocardial ischemia]. Radiologe 40:155–161 84. Wittlinger T, Voigtlander T, Rohr M, Meyer J, Thelen M, Kreitner KF, Kalden P (2002) Magnetic resonance imaging of coronary artery occlusions in the navigator technique. Int J Cardiovasc Imaging 18:203–211 85. Zerhouni EA, Parish DM, Rogers WJ, Yang A, Shapiro EP (1988) Human heart: tagging with MR imaging–a method for noninvasive assessment of myocardial motion. Radiology 169:59–63

5.3.8  Cardiac Tumors B.J. Wintersperger 5.3.8.1  Basic Consideration in MRI Cardiac neoplasms are to be considered as rare entities in comparison to other tumors but may cause significant morbidity and mortality. Echocardiography represents the most common screening modality for suspected cardiac or pericardiac masses but is limited by its limited acoustic window and field of view (FOV). For further workup of possible disease, other modalities have to be considered. Magnetic resonance imaging (MRI) has been evaluated in the assessment of cardiac masses within recent years and has been proven to allow conformation or exclusion of suspected masses. MRI enables evaluation not only of the exact extent of masses, but also of their possible impact on cardiac function. Identifying characteristic features of tumors on MRI enables narrowing of possible differential diagnosis and, in selected cases, a final diagnosis. Especially in evaluation of the tumor soft

5.3  Heart

tissue characteristics MRI has shown great benefits and thus has been established as the standard modality in the workup of cardiac masses. This section discusses the application of MRI in the assessment of cardiac masses and the imaging features of the most common cardiac masses. The techniques employed are typically anatomic T1-weighted and T2-weighted techniques that may also be accompanied by dynamic techniques such as perfusion or cine imaging. These techniques give information about specific features and soft tissue differentiation. Further details of the MR imaging techniques are highlighted in Sect. 5.3.1. 5.3.8.2  Epidemiology of Cardiac Masses Cardiac neoplasms may arise as primary and secondary tumors and thus have to be differentiated. While primary cardiac tumors arise within the heart and its various structures, secondary tumors are due to metastasis of malignancies primarily located outside the heart (most commonly) or due to direct tumor spread and invasion of masses located adjacent to the heart (e.g., bronchogenic carcinoma) or growth along tubular vascular structures (e.g., renal cell carcinoma, hepatocellular carcinoma). Therapeutic strategies differ substantially for primary neoplasms, and secondary tumor spread to the heart because of the underlying malignant disease in the latter case. Primary tumors of the heart are far less common that secondary tumors. The prevalence of primary cardiac masses is reported in the range of 0.001 to 0.5% (Lam 1993; Burke 1996) while secondary cardiac tumors are ~20–40 times more frequent than primary ones (Lam 1993; Burke 1996). Modern non-invasive imaging modalities such as echocardiography in particular, but also MRI, allow diagnosis of a cardiac mass before death and facilitate therapy planning. In addition to real neoplasms, tumor-like lesions and pseudotumors can be evaluated by MR imaging techniques, and cross-sectional imaging techniques are increasingly being performed when cardiac thrombi are suspected. 5.3.8.3  Benign Cardiac Tumors About 60–85% of primary cardiac neoplasms are benign (Grande 1993; Burke 1996; Endo 1996; Perchinsky 1997; Meng 2002). Cardiac myxoma is the most common benign cardiac mass followed by lipoma, papillary fibroelastoma, rhabdomyoma, and fibroma (Burke 1995) (Table 5.3.16). However, the incidence of the various benign cardiac tumors differs with age. While myxoma represents the most common cardiac tumor entity in adults, rhabdomyoma accounts for the majority of primary tumors in infants and children (Freedom 2000). Besides the

most common benign primary cardiac tumors that are described here in more detail, other varieties are known that have an even lower incidence. 5.3.8.3.1  Myxoma Cardiac myxomas account for ~50% of all primary cardiac neoplasms, and although myxomas can be found in all cardiac chambers, in ~75% of the cases the mass is located within the left atrium (Burke 1996; Sparrow 2005). The majority of the remainder is located in the right atrium (~20%), followed by the ventricles (5%, right and left ventricle). Seldom can myxomas be found at multiple locations within the heart. While most cases of cardiac myxoma are sporadic (>93%), they may also be familial (7%) and potentially also a part of the Carney complex, an autosomal dominant syndrome of cardiac myxomas that are associated with hyperpigmented skin lesions (Carney 1985; Carney 1986). Myxomas in an atrial location typically arise from the interatrial septum in the region of the fossa ovalis though other locations are possible (Tazelaar 1992; Burke 1996). Although cardiac myxomas are histologically benign, they tend to form emboli and cause intracardiac obstruction, so that they need to be classified as potentially fatal neoplasms of the heart (Braun 2005). The majority of myxomas are found in adulthood between the fourth and seventh decade (Burke 1996). Myxomas are seldom diagnosed on the basis of the history, physical findings, electrocardiogram, or chest X-ray because they are notorious for mimicking a great variety of cardiovascular diseases. Patients can be symptomatic for a long time before myxoma is considered as a potential differential diagnosis (Burke 1993). The most frequent initial manifestations of myxomas include dyspnea and symptoms of congestive heart failure by cardiac obstruction (myxoma prolapse), palpitations, chest pain, potential signs of embolism and systemic rheumatism, depending on the localization, size, and humoral activity of the tumor (Meng 2002; Braun 2005). The target vasculature of embolic events varies and is dependent on the location of the tumor. Myxomas show a predilection for the interatrial septum, with the majority arising from an area immediately adjacent to the fossa ovalis (Burke 1996) (Fig. 5.3.47). In addition they often show a pedunculated appearance and may even prolapse into the atrioventricular valves, thus affecting ventricular filling. The vast majority of myxomas will demonstrate heterogeneous signal intensity in MRI, reflecting the underlying tissue heterogeneity. In plain T1-weighted images they appear hypointense to myocardium but show high signal intensity on T2-weighted images due to the high extracellular water content. Fibrous tissue parts and/or tumor hemorrhage may show a different appearance. Following contrast administration myxomas show a predominately heteroge-

779

780

5  Thorax and Vasculature Table 5.3.16  Criteria for differential diagnosis of the most common benign and malignant cardiac tumors Tumor

Epidemiology

Location (most common)

T1

T2

weighting1

weighting1

a

Typical/specific features

Myxoma

Most common primary tumor and most common benign mass (~75%)

Left atrium

Hypointense

Hyperintense

Heterogeneous enhancement

Arises within the area of the oval fossa; may be pedunculated and show mass mobility (valve protrusion); may show signs of hemorrhage in GRE2

Lipoma

Second most common benign primary mass (~4-5%)

Often epicardial

Hyperintense

Hyperintense

No contrast uptake

Resembles the features of subcutaneous fat; fat suppression techniques are helpful; differentiate from LHIS

Rhabdomyoma

Most common primary tumor in children/infants (~4-5% in total)

Ventricles

Isointense (slightly hyperintense)

Hyperintense

Hypointense to surrounding myocardium

Often multiple tumors;  50% are associated with tuberous sclerosis; may regress spontaneously

Fibroma

Second most common mass in infants(children)

Ventricles

Iso- or hypointense

Homogeneously hypointense

Heterogeneous with varying uptake level

“congenital tumor” that is often diagnosed at birth or shortly after

Angiosarcoma

Most common malignant primary tumor

Right atrium

Hypo-, iso- and hyperintense (tumor, necrosis, methemoglobin)

Hyperintense

Typically sunray appearance; marked surface enhancement+central necrosis

“Cauliflower” appearance

Rhabdomyosarcoma

Most common primary malignancy in children/infants

No predominant location

Isointense

Hyperintense

Heterogeneous enhancement; may show central necrosis

Tendency to invade valvular structures

Osteosarcoma

~9% of primary cardiac malignancies

Left atrium

Hypointense

Hyperintense

Heterogeneous enhancement

CT recommended to confirm calcifications

Lymphoma

More common in immunocompromised patients

Right heart

Isointense

Hyperintense

Heterogeneous, typically more peripheral enhancement

Tendency to invade more than one chamber and pericardium

Typically describes the MR signal in comparison to normal myocardium Gradient-recalled echo techniques

a

CA behaviour

5.3  Heart

Fig. 5.3.48a,b  Typical appearance of a lipomatous hypertrophy of the interatrial septum (LHIS) (arrows) with signal isointense to subcutaneous and pericardial fat in T1-weighted imaging (a) that diminishes with applied fat saturation (b)

Fig. 5.3.47a–d  Myxoma within the right atrium arising from the fossa ovalis. The mass shows a typical appearance with a stalk (a,b) and slight heterogeneity in T1-weighted images with fat saturation (c). In SSFP techniques myxomas tend to be only slightly hypointense to the blood pool (d). T tumor, LA left atrium, LV left ventricle, RV right ventricle

neous enhancement and enhancement depends on the tumor vascularity (Fig. 5.3.47). Cine gradient-echo (GE) sequences demonstrate tumor mobility and may show a tumor prolapse into the ventricles. In SGE sequences myxomas usually appear dark based on potential hemorrhage-based iron deposits, while in SSFP techniques the tumor may be only slightly hypointense to the blood pool and hyperintense to myocardium (Fig. 5.3.47). The main differential diagnoses for myxoma include atrial thrombus, both showing a rather heterogeneous appearance with areas of calcification. Whereas thrombus is more likely located posteriorly in the left atrium, myxomas are more likely to arise anteriorly from the interatrial septum. In addition, myxomas will enhance after gadolinium application, whereas thrombus, in most cases, should not (Gomes 1987; Funari 1991). 5.3.8.3.2  Lipoma Lipomas are the second most common benign primary tumor of the heart (Burke 1996) (Table 5.3.16). They consist of an accumulation of encapsulated adipose cells, and the majority arises in an epicardial location expanding into the pericardial space. However, they may also be found in an endocardial location and thus may lead to blood flow obstruction symptoms. The diagnosis of cardiac lipoma is easily made on its characteristic imaging features in MRI. They usually show signal behavior simi-

lar to that of subcutaneous or mediastinal fat with hyperintense appearance in T1-weighted and T2-weighted images. Fat suppression techniques may also help in diagnosis and lipomas do not show contrast agent uptake (Hananouchi 1990). A potential differential diagnosis is lipomatous hypertrophy of the interatrial septum (LHIS) (O’Connor 2006) (Fig. 5.3.48). This entity is characterized by an accumulation between the atrial muscle cells and is defined as any deposit of fat in the atrial septum exceeding 2 cm in transverse. It occurs in older obese patients with potential atrial fibrillation. In MR imaging the usual sparing of the fossa ovalis is clearly visible (Heyer 2003). Whether LHIS has to be considered as a neoplastic or an anomalous developmental lesion has not yet been clarified (O’Connor 2006). 5.3.8.3.3  Papillary Fibroelastoma Although papillary fibroelastomas are approximately as common as lipomas they are rarely detected with MR imaging. They usually appear at endocardial surfaces and are most commonly located (90%) on valve surfaces (Edwards 1991). The vast majority of these tumors are less than 1cm in diameter (Abu Nassar 1971). Their typical appearance on the rapidly moving valves and their typical small size makes it rather difficult to identify them using MR imaging or CT, although they may easily be detected at echocardiography. If MRI is able to depict these valvular masses, it is mostly likely to be because of the use of cine techniques showing the rapid valve motion and the hypointense mobile fibroelastomas (Sparrow 2005). Contrast-enhanced techniques are not helpful in diagnosis of these masses as they consist of avascular connective tissue (Sparrow 2005). Potential differential diagnosis might be thrombus or valve vegetations. The latter often comes along with valve incompetence based on destruction of valvular leaflets. Although symptoms are less fre-

781

782

5  Thorax and Vasculature

quent than in cases of myxoma, papillary fibroelastomas may also lead to embolic events mostly based on adherent thromboembolic material. 5.3.8.3.4  Rhabdomyoma This tumor is the most common primary neoplasm of the heart in neonates, infants, and children (Freedom 2000) (Table 5.3.16). It represents a benign tumor of the cardiomyocytes and is considered by some as a hamartoma occurring exclusively in the heart (Freedom 2000). Rhabdomyomas have a distinct preference for the ventricles, and the majority show multiple lesions, although they are not apparent immediately (Fenoglio 1976; Becker 2000). Rhabdomyomas occur in patients with tuberous sclerosis (30–50%) but may also be sporadic or in conjunction with congenital cardiac malformations (Becker 2000; Freedom 2000; Grebenc 2000). The majority of patients are typically asymptomatic, but depending on the exact tumor location some may develop cardiomegaly with cardiac failure due to left ventricular outflow tract obstruction or episodes of arrhythmia possibly requiring surgical intervention. Compared to the surrounding myocardium the tumors show an isointense to marginally hyperintense signal on T1-weighted images and show hyperintense signal on T2-weighted imaging (Winkler 1987; Luna 2005; Sparrow 2005). After contrast administration rhabdomyomas may present as hypointense in comparison to surrounding myocardium on T1-weighted images. Spontaneous regression of cardiac rhabdomyomas may occur. 5.3.8.3.5  Fibroma Fibroma (some authors refer to it as a hamartoma) represents the second most common primary neoplasm in childhood (Feldman 1976) (Table 5.3.16). In rather rare instances this tumor might be associated within malignant skin tumors, i.e., the Gorlin syndrome (Cotton 1991). They are typically diagnosed in children under the age of 1 year, and the clinical presentation and symptoms depend largely on the size and the location of the tumor (Becker 2000). The leading sign is cardiomegaly, but symptoms may include heart failure, arrhythmias, sudden death, cyanosis, and chest pain (Burke 1994). Occasionally, a cardiac fibroma may grow to an enormous size while the patient remains asymptomatic (Tahernia 1990). The tumor arises within the myocardium and may even obliterate the ventricular cavity once growing to excessive size. The tumor is composed of fibroblasts and may also show calcification. Necrosis or hemorrhage though is rather uncommon. Fibromas show a characteristic signal behavior being iso- or hypointense on T1-weighted images and homogeneously hypointense on T2-weighted

Fig. 5.3.49a–d  Multiple short-axis orientations in a 4-year-old child with tachyarrhythmia. The tumor within the left ventricle has been known almost since birth and shows the typical appearance of a cardiac fibroma. In plain T1-weighted images the fibroma is hypointense to myocardium (a) and also shows the typical low signal in T2-weighted STIR (b). After contrast administration the tumor shows slightly heterogeneous enhancement (c). Inversion-recovery (IR) turbo GRE techniques may also be used for confirmation of contrast uptake (d). T tumor, RV right ventricle, LV left ventricle

techniques (Brechtel 1999; Hoffmann 2003) (Table 5.3.16; Fig. 5.3.49). The contrast enhancement pattern of fibromas may be variable and can be evaluated either using T1-weighted TSE techniques or IR-turboFLASH techniques with heavy T1 weighting. The latter actually allows for a faster data acquisition and also shows a consistent blood pool signal without flow artifacts (Fig. 5.3.49). 5.3.8.4  Malignant Cardiac Tumors 5.3.8.4.1  Primary Cardiac Malignomas Metastases are the most common cardiac neoplasms and as such, also far more common then primary malignant cardiac tumors. Primary malignancies of the heart account for ~25% of all primary cardiac tumors (Burke 1996). The distribution and frequency of different entities varies within published data (Burke 1996) (Table 5.3.16). Primary cardiac malignancies represent a clinical dilemma. They are often asymptomatic until they become large, and even then they produce nonspecific symptoms (Araoz 1999). Before the advent of cross-sectional imaging, primary cardiac malignancies were rarely diagnosed before death. Nowadays they are being diagnosed within living patients allowing for conservative or even surgical treatment including heart transplantation (Uberfuhr

5.3  Heart

2002). However, based on the usual delayed diagnosis of then extended disease including metastasis, these tumors have a rather bad outcome. MRI allows accurate assessment of the heart and the surrounding mediastinum and therefore enables evaluation of the extent of disease. The combination with whole-body MR approaches would also enable screening for metastasis. Angiosarcoma Angiosarcoma represents the most common primary cardiac malignancy of adulthood that accounts for ~37% of all cases (Burke 1996) (Table 5.3.16). The tumor consists of ill-defined vascular spaces lined by endothelial cells but also may exhibit large vascular areas of spindle cells (Burke 1996). The tumor tends to occur in the right atrium (~75%), and patients usually present with rightsided heart failure or pericardial tamponade based on pericardial invasion (Janigan 1986; Burke 1996; Araoz 1999). The prognosis of angiosarcoma is poor which is mainly related to a typically late diagnosis when metastases, most commonly to the lung, are already present (Janigan 1986; Dichek 1988). On MRI angiosarcomas typically present as large heterogeneous masses with disruption of fat planes and/or pericardial thickening (sign of pericardial involvement). T1-weighted images usually show a heterogeneous appearance with low-, intermediate-, and high-signal areas, reflecting necrosis, tumor tissue, and possible hemorrhage (Fig. 5.3.50). Typically, on T2-weighted images, angiosarcomas have a predominantly hyperintense signal and some heterogeneity (Fig. 5.3.50). They usually present with a heterogeneous contrast uptake of a sun ray appearance with marked surface enhancement and central necrosis (Araoz 1999; Kaminaga 2003; Sparrow 2005) (Fig. 5.3.50). Undifferentiated Sarcoma This group of tumors refers to sarcomas that do not show any specific histological feature; they are the second most common primary cardiac malignancy (Sparrow 2005). The tumors typically arise in the left atrium (~80%) and may present with various imaging features on MRI. They may present as polypoid masses or as infiltrating cardiac lesions (Luna 2005). The advent of immunohistochemical staining and electron microscopy led to a decrease in their incidence. Rhabdomyosarcoma Different from angiosarcoma and undifferentiated sarcoma, rhabdomyosarcoma do not show a preferred location within the heart. Overall they account for only 4 to 7% of cardiac sarcomas though they represent the most common primary cardiac malignancy in infants and children (Sparrow 2005). Two entities of this tumor can be differentiated, the embryonal type that mainly occurs in children and a pleomorphic type that occurs in adulthood though it is much less frequent (Burke 1996). Rhab-

Fig. 5.3.50a–d  Large angiosarcoma of the heart within the right atrium almost occluding the superior vena cava (*). The mass shows a heterogeneous signal behavior in both T1-weighted (a) and T2-weighted images (b). After contrast administration there is heterogeneous uptake of non-necrotic parts (c) that is even better delineated by employing fat-suppression techniques (d). Note that pericardial effusion is also present (arrowheads). H hemorrhage, AA ascending aorta, RVOT right ventricular outflow tract

Fig. 5.3.51a,b  Recurrent rhabdomyosarcoma invading the mitral valve shown with cine SSFP technique in (a) systole and (b) diastole. This technique shows the mobility of the mass protruding into the left ventricle

domyosarcoma is more likely than other malignancies to involve the cardiac valves or even arise from them but typically shows a site of myocardial involvement (Hajar 1986; Watanabe 1989) (Figs. 5.3.51, 5.3.52). They also may present with multiple areas of involvement (Sparrow 2005). In plain T1-weighted imaging the tumors typically show an isointense signal to myocardium but occasionally may exhibit heterogeneous signal behavior (Fig. 5.3.52). After contrast administration they show a more or less homogeneous enhancement (Sparrow 2005).

783

784

5  Thorax and Vasculature

Fig. 5.3.52a,b  Same patient as in Fig. 5.3.41, with rhabdomyosarcoma that shows typical isointense signal to myocardium in T1-weighted imaging without contrast (a). After contrast administration the involvement of the interatrial septum can clearly be identified based on the heterogeneous enhancing tumor nodule (b arrows). T tumor

Osteosarcoma Besides metastatic involvement of the heart in peripheral osteosarcoma, primary cardiac osteosarcoma is known as a separate entity accounting for 3 to 9% of cardiac sarcomas (Burke 1996). Primary cardiac osteosarcomas are typically located within the left atrium, while metastases of peripheral tumors typically involve the right atrium (Table 5.3.16) (Fig. 5.3.53). In early stages with a lack of calcification they may be mistaken for myomas. However, they typically arise away from the fossa ovalis and do show invasive tumor behavior. For confirmation of the presence of potential calcification, cardiac CT is recommended in addition to MRI. In T1-weighted techniques the tumors may exhibit a heterogeneous hypointense signal, whereas T2-weighted techniques they may be hyperintense (Fig. 5.3.53). Cardiac Lymphoma Primary cardiac lymphomas are exceedingly rare and usually of a non-Hodgkin’s B-cell type. They are typically confined to the heart and pericardium, a feature that also allows differentiation from secondary lymphoma involvement which is much more common (Burke 1996; Grebenc 2000). The incidence of primary cardiac lymphomas is higher in immunocompromised patients (Araoz 1999). Patients typically present with rapidly worsening heart failure or arrhythmias. Lymphomas arise from the right side of the heart more frequently, particularly the right atrium, but also tend to involve more than one chamber (Table 5.3.16; Fig. 5.3.54). Pericardial effusion may be the only finding in some patients with cardiac lymphoma. In MR imaging, cardiac lymphomas typically present as isointense to myocardium on T1-weighted imaging and as heterogeneously hyperintense on T2-weighted imaging. The absence of central necrosis and a less common involvement of cardiac valves may help in differentiating

Fig. 5.3.53a–d  Proven primary osteosarcoma of the heart within a typical location at the posterior aspect of the atrial wall (arrows). The tumor exhibits a typical high signal in T2-weighted (STIR) images (a) while being isointense to myocardium in T1weighted imaging (b,c). After contrast administration the tumor enhances heterogeneously (d)

Fig. 5.3.54a–d  Extensive lymphoma primarily located within the right atrium but invading also other cardiac structures including the pericardium owing to pericardial effusion (arrowheads) Fast T2-weighted imaging (HASTE) allows a rapid overview of the tumor extent (a,b) while inversion-recovery (IR) Turbo SGE imaging allows the depiction of multiple tumor locations (arrows) after contrast without flow artifacts. T tumor, LV left ventricle

5.3  Heart

them from cardiac sarcomas. After contrast administration they demonstrate a rather heterogeneous uptake with lower enhancement in the center compared to the tumor periphery (Sparrow 2005). Other Primary Cardiac Malignancies The aforementioned entities of cardiac malignancies are known to be the most common ones, although others may be found as well. They tend to be very rare and generally have a non-specific appearance at MR imaging. They mostly arise within the left atrium and may invade various cardiac structures. Malignant fibrous histiocytomas and leiomyosarcomas tend to invade the pulmonary veins. Also fibrosarcomas, liposarcomas, and pericardial mesotheliomas may affect the heart or its components.

5.3.8.4.2  Secondary Cardiac Malignancies Secondary malignancies involving the heart are 20–40 times more frequent than are primary ones. Based on autopsy studies, metastases have been reported to involve the heart in 10–12% of patients with known malignant neoplasms (Abraham 1990; Klatt 1990). The most common malignancies that lead to cardiac or pericardial involvement are tumors that primarily arise within the chest such as bronchogenic carcinoma, lymphoma, leukemia, and carcinomas of the breast or esophagus (Sparrow 2005). The heart may be involved by various ways: by distant metastatic spread to the heart (Fig. 5.3.55), by direct tumor involvement or invasion or by transvascular extension of tumors to the heart (Fig. 5.3.56). Tumor spread is typically through mediastinal lymphatics leading to implantation in the epicardial myocardium, although some tumors may also lead to hematogenous spread (e.g., melanoma, sarcoma) (Fig. 5.3.55). Direct extension of tumors with pericardial or cardiac affection may occur in bronchogenic carcinoma or carcinomas of the breast or esophagus due to their proximity to the heart. Involvement of the heart based on transvascular or more likely transvenous tumor extension can occur with renal cell carcinoma, hepatocellular carcinomas, or adrenal malignancies. MRI features of secondary cardiac tumors are heterogeneous and vary with the imaging features of their primary tumors. In case of transvascular tumor involvement the mass might be a combination of tumor and thrombus formation, which may also result in different imaging features. 5.3.8.5  Cardiac Pseudotumors or Tumor-Like Lesions

Fig. 5.3.55a–d  A patient with a known malignant melanoma presented with cardiac failure. MRI showed an extensive metastasis within the right ventricle and right ventricular outflow tract (a). Again inversion-recovery (IR) turbo SGE imaging allows adequate depiction of the real tumor extent (arrows). M metastasis, RV right ventricle, LV left ventricle

There are several entities that may mimic cardiac neoplasms. Besides those of congenital origin (e.g., pericardial cysts, bronchogenic cysts), pseudotumors such as cardiac thrombus formation may lead to the suspicion of a cardiac neoplasm.

Fig. 5.3.56  The female patient originally presented with suspicion of adrenal carcinoma. After CT imaging the patient was referred to MRI for assessment of the tumor thrombus. MRI (in-

version-recovery turbo SGE) shows a continuous mass (arrows) extending from the abdomen along the inferior vena cava into the right atrium and even into the atrioventricular valve plane

785

786

5  Thorax and Vasculature

5.3.8.5.1  Cardiac Thrombus Thrombotic deposits within the heart are often referred to as pseudotumors or pseudomasses. Intracardiac thrombi may be caused by a variety of different pathologies and may occur in any cardiac chamber. They are responsible for about 15% of all ischemic strokes and put patients at major risk for stroke. Therefore early identification with subsequent therapy is of paramount importance. MR imaging is also increasingly performed for differentiation of thrombus and real cardiac neoplasm. Some characteristic features may help to differentiate thrombus of neoplastic disease. Within the left atrium thrombi are common in cases of atrial fibrillation and are typically located at the posterior wall of the left atrium or the atrial appendage (Figs. 5.3.57, 5.3.58). However, they may also be located within the left ventricle in case of severe global or regional dysfunction (e.g., after myocardial infarction) or even in the right atrium. MR imaging features are age dependent and are mainly based on the status and content of hemoglobin and its derivates. While in acute formation, thrombus typically shows a high signal in T1-weighted and T2-weighted techniques, subacute thrombus appear bright on T1-weighted images and of low intensity on T2-weighted images based on the paramagnetic effects of methemoglobin. A chronic formation though typically appears with low signal intensity on both T1-weighted and T2-weighted images. Differentiation of tumor and thrombus may also be alleviated by the use of contrast agent, as the latter typically does not show enhancement whereas the former usually shows contrast uptake (Paydarfar 2001) (Figs. 5.3.57, 5.3.58). Inversion-recovery techniques typically used for delayed enhancement imaging of myocardial viability may be especially useful in delineation and differentiation of thrombus formations (Barkhausen 2002) (Fig. 5.3.58).

Fig. 5.3.57a–c  Male patient after cardiac transplant with suspicious mass within the left atrium in transthoracic echocardiography. MRI shows a large, dilated left atrium with a thrombus at the posterior wall. The thrombus is a slightly hyperintense mass in T2-weighted (a) and iso-to-hypointense mass in T1-weighted images (b). After contrast enhancement the mass does not show any contrast uptake (c). Note the extensive signal and flow artifacts within the atria based on a failure of the black-blood preparation

Fig. 5.3.58a,b  Inversion-recovery turbo SGE technique in four-chamber (a) and two-chamber (b) views clearly demonstrates the thrombus. Phase-sensitive techniques, especially, allow excellent delineation of the thrombus without the need for optimization of the inversion time

References 1.

Abraham KP et al. (1990) Neoplasms metastatic to the heart: review of 3314 consecutive autopsies. Am J Cardiovasc Pathol 3:195–198 2. Araoz PA et al. (1999) CT and MR imaging of primary cardiac malignancies. Radiographics 19:1421–1434 3. Barkhausen J et al. (2002) Detection and characterization of intracardiac thrombi on MR imaging. AJR Am J Roentgenol 179:1539–1544 4. Becker AE (2000) Primary heart tumors in the pediatric age group: a review of salient pathologic features relevant for clinicians. Pediatr Cardiol 21:317–323 5. Braun S et al. (2005) Myocardial infarction as complication of left atrial myxoma. Int J Cardiol 101:115–121 6. Brechtel K et al. (1999) Cardiac fibroma in an infant: magnetic resonance imaging characteristics. J Cardiovasc Magn Reson 1:159–161 7. Burke AP, Virmani R (1993) Cardiac myxomas: a clinicopathologic study. Am J Clin Pathol 100:671–680 8. Burke AP, Virmani R (1996) Tumors of the heart and great vessels. Armed Forces Institute of Pathology, Washington, D.C. 9. Burke AP et al. (1994) Cardiac fibroma: clinicopathologic correlates and surgical treatment. J Thorac Cardiovasc Surg 108:862–170 10. Carney JA et al. (1985) The comlex of myxomas, spotty pigmentation and endocrine overactivity. Medicine 64:270–283 11. Carney JA et al. (1986) Dominant inheritance of the complex of myxomas, spotty pigmentation, and endocrine overactivity. Mayo Clin Proc 61:165–172 12. Cotton JL et al. (1991) Cardiac tumors and the nevoid basal cell carcinoma syndrome. Pediatrics 87:725–728

5.3  Heart 13. Dichek DA et al. (1988) Angiosarcoma of the heart: threeyear survival and follow-up by nuclear magnetic resonance imaging. Am Heart J 115:1323–1324 14. Endo A et al. (1996) [Clinical incidence of primary cardiac tumors]. J Cardiol 28:227–34 15. Feldman PS, Meyer MW (1976) Fibroelastic hamartoma (fibroma) of the heart. Cancer 38:314–323 16. Fenoglio JJ Jr et al. (1976) Cardiac rhabdomyoma: a clinicopathologic and electron microscopic study. Am J Cardiol 38:241–251 17. Freedom RM et al. (2000) Selected aspects of cardiac tumors in infancy and childhood. Pediatr Cardiol 2:299–316 18. Funari M et al. (1991) Cardiac tumors: assessment with Gd-DTPA enhanced MR imaging. J Comput Assist Tomogr 15:953–958 19. Gomes AS et al. (1987) Cardiac tumors and thrombus: evaluation with MR imaging. AJR Am J Roentgenol 149:895–899 20. Grande AM et al. (1993) Primary cardiac tumors. A clinical experience of 12 years. Tex Heart Inst J 20:223–230 21. Grebenc ML et al. (2000) Primary cardiac and pericardial neoplasms: radiologic-pathologic correlation. Radiographics 20:1073–1103; quiz 1110–1111, 1112 22. Hajar R et al. (1986) Embryonal botryoid rhabdomyosarcoma of the mitral valve. Am J Cardiol 57:376 23. Hananouchi GI, Goff WB II (1990) Cardiac lipoma: sixyear follow-up with MRI characteristics, and a review of the literature. Magn Reson Imaging 8:825–828 24. Heyer CM et al. (2003) Lipomatous hypertrophy of the interatrial septum: a prospective study of incidence, imaging findings, and clinical symptoms. Chest 124:2068–2073 25. Hoffmann U et al. (2003) Usefulness of magnetic resonance imaging of cardiac and paracardiac masses. Am J Cardiol 92:890–895

26. Janigan DT et al. (1986) Cardiac angiosarcomas. A review and a case report. Cancer 57:852–859 27. Kaminaga T et al. (2003) Role of magnetic resonance imaging for evaluation of tumors in the cardiac region. Eur Radiol 13 Suppl 6:L1–L10 28. Klatt EC, Heitz DR (1990) Cardiac metastases. Cancer 65:1456–1459 29. Lam KY et al. (1993) Tumors of the heart. A 20-year experience with a review of 12,485 consecutive autopsies. Arch Pathol Lab Med 117:1027–1031 30. Luna A et al. (2005) Evaluation of cardiac tumors with magnetic resonance imaging. Eur Radiol 15:1446–1455 31. Meng Q et al. (2002) Echocardiographic and pathologic characteristics of primary cardiac tumors: a study of 149 cases. Int J Cardiol 84:69–75 32. O’Connor S et al. (2006) Lipomatous hypertrophy of the interatrial septum: an overview. Arch Pathol Lab Med 130:397–379 33. Paydarfar D et al. (2001) In vivo magnetic resonance imaging and surgical histopathology of intracardiac masses: distinct features of subacute thrombi. Cardiology 95:40–47 34. Perchinsky MJ et al. (1997) Primary cardiac tumors: forty years’ experience with 71 patients. Cancer 79:1809–1815 35. Sparrow PJ et al. (2005) MR imaging of cardiac tumors. Radiographics 25:1255–1276 36. Tahernia AC et al. (1990) Intracardiac fibroma in an asymptomatic infant. Clin Cardiol 13:506–512 37. Tazelaar HD et al. (1992) Pathology of surgically excised primary cardiac tumors. Mayo Clin Proc 67:957–965 38. Watanabe AT et al. (1989) Magnetic resonance imaging of cardiac sarcomas. J Thorac Imaging 4:90–92 39. Winkler M, Higgins CB (1987) Suspected intracardiac masses: evaluation with MR imaging. Radiology 165:117–122

787

788

5  Thorax and Vasculature

5.4  MR Angiography

5.4.1.2.1  Phase-Contrast MRA

5.4.1  MRA Techniques and Acquisition Techniques

Phase-contrast (PC) MRA (Figs. 5.4.1, 5.4.2, 5.4.3) depends on the flow-induced phase shift of moving spins in the blood along a magnetic gradient (Dumoulin et al. 1990). With the application of bipolar gradients the signal from non-moving spins is canceled out after the second lobe of the gradient. The almost unchanged phase of stationary tissue leads to very good background suppression. Spins moving along the readout gradient however experience a phase shift that is proportional to the flow velocity. For a correct PC MRA the correct choice of the sensitivity, the so-called velocity-encoding gradient (VENC) is required. As this VENC has to be chosen manually, it is a typical source of errors. The VENC should always be slightly higher than the expected maximal blood flow velocity. If the VENC is chosen too low, aliasing (i.e., a phase wraparound) will occur. With a VENC chosen too high, the overall sensitivity to detect slow flow decreases as does the signal-to-noise ratio of the resulting images. Typical VENCs for different application areas are given in Table 5.4.1. In contrast to TOF MRA, PC MRA is very insensitive to saturation effects. PC MRA can be acquired quickly as a 40-mm-thick slab in just 23 s at 1.5 T with a VENC of 20 cm/s. These projection images are suitable as vessel scouts for the identification of slow flow in the brain in case of suspected thrombosis of the venous sinuses. If three-dimensional images are requested, bipolar gradients have to be applied out along the three orthogonal axes. An interleaved readout is often employed in 3DPC MRA to reduce artifacts from patient motion. Also, EKG triggering can be applied with PC MRA to reduce the pulsation-related artifacts. Alternatively, the EKG triggering can be used to obtain dynamic flow images over the cardiac cycle. Theoretically, 3D-PC MRA allows a relatively high spatial resolution in particularly for ves-

H.J. Michaely and S.O. Schönberg 5.4.1.1  Introduction

Since its introduction in the mid 1980s (Dumoulin and Hart, Jr. 1986; Potchen 1992) magnetic-resonance angiography (MRA) and contrast-enhanced (CE) MRA in particular (Prince et al. 1993) have evolved into a standard clinical exam for the evaluation of virtually all vascular body regions including the supra-aortic and intracranial vessels, the thoracoabdominal vessels and the peripheral arteries (Krinsky and Rofsky 1998; Vosshenrich and Fischer 2002: Bongartz et al. 1997; Nael et al. 2005; Becker et al. 1999; Swan et al. 1992). The main advantages of MRA are its non-invasiveness, the combination of morphologic and functional information in a single exam, the high spatial resolution (HR) of the threedimensional (3D)-data sets and (if applied) the lacking nephrotoxicity of the gadolinium (Gd) chelates. Over the years, the applied MRA sequences were subject to continuous development and change. Early techniques such as dark-blood MRA, phase-contrast (PC) MRA, and time-of-flight (TOF) MRA have almost completely disappeared or have remained in use for a few dedicated applications. Since its first presentation in the literature contrast-enhanced three-dimensional (3D) CE MRA has become the most widespread and widely applied technique for almost all body regions. Today, PC MRA and TOF MRA are mainly applied for imaging of the cerebral vessels. Arising MRA techniques include steady-statefree-precession (SSFP) MRA, which is currently under investigation for cardiac and abdominal imaging. This section explains the technical basis for all MRA techniques, with a focus on 3D CE MRA as the work- Table 5.4.1  Typical VENC settings for PC MRA horse of all clinical MRA applications. The important Optimal sequences properties, potential pitfalls, and artifacts are Vessel VENC (cm/s) Application discussed. In a last section guidelines for optimal sequence protocols for different body areas are provided. 5.4.1.2  MRA Techniques Three MRA techniques—PC MRA, TOF MRA, and CE MRA—are explained in the following with a strong emphasis on CE MRA as it is considered as standard techniques for most applications nowadays. Specific application areas for PC MRA and TOF MRA are also outlined where indicated. Newer techniques such as SSFP MRA are not considered in depth due to their lacking clinical relevance.

Aorta

100

Flow

Aortic stenosis

150–500

Flow

Carotid arteries

70

Flow

Carotid artery stenosis

100–300

Flow

Renal arteries

75

Flow

Cerebral arteries

40–60

MRA

Cerebral veins

5–20

MRA

Portal vein

5–10

Flow

5.4  MR Angiography Fig. 5.4.1a,b  Sagittal 2D PC MRA measured at 3 T without parallel imaging. The slab thickness was 40 mm, image acquisition time was 27 s, and the VENC was 40 cm/s. This image provides sufficient information for positioning of further MRA sequences. Also the sagittal sinus can be clearly depicted. For assessment of the transverse sinus an additional axial or coronal sequence is required

Fig. 5.4.2  a MIP of a 3D PC MRA of the brain acquired at 3 T (TR/TE – 29/6.2 ms, FA = 10°) without parallel imaging with a spatial resolution of 1 × 0.8 × 3 mm3 in 4 min without contrast agent. The image clearly depicts the peripheral arterial vessels of the brain as well as the main venous system. In this patient a dominant left transverse sinus and a hypoplastic right transverse sinus are seen. b Source images of the above mentioned 3D PC MRA demonstrating again the clear vessel depiction and the good vessel-background contrast even for the thin, distal parts of the middle cerebral artery

789

790

5  Thorax and Vasculature

are typically in the order of 40 ms, should be smaller than the T1 of the stationary tissue saturate the spins and thus suppress the signal from the background. The echo time (TE) should be chosen as short as possible (about 5 ms) to reduce intra-voxel dephasing. When choosing TE, an in-phase TE is preferable to out-of-phase TE to avoid signal cancellations at the vessel wall. Blood flowing into the imaging plane is not yet saturated and hence yields the main signal determining the image contrast. As the blood travels through the acquisition volume and is repetitiously excited, the resulting signal becomes gradually weaker. This problem can be avoided by acquisition of small slabs in which the blood spins are not saturated in the distal part of the imaging volume. If a large imaging volume is needed, as in the case of the cerebral arteries, a technique of gradually increasing flip angles over the imaging volume (tilted optimized non-saturating excitation [TONE]) can be employed. Alternatively, the 3D volume can be segmented into small overlapping thin slabs, a technique called multiple overlapping thin-slab acquisition (MOTSA). The 3D volume is reconstructed Fig. 5.4.3a,b  agittal (a) and coronal (b) 2D-PC MRA se- from the center slices of the single slabs only discarding quences acquired as vessel localizers for a peripheral MRA. The the peripheral parts. However, this technique is suscepacquisition of such data can be accomplished in as little as 30 s tible to signal cancellations at the overlapping areas—the per 2D-slab. The spatial resolution and information provided is so-called Venetian blind artifact—and requires longer not sufficient for diagnostic reading but allows the technicians a acquisition times. In order to increase the vessel-background contrast proper positioning of the 3D MRA volume magnetization transfer contrast technique (MTC) is often applied. In this technique, the macromolecule-bound protons are saturated by a dedicated off-resonance RF pulse, sels with slow flow. Therefore one main indication for yielding a signal reduction of 15–40% in the gray and 3D-PC MRA is to detect or rule out intracerebral venous white matter of the brain, whereas the moving protons in thrombosis in pregnant women, where the use of a MR the blood are not affected by this pulse. This results in a contrast agent is contraindicated. However, the admin- net contrast gain between the strongly suppressed backistration of contrast agents can improve the depiction of ground and the vessels. As fat-protons are not saturated smaller vessel significantly. At 1.5 T, a 128-mm-thick slab by the MTC pulse, additional fat-suppression is needed. with a voxel size of 0.8 × 0.8 × 2 mm3 can be acquired in TOF MRA is not dependent on the application of contrast 5.03 min. Due to reduced acquisition times with paral- agents. The administration of contrast agents can, howlel imaging in combination with high-field scanners, PC ever, increase the visibility of small vessels. But this adMRA may experience a renaissance for the intracerebral vantage is accompanied by a stronger venous signal after vessels. In other body regions, PC MRA serves as vessel contrast administration. A major drawback of TOF MRA is the insensitivity to in-plane flow. This is of particular localizer sequence with no other main application. interest for pathologies of the carotid siphon, where signal voids due to saturation and turbulent flow can occur resulting in non assessable vessels segments. A signal loss 5.4.1.2.2  Time-of-Flight MRA can also be a consequence of a stenotic lesion. The turbuToday, time-of-flight (TOF) MRA is mainly used for MRA lent flow distal to the stenosis may partly cancel out the of the intracerebral vessels (Michaely et al. 2004a). It has signal and hence lead to an overestimation of the stenosis. the advantage of not being dependent on contrast agent, Similarly, large aneurysms can be critical for TOF MRA and it allows a high spatial resolution. On new 1.5- and 3- when they exhibit a turbulent inside the aneurysm. AnT scanners even 1,024 matrices can be acquired resulting other problem with TOF MRA is that tissues with short T1 in an isotropic resolution of up to 0.3 mm (Willinek et times such as thrombus or hemorrhage (methemoglobin) al. 2003a). In TOF MRA, the vessel-background contrast are not saturated like the background and may simulate is based on relative enhancement of the inflowing blood a blood vessel. This is of particular importance as TOF while the stationary background tissue is saturated. The MRA is often used in patients with suspected intracranial fast repetitive excitation with short repetition times which arteriovenous malformations or developmental aneu-

5.4  MR Angiography

alternating radiofrequency, Edelman et al. 1994): blood flowing into the imaging slab is inverted and hence labeled by a 180° inversion RF pulse. After a variable inflow time (TI) during which the labeled blood can flow into the imaging slab, several lines of k-space are read out using a FLASH-sequence. The same measurement is then repeated without blood labeling. Subtraction of the raw data (with and without labeling) before image reconstruction yields images that show the blood flow in the readout slice with very good background suppression. This spin-labeling technique benefits greatly from prolonged T1 relaxation times at 3 T, with the subsequent slower decay of the blood labeling. To minimize image artifacts from pulsatile flow the measurement can be synchronized with the cardiac cycle (EKG triggering). By this means the blood flow in the readout slice can be demonstrated as a function of the R-R interval (Weber et al. 2004). 5.4.1.2.4  Contrast-Enhanced MRA Fig. 5.4.4  a Blood flowing into the imaging plane in TOF MRA demonstrates a slow saturation leading to decreased signal intensity in the distal parts of the 3D volume. Blood flowing in-plane exhibits a faster saturation as the spins are subject to continuous RF excitation. b Scheme demonstrating the slab positioning for an arterial TOF MRA of the brain. The slab is tilted to contain the main cerebral arteries and the circle of Willis. Distal to the imaging volume a saturation slab is positioned to saturate venous blood flowing into the imaging volume

rysms which are subject to rupture and bleeding. Ghost artifacts from pulsatile flow can be reduced by employing EKG gating at the cost of increased imaging time. In 3D TOF MRA, an additional presaturation slab is positioned to suppress either the venous or the arterial signal. If the venous signal is to be suppressed, then the saturation slab is positioned above the arterial imaging slab. By this means the venous blood flowing into the imaging slab from distal is already saturated. Vice versa, when a venous 3D TOF MRA is to be acquired the presaturation slab is positioned upstream. To avoid signal saturation in two-dimensional (2D) venous TOF MRA the slab orientation—which is sagittal in most cases— should be slightly tilted in axial and coronal orientation. In so doing, particularly the signal from the sagittal sinus will not be saturated (Figs. 5.4.4, 5.4.5).

Contrast-enhanced (CE) MRA was introduced into the clinical routine in the mid-1990s (Prince et al. 1995). It is currently considered state-of-the-art for most MRA applications because of its relatively easy application, fast three-dimensional image acquisition, and capacity to provide high spatial resolution over a large field of view (FOV). For CE MRA acquisition, the patient needs to have an intravenous access, typically a 20-ga line. CE MRA is dependent on the combination of fast imaging and the application of a paramagnetic contrast agent (i.e., typically gadolinium chelates), which shortens the relaxivity of blood significantly. 1 1 = + R1[G d] (5.4.1) T 1 T 10

(T1 – T1 is the value of contrast-enhanced blood; T10 – T1 is the value of non-enhanced blood; R1 is the field-dependent relaxivity of the contrast agent) From Eq. 5.4.1 it can be inferred that the higher the relaxivity R1 of the contrast agent the shorter the resulting T1 of the tissue after contrast administration will be. For an optimal image quality the readout of central k-space lines and the arrival of the contrast agent bolus in the target vessel have to coincide. Typically spoiled gradient-echo (GRE) sequences are used for CE MRA with a TR of 5 ms or less. As this short TR is below the tissue specific T1 time, the resulting images reveal almost no contrast before the administration of contrast agent. After the administration 5.4.1.2.3  Time-Resolved Non-Contrast-Enhanced of contrast agent however, the T1 time of blood is markedly reduced (Eq. 5.4.1; Table 5.4.2), so that fast imagMRA (Tagging) ing with the abovementioned TR times now yields images Dynamic MRA is feasible with tagging techniques where with enhanced vessels while suppression of background the labeling of blood upstream is used for contrast gener- tissue is maintained. The technical demands for 3D CE ation. A typical approach is STAR (signal targeting with MRA can be summarized by the statement “timing is ev-

791

792

5  Thorax and Vasculature Fig. 5.4.5  a Single slice taken from the raw data of a 3D TOF MRA at 1.5 T. There is marked enhancement of the vessels; the background is suppressed by a MTC preparation. b Coronal MIP view of the same 3D TOF MRA demonstrating the cerebral vessels. In this MIP view there is almost no disturbing background signal. c Venetian blind artifacts in a 3D TOF MRA, which can be clearly recognized as the dark bands running parallel through the image. These artifacts are a result of the overlapping reconstruction algorithm used in the MOTSA technique. d Axial MIP view of a patient suffering from a large left-sided ateriovenous malformation (AVM). The 3D TOF MRA acquired at 3 T with a voxel size of 0.3 × 0.3 × 0.3 m³ clearly displays the numerous, small, tortuous vessels of the AVM

Table 5.4.2  Contrast agent characteristics

erything.” The time during which arterial imaging can be performed is usually short (less than 20 s). In the following time, veins and background tissue show considerable enhancement, which makes arterial imaging impossible. Bolus timing is essential for artifact-free MRA images (see following section). The vessel depiction in 3D CE MRA is only dependent of the shortening of the blood T1. Depending on the timing arterial or venous images can be acquired. The TE should be chosen to be as small as possible to reduce intra-voxel dephasing and susceptibility effects. However, care should be taken not to choose the TE to be out of phase, as this may lead to increased chemical shift artifacts at the vessel borders. The flip an-

1.5 T

3T

T10 arterial blood

1,250 ms

1,650 ms

T1 fat (i.e., background)

343 ms

382 ms

R1 Magnevist

4.1 l/mmol/s

3.7 l/mmol/s

T1 blood + Gad (5 mmol/l)

47 ms

52 ms

Signal ratio blood/fat

5.15

5.3

gle is typically chosen between 15 and 30°. Higher flip angles lead to saturation effects; lower flip angles result in an insufficient SNR and vessel-to-background contrast. Basically, T1 and R1 are dependent on the field strength and rise with increasing field strength. While this implies longer TR times for T1-imaging at 3 T, this also implies a better background suppression for 3-T MRA. Thereby, a net gain in vessel-to-background contrast is achieved since the relaxivity of standard extracellular, non-intravascular gadolinium chelates (e.g., Magnevist; see inline table below) is relatively unchanged. In the past 5 years parallel imaging (PI) has been widely applied to MRA. Particularly in combination with

5.4  MR Angiography

3-T imaging PI allows for submillimeter isotropic resolution with acquisition times of less than 20 s for first-pass abdominal MRA. More detailed explanations on PI will be given in the next section. A more recent approach to CE MRA is time-resolved MRA, which yields dynamic information on blood flow. Typical applications are the characterization of dissections, aneurysms, arteriovenous malformations, detection of subclavian steel syndromes, or detection of aberrant vessels such as MAPCAs (major aortopulmonary collateral arteries with different enhancement from that of the main pulmonary artery): a basic multiphasic technique—which simply acquires multiple 3D phases in consecutive order—has been in use for almost 10 years now. More recently, techniques using view-sharing and keyhole techniques have been introduced (Korosec et al. 1996). Depending on the spatial resolution chosen they allow for up to 0.5 s acquisition per entire 3D volume by sharing lines of peripheral k-space between adjacent frames.

5.4.1.3  Contrast Agents for MRA Basically all paramagnetic contrast agents are suitable for MRA. In Europe, only Gadovist (gadobutrol, Schering AG, Berlin, Germany) and Vasovist (Schering) have an official MRA label. Gadobutrol has a MRA label for all vascular beds, while gadofosveset is only approved for run-off studies (renal arteries and lower). For all other contrast agents the MRA application is an off-label use. MultiHance (gadobenate dimeglumine, Bracco SpA, Milan, Italy) is currently in phase III for MRA approval (Table 5.4.3). The image contrast in MRA sequences is dependent on the presence of the contrast agent during the acquisition of the central k-space lines. Hence, for MRA a well-defined amount of contrast agent at a clearly defined point in time is required for good image quality. Therefore for MRA, the use of automated injector pumps that allow an exact injection of contrast agents in terms of amount and flow rate is strongly recommended. Also, as the spatial resolution is increased, the scan time shortened and par-

Table 5.4.3  Characteristics of the different MRA techniques TOF MRA

PC MRA

CE MRA

Advantages

- No contrast agent needed - Selective depiction of arterial or venous system feasible - Multiple acquisitions feasible - High-spatial resolution feasible

- No contrast agent needed - Sensitive to slow flow - Superior background suppression - No saturation - Multiple acquisitions feasible

- Dynamic acquisition possible - Fast and robust technique, less susceptible to intra-voxel dephasing - Large field of view acquisition feasible - High contrast and SNR - High spatial resolution feasible

Disadvantages

- Susceptible to saturation effects (3D TOF) and turbulent flow - Decreased sensitivity to in-plane flow - Less sensitive to slow flow (aneurysms) - Tissue with short T1 times (e.g., thrombus) may simulate blood - Long acquisition times

- Long acquisition times, particularly for 3D acquisitions - Dependent on correct estimation of VENC - Susceptible to phase errors in acquisitions with low signal-to-noise ratio

- Contrast agent required - Depends on correct timing - Technically demanding (strong and fast gradients)

Applications

- Intracerebral vessels - Carotid vessels

- Intracerebral vessels, particularly venous system - Portal vein - CSF studies

- Supra-aortic MRA - Thoracic and abdominal MRA - Peripheral MRA

Current development

- Increased spatial resolution with moderate scan times feasible for cerebral MRA with parallel imaging at 3 T

- 3D venography of the entire brain at 3 T with parallel imaging and high factors of acceleration in two dimensions

- Evolving into the standard technique for all applications due to direct benefit from parallel imaging and high-field imaging. - Dynamic studies increasingly used

793

794

5  Thorax and Vasculature

allel imaging is applied the signal-to-noise ratio of the sequences is lowered (see Sect. 5.4.1.4). Therefore, the lower the SNR of the MRA sequence, the greater the requirements are for the applied contrast agent. The following table provides an overview of the different classes of contrast agents and of their applications in MRA.

where Ny is the number of phase-encoding steps in the y-direction, Nz the number of partitions or the number of phase-encoding steps in the z-direction in case of threedimensional sequences, and NEX the number of excitations. The SNR for three-dimensional GRE sequences can be calculated as

5.4.1.4  General Technical Considerations

SNR ∝

B02 × Ny × Nz × NEX , BW

(5.4.3)

where BW is the readout bandwidth and B0 the static magnetic field of the scanner. Basically speaking, the In MRA and particularly CE MRA fast acquisition of the SNR is increased with larger voxel size, multiple averdata is mandatory to obtain as few motion-related arti- ages, and small readout bandwidths. As can be seen in facts as possible. At the same time a high signal-to-noise Eq. 5.4.2, the acquisition time can be decreased by shortratio (SNR) with concomitant high spatial resolution is ening any of the three factors TR, Ny, and Nz. Technically desired. These three aims are partly contradictory and speaking, the easiest approach is to minimize TR, which cannot be achieved to their full extent at the same time. largely depends on the technical capabilities of the MR Therefore, it is essential to optimize the scan parameters scanner in use. Reduction of TR is limited, however, as extensively. Due to mandatory respiratory excursion and with a shorter TR a larger amount of energy per second because of bolus timing issues the time-frame for MRA is deposited in the patient. As legal restrictions for the acquisition is limited. The total acquisition time for a maximal amount of energy deposition exist (specific abgiven 3D MRA sequence can be calculated as sorption rate), TR cannot be reduced ad infinitum. Also a short TR may lead to peripheral nerve stimulation. Acquisition time = TR × Ny × Nz × NEX, (5.4.2) Therefore most scanners do not allow shortening of the 5.4.1.4.1  SNR and Acquisition Time

Trade Name

Generic Name

Comments

Standard agents

Magnevist Omniscan Dotarem Prohance

Gadopentetate dimeglumine Gadodiamide Gadoterat-meglumin Gadoteridol

Standard contrast agents that can be used for MRA and are sufficient for MRA without parallel imaging and larger voxel volumes. Due to their widespread use they have a quasi “pseudo-label” for MRA

“Advanced” agents

Gadovist* MultiHance

Gadobutrol Gadobenate Dimeglumine

Double concentration contrast agent (gadobutrol) and slightly protein binding contrast agent with higher relaxivity (gadobenate dimeglumine), which allow for a very tight contrast agent bolus. Both yield better enhancement than do standard agents and are well suited for MRA with higher factors of parallel imaging or with smaller voxel sizes. Venous MRA is feasible with gadobenate dimeglumine with moderate venous enhancement

Intravascular Agents

Vasovist*

Gadofosveset

Strongly protein-binding contrast agent with highest available relaxivity. First-pass MRA is feasible as with any other contrast agent. Due to the strong protein interaction MRAs can be acquired in the steady state for up to 1 h after the initial contrast administration with strong venous and arterial contrast

*Official EMEA (European Medicines Agency) label for MRA (gadobutrol: whole body, gadofosveset: infrarenal aorta and peripheral vessels)

5.4  MR Angiography

Fig. 5.4.6a–d  Properties of k-space reconstruction. a From fully acquired k-space the image is reconstructed using a Fourier transformation. However, acquisition of the entire k-space is time-consuming as each line in k-space requires a phase encoding step. b–d Due to the Hermetian symmetry of k-space, it is sufficient to acquire one part of k-space without losing the possibility of reconstructing the image from this under-sampled k-space. The less k-space is used for reconstruction the more

artifacts (due to the non-perfect symmetry and due to phase errors) are introduced into the image which present as blurring. This can be particularly well appreciated in c and d, where only 40 and 20% of k-space, respectively, were used for reconstruction. In clinical routine, partial-Fourier techniques which acquire 60–80% of k-space as show in b are used. They offer the best compromise between increased imaging speed and hereby induced artifacts

TR below a certain threshold. Theoretically, decreasing TR, which implies an increased readout bandwidth, also decreases SNR. Yet, for 3D CE MRA this effect is less pronounced as due to the contrast agent there is sufficient SNR in the vessels so that a slight SNR reduction can be neglected. In addition, decreased TR also leads to a better suppression of the background tissue that has longer T1 relaxation times than do the enhanced blood vessels. Therefore, even though fast imaging with increased readout bandwidth and decreased TR reduces the SNR, the vessel-to-background contrast does not necessarily suffer or may even increase. This effect is even more pronounced at 3 T, where the T1 relaxation times of most tissues are prolonged while the contrast agent’s relaxivity is relatively unchanged. Another way to decrease the acquisition time is reducing the acquired k-space lines (i.e., Ny) (Eq. 5.4.2) by using the inherent conjugate (Hermetian) symmetry of k-space. The symmetry of k-space is deployed in halfFourier (also known as ½ NEX) imaging, where only half of k-space and some additional extra lines to avoid phase errors are acquired thus reducing the image acquisition time by roughly 50% but also reducing the SNR by a factor of √2. As images reconstructed from less than 50% of

k-space show an increasing amount of blurring, normally 60% of k-space is read out. However, it is important to include the central lines of k-space completely as they encode for the image contrast while the peripheral lines encode for the image details (Fig. 5.4.6). 5.4.1.4.2  Bolus Timing Because the central lines in k-space encode for the main image contrast bolus, timing is of the utmost importance for CE MRA. The arrival of the contrast agent bolus has to be synchronized with the acquisition of the central lines of k-space. Mis-timing leads to either ringing artifacts when the acquisition of the central lines of k-space started before the contrast agent bolus has arrived. These ringing artifacts result from steep increase of the contrast agent concentration during the readout of central k-space and can be seen in the image as hyperintense and hypointense lines parallel to the course of the vessels while the vessels themselves are not yet fully opacified. When the acquisition of central k-space occurs too late, venous overlay of the images results (Figs. 5.4.7, 5.4.8, 5.4.9). Bolus timing can be done via the “classic way” using a test

795

796

5  Thorax and Vasculature

Fig. 5.4.7a,b  Even though each point in k-space encodes for each image pixel, the k-space can be roughly subdivided into two complementary parts, k-space center and k-space periphery. The first encodes largely for the image contrast, the latter for the image details. a Image reconstructed via Fourier transformation from the central k-space lines only. The resulting image

contains all necessary contrast information (arteries and kidneys can be easily marked off) but detail structures cannot well be assessed. b Image reconstructed via Fourier transformation from the peripheral k-space lines. The resulting image presents with all detail information such as vessel wall borders etc but is poorly contrasted

Fig. 5.4.8a,b  After the bolus injection the contrast agent bolus travels through the arterial and venous system. During this time it becomes increasingly spread out, the bolus coherence is lost, and a certain amount of contrast agent leaves the vessel bed (at least in case of standard Gd chelates). a For the depiction of the arterial vessels which are most often the target of the MRA examination, it is essential to synchronize the acquisition of the central k-space lines that encode for the image contrast with the arrival of the contrast agent bolus in the target vessel. If the acquisition is started too early during the steep increase in contrast agent concentration ringing artifacts will result. On the other hand, when the acquisition is delayed a mixed arteriovenous contrast will result which may hinder diagnostic reading of the images. b Different k-space trajectories require different bolus timing approaches. In standard Cartesian k-space readout, the acquisition has to be started before the arrival of the contrast agent bolus as peripheral lines in k-space are filled before the center of k-space is reached. Central (spiral, ellipticcentric) k-space trajectories start at the center of k-space. Therefore, the sequence has to be started at the time of maximal vessel enhancement. As the vessel enhancement can be monitored visually or semiautomatically, these sequences do not require a test bolus. They can be (semi-) automatically started when the arrival of the contrast agent is detected

bolus, from which the circulation time can be calculated. More modern approaches apply semiautomatic bolus detection techniques with automated start of elliptic centric MRA sequences. Both the classic and more modern approaches are appropriate for achieving well-enhanced MRA. A third approach is to monitor the arrival of the

contrast visually and to start the MRA sequence manually. This approach is however prone to timing errors. In patients with congestive heart failure there is commonly delayed arrival of the contrast agent bolus, but good enhancement as the bolus is still relatively coherent. In contrast, in case of a hyperdynamic circulation situation

5.4  MR Angiography

Fig. 5.4.9a,b  Fat-saturated 3D GRE (VIBE) pre contrast (a) and 3D CE MRA (b) of a dissecting aortic aneurysm acquired with 1 mm isotropic resolution in 19 s at 3 T. The MRA timing was calculated using conventional bolus timing with a test-bolus slice positioned at the level of the diaphragm. The MRA was timed from the contrast arrival in the true lumen and demonstrates strong enhancement of the entire true lumen all the way through the field of view. Also the main abdominal branches of the aorta (arrow) arising from the true lumen are already enhanced. In contrast, the false lumen reveals only partial enhancement with the contrast agent dimming out at the beginning of the abdominal aorta (arrowheads). The abdominal part of the false lumen is not yet enhanced

Table 5.4.4  Typical parameters for contrast agent application using standard 0.5 M Gd chelates Region

Contrast volume

Flow rate

Test bolus ROI

Supra-aortic vessels

0.4 ml/kg body weight

1–1.5ml/s

Proximal common carotid artery

Pulmonary arteries

0.4 ml/kg body weight

1.5 ml/s

Pulmonary main stem

Thoracic/abdominal aorta (high-resolution)

0.4 ml/kg body weight

1.5–2 ml/s

Aortic arch/ subdiaphragmatic aorta

Thoracic/abdominal aorta (time-resolved)

0.2 ml/kg body weight

3–4 ml/s

No bolus timing required

Peripheral MRA (bolus chase)

0.4 ml/kg body weight

0.5 ml/s

Infrarenal aorta

(AV shunts, Osler’s disease) good enhancement can be serve as a rough overview, as depending on the scanharder to achieve as the contrast agent is diluted in blood, ner, sequence, and user preference, different application and the bolus coherence is destroyed by the cardiac con- schemes for the contrast agent may be used. Each contractions. Similarly, larger abdominal aneurysms can lead trast injection should be followed by a sufficient saline to enhanced bolus dispersion, which degrades the quality chaser of at least 20 ml. The bolus length and hence the of the bolus in the peripheral vessels. In the case of abnor- injection parameters should be adapted to the anticimal vessel anatomy or in cases of dissection aneurysms, pated scan time. For very fast abdominal MRA-acquisiit is important to choose the correct vessel or the correct tions (less than 20 s) a faster bolus injection rate of 2–3 lumen in order to finally have the correct vessel enhanced. ml/s may be chosen to make sure that the majority of the In dissecting aneurysms of the aorta, the true lumen with contrast agent reaches the volume of interest during the the higher flow is most often located anteriorly. image acquisition. This is of particular interest for multiTypical parameters for the amount of contrast agent phasic or time-resolved MRA. As an estimate, the bolus is and flow rates are given in Table 5.4.4. Table 5.4.4 should prolonged by about 6–8 s during the pulmonary passage.

797

798

5  Thorax and Vasculature

5.4.1.4.3  Voxel Size, Matrix, and Zero Filling Due to its three-dimensional acquisition, 3D CE MRA in principle holds the potential to perform reconstructions along any desired imaging plane, thus allowing the determination of the area stenosis (Figs. 5.4.10, 5.4.11). However, it is well known that due to the short imaging time required to perform the entire 3D CE MRA, spatial resolution is usually not equal along all three scan orientations—a fact known as anisotropic voxel size. Typically spatial resolution is the highest in the frequency encoding head–feet direction, followed by the phase encoding left–right direction, while the lowest resolution is acquired in the slice encoding anterior–posterior direction. For coronal display of the image in maximum intensity projection, the relatively poor resolution in slice encoding direction is not visualized; however, it becomes immediately evident when sagittal or oblique reconstructions are performed. In this case, blurring in the direction of the lowest resolution from anisotropic voxel sizes results in a decreased accuracy of the vessel area measurements. The use of isotropic voxel sizes enables a distortion-free reconstruction of the vessel area in any desired three-dimensional orientation. As described above, a high isotropic resolution is desirable to exactly assess vascular pathologies. However increasing the matrix to increase the spatial resolution has to disadvantageous effects. First, with the increasing number of phase and frequency encoding steps the image acquisition time is increased as well. Second, a higher matrix also leads to a SNR loss as outlined in the below given equations: SNR ∝ Voxel volume

Voxel volume =

(5.4.4)

FOV x × FOV y × ∆z Nx × N y



(5.4.5)

FOVx and FOVy denote the field of view in x- and y-directions, ∆z stands for the number of partitions, and Nx and Ny for the number of frequency and phase encoding steps. From Eqs. 5.4.4 and 5.4.5, it becomes clear that the SNR is reduced with increasing matrix. Therefore, dedicated MRA protocols have to be employed that take into consideration the desired spatial resolution, the available scan time, and the available SNR for an optimal protocol. Due to the high SNR in CE MRA frequency-encoding matrices with 448 to 512 frequency-encoding steps can be applied. The scan time can be kept reasonably short when half-Fourier techniques (described above) and parallel imaging (described in the following paragraph) are applied. Another way to increase image matrix without gain in true image content is the so-called zero filling, where additional lines in the periphery of k-space are filled with

Fig. 5.4.10  68 year-old male patient with a high-grade renal artery stenosis of the left renal artery. a The plane of view is oblique axial. The axial targeted MIP image of the 3D CE MRA data set with high spatial resolution (slice thickness = 0.9 mm, in-plane resolution = 0.8 × 0.8 mm) shows that the stenosis (arrow) involves the ostium of the renal artery. b The plane of view is oblique sagittal. Cross-sectional reformats of the 3D CE MRA data set with high spatial resolution (upper row) clearly show the residual lumen at the stenosis site (left) as well as in the normal segment of the renal artery (right). The calculated degree of reduction of vessel area (= area stenosis) of 72.7%. These results are in excellent agreement with the correspon­ding intravascular ultrasound images (IVUS, lower row). On IVUS a 75.9% area stenosis was measured. Note that also the eccentric shape of the lumen at the stenosis site (left) and in the reference segment (right) is well resembled by the 3D CE MRA reformats (arrowheads). The large, open arrow points to an area with complete extinction of the ultrasound signal due to a large atherosclerotic plaque intruding from the aorta into the renal ostium (Reprinted with permission from Schönberg SO, Rieger J, Weber CH, Michaely HJ, Waggershauser T, Ittrich C, Dietrich O, Reiser MF [2005] High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 35:687–98)

5.4  MR Angiography

image detail. However, studies suggest that the resulting higher matrix allows an easier and more accurate image reading with decreased partial volume effects (Du et al. 1994). 5.4.1.4.4  k-Space Trajectories In conventional or so-called Cartesian readout of the k-space, the k-space lines are filled sequentially, starting and ending in the periphery. Hereby, the center of k-space is filled roughly in the middle of the scan. Cartesian filling is sufficient for standard imaging like orthopedic morphologic imaging and can also be used for MRA. In this case, the contrast agent bolus has to be timed to coincide with the acquisition of the central k-space regions. Most scanners provide information on when the center of k-space is reached so that the users can adapt the bolus timing and the sequence start correspondingly. As many users consider the circulation time calculation and acquisition timing as disadvantageous, they prefer semiautomatic techniques where the arrival of the contrast agent bolus is monitored by the scanner in real time, and the MRA sequence is started when a certain enhancement threshold is reached. As the MRA sequence then has to be started at the time of the maximal enhancement, starting to fill peripheral k-space as with Cartesian trajectories would result in poorly contrasted MRA images. Therefore, a different k-space scheme can be applied where the center of k-space is filled first. There are different approaches to this concept: spiral imaging where the acquisition vector spirals out from the center of k-space to the periphery, centric or elliptic-centric imaging where equidistant points are sampled at the same time from k-space center towards the periphery, and the Fig. 5.4.11  a Scheme of k-space with a centrally acquired part so-called CENTRA (Philips’ acronym for a specific k(plain color) and a “zero-filled” outer part. The zero filling takes space sampling technique) approach where the center of place after the data acquisition and is neither time- nor SNR- k-space is filled in a random order (Fig. 5.4.12) (Willinek consuming but does not increase true image content. b Nev- et al. 2002). All these different approaches have their adertheless, the resulting images show a smoothened aspect and vantages and drawbacks. CENTRA is a robust approach allow for a better delineation of fine structures. In this example for k-space filling; however, it is limited to a certain ventwo images acquired with a matrix of 128 × 128 and 256 × 256 dor. Spiral imaging is very fast, has sufficient SNR, and are zero filled and interpolated to the next matrix size. The re- is increasingly often used for time-resolved imaging as it sulting images are smoother; no increased image noise can be conveniently allows easy k-space under-sampling. Both, seen. Please note how well the renal artery stenosis on the upper spiral and (elliptic) centric trajectories are insensitive panel can visualized in the zero filled image despite the fact that to venous overlay as the acquisition of the k-space centhe true image content is unchanged between the left and right ter and hence, the determination of the image contrast takes place at the beginning of the initial arterial vessel image enhancement prior to venous return. On the other hand, they are even more susceptible to missed bolus timing. Radial imaging, a newly arising imaging technique, zeros before Fourier transformation of the data. Zero filling affects neither the image SNR nor the image acquisi- allows for very fast imaging and has inherent undertion time. No extra lines in k-space are acquired with this sampling properties as central and peripheral parts of technique. This technique basically only interpolates the k-space are acquired at the same time. An image can image data to the size of the next matrix. The resulting be reconstructed from just a few sampled views. Howimages are almost doubled in size, without increase of ever, with too few views acquired there are characteris-

799

800

5  Thorax and Vasculature

Fig. 5.4.12a–c  Schematics for the most common k-space trajectories: Cartesian, spiral, and radial. a In Cartesian k-space readout the data are read out line by line where usually each line is read out per single echo. In turbo spin-echo sequences multiple lines are filled per echo train. b In spiral readout the readout vector spirals out from the center of k-space following a spiral

readout gradient, which allows for fast imaging. In time-resolved imaging interleaved time-resolved techniques are often used. c In radial imaging different views of k-space are acquired, which all extent through the center of k-space. As a large part of k-space can be covered with just a view views acquired radial techniques are often used for fast imaging such as real time cardiac MRI

tic reconstruction artifacts, the so-called streak artifacts, PI offers to either shorten scan time with unchanged which can only be alleviated with an increased number of spatial resolution, to increase spatial resolution with unsampled views. Currently, centric approaches and Carte- changed scan time, or to combine both. Increasing the sian sampling are the main techniques used. With the use spatial resolution in combination with shortened scan of parallel imaging, which further decreases image acqui- time will however further decrease the SNR (Eqs. 5.4.4, sition times, timing becomes even more crucial with cen- 5.4.5). The images are reconstructed from the undertric approaches. However, initial papers have reported sampled data set either in the image domain using either that image contamination from venous signal decreased the known coil sensitivities (i.e., using sensitivity encodwhen centric approaches and parallel imaging were com- ing [SENSE]) or in the k-space domain using reconstrucbined and when image acquisition was timed correctly. tion weights (as in the GRAPPA algorithm). GRAPPA seems hereby somewhat superior, as it is less susceptible to patient motion and aliasing artifacts, which are typical for MRA with a limited FOV in the left–right direction 5.4.1.4.5  Parallel Imaging (Griswold et al. 2004; Goldfarb 2004). Due to the high Since its introduction in the late 1990s (Pruessmann et al. available SNR, contrast-enhanced MRA is well suited for 1999; Griswold et al. 2002), parallel imaging (PI) has rev- the application of PI (Heidemann et al. 2004). Various olutionized the MRA technique. A technical prerequisite studies have been published that examined the impact of for PI are coil systems with multiple receive coil elements PI on MRA ranging from feasibility to image quality and providing spatially differing sensitivity profiles: acquir- superior spatial resolution (Born et al. 2005; Michaely et ing data with all coil elements simultaneously, PI system- al. 2005a; Schönberg et al. 2005). As the SNR may drop atically under-samples the k-space and thus reduces the significantly when PI is employed and the spatial resoimage acquisition time (see Eq. 5.4.3) (Figs. 5.4.13, 5.4.14, lution increased, the application of improved contrast 5.4.15). With an acceleration (or reduction) factor of R, agents may be beneficial in such cases: either the 1 M only every Rth line in k-space is acquired. For PI with gadobutrol (Schering), the higher-relaxivity gadobenate acceleration factor R, the following relationships between dimeglumine (Bracco) or the intravascular protein-bindscan time and SNR can be found: ing gadofosveset (Schering) seem to offer better enhancement during the first pass of the contrast agent. Another ScanTime NoPI ScanTime PI = (5.4.6) very promising combination is to use PI at 3 T. The R doubled SNR at 3 T offers enough SNR for increasing the spatial resolution in combination with higher factors of 1 , (5.4.7) parallel imaging (factor 4), with no visible degradation in SNR P I = SNR NoP I × g× R image quality. MRA sequences are largely not affected by the typical dielectric artifacts and increased susceptibility g is the so-called geometry factor a measure of the suit- at 3 T. ability of the coils for PI, which depends on the coil deApart from the good signal despite the shortened acsign and position and is always >1. quisition time, the combination of 3 T and PI results in

5.4  MR Angiography Fig. 5.4.13  In PI, k-space is underampled by acquiring only every Rth line (acceleration factor R). As this de facto is equivalent to acquiring images with a reduced FOV, the Fourier reconstructed images will reveal aliasing artifacts

Fig. 5.4.14  In PI only every Rth line of k-space is acquired leading to aliased images after the Fourier transformation. To overcome this aliasing, the spatial information derived from the position of the different coil elements is used to unfold the images. Figuratively speaking, each coil element “sees” a certain part of the image best. Based on this knowledge the aliasing can be corrected

Fig. 5.4.15  This scheme demonstrates the image quality achievable with increasing acceleration factors of PI from R = 1 (no acceleration) to R = 4. With increasing R, the image noise becomes more and more perceivable, and the SNR of the image is markedly reduced. Because of the noise, image details such as the lumbar arteries are harder to distinguish from the background. Due to the higher SNR at 3 T, acceleration with a factor of 4 would yield signal almost equivalent to a non-accelerated scan at 1.5 T with the same parameters

further benefits for MRA. While the increase of SNR at read out with parallel imaging, the SAR can be reduced 3 T is desirable, the four-fold increase in the specific-ab- and hence imaging limitations be avoided. In return sorption rate (SAR) is an undesired characteristic of 3-T however, with fewer pulses applied in a shorter time, the imaging. Due to the reduced overall number of excita- stimulation threshold may be surpassed. Thus, PI helps tions that are needed when only a fraction of k-space is effectively in reducing the SAR but choosing the short-

801

802

5  Thorax and Vasculature

est TR values is not recommended due to the above mentioned limitations. All vendors now offer multi-element array coils which allow parallel imaging with acceleration factors of at least 2 in left–right phase-encode direction. Parallel imaging with a factor of 2 can also be applied in anterior–posterior phase-encode direction when a separate spine and body coil are used. Recent developments comprise multi-element body coils which hold up to 32 elements and thus enable faster imaging with higher parallel imaging acceleration factors of at least 3 in the phase-encode direction. Since the maximum PI acceleration factor depends on the number of independent coil elements in the phaseencode direction, these coil systems allow for a much more flexible choice of image orientation (i.e., selection of the phase-encode direction) since they facilitate parallel imaging capabilities in all spatial directions. Recent developments in PI comprise parallel imaging in two directions, also known as PAT2 (parallel acquisition techniques squared) (Fenchel et al. 2006a), the phaseencode direction and the partition-encode direction. In this case, the overall acceleration factor is the product of the acceleration factor in the phase-encode direction and of the acceleration factor in the partition-encode direction. Of course with higher acceleration factors also the SNR loss is increased accordingly. PAT2 is well suited for large volumes with good differentiation of the coil sensitivity profiles. With smaller volumes the differences of the coils’ sensitivities are not sufficient for the central parts, resulting in reconstruction artifacts.

Table 5.4.5  Typical imaging parameters for a 1.5-T scanner

5.4.1.5.1  Brain

5.4.1.5.3  Pulmonary Vessels

TOF MRA can be technically acquired at any field strength, but profits from the prolonged T1 times at 1.5 or even 3 T. In return longer TR times have to be accepted as well. As can be seen in Table 5.4.5, a voxel size 0.8 × 0.6 × 0.8 mm3 can be acquired at 1.5 T. With the help of parallel imaging a spatial resolution of 0.5 × 0.3 × 0.5 mm3 can be achieved at 3 T in just 3–4 min’s acquisition time. The 3D slab is normally tilted to include the vessels of the skull base, particularly the circle of Willis. If an arterial MRA is desired it is advisable to place a

Imaging of the pulmonary circulation is an arising and demanding MRA application, as a large volume has to be covered with short imaging times. The need for short imaging times is a consequence of the high susceptibility of the lung tissue. Dedicated MRA sequences use high readout bandwidths to achieve short echo times (TE) and hence to prevent increased intra-voxel spin dephasing (Table 5.4.7). A coronal slab with a PI factor of 3 or two sagittal slabs with a PI factor of 2 are commonly used to cover the en-

Brain TOF TR/TE (ms)

36/7.15

α (°)

20

Voxel size (mm3)

0.8 × 0.6 × 0.8

Scan time (s)

3.19 min

Matrix

384 × 70%

FOV (mm2)

220 × 83%

Bandwidth (Hz/px)

73

PI

2

saturation band superior to the imaging volume to saturate venous spins before they enter the imaging volume and vice versa, if a venous MRA is to be performed the saturation-region should be placed inferiorly to the imaging volume. 5.4.1.5.2  Supra-Aortic Vessels

The slab for supra-aortic MRA is commonly positioned in a coronal orientation to include the common and internal carotid arteries as well as the vertebral arteries. As the arteriovenous transit time of the cerebral circulation is very short, exact timing of the sequence is crucial. Most 5.4.1.5  Clinical Applications users prefer semiautomatic bolus timing approaches for MRA is currently considered the gold standard for most this purpose, even though their superiority compared to diagnostic applications. Depending on the available scan- conventional MRA has not been proved so far. ner and field strengths, different protocols and imaging High spatial resolution is essential for the detection approaches are used. In the following typical sequence and grading of stenotic vessel changes and reporting of parameters for a state-of-the-art 1.5-T MRA scanner with area stenosis. Parallel imaging is an extensively investimatrix coil system and general acquisition recommenda- gated and used technique which allows for an increased tions are given. If applicable, particular advantages of 3-T spatial resolution as demonstrated by numerous authors MRA are pointed out. (Table 5.4.6).

5.4  MR Angiography Table 5.4.6  Typical imaging parameters for a 1.5-T scanner

Table 5.4.8  Typical imaging parameters for a 1.5-T scanner

Carotid MRA

Abdominal/renal MRA

TR/TE (ms)

3.69/1.23

TR/TE (ms)

3.77/1.39

α (°)

20

α (°)

25

Voxel size (mm3)

0.9 × 0.7 × 0.9

Voxel size (mm3)

1 × 0.8 × 1

Scan time (s)

0.23 min

Scan time (s)

0.20–0.26 min

Matrix

448 × 78%

Matrix

512 × 80%

FOV (mm2)

330 × 64%

FOV (mm2)

400 × 87.5%

Bandwidth (Hz/px)

360

Bandwidth (Hz/px)

350

PI

2

PI

2–3

Table 5.4.7  Typical imaging parameters for a 1.5-T scanner Pulmonary MRA TR/TE (ms)

2.39/0.89

α (°)

15

Voxel size (mm3)

1.3 × 1 × 1.1

Scan time (s)

0.20 min

Matrix

384 × 77%

FOV (mm2)

390 × 100%

Bandwidth (Hz/px)

870

PI

3

tire lungs. Again timing is crucial to obtain a solely arterial image as the arteriovenous transit times of the pulmonary vessels are only in the order of 3–6 s (Schönberg et al. 1999a). When a test-bolus technique or semiautomatic bolus detection is used, the region of interest (ROI) should be placed in the common pulmonary artery. 5.4.1.5.4  Aorta and Renal Arteries For sole imaging of the aorta—as in the case of dissecting aneurysms—an oblique sagittal slab can be prescribed, which substantially reduces total acquisition time. If the renal arteries are the focus of the examination a coronal slab is used. For the detection of subtle changes of the renal arteries a high spatial resolution is mandatory (Table 5.4.8). Coils with posterior and anterior elements are optimally suited to yield a high SNR. Imaging at 3 T in

combination with PI allows for an increased spatial resolution with decreased scan time (Michaely et al. 2005a). The ROI for test bolus or semiautomatic bolus detection is commonly placed in the aorta at the level of the celiac axis. 5.4.1.5.5  Peripheral Vessels MRA of the peripheral vessels (Table 5.4.9) almost always refers to imaging of the lower extremities, starting at the level of the renal arteries down to the feet. As imaging of the arms is rarely ordered in clinical routine, this paragraph focuses on lower-extremity MRA. If MRA of the upper extremities is requested time-resolved techniques are very helpful. Dedicated coils should be used as they allow for higher SNR and hence a better spatial resolution. There are two approaches to imaging of the lower extremities, the bolus-chase technique and hybrid techniques. In bolus-chase MRA, the contrast agent is administered and sequential MRA scans are performed at the aortopelvic level, at the level of the thighs, and finally at calf level. However, this approach is prone to venous contamination of the distal arteries. Therefore, the so-called hybrid technique is evolving to a standard technique (Meissner et al. 2005). Two separate injections of the contrast agent are performed. With the first injection the distal arteries are imaged. The second bolus is used for sequential imaging of the aortopelvic level and the thigh level. For the bolus-chase technique the ROI is positioned at the level of the renal arteries. For the hybrid technique ROIs are placed, one at the level of the proximal calf arteries and a second one at the level of the renal arteries. Alternatively, a time-resolved technique can be applied for the calf and foot levels.

803

804

5  Thorax and Vasculature Table 5.4.9  Typical imaging parameters for a 1.5-T scanner for the upper and lower legs Upper leg-MRA TR/TE (ms)

3.46/1.21

α (°)

15

Voxel size (mm3)

1.6 × 1 × 1.5

Scan time (s)

0.15 min

Matrix

512 × 60%

FOV (mm2)

500 × 75%

Bandwidth (Hz/px)

360

PI

2 Lower leg MRA

TR/TE (ms)

3.72/1.33

α (°)

15

Voxel size (mm3)

1.2 × 1 × 1

Scan time (s)

0.37 min

Matrix

512 × 80%

FOV (mm2)

500 × 75%

Bandwidth (Hz/px)

340

PI

No

5.4.2  Pulmonary MRA C. Fink 5.4.2.1  Introduction Although it is sometimes still considered the gold standard for imaging of the pulmonary vasculature, conventional catheter pulmonary digital subtraction angiography (DSA) has been largely replaced by non-invasive cross-sectional imaging. Computed tomography angiography (CTA) of the lung has practically become the firstline imaging technique for the assessment of pulmonary vascular disease (Schoepf and Costello 2004). For the assessment of the pulmonary circulation, unlike other vascular territories (e.g., the peripheral arteries of the lower legs), CE MRA is generally considered only a second-line imaging tool. The inferior spatial resolution and longer examination and/or breath-hold times are considered to be the major drawbacks of pulmonary MRA in comparison to CT. In patients with pulmonary embolism, the limited patient access in the magnet, as well as the environment of the magnetic field, requiring

dedicated MR-compatible monitoring devices, are considered potential contraindications for MRI. On the other hand, there are also potential advantages of pulmonary MRA for the evaluation of lung disease. Above all, this includes the lack of ionizing radiation, which is of major importance in congenital disease (e.g., congenital heart disease) and chronic diseases requiring frequent follow-up examinations. The favorable safety profile of MR contrast media as well as the unachieved potential to combine morphologic with functional imaging (e.g., the assessment of right heart function in pulmonary arterial disease) are further advantages of MRI. 5.4.2.2  Technical Considerations The anatomy of the pulmonary circulation has some specific implications for the imaging technique. Due to multiple air–tissue interfaces the lungs have a very high susceptibility. This substantially affects the achievable signal intensity of small peripheral lung vessels. On the other hand, the low signal of the lung parenchyma usually results in a high vessel-to-background contrast of pulmonary MRA. In general short echo times, i.e., lower than 2–3 ms should be used to eliminate susceptibility effects (Prince et al. 2003). The vascular density of the lungs is very high. In addition to the vasa publica, which consist of the pulmonary arteries and pulmonary veins, the lungs also have the vasa privata, i.e., the bronchial arteries which originate from the aortic arch. Moreover, the lungs have a very short transit time in the range of 3–5 s (Fishman 1963). As a consequence pulmonary MRA often has substantial venous contamination, potentially affecting the diagnostic accuracy. To reduce venous contamination a very compact bolus profile should be aimed by using high injection rates (e.g., 5 ml/s). Another option to improve arteriovenous separation is the use of time-resolved imaging strategies, which, depending on the acquisition time and injection protocol, will result in exclusive angiograms or venograms of the lungs (Schönberg et al. 1999a; Fink et al. 2005a). Postprocessing can also improve arteriovenous separation, including simple image subtraction or dedicated algorithms such as correlation analysis (Bock et al. 2000). Although several methods have been proposed for MRI of the pulmonary vasculature (including non-enhanced time of flight or black-blood angiography) 3D CE MRA has been established as the standard method for pulmonary MRA. There are two different approaches to pulmonary CE MRA. One approach is to acquire 3D volume data of the pulmonary vasculature with a high spatial resolution, ideally with isotropic spatial resolution in all directions. To improve the spatial resolution, the acquisition of two sagittal small-field-of-view data sets with two separate contrast injections has been proposed (Oudkerk et al. 2002). To ensure maximum vascular contrast of the pulmonary vessels the acquisition

5.4  MR Angiography

of the central k-space data has to be synchronized with the arrival of the contrast agent bolus in the pulmonary arteries. Similarly to other vascular territories this can be achieved by a test-bolus examination or automated bolus triggering. Another approach is to perform pulmonary MRA in a time-resolved fashion. In addition to an improved arteriovenous separation, time-resolved pulmonary MRA is less sensitive for incorrect bolus timing and less sensitive for motion artifacts. The latter may be relevant in dyspneic patients. In addition, and probably most importantly, time-resolved MRA also allows acquisition of functional information about the pulmonary circulation, such as the characterization of shunts or the assessment of capillary perfusion of the lung parenchyma. Several studies have shown the feasibility of pulmonary perfusion MRI using a time-resolved MRA technique (Fink et al. 2003, 2004a; Ohno et al. 2003; Nikolaou et al. 2005). In addition to a qualitative evaluation of lung perfusion, several studies have also proposed a semiquantitative evaluation of lung perfusion using time-resolved 3D CE MRA (Fink et al. 2004b, 2005b; Ohno et al. 2004a; Nikolaou et al. 2004). Although these studies have found similar lung perfusion values compared to reference data from H2O15 PET (Schuster et al. 1995), several methodological challenges have to be considered for a potential clinical application. The most important concern for a quantitative approach of lung perfusion using time-resolved 3D CE MRA is the missing linearity between the contrast agent concentration and measured signal intensity changes. However, this limitation may be overcome using optimized contrast-agent doses or dedicated dualbolus injection protocols (Nikolaou et al. 2004). The introduction of PI has substantially improved the potential of pulmonary MRA. For high-spatial-resolution MRA the acquisition time can be reduced or the spatial resolution can be increased to submillimeter isotropical resolution. For time-resolved MRA the temporal resolution can be substantially reduced without compromises in the spatial resolution. As mentioned above contrast injections should be performed with a high flow rate using automatic power injectors. For high-spatial-resolution MRA, a dose usually

in the range of 0.1 to 0.2 mmol/kg body weight is considered ideal. For time-resolved MRA the dose should be lowered, i.e., ≤0.1 mmol/kg body weight. For pulmonary MRA, blood-pool MR contrast agents are promising, as alternative imaging techniques, i.e., navigator-gated MRA might be realized, which cannot be used with conventional extracellular MR contrast agents (Abolmaali et al. 2003). Moreover, these chelates usually have a higher relaxivity that might be used to improve the spatial resolution. Finally, time-resolved MRA and high-spatial-resolution MRA might be combined after a single injection (Fink et al. 2004c). 5.4.2.3  Clinical Applications Pulmonary MRA has been increasingly used in a variety of pulmonary vascular disorders. The imaging technique can be tailored to the clinical problem, ranging from detailed morphologic evaluation using high-spatial-resolution-MRA to functional studies characterizing shunts or lung perfusion. The most important applications are reviewed in detail below. 5.4.2.3.1  Pulmonary Embolism Several studies have evaluated the feasibility and diagnostic accuracy of pulmonary MRA for the assessment of patients with suspected acute pulmonary embolism (Fig. 5.4.16). In an early study by Meaney et al. (1997) pulmonary CE MRA was evaluated in 30 consecutive patients with suspected acute pulmonary embolism. Conventional pulmonary angiography served as the gold standard. For three different readers the sensitivity ranged between 75 and 100%, with specificity from 95 to 100%. All lobar emboli were identified, whereas 16 of 17 segmental emboli were identified. The interobserver agreement was good (κ = 0.57 to 0.83 for all vessels, 0.49 to 1.0 for main and lobar vessels, and 0.40 to 0.81 for segmental vessels) (Meaney et al. 1997).

Fig. 5.4.16  Images of a time-resolved 3D CE MRA of a patient with acute central pulmonary embolism. Using PI the temporal resolution could be reduced to under 10 s per 3D data set without relevant trade-offs in the spatial resolution. The spatial resolution is sufficient to visualize peripheral emboli (arrow)

805

806

5  Thorax and Vasculature

In a subsequent study, Gupta et al. (1999) evaluated CE MRA of the pulmonary circulation in 36 patients with suspected acute pulmonary embolism. Again, conventional pulmonary angiography served as the gold standard. The overall sensitivity and specificity for pulmonary embolism was 13 of 19 (68%) and 395 of 396 (99%), respectively. If sub-segmental emboli were excluded, the sensitivity increased to 87%. In a more recent study, 141 consecutive patients with suspected acute pulmonary embolism and negative-perfusion scintigraphy were examined with two sagittalorientated 3D MRA data sets of the pulmonary arteries. Compared to conventional pulmonary angiography MRA successfully identified 27 of 35 patients with pulmonary embolism, corresponding with a sensitivity of 77%. Again, the sensitivity for the evaluation of subsegmental pulmonary embolism was poor (40%), but was excellent for lobar embolism (100%). The overall specificity of pulmonary MRA in this trial was 98% (Oudkerk et al. 2002). Also the accuracy of time-resolved MRA for the detection of acute pulmonary embolism has been evaluated. In a feasibility study in eight dyspneic patients with known or suspected pulmonary embolism, time-resolved MRA with a temporal resolution under 4 s was assessed. Subsegmental pulmonary arteries could be visualized in all subjects. In all four patients with proven pulmonary embolism there was a concordance between MRA and corroborative imaging studies. It was concluded that time-resolved MRA of the pulmonary vasculature can be obtained even in patients with severe respiratory distress (Goyen et al. 2001a). In a very recent study using PI time-resolved MRA was compared with multi-detector CT (MDCT) and ventilation-perfusion scintigraphy (VQ scan). Conventional pulmonary angiography served as the gold standard. Time-resolved MRA had a higher diagnostic accuracy for the detection of PE than CTA or VQ scan. In detail, the sensitivity of time-resolved MRA was 92% compared with 83% (CTA) and 67% (VQ scan). Similarly the specificity of time-resolved MRA was 94% compared with 94% (CTA) and 78% (VQ scan). Both the positive and negative predictive value (PPV and NPV) of time-resolved MRA was superior to CTA and scintigraphy (PPV: 85% (MRA), 83% (CTA), and 50% (VQ scan); NPV: 97% (MRA), 94% (CTA), and 88% (VQ scan) (Ohno et al. 2004b). 5.4.2.3.2  Pulmonary Hypertension Pulmonary hypertension (PH) is characterized by an elevation in pulmonary artery pressure that can lead to right ventricular failure. Although there is no cure for PH the therapeutic options have been substantially improved in the recent years. As a consequence imaging of pulmonary hypertension has become clinically relevant. One of the major goals of imaging in patients with pulmonary hypertension is the differentiation between idio-

pathic pulmonary hypertension (IPAH) and secondary forms, e.g., chronic pulmonary hypertension (CTPH), as this has major implications for the therapeutic strategy. While IPAH can be treated effectively medically with vasodilators (e.g., prostaglandin derivates), CTPH can be treated surgically by pulmonary thromboendarterectomy. As this procedure has a high mortality rate, a correct preoperative classification of the localization and extent of thrombotic material in the pulmonary arteries is mandatory. The key imaging finding in PH is dilation of the central pulmonary arteries. In addition characteristic morphologic changes of the pulmonary arteries may be observed with different types of PH. In CTPH typical findings are variation of segmental pulmonary artery diameter, vascular wall irregularity, vascular bands, webs, pouches, and complete obstruction. Moreover, dilated bronchial arteries can often be found in CTPH indicating bronchosystemic shunting (Ley et al. 2004). At perfusion MRI lobar or wedge-shaped segmental and subsegmental perfusion defects are observed with CTPH (Fig. 5.4.17). In contrast, patients suffering from IPAH will not show vessel wall irregularities or vascular obstruction. At perfusion MRI, typically a patchy perfusion pattern or a non-segmental peripheral perfusion loss is observed (Fink et al. 2004). In two recent studies pulmonary CE MRA with PI was evaluated to differentiate between IPAH and CTPH. In a feasibility study in 10 patients with IPAH and CTPH high-spatial resolution 3D CE MRA was compared to time-resolved 3D CE MRA (temporal resolution 1.5 s) (Ley et al. 2005). Visualization of the pulmonary arteries was possible down to a subsegmental level using high-spatial-resolution 3D MRA technique, while visualization of the vascular anatomy for the time-resolved MRA was only achieved to a segmental level. In general high spatial resolution outperformed time-resolved MRA in the visualization of intravascular abnormalities. However, no differences were found regarding the classification between IPAH and CTPH. It was concluded that time-resolved MRA might be useful in dyspneic patients to differentiate between different forms of pulmonary hypertension. In a more recent study by Nikolaou et al. (2005) the accuracy of a combined MR protocol of pulmonary perfusion MRI (i.e., time-resolved MRA) and high-spatialresolution MRA for the differentiation of different forms of PH was assessed. As a standard of reference for the perfusion examination scintigraphy was available in the majority of patients. It was shown that using the comprehensive MR imaging protocol, a correct differentiation of IPAH and CTPH could be made in 90% of the patients. High-resolution CE MRA showed a good agreement (72– 97%), with CTA and DSA for image findings indicating either IPAH or CTPH. MR perfusion imaging showed an agreement (i.e., identical diagnosis on a per patient basis) of 79% to perfusion scintigraphy. The interobserver agreement was good (κ = 0.63).

5.4  MR Angiography Fig. 5.4.17a–d  High-resolution MRA (a,b) and time-resolved MRA (perfusion MRI, c,d) of a patient with chronic thromboembolic pulmonary hypertension. In the highresolution MRA data set the wall-adherent thrombus can be visualized (arrows). The time-resolved MRA data shows severe segmental hypoperfusion (arrows)

Apart from the differentiation of different entities in patients with CTPH a further goal of imaging is to identify those individuals who may benefit from surgical intervention. CE MRA has been compared with CTA for this purpose in a retrospective study of 32 patients with CTPH who were studied prior to surgery. Both techniques were equally effective with regard to the visualization thrombotic wall thickening, intraluminal webs, and abnormal proximal to distal tapering (Ley et al. 2003). One of the most important advantages of MRI over other imaging modalities is that it allows a comprehensive imaging approach allowing for the evaluation of morphology and function (e.g., perfusion, blood flow velocity, etc.) in a single examination. Comprehensive MRI has also been evaluated in patients with pulmonary hypertension. In 34 patients with CTPH, MRA revealed typical signs of chronic pulmonary embolism up to the segmental level in all patients (Kreitner et al. 2004). However, with the spatial resolution of the MRA technique of this study (e.g., partition thickness 2.7 mm) MRA was still inferior to DSA. MRI could demonstrate reduced right ventricular ejection fractions and pulmonary bloodflow velocity that significantly increased after surgical thromboendarterectomy. MRI had good correlation with conventional right heart catheter measurements such as pulmonary vascular resistance or mean pulmonary arterial pressure.

5.4.2.3.3  Pulmonary Veins Catheter ablation has become a standard procedure for the treatment of atrial fibrillation (AF). CE MRA has shown a high value for the pre-interventional assessment of the pulmonary veins, as up to 40% of patients present with variant anatomy of the pulmonary veins. CE MRA has also been postulated by several authors for the assessment of pulmonary vein stenosis, which is a feared complication of AF ablation (Fink et al. 2003b; Kluge et al. 2004). In a large series 110 patients were screened with MRA before and after AF ablation therapy. Of those patients 45 showed a ≥25% reduction of the vein diameter and another 6 patients developed clinical symptoms suggestive of pulmonary vein stenosis. All those 51 patients were additionally scanned with time-resolved MRA for the assessment of perfusion changes. The comprehensive information was also used for the indication for angioplasty of pulmonary vein stenosis (Kluge et al. 2004). Anomalous pulmonary venous return is an infrequent but important anomaly of the thoracic veins. Frequently, anomalous pulmonary venous return is associated with congenital cardiac anomalies, especially atrial septal defect (White et al. 1997). Whereas total anomalous pulmonary venous return (TAPVR) usually becomes clinically apparent with cyanosis shortly after birth, the more common partial anomalous pulmonary venous return (PAPVR) may be asymptomatic. The most com-

807

808

5  Thorax and Vasculature

mon form of PAPVR is an anomalous right upper lobe vein that enters the superior vena cava or right atrium. This abnormality is associated with a high prevalence of sinus venosus type of atrial septal defect. Less common forms are anomalies of the right lower lobe vein (scimitar syndrome) or left pulmonary veins. Other venous pathologies than can be assessed by pulmonary MRA are anomalous pulmonary venous return and scimitar syndrome (Zaporozhan et al. 2005). MRI has been proposed as a tool for a comprehensive evaluation of anomalous pulmonary venous return (White et al. 1997; Ferrari et al. 2001; Festa et al. 2006; Greil et al. 2002). In addition to a 3D visualization and classification of the venous anomaly by CE MRA, cine MRI can be used for the visualization of associated cardiac anomalies. Additionally, the left-toright shunt can be quantified using PC MRI.

5.4.2.3.4  Arteriovenous Malformations Pulmonary arteriovenous malformation (AVM) is an abnormal communication between the pulmonary artery and the pulmonary vein. The majority of pulmonary AVMs are congenital (~80%), and up to 80% of these are associated with Rendu-Osler-Weber syndrome (Khurshid and Downie 2002). Pulmonary AVM usually can be well visualized by pulmonary MRA (Fig. 5.4.18). A major indication for imaging is the evaluation for potential interventional embolization. Several studies using MRA for the evaluation of patients with pulmonary AVM have been reported. Whereas static CE MRA allows for a high-resolution visualization of the AVM, time-resolved MRA allows for a functional assessment of the AVM.

Fig. 5.4.18a–f  MRA in a patient with Osler disease and pulmonary arteriovenous malformation (AVM). High-resolution MRA (a) shows a large AVM of the right upper lobe. Time-resolved MRA (b,c) shows an early venous filling of the pulmonary vein of the right upper lobe (arrow). Color-coded map of pulmonary transit time (d) calculated from time-resolved MRA shows decreased transit time in that area compared to the rest of the lung. Findings of MRA were confirmed by catheter angiography (e), which was performed for embolization (f) (Reprinted from Fink and Kauczor 2006 with permission)

5.4  MR Angiography

In a study of eight patients with suspected pulmonary arteriovenous malformations MRA was compared to conventional pulmonary arteriography or surgery. Nine of ten (90%) AVM diagnosed at MRA were confirmed at conventional angiography or surgery. The single missed malformation was small (3–4 mm) and located in the periphery of the lung. Two additional arteriovenous malformations were diagnosed in two subjects who did not undergo additional confirmatory testing (Maki et al. 2001). In a different study MRA and perfusion MRI were compared with CTA and conventional angiography in eight patients with 15 AVM. In addition patients were also studied after embolotherapy. All vessels greater than 3 mm were adequately visualized, and results were similar for the various modalities. MRA was able to demonstrate a size reduction of the AVM after interventional embolization (Ohno et al. 2002). 5.4.3  MRA of the Supra-Aortic and Intracranial Vasculature C. Fink, U. Attenberger, H.J. Michaely, and S.O. Schönberg 5.4.3.1  Introduction Diagnostic DSA is still considered as the gold standard for the evaluation of the supra-aortic and intracranial vasculature. However, DSA is expensive and still holds a certain complication rate. This includes an approximately 4% incidence of transient ischemic attack and 1% incidence of disabling stroke caused by selective arterial catheterization (Kuntz et al. 1995). Because of this, diagnostic DSA has been more and more replaced in clinical practice by non-invasive cross-sectional imaging techniques, such as Doppler ultrasonography (DUS), computed tomography angiography (CTA), and MRA. DUS has the advantage of being inexpensive and broadly available. Using DUS the cervical vasculature and proximal intracranial arteries can be assessed with a high accuracy (Derdeyn et al. 1995; Modareski et al. 1999; Patel et al. 1995; Baumgartner et al. 1995). However, the accuracy of DUS is strongly operator dependent. Another limitation is that the limited acoustic window restricts the evaluation of the proximal aortic arch branch vessels and distal intracranial circulation. Especially with the availability of multidetector scanner technology, CTA has become a more frequently used imaging technique for the assessment of the supra-aortic and intracranial vasculature (Marks et al. 1993; Cinat et al. 1999; Koelemay et al. 2004; Forsting 2005). Among the limitations of CT are the radiation exposure and the application of potentially nephrotoxic iodine-based contrast agents. Moreover, the analysis of CTA may be hampered by artifacts from calcified plaque or adjacent bone structures.

5.4.3.2  Imaging Technique Due to its superficial location the cervical and intracranial vasculature is ideally suited for MRI (Carr et al. 2001). Three different techniques are used for MRA of the supra-aortic and intracranial vasculature, the nonenhanced techniques, time-of-flight (TOF) MRA and phase-contrast (PC) MRA, and CE MRA. In TOF MRA, repetitive pulses are used to suppress stationary background tissues, while the unsuppressed protons of inflowing blood show high signal intensity (Wehrli et al. 1986). The main limitations of TOF MRA are artifactual signal loss due to spin dephasing in turbulent flow or due to spin saturation along the course of the blood through the imaging volume. To overcome the limitations of flow saturation 3D TOF MRA can be improved by a multi-slab TOF acquisition with overlapping data sets (multiple overlapping thin slab acquisition [MOTSA]) (Blatter et al. 1991). Saturation effects can also be minimized by using a variable flip-angle excitation technique (tilted optimized non-saturating excitation [TONE]). In this technique, the flip angle varies across the slab that it is set lower at the inlet side, and gradually increases as it approaches the exit side to increase the blood signal (Nagele et al. 1994). The contrast of TOF MRA is improved by the application of magnetization transfer (MT) pre-pulses, which improve the suppression of background signal of the stationary tissues (Mathews et al. 1995). The visualization of small peripheral arterial branches can be improved by intravenous injection of paramagnetic contrast material. However, this on the other hand causes enhancement of perivascular tissues, potentially reducing the image contrast (Ozsarlak et al. 2004). The other non-enhanced MRA technique used for the evaluation of the supra-aortic and intracranial vasculature is PC MRA. This imaging technique uses a bipolar phase-encoding gradient to cause a velocity-dependent phase shift of flowing blood (Dumoulin et al. 1989). Spins moving in the direction of increasing gradient strength advance in phase, while those moving in the opposite direction fall behind the phase of stationary tissue. The signed phase image is multiplied by a magnitude base image to suppress background noise resulting in an MRA. One of the major limitations of PC MRA is that blood velocities higher than the preselected velocity-encoding (VENC) will not be represented or misrepresented in the image. Therefore, the VENC has to be defined in advance and adapted to the vascular system of interest. Other limitations of PC MRA are similar to those of TOF MRA, such as intra-voxel dephasing and long acquisition times (Ozsarlak et al. 2004). Last but not least, CE MRA has become an increasingly used MRA technique, especially for the depiction of the cervical arteries. The major advantage of CE MRA over non-enhanced MRA techniques is the substantially shorter acquisition time. Moreover, the image contrast is determined by the T1-shortening effect of Gd-based MR

809

810

5  Thorax and Vasculature

contrast media. Therefore, it has the potential to overcome some of previously discussed flow-related limitations of TOF MRA. Similar to other body regions the data acquisition should be synchronized with the arrival of the contrast agent bolus using a test bolus or fluoroscopic triggering. One of the major limitations of CE MRA in the assessment of the supra-aortic and intracranial vasculature is the short bolus transit time, leading to undesired venous contamination. To reduce this problem elliptical centric encoding of the k-space can be used, in which the lower spatial frequency information that determines the image contrast is sampled first (Wilman et al. 1998; Carroll et al. 2001; Huh et al. 2004). Another option to reduce venous overlay and to improve arteriovenous separation is the use of time-resolved MRA techniques. Especially with alternative k-space sampling techniques, such as PI or view sharing, the temporal resolution of the MRA acquisition can be substantially improved, while preserving the spatial resolution (Carroll et al. 2001; Huh et al. 2004; Michaely et al. 2004; Nael et al. 2006a; Willinek et al. 2003b). Time-resolved MRA can further be used to assess hemodynamic changes in the supra-aortic and intracranial vessels, such as in subclavian steal syndrome, paragangliomas, and cerebral arteriovenous malformation (Michaely et al. 2007). In the neck, CE MRA is nowadays the most frequently used MRA technique. In contrast, TOF MRA still remains the most frequently used MRA technique for the visualization of the intracranial vasculature. In the last few years 3-T high-field MRI became clinically available. The main advantage of higher field strength for both non-enhanced MRA and CE MRA is approximately doubled signal-to-noise ratio (SNR). Using the higher SNR the spatial and/or temporal resolution may be increased without reducing the image contrast. The longer T1 relaxation times further make suppression of background tissue more efficient, thus improving the image contrast. However, there are also important disadvantages, such as more rapid saturation of slowly flowing blood, increased RF-energy deposition with SAR limitations, and stronger susceptibility effects (Willinek et al. 2003a; Nael et al. 2006a; Bachmann et al. 2006; Nael et al. 2006b).

in Western countries, extensive studies have been conducted to identify those patients who benefit from therapy of carotid artery disease. Two independent studies, the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and the European Carotid Surgery Trial (ESCT) could prove that patients with a high-grade stenosis of the internal carotid artery benefit from surgery (Barnett et al. 1998; Barer 1998). The threshold for a high-grade significant stenosis has been set at 70% diameter lumen narrowing. Although DSA is still considered the standard of reference for the assessment of atherosclerotic carotid artery disease, DSA has the substantial drawback of being a projection method of the vascular lumen. In case of eccentric disease, the measurement of percent stenosis may vary by projection which may substantially reduce the overall accuracy. Misclassification of the degree of a stenosis may result in unnecessary surgery. Since surgery is also associated with an estimated 1% incidence of disabling stroke or death, it is paramount that any diagnostic technique is accurate for differentiating surgical and non-surgical disease. Due to the limitations of DSA, MRA has emerged as one of the clinical routine techniques for a non-invasive assessment of atherosclerosis of the carotid circulation. Initially MRA was implemented using two-dimensional (2D) and three-dimensional (3D) TOF-techniques with good results (Modaresi et al. 1999; Patel et al. 1995; Polak et al. 1992, 1993; Heiserman et al. 1992). However, due to the long acquisition times of TOF MRA motion artifacts sometimes cause non-diagnostic scans. Furthermore, artifactual loss of signal intensity due to flow incoherence or saturation can be observed in vessel stenosis and vessel segments running parallel to the imaging plane, respectively (Heiserman et al. 1992; Litt et al. 1991). Due to the effect of flow saturation, the anatomic coverage of TOF MRA is limited. Thus, the aortic arch and circle of Willis are typically not included in the same acquisition as the carotid bulb. On the other hand a significant number of strokes are caused by atherosclerotic disease outside the carotid bifurcation. Furthermore, surgery may be contraindicated in the presence of significant tandem lesions in the proximal or distal carotid circulation. Therefore, it is desirable to image the carotid circulation from the aortic arch through the circle of Willis, which is difficult to accomplish with TOF MRA (Wetzel and Bongartz 1999). In the late 1990s 3D CE MRA emerged as the routine 5.4.3.3  Clinical Applications technique for the assessment of the carotids. Using highresolution CE MRA the entire carotid circulation includ5.4.3.3.1  MRA of the Supra-Aortic Vasculature: ing the aortic arch and circle of Willis can be assessed Carotid Artery Disease for atherosclerotic disease in a single acquisition. Several The most common indication for imaging of the supra- studies have compared the accuracy of TOF MRA and aortic vessels is the evaluation of atherosclerotic carotid CE MRA for the detection and grading of carotid artery artery disease. Carotid artery disease has been identi- stenosis. While some of these studies did not find signifified as a major risk and etiologic factor for 80% of all cant differences of the accuracies of TOF MRA and CE ischemic brain insults. Since stroke related to ischemia MRA (Johnson et al. 2000; Nederkoorn et al. 2003a; Fellfrom carotid emboli is one of the leading causes of death ner et al. 2005), especially more recent studies showed

5.4  MR Angiography

significant advantages for CE MRA providing better image quality, higher level of diagnostic confidence, and more interobserver agreement (Mitra et al. 2006; Willig et al. 1998; Anzalone et al. 2005). While initially TOF MRA offered a higher in-plane spatial resolution and was less frequently hampered by venous contamination, CE MRA now features both a higher spatial resolution and a shorter acquisition time. Moreover, the diagnostic window without venous contamination has been expanded by elliptical-centric kspace acquisition. With a spatial resolution of ~ 1 mm3 CE MRA can now be considered as the clinical gold standard for the evaluation of carotid artery disease. The availability of isotropic 3D MRA data further allows for

a cross-sectional analysis of stenotic lesion, which increases the accuracy and decreases interobserver variability (Fig. 5.4.19) (Schönberg et al. 1999a). The role of CE MRA as the new clinical gold standard for the evaluation of carotid artery disease is also supported by a recent meta-analysis, which demonstrated CE MRA to have a higher accuracy for the grading of relevant carotid artery stenosis than do DUS or CTA (Wardlaw et al. 2006). This meta-analysis also confirmed the superiority of CE MRA in comparison to TOF. For 70–99% carotid stenosis the sensitivity and specificity of CE MRA was 94 and 93%, compared with 88 and 84% for TOF MRA (Wardlaw et al. 2006). The improved spatial resolution achieved with 3 T and PI MRA is likely to further improve the accuracy of MRA for the assessment of carotid artery disease (Nael et al. 2006a; Bachmann et al. 2006). However, large clinical trials are still needed to prove this hypothesis. 5.4.3.3.2  Cervical Artery Dissection

Fig. 5.4.19  a Coronal MIP of a high-resolution 3D CE MRA of a patient with a proximal ICA stenosis of moderate degree (arrow). c–e For the cross-sectional analysis of the stenosis MPRs orientated orthogonal to the course of the vessel are reconstructed of the post-stenotic ICA segment (b,c) and in the stenosis (d,e)

Cervical artery dissections are assumed to cause up to 20% of ischemic strokes in young adults. A developing thrombus within the arterial wall may cause stenosis or occlusion of the artery, act as a source of emboli, or lead to a pseudoaneurysm (Bogousslavsky et al. 1987). DSA has traditionally been considered the standard of reference for the assessment of cervical artery dissection, but non-invasive cross-sectional imaging techniques such as MRA are now more frequently used in clinical practice (Bachmann et al. 2006; Levy et al. 1994; Leclerc et al. 1999). Two classical appearances of cervical artery dissection are notable on both DSA and MRA (Klufas et al. 1995). The first is an eccentric tapering stenosis or occlusion, and the second a so-called “pearl-sign” irregularity (Summers et al. 2001). MRI can further be used to visualize intramural hematoma of dissection. This can be done using fat-saturated T1-weighted spin-echo MRI, or even better ECG-triggered double-inversion recovery T1weighted turbo spin-echo MRI. Apart from an increase in the external diameter of the artery, depending on the stage, the hematoma appears isointense to the surrounding vessel wall in the very early stage (usually up to 2 days). Thus, the thickening of the vessel and narrowing of the lumen might be the only signs of dissection. In the subacute and early chronic stage the T1 shortening effect of methemoglobin causes a high signal intensity of the hematoma. Hematomas of dissections older than 2 months appear again isointense (Kitanaka et al. 1994). Depending on the location, the sensitivity of MRA for the detection of cervical artery dissection ranges between 20 and 95% (Levy et al. 1994; Keller et al. 1997; Auer et al. 1998). Due to the smaller vessel diameter and large anatomic variations MRA has been found to be less sensitive for the diagnosis of vertebral artery dissection (Levy et

811

812

5  Thorax and Vasculature Fig. 5.4.20  a Coronal MIP of a highresolution 3D CE MRA of a patient with a middle-grade, proximal left subclavian artery stenosis (arrow). b Time-resolved MRA of the same patient shows retrograde blood flow in the left vertebral artery (subclavian steal). c The retrograde blood flow in the left vertebral artery is confirmed by PC MRI flow measurements (arrow)

al. 1994). Using a higher spatial resolution at 3 T, the ac- ganglioma may present with a mass, a bruit, or a cranial curacy of MRA for the visualization of dissection is likely nerve deficit. MRA is able to visualize the feeding arteries, to improve. which is helpful in the planning of preoperative embolotherapy (van den Berg et al. 2000). Furthermore, MRI is able to further characterize the tumor (Michaely et al. 2007 van den Berg 2000; Vogl et al. 1994; van den Berg et 5.4.3.3.3  Subclavian Steal Syndrome al. 2004a). The classic appearance of paragangliomas at Subclavian steal syndrome is characterized by retrograde conventional MRI is the so-called salt-and-pepper patflow in the vertebral artery secondary to a stenosis of the tern, representing areas of high blood flow. On 3D TOF subclavian artery, proximal to the origin of the vertebral MRA these areas of high flow can be seen as areas of high arteries. Most often the etiology of the subclavian artery signal intensity within the tumor. 3D TOF MRA is more stenosis is atherosclerotic. Clinically, patients with sub- susceptible in showing these high-flow areas within the clavian steal syndrome may present with vertigo, ataxia, tumor than are conventional spin-echo MR sequences and visual disturbances (e.g., diplopia). Brainstem isch- (van den Berg et al. 2004b). Time-resolved CE MRA has emia is a complication of subclavian steal syndrome. MR been used to study the tumor hemodynamics of paraangiography allows for a non-invasive investigation of gangliomas. Paragangliomas typically show a rapid and subclavian steal syndrome (Van Grimberge et al. 2000; intense homogeneous enhancement following the intraSheehy et al. 2005). venous administration of contrast material (Fig. 5.4.21). A reversal of flow in the vertebral artery can be dem- These features can be established in tumors as small as onstrated by PC MRA or by the visualization of a signal 10 mm in diameter (Arnold et al. 2003). In a more revoid of the vertebral artery on 2D TOF MRA (Sheehy et cent publication, using a time-resolved MRA technique al. 2005). Time-resolved MRA can be used for the visual- combining PI and view-sharing (TREAT) tumor delineaization of the pathologic blood flow pattern (Fig. 5.4.20) tion with TREAT revealed to be superior to conventional (Nael et al. 2006). MRA agrees well with DSA for the grad- DSA. Moreover, using CNR TREAT allowed for a difing of subclavian artery stenosis (Randoux et al. 2003). ferentiation of glomus tumors from other head and neck tumors (Michaely et al. 2007). 5.4.3.3.4  Paragangliomas

5.4.3.3.5  Giant Cell Arteritis Paragangliomas are highly vascular tumors originating from paraganglionic tissue located at the carotid bulb, Giant cell arteritis, which is also known as temporal aralong the vagus nerve, in the jugular fossa, and tympanic teritis or Horton’s disease, is a granulomatous vasculitis cavity (van den Berg 2005). Patients with a cervical para- involving the large- and medium-sized arteries. Clini-

5.4  MR Angiography Fig. 5.4.21  a–d Coronal, sagittal and transverse MIP of a high-resolution 3D CE MRA of a patient with paraganglioma of the carotid bulb (arrow). e Time-resolved MRA of the same patient shows rapid tumor enhancement

cally, temporal headaches, localized scalp tenderness, jaw claudications, and blindness are the most frequent symptoms. High-resolution CE MRI can be effectively used for a non invasive diagnosis of giant cell arteritis. Thickening of the vessel wall and increased enhancement has been found to be a surrogate for mural inflammation (Bley et al. 2005a; Markl et al. 2006). Typically, the superficial cranial arteries with predominance of the superficial temporal artery are affected by the disease. However, giant cell arteritis can also involve other parts of the vascular system. Therefore, recent studies have proposed an MRI examination with extended coverage, including head, neck, and thorax (Markl et al. 2006; Bley et al. 2005b).

pected stroke (Scellinger et al. 2000; Derex et al. 2002). In addition to DWI a dedicated stroke MRI protocol will also include an MRA of the circle of Willis in order to identify the localization and extent of the causative arterial occlusion (Ozsarlak et al. 2004; Yang et al. 2002). The findings of MRA are relevant for the prognosis, as it has been shown that there is a strong relationship between the frequency of persistent neurological deficit and the completeness of the circle of Willis (Derex et al. 2002). To improve the ability to differentiate slow flow from vascular occlusion and to improve the visualization of small peripheral arteries TOF MRA should be acquired after contrast injection (Yang et al. 2002). MRA can also be used in the setting of screening for atherosclerotic disease of the intracranial arteries, where it competes with the other non-invasive imaging meth5.4.3.4  MRA of the Intracranial Vasculature ods, transcranial DUS and CTA. However, there are potential advantages of MRA: Compared with transcranial 5.4.3.4.1  MRA in Stroke DUS, MRA allows a non-restricted assessment of the enDiffusion-weighted MRI (DWI) is much more sensitive tire intracranial circulation and is less dependent on the than is CT in the detection of brain ischemia especially individual experience of the examiner (Baumgartner et in the first 24 h of stroke. Therefore, MRI has been pro- al. 1995). In contrast to CTA, MRA has the substantial posed as the first diagnostic test in patients with sus- advantage of only showing the vessel lumen without ar-

813

814

5  Thorax and Vasculature

tifacts from calcified plaque or adjacent bone structures. Moreover, for the purpose of a screening method, MRA is preferable due to the lack of ionizing radiation (Uehara et al. 2001). For TOF MRA a sensitivity and accuracy of almost 90% for the detection of intracranial stenosis is reported. Moreover, a high negative-predictive value and high concordance with conventional DSA has been shown (Oelerich et al. 1998; Dagirmanjian et al. 1995). A stenosis greater than 50% of the intracranial ICA is usually considered as an indication for anticoagulation. Due to the sensitivity to turbulent flow, TOF MRA may be inaccurate for the assessment of the exact vessel lumen and may overestimate the degree of stenosis. In these cases, CE MRA may be useful as an additional exam (Gottschalk et al. 2002). In patients with internal carotid artery stenosis MRA can also be used to visualize collateral blood flow from the contralateral side over the anterior or posterior communicating arteries (Hoksbergen et al. 2003). It has been shown that patients with collateral blood supply have a lower risk of transient ischemic attack and stroke. Furthermore, patients undergoing carotid endarterectomy who have collaterals supplying the operative side are less likely to develop a perioperative stroke (Henderson et al. 2000). For this purpose time-resolved MRA is a valuable tool. 5.4.3.4.2  Intracranial Aneurysms In the acute setting of subarachnoid hemorrhage DSA is still considered the gold standard for the detection of intracranial aneurysms. However, due to the limited number of projections false-negative rates of up to 5–10% are reported. MRA has been successfully used for detection of brain aneurysms, both in the acute and non-acute setting (Sankhla et al. 1996; Evans et al. 2005). Using 3D TOF MRA, aneurysms as small as 2 mm can be detected with a sensitivity of 74–98% (Bosmans et al. 1995; White et al. 2003; Unlu et al. 2005). Using the higher SNR of 3 T the spatial resolution of MRA can be increased, potentially further improving the accuracy of MRA for the detection and characterization of intracranial aneurysms (Gibbs et al. 2004). Endovascular embolotherapy has revolutionized the treatment of intracranial aneurysms, and 30–60% of all intracranial aneurysms in the United States and Europe are currently treated with an endovascular approach (Farb et al. 2005). The role of imaging in the treatment planning of these patients is to define the parent vessel and the aneurysm configuration including the size of the neck of saccular aneurysms (Adams et al. 2000). Slow and turbulent flow in the aneurysm may reduce the accuracy of TOF MRA for the assessment of the exact aneurysm dimensions, especially in a giant aneurysm. Additional limitations of TOF MRA are the potential to

misinterpret intra-aneurysmal thrombus or peri-aneurysmal hemorrhage as intraluminal blood flow (Brugieres et al. 1998; Kahara et al. 1999). Therefore, CE MRA has been proposed as an alternative to TOF MRA and has shown convincing results in animal studies (Krings et al. 2002). The only limitation of CE MRA is that aneurysms of the cavernous sinus may be rapidly masked by early venous return. Several studies have also indicated the potential role of MRA in the follow-up of cerebral aneurysms. For 3D TOF MRA sensitivity rates ranging from 71 to 100% and specificity rates ranging from 89 to 100% in ruling out residual flow in embolized aneurysms have been reported (Ozsarlak et al. 2004; Pierot et al. 2006). Again, falsenegative results can be explained by the low sensitivity of TOF MRA to slow flow, whereas false-positive results are caused by the high signal intensity from blood clot within the coil mass (Ozsarlak et al. 2004; Kahara et al. 1999; Gauvrit et al. 2006). Also, susceptibility artifacts from the coil material may reduce the accuracy of MRA (Ozsarlak et al. 2004). The patency of the parent vessel can be visualized by TOF MRA with a sensitivity of 90% (Gonner et al. 1998; Derdeyn et al. 1997). To overcome the limitations of TOF MRA recent publications have proposed CE MRA for the follow-up of embolized aneurysms (Fig. 5.4.22) (Gottschalk et al. 2002; Farb et al. 2005; Gauvrit et al. 2006). 5.4.3.4.3  Arteriovenous Malformations An arteriovenous malformation (AVM) of the brain is defined as an anastomotic network of blood vessels in which arteriovenous shunting occurs in a central nidus. Clinically an AVM is an important cause of intracranial hemorrhage, e.g., accounting for 9% of all subarachnoid hemorrhages (Ozsarlak et al. 2004). Treatment options include surgery, embolization, and radiation therapy, or a combination of these methods (Nagaraja et al. 2005). For an appropriate treatment, however, detailed information on the AVM angioarchitecture and hemodynamics is required (Duran et al. 2002). Imaging typically shows several tortuous feeding arteries of different sizes that converge toward the nidus where the arteriovenous shunting occurs (Ozsarlak et al. 2004). These feeding arteries typically originate from more than one intracranial branch of the internal carotid and/or vertebrobasilar systems. In addition imaging may demonstrate aneurysms associated with the AVM in approximately 10% of all patients (Westphal and Grzyska 2000). Several MRA techniques have been applied in patients with AVM (Ozsarlak et al. 2004, 151, 154). TOF MRA, PC MRA, and CE MRA can generate high-spatialresolution angiograms of the AVM, allowing for the assessment of the nidus size and identification of vascular feeders and venous drainage. In contrast to conventional DSA, however, conventional 3D MRA techniques have

5.4  MR Angiography Fig. 5.4.22a–c  Coronal and sagittal MIP of a high-resolution 3D CE MRA with elliptical-centric k-space acquisition of a patient with an aneurysm of the Ramus communicans anterior (arrows)

been unable to provide any hemodynamic information of the AVM. It has been shown, that the hemodynamics of AVMs are important in defining the risk for hemorrhage and for response evaluation after radiotherapy (Ozsarlak et al. 2004; Essig et al. 1996, 1999). Therefore, initially time-resolved 2D MRA techniques such as spin tagging or projection MRA have been used to acquire dynamic MRA data for the hemodynamic assessment of cerebral AVMs (Evans et al. 2005; Essig et al. 1996, 1999; Griffiths et al. 2000; Strecker et al. 2000). More, recently also time-resolved 3D CE MRA has been proposed for a combined morphologic and functional assessment of AVMs (Evans et al. 2005; Duran et al. 2002). While initially time-resolved MRA was inferior to nonenhanced spin-tagging MRA for the delineation of the nidus, it proved to be superior for the visualization of the venous drainage (Duran et al. 2002). Just by using new kspace sampling techniques for MRA (i.e., key-hole MRA, PI, view-sharing) 3D CE MRA can be acquired with a sufficient temporal to visualize the hemodynamics of cerebral AVM (Fig. 5.4.23) (Tsuchiya et al. 2004).

MRA has also been proposed for the follow-up of patients with treated AVM. Following radiation therapy complete obliteration of the AVM may occur as early as 4 months or as late as 5 years after treatment, often developing slowly and progressively (Ozsarlak et al. 2004). Because the risk of bleeding persists as long as complete obliteration is not obtained, patients have to be evaluated by close follow-up exams. As a non-invasive and radiation-free method, MRA is favorable for this task; however, due to the still superior spatial resolution DSA might still be required for a final assessment of residual nidus, when the AVM has disappeared on MRA. 5.4.3.4.4  Cerebral Venous Sinus Thrombosis Cerebral venous sinus thrombosis is responsible for 1–2% of strokes in adults; however, the true incidence is unknown (Renowden 2004). Most commonly the superior sagittal sinus, transverse sinuses, and sigmoid sinuses are affected. In approximately a third of all cases, more than

815

816

5  Thorax and Vasculature

Fig. 5.4.23  a Time-resolved 3D CE MRA (one in every four images) of a patient with an arteriovenous malformation of the left frontal lobe. Using parallel imaging and view-sharing a tem-

poral resolution of 500 ms is achieved. b High-resolution TOF MRA of the same patient acquired at 3 T showing arterial feeders mainly from the left MCA and large draining cortical veins

one sinus is involved, and in 30–40% of all cases, cerebellar and cortical vein thrombosis is associated (Renowden 2004). Headache is often the first and most frequent symptom occurring in 74–90% of all patients (Renowden 2004; de Bruijn et al. 1996). Cerebral venous sinus thrombosis should therefore be considered in the differential diagnosis in all patients with unusual or severe headache. Although CT usually will be the first imaging method performed in patients with cerebral venous sinus thrombosis, it may be negative in up to 20% of all cases. However, better results may be obtained using CE CT venography (Renowden 2004; Ozsvath et al. 1997). Besides conventional T1-weighted, T2-weighted, and T2*-weighted MRI, both, non-enhanced (i.e., TOF MRA and PC MRA) and CE MRA have been used for the visualization of the intracranial venous system. Similar to other diseases, TOF MRA has the limitation of strong in-plane saturation and intra-voxel dephasing, potentially leading to a false-positive diagnosis of cerebral venous sinus thrombosis. Moreover, the presence of blood breakdown products in the thrombus itself can create a high signal, mimicking flow and leading to a false-negative test result. Similar to the arterial system, limitations arising from flow artifacts may be reduced by using dedicated pulse sequences (i.e., MOTSA, TONE) and by the injection of MR contrast media (Renowden 2004). Compared to TOF MRA, PC MRA has the advantage that slow blood flow can also be reliably visualized (Ozsvath et al. 1997). On the other hand, it has the disadvantage of a longer acquisition time, aliasing artifacts, and

intra-voxel phase dispersion due to turbulent flow (e.g., in the transverse sinuses at the entry point of the vein of Labbe). Several studies have indicated that CE MRA may be more reliable than non-enhanced MRA in assessing the intracranial venous system (Kirchhof et al. 200; Farb et al. 2003; Mermuys et al. 2005; Wetzel et al. 2003; Lovblad et al. 2002). The major advantage of CE MRA is the much shorter imaging time and complete coverage of the entire intracranial venous system. CE MRA also allows a better detection of partial obstruction of the sinus. In addition, CE MRA can also reliably demonstrate increased venous collateralization (Lovblad et al. 2002). Several pitfalls exist in the interpretation of MRA of the cerebral venous system. Hypoplastic or aplastic sinus, which may be observed in up to 31% of normal individuals, may mimic sinus occlusion, especially in MIP projections. Therefore, source images have to be carefully analyzed in each case. Small filling defects may also result from fat, heterotopic brain, fibrous bands, and septae and arachnoid granulations. In these cases, additional conventional T1-weighted and T2-weighted MRI may improve the accuracy (Renowden 2004). 5.4.3.4.5  Other Pathologies Central nervous system (CNS) vasculitis represents a heterogeneous group of diseases that primarily affect the small leptomeningeal or parenchymal blood vessels of the brain.

5.4  MR Angiography

Frequently CNS vasculitis occurs in the context of a systemic vasculitis, systemic disease, infection (e.g., postvaricella angiopathy), or post-radiation vasculopathy. CNS vasculitis must be distinguished from noninflammatory vasculopathies, including dissection, Moyamoya disease, sickle cell disease, migraines with vasospasm, and rare metabolic vasculopathies (Ozsarlak et al. 2004; Aviv et al. 2006). MRI is highly sensitive in showing secondary manifestations of CNS vasculitis, such as cortical and subcortical infarctions; however, these changes are not specific for vasculitis (169). MRA may reveal multiple bilateral segmental stenoses and occlusion of the intracranial arteries; however, normal MRA studies may be observed in patients with proven vasculitis. Furthermore, the correlation between MRI and angiography often is only moderate (Ozsarlak et al. 2004; Aviv et al. 2006). Moyamoya disease is a rare progressive disease characterized by bilateral steno-occlusive changes of the supraclinoidal internal carotid arteries. The disease is further characterized by abnormal netlike collateral vessels (Moyamoya vessels) around the obstructed major arteries (Suzuki and Kodama 1983). The major symptoms of Moyamoya disease are related to age, with transient ischemic attack often seen in pediatric patients and intracranial hemorrhage often seen in adults. With MRI and MRA the diagnosis of Moyamoya disease can be established non-invasively (Yamada et al. 1995, 2001; Hasuo et al. 1998). When internal carotid artery occlusion and Moyamoya vessels are demonstrated at MR angiography, conventional DSA is unnecessary, particularly in pediatric patients (Fukui 1997; Fushimi et al. 2006).

cular coiling. For the follow-up of treated aneurysms, CE MRA is the modality of choice. MRA can further replace DSA for the diagnosis of cerebral AVMs. Time-resolved CE MRA with most recent k-space acquisition techniques (such as PI, view sharing) can provide a comprehensive morphologic and functional evaluation of cerebral AVMs. In the follow-up of radiation therapy, due to the still superior spatial resolution DSA might still be required for a final assessment of residual nidus, when the AVM has disappeared on MRA. Similarly, DSA still cannot be replaced in the clinical evaluation of cerebral vasculitis. However, MRI and MRA should be considered as the second-line imaging method, if DSA is negative. 5.4.4  MRA of the Renal Arteries S.O. Schönberg and H.J. Michaely 5.4.4.1  Introduction

The renal arteries can be affected by a variety of diseases ranging from atherosclerotic occlusive diseases such as renal artery stenosis, dysplastic changes of the renal artery wall, aneurysmal disease to vasculitis. Particularly, diagnostic evaluation of the renal arteries plays an essential role for the comprehensive clinical assessment of various systemic diseases such as hypertension. Recently, many efforts have been made to incorporate the exam of the renal arteries into an even more systemic assessment of disease by using newly arising techniques including whole-body CE MRA to comprehensively detect all types of systemic vascular disease manifestations (Schönberg et al. 2003a). In view of the relatively low prevalence of iso5.4.3.4.6  Remaining Role of DSA lated renal artery disease modern diagnostic algorithms Although still considered the gold standard, diagnostic for imaging of the renal arteries move to a more disease DSA has been largely replaced in clinical practice by non- specific type of imaging for both the underlying cause of invasive cross-sectional imaging techniques, including the disease as well as the potential sequelae. For example, MRA. assessment of hypertension by cross-sectional imaging One important example is the assessment of carotid requires identifying a potential renal artery stenosis to artery disease. According to recent studies, high-resolu- differentiate primary from secondary hypertension and tion CE MRA is considered as the method of choice for also providing evidence of potential sequelae from essenthe clinical assessment of carotid stenosis and can thus tial hypertension such as impaired renal perfusion from replace diagnostic DSA in clinical routine (Wardlaw et al. nephrosclerosis or cardiac hypertrophy. 2006; Nederkoorn et al. 2003). In the assessment of head and neck tumors a combination of static high-resolution CE MRA and time-resolved 5.4.4.2  Anatomy MRA is sufficient for the diagnostic work-up (Michaely et al. 2007). For this indication, DSA is only reserved for In approximately 75% of individuals, the renal arteries preoperative embolotherapy and the detection of small arise out of the aorta immediately below the superior arterial feeders as part of a diagnostic DSA during the in- mesenteric artery at the level of the L1–L2 intervertebral tervention. For the assessment of intracranial aneurysms disk space, while in the remainder they may originate anydiagnostic DSA may only be required for aneurysms where between the lower margins of T12 and L2. Coronal smaller than 2 mm, such as in mycotic aneurysmal dis- 3D CE MRA is particularly helpful to identify aberrant ease or panarteritis nodosa. Apart from this, the major renal arteries, since a large anatomic area is covered in role of DSA is the guidance of interventional endovas- the superior to inferior direction. The right renal artery

817

818

5  Thorax and Vasculature

passes behind the inferior vena cava, renal vein, duodenum, and pancreatic head. The left renal artery also lies behind the pancreatic body and renal and splenic veins. In around 60% of individuals the renal artery divides at the renal hilus, whereas in approximately 15% of cases an early bifurcation is present. In approximately 70% of individuals there is a single renal artery, and in the remainder there are multiple renal arteries. The literature differentiates between accessory renal arteries and aberrant arteries. Accessory renal arteries are supplementary vessels that enter the kidney independently at the hilus. Aberrant renal arteries enter the kidney outside the hilus. Accessory and aberrant renal arteries can originate from virtually any abdominal branch of the aorta. Other than from the aorta, accessory vessels most frequently originate from the common iliac arteries but can also originate from the superior and inferior mesentery, intercostal, lumbar and renal, inferior phrenic, right hepatic, or right colic arteries. It is important to remember that anomalous origins of renal arteries are commonly seen in patients with ectopic or horseshoe kidneys. Good results have been reported for the evaluation of kidney donors prior to transplantation in regard to the correct identification of the absolute number and location of supernummery vessels (Winterer et al. 2000). However, a recent multi-center trial reported complete agreement between MRA and DSA for the number of accessory renal arteries in only 82% of the cases (Schönberg et al. 2002). 5.4.4.3  Renal Artery Disease 5.4.4.3.1  Atherosclerotic Renal Artery Stenosis The total prevalence of renal artery stenosis (RAS) is around 4.3% in autopsy studies. However, in combination with other diseases such as diabetes mellitus and hypertension it occurs in up to 10.1% (Sawicki et al. 1991). RAS is found in up to 45% of patients with peripheral vascular disease (Missouris et al. 1994). Atherosclerotic RAS can be found in 90% of patients with RAS, fibromuscular dysplasia (FMD) accounts for the remaining 10% of cases (Textor 2000; Safian and Textor 2001; Olin 2004; Slovut and Olin 2004). If not detected and treated correctly it leads to ischemic nephropathy and end-stage renal disease (ESRD) (Safian 2001). RAS is estimated to account for 10–40% of ESRD (Scoble and Hamilton 1990) in patients without identifiable primary renal disease. Morphologic Grading of Stenosis A large number of studies have reported high accuracies for 3D CE MRA for the grading of renal artery stenosis with sensitivities and specificities of over 90%, which has been recently re-confirmed by two large meta-analyses. These studies also demonstrated the superiority of

3D CE MRA compared with non contrast-enhanced TOF techni­ques (Vasbinder et al. 2001; Tan et al. 2002). On the contrary, the Dutch multi-center RADISH trial reported only sensitivities and specificities of less than 80% for both 3D CE MRA and CTA as compared with DSA. There are several reasons for this discrepancy from earlier results. The average voxel size of the 3D CE MRA data sets was about 3–6 mm3; 3D data sets with submillimeter spatial resolution were not acquired in this study (Vasbinder et al. 2004). In addition, 3D CE MRA was not completed by phase-contrast flow measure­ments for a combined morphologic and functional assess­ment of the degree of renal artery stenosis. Data from a multi-center trial on the combined assessment of MRA and P -flow measurements in comparison to DSA revealed higher accuracies for detection of a high-grade stenosis exceeding 95% (Schönberg et al. 2002). Recent data from preliminary studies supports the fact that higher spatial resolution might also increase the accuracy of stenosis grading by 3D CE MRA. In particular, assessing the reduction of the cross-sectional vessel diameter on orthogonal cuts of the isotropic data sets appears to effectively reduce misinterpretations (Schönberg et al. 2005) (Fig. 5.4.24). Due to the still limited spatial resolution of 3D CE MRA, no larger data on the accuracy of stenosis grading of small accessory renal arteries are available. However, the significance of accessory renal artery stenosis remains controversial. Comprehensive Stenosis Grading on MRI The presence of a significant renal artery stenosis can be confirmed by adding further MR criteria to the diagnostic evaluation besides MRA. The most practical approach is to assess kidney size, cortical thickness and renal parenchymal enhancement as well as post-stenotic dilatation on the 3D CE data sets (Prince et al. 1997). The addition of PC flow measurements to the standard renal protocol has been shown to provide several diagnostic benefits. First, PC flow measurements by themselves are already sensitive in identifying a high-grade stenosis as compared to DSA, with sensitivities exceeding 90% (Schönberg et al. 1997). Second, in combination with 3D CE MRA the overall diagnostic accuracy of MRI for grading of a stenosis increases (Schönberg et al. 2002). Third, the add-on of this functional imaging modality to the morphologic stenosis grading significantly reduces interobserver variability between multiple readers and thus increases the consistency of stenosis grading. In addition, the technique helps to identify those lesions with hemodynamic significance, i.e., where the autoregulatory capacity is exceeded and loss of renal function is likely to occur (Schönberg et al. 2000). In particular, the loss of the early systolic peak marks the onset of such a hemodynamically significant lesion. In addition, the assessment of the velocity flow profile helps to early detect re-stenosis. It is important to mention that these measurements

5.4  MR Angiography

Fig. 5.4.24a–c  Renal artery stenosis with impaired renal function. 49 year old female patient with arterial hypertension. In the volume-rendered image of the CE MRA at 1.5 T after injection of 15 ml gadobutrol (Gadovist®), a proximal high-grade stenosis of the left renal artery with post-stenotic dilatation is noted (a). The multiplanar reformats perpendicular to the vessel axis (b vessel lumen in the center of the cross-hair) reveal a 90% reduction of vessel area at the site of the stenosis compared to the normal vessel lumen. Note that the residual vessel lumen within the stenosis is still visible due to the high spatial resolution. The additionally performed MR perfusion measurements (c) demonstrate a normal transit of the contrast media in the first pass with an only slightly prolonged mean transit time in the left kidney. In the late phase of the perfusion measurements, however, a substantial loss of excretory function is identified with almost no contrast media in the left renal pelvis compared with normal excretion of the right kidney. Diagnosis: functionally significant left renal artery stenosis

can be also performed after stent placement while 3D CE MRA is often non-diagnostic due to signal loss at the stenosis site. Care needs to be taken to position the PC-flow measurement plane approximately 1 cm downstream of the stenosis site to avoid phase errors from turbulent flow. Based on the combination of PC flow measurements or 3D phase-contrast angiography, a modified grading scheme of renal artery stenosis can be advised that is less sensitive to errors in the numerical measurement of the morphologic degree of renal artery stenosis (see Table 5.4.10). Further characterization of the functional effects of renal artery stenosis can be carried out by means of perfusion measurements. Measurements of perfusion using intravascular contrast agents have been shown to successfully quantify cortical perfusion with good agreement with invasively determined data (Aumann et al. 2003). Mean cortical perfusion of normal kidney is

about 400 ml/100 g tissue/min. In cases of high-grade steno­sis exceeding 80% these values can drop to less than 200 ml/100 g tissue/min. For semiquantitative perfusion measurements with standard gadolinium chelates, significant differences between patients without stenoses or low-to-intermediate-grade stenoses and those with high-grade stenoses could be found for the mean transit time (MTT), the maximum upslope (MUS), and the time to peak (TTP) measurements in one study (Michaely et al. 2006a). In patients with renal artery stenosis MTT/TTP/MUS was 37.2 s/25.4 s/10.7, while the MTT/ TTP/MUS of healthy patients was 21.4 s/15.5 s/ 21.3 (Michaely et al. 2006a) (Fig. 5.4.24). Recent developments in post-processing also allow determining quanti­tative perfusion parameters of plasma and tubular flow from these measure­ments (Fig. 5.4.25). In addition, perfusion measurements are valuable to identify segmental perfusion defects resulting from distal renal artery stenosis in

819

820

5  Thorax and Vasculature Table 5.4.10  Modified grading scheme of renal artery stenosis based on comprehensive assessment of morphologic and hemodynamic changes Morphologic degree of stenosis

Findings on 3D gadolinium MR angiography

Findings on 2D MR phasecontrast flow measurements

Normal

Normal

Normal flow profile

Mild

Mild stenosis <50%

Partial loss of early systolic velocity peak

Moderate

Stenotic ≥50%

Complete loss of early systolic peak and beginning decrease of midsystolic velocity components

Severe

Stenosis >75%

Flattened flow profile with no distinct systolic velocity maximum

Fig. 5.4.25a,b  Ischemic nephropathy with loss of renal function 40-mm-thick MIP image of a high-resolution MRA acquired at 1.5 T in a 53-year-old patient with arterial hypertension. A high-grade filiform stenosis of the left renal artery is present which resulted in a small ischemic left kidney (a). Quantitative evaluation of the MR perfusion measurements reveals a decreased plasma flow of 192 ml/100 ml tissue/min in the left kidney compared with 237 ml/100 ml tissue/min in the right kidney (b). This difference in renal function is further enhanced if the excretory function is analyzed quantitatively based on the data from the same MR perfusion measurements. While the normal right kidney demonstrates a tubular flow of 12 ml/100 ml tissue/min, this value is reduced to 3 ml/100 ml tissue/min in the left kidney. Diagnosis: chronic ischemic nephropathy from long-standing high-grade renal artery stenosis

segmental arteries, which are frequently obscured by ve- and to correlate with cardiac risk factors (Gandy et al. nous overlay on the high-resolution CE MRA images or 2004). The importance of renal perfusion is underlined from stenoses of small accessory renal arteries in which by the fact that patients with impaired renal perfusion the exact degree of stenosis is difficult to determine due parameters have an increased morbidity and mortality. to limits in spatial resolution (Fig. 5.4.26). Renal perfu- Newer CE MRA techniques allow for time-resolved data sion parameters have been found in previous studies to acquisition with high spatial and temporal resolution, reflect changes in the renal function (Shariat et al. 1998) permitting the renal perfusion data and an MRA data set

5.4  MR Angiography Fig. 5.4.26a,b  Segmental renal perfusion defect due to accessory renal artery stenosis. High-resolution CE MRA of the renal artery acquired at 1.5 T in an 80-year-old male patient with arterial hypertension. There are marked atherosclerotic changes and elongation of the abdominal aorta present. On the right side, the kidney is supplied by two renal arteries of which the accessory renal artery supplies the lower pole and reveals a focal high-grade stenosis greater 90%, whereas the main renal artery demonstrates only a 50% stenosis (a). In the MR perfusion measurements (b) a normal contrast enhancement of the left kidney with regular transit of the contrast media bolus is seen. In the right kidney, the high-grade stenosis of the accessory renal artery causes a delayed and decreased perfusion of the lower pole. In comparison with Fig. 5.4.24, the overall contrast enhancement of the kidneys is decreased, consistent with chronic renal insufficiency. Diagnosis: chronic renal insufficiency with additional hemodynamically significant renal artery stenosis

with good spatial resolution (1.5 × 1.5 × 3 mm3) to be acquired at once (Michaely et al. 2006b). So far, 3D CE MRA has not pro­ven to have any higher impact in curing renal hypertension or renal insufficien­cy than ultrasound does. Due to the relatively high costs, aggressive search for renal artery stenosis by angiographic techniques in patients with hypertension is therefore controversial. Quantitative or semiquantitative perfusion measurements offer an independent measure of parenchymal blood flow in the renal cortex as well as the medulla (Aumann et al. 2003; Michaely et al. 2004b, 2006a), allowing renal function to be assessed independently in the presence of renal artery stenosis (Michaely et al. 2004b, 2006c; Schönberg et al. 2003b). Changes in these quantitative parameters have been found in patients with primary parenchymal disease in whom no underlying

renal artery stenosis is present. Thus, these techniques appear to be a promising additional factor to separate renovascular disease from parenchymal diseases. For several reasons, it appears highly important to accurately select those patients with a renal artery stenosis who may benefit from an intervention (Fig. 5.4.27). First, renal artery stenosis is an important independent factor for 5-year patient survival (Connolly et al. 1994). Second, renal interventions with stent placements are costly procedures. Third, in a certain percentage of patients renal function may also deteriorate after an intervention. Recently, the use of the resistive index (RI) in ultrasound has been shown to be a highly valuable parameter for the prediction of patient improvement after interventional balloon angioplasty of renal artery stenosis (Radermacher et al. 2001). However, other groups found contradictory

821

822

5  Thorax and Vasculature Fig. 5.4.27a–e  Improvement of renal function after percutaneous transluminal angioplasty (PTA). 59-year-old male patient with arterial hypertension. High-resolution CE MRA (a) acquired at 1.5 T after administration of 15 ml gadobutrol (Gadovist®) demonstrates a high-grade stenosis of the right renal artery proximal to an early bifurcation. The left renal artery is unremarkable. In the additionally performed MR perfusion measurements a markedly delayed arrival of the contrast media bolus is noted in the affected right kidney with retention of the contrast media in the late phase (b). After successful PTA of the stenosis with placement of a stent, the control CE MRA and MR perfusion exam shows normalization of the perfusion in the right kidney with bilaterally regular transit of the contrast media into the medulla (d). Diagnosis: Successful stenting of a high-grade renal artery stenosis with post-interventional improvement of renal function. e see next page

5.4  MR Angiography

ease activity. In case of suspected dissections as in Marfan disease the addition of time-resolved MRA sequences may be indicated. Time-resolved techniques overcome the problems associated with bolus timing when the true and false lumens reveal markedly different blood flow velocities. They also ease the detection of the origin of the renal arteries, which may be hard to identify on a singlephase CE MRA scan. 5.4.4.3.3  Fibromuscular Dysplasia

Fig. 5.4.27a–e  (continued) Improvement of renal function after percutaneous transluminal angioplasty (PTA). Quantitative assessment of MR perfusion demonstrates a substantial improvement of renal blood flow in the right kidney with a reduction of the mean transit time from 25 s pre-interventional to 19 s post-interventional. This can be also detected visually from the increased upslope of the signal intensity–time curves with a better delineation of the early perfusion peak (e). Diagnosis: Successful stenting of a high-grade renal artery stenosis with postinterventional improvement of renal function

results for the RI as a predictor of successful outcome (Zeller et al. 2001, 2002). The role for the prediction of outcome after inter­ven­tio­nal or operative revascularization has yet not been fully established for MRI. The potential to differentiate patients with predominant renovascular or renoparen­chymal disease based on a single comprehensive MR exam could represent an important adjunct to the current morphologic evaluation of renal artery stenosis by 3D CE MRA. Currently, comparative studies using both DUS and MRI prior to and after revascularization are ongoing. 5.4.4.3.2  Grading of Renal Artery Stenosis as Part of a Systemic Vascular Assessment In case of other systemic vascular diseases involving the renal arteries such as Takayasu’s arteritis, Marfan syndrome, or Ehlers-Danlos syndrome whole-body approaches that were initially developed for cardiovascular screening can be successfully applied. While these protocols mostly consisted of angiography sequences, here the application of additional sequences may be indicated. For inflammatory diseases, T2-weighted spin-echo sequences and fat-saturated T1-weighted gradient-echo sequences should be added. These sequences allow detecting vessel wall edema (T2 weighted), hemorrhage (T1 weighted), and vessel wall contrast-agent uptake (T1 weighted), three parameters that characterize the inflammatory dis-

Detection of fibromuscular dysplasia (FMD) is of high clinical importance for two reasons. First, these patients are usually young at initial presentation and may require high doses of antihypertensive medications to control blood pressure. Second, fibromuscular dysplasia responds particularly well to angioplasty with mean cure rates for blood pressure of at least 50% and mean improvement rates of more than 40%. Fibromuscular dysplasia can be subdivided into intimal fibroplasia, medial fibromuscular dysplasia and adventitial fibroplasia. Whereas the intimal type of FMD is predominately found in infants, the medial type is usually seen in younger women and makes up about 90% of all types of FMD. The etiology of FMD is unknown, but it is considered to be a developmental disease that may be progressive over time. Involvement of the distal main renal artery as well as the segmental renal arteries is common, whereas the proximal segments are frequently spared. So far, DSA has been the standard of reference for diagnostic evaluation of FMD since high spatial resolution combined with short acquisition times is mandatory in order to detect and grade the often subtle irregularities. Angiographically, a classic pattern of alternating web-like stenoses in conjunction with aneurysmal segments is found, accounting for the so-called string-of-beads appearance. Stenotic segments can be long or short in length. In approximately 40% of the cases the disease is bilateral. In addition, other vascular territories can be affected by FMD such as the carotid, vertebral, splanchnic, or iliac arteries. No larger study on the accuracy of 3D CE MRA for the assessment of fibromuscular dysplasia exists. In most MRA studies that reported high accuracies for stenosis grading, the prevalence of FMD was either low or cases with FMD were excluded from the evaluation. The most recent publication reports a sensitivity and specificity for the detection of FMD of 22 and 96% for MRA (Vasbinder et al. 2004). Therefore, right now the accuracy of 3D CE MRA for assessing FMD is not completely established but is expected to be less than for atherosclerotic stenoses. Future improvements in spatial resolution at 3 T promise a better depiction of the distal renal arteries (Fig. 5.4.28). Initial reports on the achievable resolution and image quality are promising (Michaely et al. 2005).

823

824

5  Thorax and Vasculature

is also quite common with the invasion of renal cell carcinoma into the renal vein; RVT with Wilms’ tumor is far less common. RVT also occurs in the posttraumatic state, with pancreatitis (Ma et al. 2002), secondary to renal vein compression by lymph nodes or abscesses or by continuous spread of gonadal vein thrombosis into the renal vein. RVT is common in patients with membranous glomerulonephritis, which occurs in 30% of patients suffering from systemic lupus erythematosus. The thrombus may extend into the inferior vena cava up to the right atrium. The classic diagnostic triad of a flank mass, gross hematuria, and thrombocytopenia is only present in 13% of all patients with RVT. Therefore, radiology is often required to establish the diagnosis. If ultrasound is not sufficient, MRA or CTA can be used to detect RVT. In the setting of decreased renal function, however, MRA with comprehensive morphologic and functional assessment is preferred (Fig. 5.4.29). Typical radiographic findings of RVT include swelling of the affected kidney in the acute phase and cirrhosis of the kidney with longstanding RVT. MR imaging reveals low signal intensity of the renal parenchyma in both T1weighted and T2-weighted sequences as well as a compression of the collecting system. The signal intensity loss of the renal cortex in T2-weighted sequences is present from the first day on. Especially in acute RVT, a lowsignal-intensity band at the outer part of the medulla can be seen, which is considered to represent hemorrhage secondary to the impaired blood drainage. This imaging finding resembles intrarenal changes seen in patients suffering from hemorrhagic fever with renal syndrome. Extensive venous collateral vessels as well as dilatation of the left gonadal vein may be seen in chronic RVT. The collaterals usually arise from the renal hilum around the proximal ureter and of the capsular vessels. The corticomedullary differentiation vanishes from the 15th day on T1-weighted images, and the kidney becomes atrophic after 1 month. Fig. 5.4.28a,b  Female patient with hypertension. High-spatialresolution MRA acquired at 3 T in a 52-year-old female patient with hypertension unresponsive to medical therapy (a). Due to the small voxel size of 0.9 × 0.9 × 0.9 mm, the typical string-ofbeads appearance in the distal segment of the right main renal artery is well visualized in the magnified view (b). Diagnosis: fibromuscular dysplasia

5.4.4.5  Transplantation 5.4.4.5.1  Preoperative Planning

Preoperative imaging of potential kidney donors is an increasing indication for renal imaging. Depending on the hospital, CTA or MRA may be preferred for this kind of study (Subramaniam et al. 2004). Studies comparing MRA with CTA showed comparable inter-reader agreement for both readers as well as similar performance of 5.4.4.4  Renal Vein Thrombosis these modalities compared to DSA for the detection of Renal vein thrombosis (RVT) most commonly affects accessory vessels. MRA has the advantage of avoiding the left renal vein, probably due to its longer course to radiation exposure and using non-nephrotoxic contrast the IVC. The most common reason for RVT is a hyper- agent, but the examination requires a longer period of coagulable state like dehydration (Jeong et al. 2002), ne- time and is more expensive than CTA. The imaging prophrotic syndrome, and clotting factor imbalances. RVT tocol should include an angiography, morphologic im-

5.4  MR Angiography Fig. 5.4.29a–c  Renal vein thrombosis. 27-year-old female patient with abdominal discomfort and increasing serum creatinine. In the initially performed MR perfusion measurements at 1.5 T after injection of 7 ml of gadobutrol (Gadovist®), a substantial difference of perfusion is found between the left and the right kidney. The left kidney reveals a minimum enhancement during the arterial phase of the contrast media transit and no enhancement of the medulla or contrast media excretion in the later phases. In the unenhanced mask acquisition of the CE MRA scan the right ureter is already well visualized due to excreted contrast media from the previous MR perfusion scan whereas no excretion is present on the left (b). In the venous phase of the CE MRA scan a completely thrombosed left renal vein is noted with a fresh, non-vascularized thrombus as the cause for the functional impairment (c). Diagnosis: renal vein thrombosis due to nephrotic syndrome

aging of the kidneys to be transplanted and also include a urographic phase to evaluate the proximal ureters for anatomic variants such as ureter duplex or ureter fissus. The MRA in the preoperative planning of renal transplant donors should cover the entire abdominal aorta from the diaphragm down to the external iliac arteries (Pozniak et al. 1998). Aberrant origins of renal arteries can hereby be detected. The origin and course of all renal arteries needs to be described in detail for operation planning. This is of importance since the operation time increases by more than 30% if more than two renal arteries are present. Different operation techniques such as inclusion of an aortic patch with multiple arteries may also be considered in these patients. If reporting on the renal arteries it is important to look for early bifurcation of the vessels. Early bifurcation is defined by a bifurcation of the renal artery into the segmental arteries within a distance of 1.5 cm or less from the aorta. These patients may be not suitable as renal donors. The renal veins also need to be described in detail. Multiple, duplicated retroaortic or bifurcated renal veins

as well as anomalies of the gonadal and lumbar vein anastomoses need to be mentioned. The distance of the orifice of the gonadal vein should be measured and reported. Intravascular contrast agents such as gadofosveset trisodium may be helpful for stronger opacification of the renal veins. The urographic phase of the examination focuses on the collecting system. Anomalies such as ectasia, duplication of the collecting system or malrotation of the kidney must be described. 5.4.4.5.2  Postoperative Surveillance and Symptomatic Assessment CT and MR are commonly used in the surveillance of renal transplants (Sebastia et al. 2001). While CT offers shorter examination times and allows easy detection of postoperative complications such as seroma, haematoma, and abscesses, MRI may be more suitable for the vascular assessment due to MR contrast agents’ lack of nephrotoxicity.

825

826

5  Thorax and Vasculature

The most common surgical approach for renal transplant vessels is an end-to-side anastomosis with the external or common iliac artery. In the case of multiple renal arteries an aortic patch is often used. The transplant renal vein is anastomosed end to side with the recipient’s external iliac vein. Normally, the donor’s left kidney is used since it provides a loner renal vein. If the right kidney has a sufficiently long renal vein it can be harvested as well. If the right renal vein is not long enough, a venous patch is used to compensate for the shorter right renal vein. Mainly because of its lack of nephrotoxicity, 3D CE MRA has become a standard tool for the evaluation of renal transplants. The ability to reformat MR 3D data sets in any desired way is another advantage for the assessment of the often tortuous transplant arteries. For the renal transplant MRA, a standard 3D CE MRA sequence can be utilized. Due to the pelvic position of the transplant organ and the decreased respiratory motion, longer scan times can be considered to achieve a higher spatial resolution. A prevalence of RAS between 1 and 25% has been reported occurring typically one month to two years post transplantation. While proximal RAS at the anastomosis site is mainly a consequence of the surgical procedure, distal RAS is considered to be a consequence of chronic rejection (Buturovic-Ponikvar 2003). Therefore, proximal RAS tends to occur in the immediate postoperative period. In order to detect possible surgery-induced intimal flaps causing RAS, high resolution is essential. MRA has been found to be a highly accurate modality for detection and grading of transplant RAS. Compared to DSA, sensitivities of 100% and specificities ranging from 93 to 97% have been found (Chan et al. 2001). Since the hemodynamic relevance is characterized by impaired blood flow to the organ, MR flow measurements may be indicated in ambiguous cases. However, they may be hard to position perpendicular to the vessel axis due to the often short and tortuous arterial segment. A possible solution to this dilemma is provided by MR perfusion measurements which can be positioned over the kidney itself and are not restricted by a short transplant renal artery. They allow displaying pathologic perfusion parameters in the affected kidney or even in an affected kidney segment (Michaely et al. 2004b). The combination of these perfusion measurements with CE MRA in one protocol also allows the differentiation between those perfusion defects induced by proximal or distal renal artery stenosis and those from chronic allograft rejection (Fig. 5.4.30). Apart from the transplant renal artery itself it is also necessary to include the iliac arteries in the field of view. In rare cases, stenoses of the iliac arteries can lead to transplant dysfunction as well. They may either occur as a result of vessel wall damage from intraoperative clamping while suturing the anastomosis or from chronic atherosclerotic disease. A mild narrowing at the anastomosis site can often be found, which presents suture-related narrowing without

Fig. 5.4.30a,b  Segmental perfusion impairment of renal transplant. 40-year-old female patient status 5 years post-renal transplantation, with rising serum creatinine. In the arterial phase of the MR perfusion measurement a normal contrast enhancement of the entire renal cortex is present, whereas in the later phases, accumulation of contrast media in the medulla as well as excretion into the renal pelvis takes only place in the upper pole of the kidney (a). The subsequently performed CE MRA reveals completely normal proximal and peripheral renal arterial vasculature (b), but confirms the findings from the perfusion measurements. Only the upper pole reveals contrast media excretion. These findings were also confirmed by renal scintigraphy which also demonstrated impaired excretory function of the lower pole. Diagnosis: segmental chronic allograft nephropathy, CAN (confirmed by biopsy)

hemodynamic relevance. When bladder augmentation techniques are used for surgery, the incidence of RAS is increased. RAS and kinking of the transplant renal artery may lead to impaired renal blood supply of the allograft, particularly in the early postoperative days (Wong-YouCheong et al. 1998). Renal allograft torsion with rotation of the vascular pedicle is mainly seen in children with intraperitoneal renal transplant. Rotation of the renal transplant can be detected by changes in graft axis orientation and vascular kinking. Immediate therapy is indicated in these cases to salvage the renal transplant.

5.4  MR Angiography

Renal artery thrombosis is a very rare and catastrophic early complication in renal transplants. Renal artery thrombosis presents as missing or delayed and reduced enhancement of the transplant renal artery. Extreme atherosclerotic changes in the aorta and iliac arteries should raise the index of suspicion for renal artery thrombosis due to emboli. In renal transplants, it is also particularly important to start the venous phase scan immediately after the arterial phase to enable evaluation of the renal vein for pathologies. Renal vein thrombosis in renal transplant patients may arise due to kinking of the renal vein as well as acute or chronic rejection (Yang et al. 1996). Imaging findings are the same as with native renal vein thrombosis. 5.4.4.6  Challenges for Renal MRA for Replacement of DSA 5.4.4.6.1  Controversies in Grading of Renal Artery Stenosis Pathology studies have shown that atherosclerosis does not occur concentrically in the wall of the vessel but is a non-uniform eccentric process leaving an irregularly shaped residual lumen of the remaining artery (Jeremias et al. 2000). Therefore, the only exact measure of the true reduction of vessel lumen is by cross-sections perpendicular to the long axis of the vessel. However, due to the short imaging time required to perform the entire 3D CE MRA scan within a single breath hold of less than 25 s, spatial resolution is usually not equal along all three scan orientations. In this case geometric distortions from non-isotropic voxel sizes result in a decreased accuracy of the vessel-area measurements. The use of isotropic voxel sizes, as shown in one study, enables a distortion-free reconstruction of the vessel area in any desired three-dimensional orientation. Good results could be demonstrated in correlation to invasive reference measurements by intravascular ultrasounds of the renal artery (Schönberg et al. 2005). Measurement of area stenosis is subject to a much lower interobserver variability than assessment of diameter stenosis (Schönberg et al. 2005). One of the most difficult problems to solve in 3D CE MRA is random motion of the distal renal arteries induced by involuntary contractions of the diaphragm, which cannot be suppressed by breathing suspension or ECG gating. One study therefore concluded that the accuracy for grading of distal renal artery stenosis is inherently limited (Vasbinder et al. 2002). Since these contractions occur somewhat periodically in the order of several s, one possibility would be to reduce the acquisition to less than 10 s for a three-dimensional data set and to use a time-resolved approach. This time-resolved approach has also proven to be superior in another study in terms

of vessel visibility in the distal main renal artery as well as the proximal segmental intrarenal arteries due to the absence of overlaying enhancing renal parenchyma (Schönberg et al. 1999b). The down side of this approach, however, is that in current acquisitions scan times in the order of 20 s are required to obtain the mandatory isotropic spatial resolution of 1 mm3 or less. One study has demonstrated the use of spiral echo-planar imaging (EPI) for accelerating the acquisitions in the x–y-plane for time-resolved 3D CE MRA of the renal arteries; nevertheless, the spatial resolution could not be increased to voxel lengths of less than 1.5 × 1.5 × 2 mm3 (Amann et al. 2002). 5.4.4.6.2  Optimized Implementations for Renal 3D CE MRA The use of parallel imaging (PI) with the GRAPPA algorithm with an acceleration factor of 2 for the first time resulted in isotropic voxel lengths of 0.9 mm and a scan time of 23 s. The typical artifacts of PI, in particular aliased signal being propagated into the center of the image, can be kept at a minimum (Griswold et al. 2002; Schönberg et al. 2005). With the recent introduction of multi-channel MRI systems, acceleration factors of 3 can be routinely used with no visible loss of image quality despite a numerically higher noise level (Michaely et al. 2006b). Acquisition time can be reduced to 16 s for a 3D CE MRA scan with a spatial resolution of 1 mm3. One recent study has found better image quality for scans with a threefold acceleration compared with standard twofold acceleration in terms of vessel visibility in particular in the distal segments of the renal vascular bed (Michaely et al. 2006b). The loss of signal to noise from the use of PI can be effectively counterbalanced to a certain degree by the use of contrast agents with higher relaxivities such as 1 M agents like gadobutrol (Gadovist) or weakly proteinbinding contrast agents such as gadobenate-dimeglumine (MultiHance), or strongly protein-binding agents such as gadofosveset trisodium (Vasovist). Both Vasovist and Gadovist are now approved for the use of 3D CE MRA of the renal arteries; the approval for MultiHance is expected in the near future. At 3 T, one preliminary study has shown the feasibility of a high-resolution 3D CE MRA scan with isotropic voxel sizes of 0.9 × 0.9 × 0.9 mm3 in a 16-s scan time (Michaely et al. 2005a). The authors could demonstrate an excellent vessel visibility up to the level of the segmental arteries with almost no artifacts from respiratory motion. Overall these improvements can be considered a sufficiently robust measure to eliminate the use of diagnostic DSA in clinical routine for evaluation of atherosclerotic RAS.

827

828

5  Thorax and Vasculature

5.4.4.6.3  Remaining Limitations for Renal 3D CE MRA In fibromuscular dysplasia, accurate detection of stenosis is challenging for several reasons. First, the dysplastic changes of the vessel wall with the typical string-of-beads appearance often cause multiple stenoses of varying grade. Second, the changes can appear very subtle when the vessel is looked at in a coronal view, leading to underestimation of the true reduction of lumen diameter. Therefore, it is very important to reconstruct the images perpendicular to the long axis of the vessel to truly see the multiple reductions of the vessel lumen and correctly identify the target site of the maximum degree of stenosis. Third, FMD typically involves the distal main artery as well as the intrarenal segments, which are not only smaller, and thus require higher spatial resolution to accurately assess stenotic changes, but also require shorter scan times to reduce the amount of overlaying enhancing renal parenchyma and random motion from diaphragmatic contractions.

Higher acceleration factors are theoretically possible when using PI with acceleration in two dimensions, i.e., an acceleration factor of 4 in the phase-encoding direction and an acceleration factor of 2 in the partition-encoding (3D) direction totaling up to an overall acceleration of R = 8. The problem of this PI acceleration in 2D for renal 3D CE MRA is that due to the relatively thin coronal slab positioned in the core of the human body, the sensitivity profiles of the different coil elements located anteriorly and posteriorly are not distinct enough to allow reliable PI in the partition-encoding direction Therefore, images with three- to fourfold acceleration at 3 T using voxel lengths of 0.8 or less and acquisition times of approximately 15 s are currently considered the maximum performance 3D CE MRA (Fig. 5.4.31). For the same reasons, the assessment of vascular disease primarily involving the intrarenal branches is still a domain of DSA including vasculitis such as polyarteritis nodosa. Time-resolved 3D CE MRA with short acquisition times reduces the amount of overlay from enhanced renal parenchyma, thus allowing a better visualization of

Fig. 5.4.31a–d  Advances for renal contrast-enhanced MRA at 3 T. Ultrahigh-resolution MRA acquired in 16 s with a PAT acceleration factor of 3 in the left-to-right phase-encoding direction and a measured temporal resolution of 0.8 × 0.8 × 0.8 mm (a). Time-resolved dynamic MRA of the renal arteries with high temporal resolution (1.8 s per time frame) and a spatial resolution of 1.3 × 1 × 4 mm (b). Patient status post renal transplant with presence of extensive kinking of the renal artery and a proximal high-grade renal artery stenosis (c). Due to the high spatial resolution the residual lumen at the site of the stenosis is still visualized in the magnified view

5.4  MR Angiography

the intrarenal branches, but demonstrates constraints in spatial resolution. With the introduction of TREAT (timeresolved echo-shared angiographic technique), frame rates in the order of 1.8 s are feasible while maintaining a spatial resolution of 1.3 × 1 × 4.3 mm3 (Nael et al. 2006b); however, no systematic studies have been published so far on the overall value of this technique for assessment of renal artery stenosis (Fig. 5.4.31). 5.4.5  Diseases of the Aorta D. Theisen 5.4.5.1  Introduction Diseases of the aorta can present with a broad clinical spectrum of symptoms and signs. Their prevalence appears to be increasing in western populations, most likely corresponding to aging and heightened clinical awareness but also because of the progress of high-resolution, non-invasive imaging modalities. Compared to other non-invasive imaging modalities, MRI combines highspatial-resolution imaging with unrivalled temporal resolution. During the last decade, it has become the preferred imaging modality in many cardiovascular centers for the evaluation of congenital anomalies and acquired diseases of the aorta.

Fig. 5.4.32  Postductal aortic coarctation; sagittal MIP reconstruction of 3D CE MRA shows collateralization via prominent intercostal arteries

gered double-inversion recovery dark-blood TSE images in oblique and oblique-sagittal planes provide excellent information on the morphological aspects of the stenotic segment, although partial volume effects may lead to a slight underestimation of the severity of coarctation. 3D CE MRA can display the severity and extent of involvement without spin dephasing artifacts and partial volume effects. However, the assessment of the clinical significance of aortic coarctation depends on its hemo5.4.5.2  Congenital Anomalies dynamic relevance. On cine GRE images, the stenotic jet appears as signal dropout. This is less marked with SSFP 5.4.5.2.1  Aortic Coarctation sequences because of short echo times (TE) in combinaAortic coarctation (Fig. 5.4.32) accounts for 6–8% of con- tion with complete refocusing of transverse magnetizagenital heart anomalies and refers to area of narrowing of tion. PC-MR flow mapping using the modified Bernoulli the thoracic aorta in the region of the insertion of the ar- equation for an estimation of the pressure gradient at the terial duct with or without additional abnormalities of the level of the coarctation can provide further functional inaortic arch (Brierly et al. 2001). An abnormal plication of formation: the tunica media of the posterior aortic wall leads to a (5.4.8) formation of a fibrous ridge that protrudes into the aortic ∆P = 4V2, lumen and causes obstruction. The stenotic segment can be focal (aortic coarctation), diffuse (hypoplastic aortic where P is the pressure drop across the stenosis (mmHg) isthmus), or complete (aortic arch interruption). Recent and V is velocity (m/s). The prediction of the severity of studies have shown that one third of patients remain or a stenosis with this technique shows a high sensitivity and specificity compared with DSA (95 and 82%, respecbecome hypertensive despite corrective surgery. DSA has been the standard of reference for the diag- tively). An estimation of the collateral circulation can be nosis of aortic coarctation, but MRI has the advantage of carried out by measuring the flow in the proximal and high diagnostic accuracy in the pre- and postoperative descending aorta (Julsrud et al. 1997). evaluation of the disease without the need for ionizing MRI can also be used to assess secondary pathologies radiation and invasiveness. For management decisions, in patients with coarctation, including the evaluation it is crucial to determine the exact location and degree of the aortic root for dilatation secondary to a bicuspid of the stenosis, the length of the obstructed segment, as- aortic valve, which has a frequency of 15% in coarctation sociated aortic arch involvement, the collateral pathways and the assessment of aortic regurgitation. Furthermore (internal mammary and posterior mediastinal arteries), LV function and ventricular mass can be assessed by adand the course of the left subclavian artery. ECG-trig- ditional CINE imaging of the heart. Recently, interven-

829

830

5  Thorax and Vasculature

tional methods such as balloon angioplasty have come A very rare anomaly associated with a right aortic arch is into wide use providing good long-term results espe- an isolated left subclavian artery, which is supplied retrocially in lower-grade focal stenoses (Paddon et al. 2000), gradely via the left vertebral artery, and can cause vertewhereas long transverse arch or tubular stenosis of the bral steal syndrome (Van Grimberge et al. 2000). A right isthmus is frequently referred for surgical repair. aberrant subclavian artery originating from a left aortic arch does not usually cause a complete vascular ring, but may cause dysphagia by posterior esophageal compression and dyspnea by tracheal stenosis. 5.4.5.2.2  Aortic Arch Anomalies During the fetal period, six pairs of aortic arches form to join the two dorsal aortae with the aortic sac that will become the ascending aorta. At the end of development, some of the arches disappear; the fourth right arch forms the right subclavian artery as far as the origin of its internal mammary branch, while the fourth left arch constitutes the arch of the aorta between the origin of the left carotid artery and the termination of the ductus arteriosus. Aortic arch anomalies (Table 5.4.11) result either from abnormal regression of an embryonic arch that normally remains patent, or from persistent patency of a structure that normally regresses (Thiene and Frescura 1999). TSE imaging in axial or coronal orientation provides an excellent visualization of the vascular structure and its relationship with mediastinal organs, while 3D CE MRA is very useful to define the complex anatomy of the aortic arch and supra-aortic vessels. The challenge for MR imaging is to identify the formation of complete vascular rings that require corrective surgery. Aberrant right subclavian artery (arteria lisoria) and the double aortic arch are discussed in sect 5.3.2 (T. Johnsons section on congenital heart disease). A right aortic arch that passes to the right of the trachea and may descend either to the left or to the right of the thoracic spine can be observed in about 0.1% of the population. It can either occur with a mirror-image branching pattern or an aberrant left subclavian artery. In case of a mirror-image branching pattern, it is usually not associated with vascular ring formation. Association with an aberrant left subclavian artery can result in posterior esophageal compression and, if the arterial ligament is on the left side, in a complete vascular ring formed by the right aortic arch, anterior left common carotid artery, arterial ligament and retroesophageal left subclavian artery.

Table 5.4.11  Aortic arch anomalies Right aortic arch Aberrant left subclavian artery Mirror-image branching Isolated left subclavian artery Left aortic arch Aberrant right subclavian artery Right descending aorta Double aortic arch

5.4.5.3  Acquired Diseases 5.4.5.3.1  Aortic Aneurysm Aortic aneurysm is defined as a localized or diffuse dilatation involving all layers of the aortic wall and exceeding the normal aortic diameter by a factor of 1.5 or more (Fattori and Nienaber 1999). It has an incidence of 10.9 in 100,000 individuals/year (Clouse et al. 1998). The most common pathologic condition associated with aortic aneurysm is atherosclerosis. Atherosclerotic aneurysms usually have a long fusiform shape. Saccular aneurysms (Fig. 5.4.33) may also develop in atherosclerotic patients as a consequence of penetrating aortic ulcer (Sect. 5.4.5.2.7). It remains controversial whether atherosclerosis itself actually causes aneurysms or whether atherosclerosis develops in the dilated aorta. Clinical and basic research studies suggest that atherosclerosis should be considered a concomitant process rather than a direct cause of aneurysm formation and growth (Xu et al. 2001; Carrell et al. 2002). Seventy-five percent of atherosclerotic aneurysms are located in the distal abdominal aorta, below the renal arteries. Aortic medial degeneration has been demonstrated in most aneurysms, regardless of their cause and location. In the ascending aorta, gradual degenerative changes of the media can be related to congenital disorders of the extracellular matrix associated with an alteration of the elastic fibers such as Marfan syndrome (Sect. 5.4.5.2.7). Evaluation of aortic aneurysms should include quantification of the transluminal diameter of involved segments and comparison with uninvolved segments. For consistent results, vessel dimensions should be measured at the same anatomical location each time. In AAA, the aortic diameter has to be measured at the level of the celiac artery and the renal arteries. Furthermore, the maximal diameter and the diameters of the iliac and femoral arteries have to be assessed. The exact relation of the aneurysm to both renal and iliac arteries (“proximal and distal neck”) has to be determined, since the most accepted contraindication for endovascular repair is a proximal “neck” either shorter than 15 mm or absent. 3D CE MRA is most useful for depicting location and extent. Moreover, stenoses of aortic branches can be detected with high sensitivity. Since MIP images represent a cast of the lumen, diameters should be obtained from the original source images. SSFP sequences

5.4  MR Angiography Fig. 5.4.33a,b  Fusiforme aneurysm of the ascending aorta an the aortic arch. a Coronal MIP reconstruction of 3D CE MRA; b axial T1weighted TSE images provide best anatomic detail of the aortic wall

Fig. 5.4.34a,b  Partially thrombosed AAA. a MIP reconstructions of 3D CE MRA represent a cast of the lumen, while the surrounding thrombus remains invisible. b Exact diameter measurements should be obtained from the original source images

and monitoring of the expansion rate in subsequent examinations. In asymptomatic patients, whose aneurysms are too small to justify surgery or intervention, imaging is recommended at least every 6–12 months to monitor expansion. Open surgical treatment of the aneurysm and insertion of a stent-graft is indicated for aneurysms of any size expanding rapidly (>10 mm/year) or associated with symptoms. In asymptomatic patients, conventional or endovascular repair is indicated for TAA >6 cm or AAA >5.5 cm. In Marfan syndrome patients, TAA >5 cm should be considered for surgery. Early surgery is not associated with improved long-term survival. The capability of CE MRA to visualize the Adamkiewicz artery represents an important advance to avoid postoperative neurological deficit secondary to spinal cord ischemia (Yamada et al. 2000). However, this requires high spatial resolution scans with voxel sizes of less than 1 mm3. 5.4.5.3.2  Mycotic Aneurysm

A mycotic aneurysm is a rare condition that develops as a result of staphylococcal, streptococcal, or Salmonella inprovide a fast and precise overview of the aortic lumen fections of the aorta, usually at an atherosclerotic plaque. and surrounding tissues. TSE sequences are helpful to These aneurysms are typically saccular. Blood cultures assess alterations in the aortic wall and periaortic space. are often positive and reveal the nature of the infecting Periaortic hematoma and areas of high signal intensity agent. Therapy often consists of transient stenting of the within the thrombus may indicate instability of the aneu- aneurysm as a bridging procedure prior to reconstrucrysm. Atherosclerotic lesions are visualized as areas of in- tive surgery after completed antimicrobial therapy. creased thickness with high signal intensity and irregular shape. With fat-suppression techniques, the outer wall of the aneurysm can be easily distinguished from periad- 5.4.5.3.3  Congenital Aneurysm ventitial fat tissue. MRI is frequently used as follow-up tool to monitor the course of the disease. The estimated Families with a genetic predisposition to the developrisk of rupture substantially increases for thoracic aortic ment of thoracic or abdominal aortic aneurysms without aneurysms (TAA) >6 cm and abdominal aortic aneu- evidence of collagen-vascular disease have been docurysms (AAA) >5.5 cm (Fig. 5.4.34). Since the only well- mented. Congenital aortic aneurysms may be primary or documented risk factor for aortic rupture is increasing associated with anomalies such as bicuspid aortic valve size, the major diagnostic goal is accurate measurement or aortic coarctation.

831

832

5  Thorax and Vasculature

Fig. 5.4.35  Inflammatory abdominal aortic aneurysm. Periaortic fibrosis can be differentiated from surrounding abdominal fat tissue on CE T1-weighted TSE images with fat saturation such as this

5.4.5.3.4  Inflammatory AAA Inflammatory abdominal aortic aneurysm (IAAA) (Fig. 5.4.35), also known as periaortic fibrosis is a subentity with an incidence ranging from 2.2 to 18.1% of abdominal aortic aneurysms (Stella et al. 1993). It is characterized by periaortic retroperitoneal fibrosis with lymphocytic infiltration of the aortic adventitial layer and callosity of the aortic wall. The disorder almost exclusively affects male patients over the age of 60 years. If untreated, the rate of rupture is about 20%. The cause of these inflammatory changes remains unclear. They are probably the result of a local autoallergic reaction to certain components of atherosclerotic plaques (Leseche et al. 1992). Surgical repair is associated with a higher morbidity and mortality than in simple aortic aneurysms; therefore a precise evaluation of the involvement of adjacent structures is important for appropriate planning of preoperative and surgical therapy. Patients with ureteral involvement and obstruction may benefit from preoperative stent graft implantation.

Fig. 5.4.36a,b  DeBakey type III/Stanford B aneurysm; sagittal MIP (a) and VRT (b) reconstructions of 3D CE MRA acquired at 3 T; celiac and superior mesenteric artery arise from the true lumen

5.4.5.3.5  Aortic Dissection Aortic dissection is caused by laceration of the aortic intima and inner layer of media that allows blood to course through a false lumen in the outer third of the media (Fig. 5.4.36). Most commonly, it occurs secondarily to hypertension, but there are a few other causes that have to be borne in mind, especially in young patients (Table 5.4.12). Classification of aortic dissection (Table 5.4.13)

Table 5.4.12  Causes of aortic dissection Hypertension Aortic coarctation Valvular disease Marfan syndrome Iatrogenic injury

Table 5.4.13  Common classifications of aortic dissection DeBakey et al. (1965)  Type Ia Intimal tear in ascending aorta; terminates distal to innominate artery Type IIa Intimal tear in ascending aorta; terminates proximal to innominate artery Type IIIb Intimal tear at or distal to left subclavian artery Requires urgent surgery Can be treated medically if uncomplicated

a

b

Stanford Type Aa Involves ascending and descending aorta Type Bb Involves only descending aorta

5.4  MR Angiography

Fig. 5.4.37  Intimal flap in a dissection of the thoracic aorta; axial T1-weighted image also reveals intramural hematoma by its high signal intensity

7 Fig. 5.4.38  Coronal MIP of 3D CE MRA in a Marfan patient with persistent aortic dissection after replacement of the ascending aorta clearly depicts the intimal flap involving both carotid arteries

is commonly based on anatomical location and extension of the intimal flap (Fig. 5.4.37). Acute aortic dissection is a life-threatening condition requiring prompt diagnosis and treatment (Coady et al. 1999). The 14-day period after onset has been designated as an acute phase because morbidity and mortality (15–25%) rates are highest during this period: 1–2% per hour in the first 24 h after onset and 80% within 2 weeks. Surviving patients usually stabilize during this period. Physical findings may be absent at clinical presentation or may mimic other disorders such as acute myocardial ischemia or stroke. Therefore, the diagnostic goal for successful clinical management is a clear anatomic delineation of the intimal flap, the detection of the entry and re-entry sites, and the presence and degree of flow in the aortic branches. While transesophageal echocardiography (TEE) and multi-slice spiral CT provide high sensitivity and specificity in the detection of aortic dissection, the high degree of contrast and spatial resolution along with its capability for multiplanar acquisition have made MRI increasingly important. Moreover, it has been found the most reliable imaging modality in the evaluation of important complications such as aortic valve insufficiency, myocardial infarction in coronary artery involvement and hemodynamic consequences of arterial branch vessel dissection or occlusion. In suspected cases, MRI should cover the whole aorta from the valve to the bifurcation. Standard protocols include ECG-triggered T1-weighted TSE black-blood

Fig. 5.4.39  TREAT MRA of a patient with replacement of the ascending aorta and acute Stanford A dissection allows for a clear identification of the patent false lumen and the entry site; the dissection involves both innominate arteries

sequences in axial and sagittal planes. The intimal flap is detected as a straight linear image inside the aortic lumen. The true and the false lumen can be differentiated either by anatomic features or flow pattern. The true lumen shows a signal void, whereas the false lumen has higher signal intensity. However, it should be borne in mind that artifacts caused by imperfect triggering, respiratory motion or slow blood flow can simulate or obscure an intimal flap on TSE images. A high signal intensity of pericardial effusion indicates a hemorrhagic component and is considered to be a sign of impending rupture of the ascending aorta. ECG-triggered SSFP sequences allow an accurate delineation of the true and false lumen and are therefore very useful for classification and diagnosis of aortic dissection (Pereles et al. 2002; Carr and Finn 2003). With ECG-triggered 3D CE MRA the intimal flap is easily detected and the relationship to the aortic branches clearly depicted (Fig. 5.4.38). Entry and reentry sites appear as segmental interruption of the linear intimal flap on axial or sagittal images. The invention of time-resolved multiphase techniques was a milestone in 3D CE MRA since they allow for a clear differentiation between the true and the false lumen and provided detailed information on entry and reentry sites in one sequence (Schönberg et al. 1999c). In combination with PI and view sharing techniques (TREAT), the total acquisition time can be reduced to a few minutes (Fig. 5.4.39).

833

834

5  Thorax and Vasculature

observed calcifications cause susceptibility artifacts on GRE images. There is a considerable clinical and imagIntramural hematoma (IMH) was first described in 1920 ing overlap between thrombosed false lumen in localas “dissection without intimal tear.” Spontaneous rupture ized dissection, intramural hematoma, and penetrating of aortic vasa vasorum of the media layer is considered atherosclerotic ulcer. A thrombosed false lumen in localto be the initiating process, which results in a circumfer- ized dissection has often a more eccentric or asymmetric entially orientated blood-containing space in the aortic appearance, while the spread of blood in an intramural wall. IMH may occur spontaneously or as a consequence hematoma frequently affects the aortic wall more cirof penetrating aortic ulcer in intrinsically diseased media. cumferentially. As in aortic dissection, arterial hypertension is the most frequent predisposing factor. Moreover, clinical symptoms and outcome do not differ from classic aortic dis- 5.4.5.3.8  Marfan Syndrome section. Therefore, treatment is similar in both conditions. This means urgent surgery for Stanford type A intramural Aortic dilatation and dissection are the major criteria in hematoma and medical treatment for type B hematoma. the cardiovascular system for the diagnosis of Marfan In a comparison of different imaging modalities, MRA syndrome (MFS). The disorder has an incidence of about demonstrated the highest sensitivities in the detection of 1 in 10,000 in most racial and ethnic groups. It is inherIMH. The diagnosis with MRI relies on the visualization ited as an autosomal dominant trail; one-fourth of cases of blood in the thickened aortic wall. T1-weighted images are probably due to new mutations. Severe MFS is charreveal a crescent-shaped area of high signal intensity in acterized by a triad of features: (1) long, thin extremities the aortic wall, which can affect the entire aortic circum- frequently associated with other skeletal changes, such ference. Precontrast fat-saturation images can be helpful as loose joints and arachnodactyly; (2) reduced vision in differentiating intramural hematoma from surround- as the result of dislocations of the lenses (ectopia lening mediastinal fat. Moreover, MRI is the only imaging tis); and (3) aortic aneurysms that typically begin at the modality capable of estimating the age of the hematoma sinuses of Valsalva (Fig. 5.4.41). Cardiovascular abnoron the basis of different signal intensities of hemoglobin malities are the major source of morbidity and mortality. degradation products. Mitral valve prolapse develops early in life and in about a quarter progresses to mitral valve regurgitation of increasing severity. MRI provides definitive measurements of aortic dimensions, which are essential for the diagno5.4.5.3.7  Aortic Ulcers 5.4.5.3.6  Intramural Hematoma

Penetrating atheromatous ulcer of the aorta is characterized by ulceration of an atheromatous plaque that penetrates through the internal elastic lamina into the media of the aortic wall, causing intramural hematoma or localized intramedial dissection. The plaque may also break through to the adventitia forming a saccular pseudoaneurysm. Aortic ulcers occur almost exclusively in the descending aorta, but a few cases with location in the ascending aorta (Fig. 5.4.40) and the aortic branch have been reported. Although clinical features may be similar to those of aortic dissection, penetrating aortic ulcers should be considered a distinct entity with a different prognosis and management strategy. Persistent pain, hemodynamic instability, and signs of expansion should trigger surgical treatment, whereas asymptomatic patients can be treated medically with a regular imaging follow-up. Incidence of transmural rupture has been reported in literature ranging from 8 to 42%. Diagnosis is based on the visualization of a crater-like ulcer in the aortic wall. Mural thickening with intermediate or high signal intensity on TSE sequences may indicate extension into the media and formation of an intramural hematoma. In 3D CE MRA, aortic ulcers are detected as contrast-filled outpouching of the aortic lumen, which Fig. 5.4.40  Sagittal MIP of a 3D CE MRA shows multiple ulmay even result in large pseudoaneurysms. Frequently cers of the descending aorta

5.4  MR Angiography Fig. 5.4.42  Coronal MIP reconstruction of 3D CE MRA in a patient with Takayasu’s arteritis shows severe stenosis at the origin of the left common carotid artery with post-stenotic dilatation

Fig. 5.4.41  Dilatation of the aortic root in a patient with suspected Marfan syndrome, coronal MIP reconstruction of 3D CE MRA; ECG triggering is essential to avoid motion artifacts resulting in double lining of the aortic wall

loss, and other systemic symptoms may be evident. The chronic stages of the disease present with symptoms related to large artery occlusion, including upper extremity claudication, cerebral ischemia, syncope, and visual impairment. The chronic disease is intermittently active. sis of aortic involvement and monitoring the course of High-dose corticosteroids are the basis of therapy. Treatthe disease in Marfan patients. Diameters must be related ment of symptomatic stenoses (Fig. 5.4.42) or occlusions to normal values for age and body surface area (Mohiad- requires either interventional or surgical therapy. This can din et al. 1990). MRI also enables the assessment of aortic be achieved by angioplasty with or without stenting and, compliance. Recent studies have shown decreased aortic in severe cases, by vascular resection and surgical placedistensibility and increased flow wave velocity distal to ment of composite grafts. DSA has traditionally been the the aortic root in MFS patients without aortic dilatation method of choice in the diagnostic evaluation of TA and (Nollen et al. 2003). Furthermore, secondary pathologies is still useful in guiding interventional procedures. Besuch as aortic regurgitation, dissection of the aorta, and cause of the lack of need for ionizing radiation and iorupture can be detected in one examination. dinated contrast material, MRI is ideal for serial evaluation of patients with TA under treatment. Moreover, it is very useful for early diagnosis because of its ability to evaluate wall thickness rather than just luminal narrow5.4.5.3.9  Inflammatory and Infectious Diseases ing. Imaging findings in TA include mural thrombi, signal alterations within and surrounding inflamed vessels, Nonspecific Etiology fusiform vascular dilation, thickened cusps of the aortic Takayasu’s Arteritis valve, multifocal stenoses, and concentric thickening of Takayasu’s arteritis (TA) is an idiopathic inflammatory the aortic wall. Additional ECG-gated SSFP imaging of vascular disorder that involves the thoracoabdominal the heart may also reveal pericardial effusion and signal aorta and its branches as well as the pulmonary arteries. alterations within the pericardial sac, representing granIt is also known as pulseless disease because of frequent ulation tissue. 3D CE MRA provides detailed vascular inocclusion of large arteries originating from the aorta formation including the location, degree, and extent of in chronic stages. TA causes arterial media destruction, stenoses and dilation, as well as patency of collateral vesleading to aneurysm formation and, uncommonly, rup- sels and bypass grafts. The introduction of new bloodture of involved arteries. As a result of the predominant pool contrast agents provides an opportunity to visualize affection of large vessels, histopathology is seldom avail- even small branch vessels of the aorta due to increased able and can mimic other types of arteritis. Therefore, the spatial resolution. diagnosis is largely based on the combination of clinical information, laboratory findings, and diagnostic imaging. Rheumatic Aortitis The disease is most prevalent in young females of Asian Rheumatoid arthritis, ankylosing spondylitis, psoriatic descent, although it also occurs in Europe, North Amer- arthritis, Reiter’s syndrome, Behçet’s syndrome, relapsica, Africa, and in the Middle East, with a female pre- ing polychondritis, and inflammatory bowel disease dominance. During the acute stage, fever, malaise, weight may all be associated with aortitis involving the ascend-

835

836

5  Thorax and Vasculature

ing aorta. The inflammatory lesions may extend to the sinuses of Valsalva, the mitral valve leaflets, and adjacent myocardium. Clinical manifestations include aortic aortic dissection (Fig. 5.4.43), aneurysm formation, aortic regurgitation, and involvement of the cardiac conduction system. Giant Cell Arteritis This vasculitis occurs in older individuals and affects women more than men. Primarily large- and mediumsized arteries are affected by focal granulomatous lesions involving the entire arterial wall. It may be associated with polymyalgia rheumatica. Obstruction of major branches of the aorta and development of aortitis and aortic regurgitation are common complications. Highdose glucocorticoid therapy may be effective in early stages of the disease. Specific Aortitis: Syphilitic Aortitis This late manifestation of luetic infection usually affects the proximal ascending aorta, particularly the aortic root, resulting in aortic dilatation and aneurysm formation. Involvement of the aortic arch and the descending aorta may occasionally occur. The usually asymptomatic aneurysms can be saccular or fusiform. Compression and erosion of adjacent structures may cause symptoms; rupture also occurs. The pathogenesis of aortic involvement is that of obliterative endarteritis of the vasa vasorum caused by inflammatory response to spirochetes, which results in medial destruction followed by arterial dilation, scar formation, and calcification of the aorta. These changes account for the characteristic radiographic appearance of a calcified ascending aortic aneurysm. 5.4.5.4  Aortic Trauma Traumatic aortic rupture or transection occurs following rapid deceleration forces as in car accidents and is associated with a significant mortality. Any delay in treatment and diagnosis can lead to death. Almost 90% of aortic ruptures appear in the region just beyond the aortic isthmus, where the relatively mobile thoracic aorta is joined by the arterial ligament. The ascending aorta and the aortic root are only involved in 10% of all cases. Other less common sites are the distal ascending aorta, distal segments of the descending aorta, or the abdominal infrarenal segment. The transverse lesion involves the arterial circumference to a variable degree, penetrating through the aortic layers with the formation of a false aneurysm. MRI is most frequently used in the subacute phase. A combination of TSE and 3D CE MRA seems to be the best imaging strategy. The hemorrhagic component of a lesion can be detected by its high signal intensity on nonenhanced T1-weighted TSE images with fat saturation. Sagittal planes help to differentiate between partial (a tear

Fig. 5.4.43a,b  Aortitis and aortic dissection in a patient with rheumatoid arthritis. a Coronal MIP reconstruction of 3D CE MRA allows for a clear differentiation of the true and the false lumen. b Ring enhancement of the aortic wall on T1-weighted TSE image with fat saturation

limited to the anterior or posterior wall) and complete lesions which encompass the entire aortic circumference (Fattori and Nienaber 1999). This discrimination is of prognostic significance, since circumferential lesions are more likely to rupture (Fattori et al. 1998). The presence of periadventitial hematoma and pleural or mediastinal hemorrhagic effusion are considered signs of acute or impending rupture. 3D CE MRA provides an excellent display of the aortic lesion and its relationship to the major aortic branches. 5.4.5.5  Aortic Stent-Graft Imaging Endovascular repair with aortic stent-grafts becomes increasingly important in the therapy of aortic pathologies (Fig. 5.4.44). These metallic implants are safe for MR imaging; however, GE sequences are prone to field inhomogeneities induced by metal components. ECG-gated TSE imaging provides the solution to this, since the T2 star effect is counterbalanced by the 180° refocusing pulse. MRI in patients with stent-graft includes the evaluation of the grafts morphology, position, and configuration of metallic frames, kinking or rotation, and the diameter of the proximal and distal neck (Fattori et al. 2003). Long-term graft failure is usually due to progression of native disease in the outflow vessels. Graft failure in the first year is often due to intimal hyperplasia progressing to graft thrombosis. For this reason, periodic graft surveillance within the first year with duplex ultrasound has become routine. MR angiography may be requested for further evaluation prior to graft revision or for detection of endoleaks (Table 5.4.14). An endoleak is the persistence of blood flow outside the lumen of the endoluminal graft but within an aneurysmal sac or adjacent vascular segment. It is usually due to incomplete sealing or exclusion of the aneurysmal sac or vessel, but it can also be caused by retrograde blood flow into the aneurysmal sac.

5.4  MR Angiography

Fig. 5.4.44a,b  Stent-graft in the descending aorta in a patient with DeBakey type III/Stanford B dissection; a Sagittal MIP reconstruction; b axial MPR of 3D CE MRA reveal wall irregularities but no endoleak

Table 5.4.14  Endoleak classification Type 1 Blood flow between stent-graft and proximal or distal aortic wall Type 2 Persisting retrograde blood flow into the aneurysmal sac via collaterals Type 3 Graft defect (primary and secondary) or disconnection

5.4.6  Peripheral MRA H.J. Michaely and H. Kramer 5.4.6.1  Peripheral Arteries The normal structure of a healthy arterial vessel consists of three layers: intima, media, and adventitia. The intima is a continuous layer of endothelial cells lining each arterial vessel. The media consists of smooth muscle cells which are liable for the elasticity of the vessel and the capability to contract and thus to propagate the blood to the periphery. The adventitia contains the vasa vasorum and nerves.

Changes of one or more of these layers can lead to vessel diseases like stenosis, occlusion, or aneurysmatic changes. The aging of the vessel wall is reflected by thickening of the intima and rigidification of the media. This leads to elongation and dilatation of the greater arteries, whereas small arteries demonstrate a decreasing lumen. Aneurysmatic changes are defined as focal enlargement of an artery to greater than 1.5 times its normal diameter. For the abdominal aorta the cutoff value normally is 3 cm, for the iliac arteries 1.8 cm, and for the popliteal arteries 0.7 cm (Leiner 2005, Leiner et al. 2005a). The radiological workup of an aneurysm includes full morphologic description of the aneurysm itself. This means description of the exact location, diameter, presence of thrombus or calcifications, and relationship to branch vessels and surrounding organs. Because of the potential limb-threatening complications the most important task for the radiologist is the differentiation between aneurysms that are suitable for endovascular or surgical repair and those that do not require treatment at all. If there is a strong suspicion or definite diagnosis of a peripheral aneurysm, imaging of the complete arterial vasculature from the diaphragm down to the feet should be performed. In 70% of patients with peripheral aneurysms there is also an aneurysm of the abdominal aorta. Moreover complete angiography of the lower body part should be provided for treatment planning, e.g., to clarify the interventional approach. If on the other hand there is the diagnosis of an aneurysm of the infrarenal aorta it is mandatory to image the iliac and thigh vessels as well to estimate the extent of the aneurysm. If the iliac arteries are involved in the aneurysm and stent grafting is performed, the internal iliac arteries have to be occluded by coils in advance to avoid an endoleak type II. 5.4.6.2  Arteriosclerosis Arteriosclerosis is an umbrella term for thickening and rigidification of the arterial wall. The most prevalent subgroup of arteriosclerosis is atherosclerosis, which is the most common cause of coronary heart disease, aneurysmatic changes of arterial vessels, and peripheral occlusive disease. In morbidity and mortality statistics atherosclerosis ranks number one in nearly all age-groups and in both genders. Other subgroups of arteriosclerosis are the focal calcifying arteriosclerosis (Mönckeberg) and the arteriolosclerosis. 5.4.6.3  Atherosclerosis As yet there is no definite explanation for the development of atherosclerosis but a widely used hypothesis is that it is a “reaction to injury.” The endothelial cells of the intima experience repeated or continuous injuries due to

837

838

5  Thorax and Vasculature

hypertension, metabolic changes like chronic hypercho- nerable vessel regions for this kind of atherosclerotic lelesterolemia or stress to the immune system as it is known sion are vessel bifurcations or orifices. As a result of these after transplantation. These injuries lead to an adsorption initial changes the blood flow is altered, which again of monocytes, which can migrate through the vessel wall leads to micro injuries of the vessel wall. Thus, these iniand there lead to proliferation of macrophages. Further- tial microchanges are the starting point of a vicious cycle more these injuries allow blood plasma components to until ultimately symptomatic lesions occur. These micropenetrate the intima, which leads to thrombocyte aggre- structural changes are influenced by some other factors. gation as well as stimulation of the smooth muscle cells. The most common risk factors are a positive family hisProliferation of muscle cells can cause narrowing of small tory for atherosclerosis, increased lipoproteins, cigarette vessels and emplacement of lipoproteins. The most vul- smoking, diabetes, and obesity. Symptoms like pain, paleness, or hypoesthesia are often the first signs for atherosclerotic changes of the arterial vasculature because there is no easy-to-perform screening test such as a blood sample that could be used as an indicator for early stages of atherosclerosis. One arguable method to assess the likelihood of peripheral atherosclerotic changes is the ankle-brachial index (ABI). Here blood pressure is measured at the ankle and at the arm. The normal ratio should be 1 to 1.1. If the index is less than 0.95, significant narrowing of one or more blood vessels in the legs is indicated. If it is less than 0.8, pain in the lower extremity may occur during exercise, if less than 0.4, pain may occur during rest. An ABI of 0.25 or less indicates severe limb-threatening PAOD. But this test can be significantly influenced by unknown pathologic changes of the vasculature of the upper extremity. There are several imaging methods that can detect the described changes, but all of them have in common that they are not used as a method of early detection. Today DSA is still the standard of reference for detection of atherosclerotic changes (Fig. 5.4.45). It offers an excellent spatial and temporal resolution like no other modality. But certainly there are the well known disadvantages of ionizing radiation as well as nephrotoxic contrast agents (CA). Moreover there are some risks resulting from the invasiveness of the method such as bleeding after arterial puncture or vessel wall dissection. There are three noninvasive methods of imaging the arterial vasculature: computed tomography angiography (CTA), magnetic resonance angiography (MRA), and Doppler ultrasound (Bezooijen et al. 2004; Leiner et al. 2005b) (Fig. 5.4.46). All these modalities have a reduced spatial resolution compared to DSA. CTA has the best spatial resolution of these alternatives but it also has to deal with the disadvantages of ionizing radiation and nephrotoxic CA as well as the problem of wall calcifications that compromise the graduation of stenoses. Doppler ultrasound can

Fig. 5.4.45  Digital subtraction angiography (DSA) of the left leg with multiple atherosclerotic lesions in the popliteal artery and in the proximal calf (arrows). Even more distal vessel occlusions can be detected (arrowheads). Because of its excellent spatial resolution DSA is still the standard of reference in angiography but it also still suffers from some major drawbacks, e.g., its invasiveness and the amount of potentially nephrotoxic CA needed as well as the lack of a 3D display

5.4  MR Angiography

Fig. 5.4.46a,b  MRA (a) and CTA (b) of the peripheral arterial vasculature. Both are contrast-enhanced scans and allow spatial resolution of 1 mm3. Acquisition time for CTA is about 15 s. For MRA acquisition time of the arterial phase ranges around 45–60 s. But complete examination time for MRA is

significantly longer compared than for CTA because of patient positioning and use of native scans as subtraction masks. Advantages of MRA are better definition of the small vessels in the distal calf and feet as well as the absence of image quality constraints due to calcifications

provide a very good spatial resolution as well as temporal resolution in vessels with superficial location, but in vessels located more deeply the sonographic access is markedly reduced and sonic transducer probes with lower frequency, i.e., poorer spatial resolution, have to be used. In the past MRA clearly had reduced spatial resolution compared to the other methods and dynamic information was not available. But the latest developments in hard- and software have made it possible to increase spatial resolution to a comparable level and also to offer dynamic information. All imaging modalities show that the volume of calcifications in the vessel wall does not correlate with the degree of lumen narrowing. The presence of calcifications certainly is an indicator for the presence of atherosclerotic changes but it is also possible to find high-grade atherosclerotic stenoses without any calcifications. The diameter of the calf arteries measures about 5–6 mm distal to the popliteal artery and 2–3 mm in the distal calf next to the foot. Spatial resolution therefore should be 1 mm3 isotropic or less. There are several methods to avoid venous enhancement. The two most common and easy to realize methods are to place soft foam padding in the popliteal fossa in order to eliminate compression of the calves and the superficial veins of the calves, and, second, to perform venous compression at the thigh level. Therefore a blood pressure cuff is placed just above the knee and inflated to sub-systolic pressure values of 50–60 mmHg (Herborn et al. 2004a, b). Both lead to a reduced venous backflow and a delayed CA inflow in the venous system.

the abdominal aorta and the iliac vessels are affected in 30% of cases. In 80–90% atherosclerotic changes are found in the femoral and popliteal arteries and in 40–50% in the arteries of calves and feet (Fenchel 2006b) (Fig. 5.4.47). Studies have shown that nearly 50% of the patients with symptomatic peripheral vessel occlusive disease also suffer from significant coronary artery disease. The mortality statistics of recent years show that the 5-year survival rate of patients with intermitting claudication is 70%, and the 10-year survival rate is 50%. It is of interest that the likelihood of a progression of the peripheral occlusive disease is lower than that for a lethal myocardial infarction. The morphologic pattern of atherosclerotic changes of the peripheral vasculature is subdivided in three different types. Type 1 is limited to involvement of the distal abdominal aorta and the common iliac arteries. Type 2 includes atherosclerotic changes limited to suprainguinal peripheral arteries, and type 3 stands for central and peripheral, so-called multistation, occlusive disease. Another important point is to differentiate between a relatively short, focal stenosis or occlusion and an occlusion over a long segment. The most common graduation of peripheral arterial occlusive disease (PAOD) is the Fontaine’s classification. Grade I stands for stenosis or occlusion of one or more arterial vessels of the lower extremity without any symptoms. Pain in the legs/claudication from a walking distance of more than 200 m is classified as grade IIa, whereas occurrence of pain within the first 200 m of walking indicates grade IIb. Grade III stands for leg pain at rest and necrosis or gangrene is defined as Fontaine grade IV. A very important disease pattern in PAOD is the so-called critical ischemia, which is defined as symptomatic peripheral vessel high grade stenosis or occlusion (Fontaine grade III), which needs contemporary treatment. The results of several studies evaluating the arteries of the lower body part with different MRA techniques are

5.4.6.3.1  Atherosclerosis of the Lower Extremities Atherosclerosis mostly occurs in the great and medium arteries of the lower body part. In symptomatic patients

839

840

5  Thorax and Vasculature Fig. 5.4.47  Multiple atherosclerotic changes from the abdominal aorta down to the feet. Occlusion of the right superficial femoral artery as well as multiple mediumto high-grade stenoses in the calves in a patient with PAOD Fontaine IIb

Table 5.4.15  Results of several studies evaluating the arteries of the lower body part with different MRA techniques No. of patients

Target value

Sensitivity (%)

Specificity (%)

Lower extremity

25

Stenosis >70%

90

96

DSA

Ankle/foot

24

Vessel patency

87

95

2D TOF

DSA

Lower extremity

23

Stenosis >50%

89

98

1996

2D TOF

DSA

Calf/foot

31

Stenosis >75%

98

96

Ho et al.

1998

3D CE MRA

DSA

Lower extremity

28

Stenosis >50%

93

98

Sueyoshi et al.

1999

3D CE MRA

DSA

Lower extremity

23

Stenosis >75%

100

99

Schmitt et al.

2005

3D CE MRA

DSA

Lower extremity

69

Stenosis >50%

92

98

Huegli et al.

2006

3D CE MRA

DSA

Lower extremity

20

Stenosis >50%

93

86

Pereles

2006

3D CE MRA

DSA

Lower extremity

45

Stenosis >50%

95

95

Group

Year

Technique

Reference

Region

Yucel et al.

1993

2D TOF

DSA

McDermott et al.

1995

2D TOF

Glickerman et al.

1996

Cortell et al.

5.4  MR Angiography

listed in Table 5.4.15. All these studies show an increase in accuracy compared to DSA over the years. In particular, the introduction of contrast enhanced techniques helped to increase the sensitivity and specificity of MRA. Thus MRA should be regarded as the new standard of reference not only in the iliac and thigh station, but also in the calf and foot, where sensitivities and specificities of 98 and 96%, respectively, could be achieved. To get this high-grade accuracy the imaging protocol should be adjusted to the clinical question and appearance of the patient. If symptoms correspond to Fontaine’s stadium I to IIb a standard runoff MRA is sufficient to assess the atherosclerotic changes. In Fontaine’s stadium III and IV, a hybrid MRA approach with primary imaging of the calf and foot, followed by a standard runoff MRA is needed to assess the most distal vasculature with an excellent image quality which is suitable for treatment planning. When using this technique MRA is even superior to DSA in the detection of patent pedal vessels. One explanation for that is the ability of MRA to image blood flow at velocities as slow as 2 cm/s. These velocities represent a major problem in DSA because of the dilution of CA due to more proximal located multiple stenosed vessel segments.

5.4.6.3.3  Atheroembolism Atheroembolism is a subgroup of the acute arterial vessel occlusion. It mostly occurs in small arterial vessels due to rupture of more proximally located atherosclerotic lesions or aneurysmatic changes as well as after an arterial puncture. Multiple small particles like fibrin, thrombocytes, or cholesterol particles occlude small arterial muscle and cutaneous vessels. Distal pulses are most often palpable; clinical symptoms are circumscribed pain and paleness. Due to the fact that most often small arterial branches are affected an invasive therapy is impossible in the majority of cases. A surgical intervention may be the cause of further atheroembolism. 5.4.6.4  Thoracic Outlet Syndrome

This syndrome arises not from any arteriosclerotic or degenerative changes but from a compression of the vesselnerve bundle in the neck and shoulder region. The first rib, the anterior scalenus muscle, or the minor pectoral muscle can lead to a stricture of the subclavian artery or the brachial plexus. Symptoms only occur at certain movements and are often provocable. In these cases pain of shoulder and arm, “all gone feeling,” paresthesia, and 5.4.6.3.2  Acute Arterial Embolism necrosis can occur. If symptoms are not provoked clinical examination normally shows no noticeable problems. An acute arterial embolism leads to a reduced periph- Symptoms can be provoked by abduction and outer roeral blood supply. The consequences of an arterial oc- tation of the arm as well as rotation of the head to the clusion depend on the localization and the size of the side showing symptoms. If any imaging procedure is perischemic area as well as the potentially present collateral formed these maneuvers should be done as well (Charon vessels. The most common sources of arterial emboli are et al. 2004; Dymarkowski et al. 1999; Hagspiel et al. 2000; the heart, the aorta, and the great vessels. Chronic atrial Yanaka et al. 2004). MRA offers the possibility to perform fibrillation, acute myocardial infarction, myocardial dynamic studies in a standard position as well as in a proaneurysms, endocarditis, and valve prosthesis are risk voking position. Another advantage of MRI as diagnosfactors for emboli coming from the heart. Distal arte- tic modality in this disease is the feasibility to clarify the rial occlusions can also emerge from atherosclerotic or other differential diagnoses like herniated vertebral disk aneurysmatic changes in more proximally located great or brachial plexus lesion causing these symptoms. vessels. Arterial emboli often occur at bifurcations because the vessel lumen decreases in the periphery. In the lower body part most embolic vessel occlusions arise in 5.4.6.5  Raynaud’s Syndrome the superficial femoral arteries, directly followed by the iliac arteries. Furthermore, embolic occlusions in the This syndrome is characterized by ischemic attacks of the lower extremities occur in the popliteal arteries and the fingers or toes because of a spastic occlusion of the small tibiofibular trunk. Clinical symptoms of an acute arte- arteries. The attacks are typically divided in three phases: rial embolism are pain, paresthesia, paleness, pulseless- first complete paleness of the affected extremity, then ness, prostration, and coldness. At present, CTA and cyanosis, and finally redness of fingers or toes. Paleness DSA are superior to MRA in the acute diagnosis of appears because of the vasospasm and the missing blood arterial embolism because CTA is easier and faster to supply of the extremity. Cyanosis manifests because of perform and DSA offers the interventional possibilities dilatation of capillaries and veins filled with deoxygenof dilatation or local lysis therapy. On the other hand ated blood. Redness occurs reflective when vasospasm MRA of the peripheral arteries provides excellent image ends and the small arteries are hyperperfused. The first quality which is not impaired by calcifications and does two phases are characterized by paresthesia and coldness; not suffer from the drawbacks of ionizing radiation and the last phase leads to hot flashes and pain in fingers or iodinated CA. toes (Kransdorf et al. 1998).

841

842

5  Thorax and Vasculature

eters via the subclavian vein even thromboses of the arms appear. Peripheral veins especially in the lower extremity can be Thrombosis of a superficial vein is innocuous and divided in superficial and deep veins. The branches of may be treated by symptoms. An exclusively superficial the saphenous vein represent the superficial venous sys- thrombosis cannot lead to pulmonary embolism. tem; the deep veins follow the arteries paired. Perforating veins connect both systems; bicuspid valves direct the blood flow to the deep venous system. 5.4.6.7  Vascular Malformations A thrombosis in a superficial or deep vein and the accompanied inflammatory reaction of the vessel wall are Arteriovenous malformations (AVMs) and hemangiocalled thrombophlebitis or vein thrombosis respectively. mas (Fig. 5.4.48) are benign soft-tissue neoplasms that Initially the thrombus contains thrombocytes and fibrin. most often are congenital. They can cause complications Erythrocytes agitate with fibrin and drive the thrombus through local extension or be considered a cosmetic growth in the direction of blood flow. problem particularly in young women. MRA is perPredisposing factors for deep vein thrombosis (DVT) formed to describe the exact local extension and to detect are stasis, injuries of the vessel wall, and hypercoagulabil- further AVMs in case of multiple AVMs. As patients with ity. DVT may occur in 50% of patients with orthopedic AVMs are relatively young, CE MRA is an appropriate surgery (hip and knee) and in 10–40% of patients with diagnostic tool. AVMs may be missed on a conventional abdominal or thoracic surgery. Moreover DVT often oc- MRA due to slow filling. Therefore, multi-phased or timecurs in combination with malignant disease; 10–20% of resolved MRA exams are indicated (Herborn et al. 2003). patients with the diagnosis of DVT suffer from an un- Time-resolved studies allow differentiating between highknown malignancy. flow AVMs and slow-filling lesions such as hemangiomas The most important aftereffect of a DVT is pulmonary or capillary malformations. If in doubt whether a high embolism. Swelling, warmth, and redness of one extrem- spatial or temporal resolution is to be chosen, the high ity can be a leading sign for deep vein thrombosis. Clini- temporal resolution should be preferred. By this, feeding cal diagnosis of a DVT may be difficult because only one and draining vessels may be identified, providing imporvein can be occluded and there are sufficient collateral tant information for treatment planning. To round off vessels. DVT was nearly exclusively known in the lower the exam T1-weighted fat-saturated sequences after conextremity but since the introduction of central vein cath- trast agent administration and STIR sequences should be 5.4.6.6  Peripheral Veins and Venous Thrombosis

Fig. 5.4.48  a Time-resolved (TREAT) MRA at 1.5 T of a young male patient suffering from multiple hemangiomas of the foot. The temporal resolution was 3 s/3D frame. 6 s after the initial contrast arrival at the foot the hemangiomas at the level of the calcaneus become gradually visible (arrows). b High-spatial-resolution MRA of the same patient acquired (from left to right) in the arterial phase, and three venous phases. In the arterial phase none of the hemangiomas is visible. Throughout the venous phases the hemangiomas become visible. There are hemangiomas at the level of the calcaneus as well as at the distal forefoot. In a solely arterial scan these findings would have been missed

5.4  MR Angiography

added (Herborn et al. 2003). They allow for detection of partial thrombosis in a hemangioma and for a good differentiation between non-affected soft tissue and AVM or hemangioma. Despite these positive results conventional X-ray angiography is still considered the gold standard for detection of assessment of peripheral AVMs (Leiner 2005; Herborn et al. 2003). In contrast to AVMs and hemangiomas, arteriovenous fistulas (AVF) are most often acquired iatrogenically. They are therefore located at typical puncture sites such as at the common femoral artery. AVFs can be easily examined with Doppler ultrasound. In case of a non-diagnostic ultrasound exam, a time-resolved MRA with the highest possible temporal resolution can be performed. AVFs will present with premature venous return at the AVF site. Due to pulsation artifacts and the limited spatial resolution of time-resolved MRAs the fistula itself is rarely visible.

essential to do the bolus timing at the extremity of interest to prevent venous overlay from early venous shunting (Leiner 2005). 5.4.6.8.1  Specific Diseases

Thromboangiitis Obliterans: Buerger’s Disease A relatively frequent manifestation of non-atherosclerotic peripheral arterial disease is thromboangiitis obliterans (TAO), a segmental inflammatory disease most commonly affecting the small and medium-sized arteries, veins, and nerves of the arms and legs of young heavy male smokers. The disease is characterized by an early onset, most often before the age of 40. Patients exhibit the typical symptoms of peripheral arterial disease: claudication of the feet, legs, hands, or arms, which may progress to ischemic rest pain and ischemic ulcerations. Angiographically, TAO is characterized by segmental occlusive lesions in small and medium-sized vessels of the extremities (palmar, plantar, tibial, peroneal, radial, ulnar, digi5.4.6.8  Vasculitis and Inflammatory Diseases tal arteries). The vessel affection is distally pronounced; Peripheral MRA is often ordered to detect and grade in- proximal vessels are rarely involved. Collaterals present flammatory changes either in the setting of vasculitis or with a typical corkscrew pattern bridging occlusions. inflammatory changes associated with diabetes or non- There is no apparent source of emboli. The only definidiabetic osteomyelitis. tive therapy for TAO is complete cessation of smoking; Acute or chronic inflammatory changes of small, me- surgery is not indicated (Leiner 2005; Olin 1994, 2000; dium, and large arteries in particular are the hallmarks of Olin et al. 1990). vasculitis. Patients may present with systemic symptoms such as fever, fatigue, and rheumatic complaints. In these Giant Cell Arteritis patients peripheral MRA is ordered to assess the involve- Patients with giant cell arteritis usually suffer from aorment of the peripheral arteries in particular. In rare cases tic and aortic branch stenoses (Takayasu’s disease) or the venous system is also affected. The MR-examination occlusive extracranial lesions (temporal arteritis). Norfor these patients should include MRA sequences but mally, the peripheral vessels are not affected. Inflammaalso some morphological sequences to determine the tory changes from giant cell arteritis are characterized extent of vessel wall and soft tissue involvement. As vas- angiographically by smooth, tapering stenoses (which culitis often affects medium-sized and small vessels the are often symmetric and bilateral) with post-stenotic anmaximally achievable spatial resolution is indicated. For eurysmal dilatations and multiple collateral vessels due the involvement of the vessel wall dark-blood EKG-trig- to the chronic nature of the disease. In particular, MR gered sequences are required. They allow for artifact-free imaging at 3 T allows clear depiction of inflammatory high-contrast depiction of the wall. T2-weighted fat-satu- vessel wall changes (Markl et al. 2006). Also, a persistent rated sequences for detection a vessel wall edema and T1- enhancement of the vessel wall can be seen with delayed weighted fat-saturated sequences after contrast admin- T1-weighted sequences, which is indicative of vessel wall istration that show inflammatory contrast agent uptake inflammation (Bley et al. 2005a). into the vessel wall should be performed. Diabetic soft tissue infections are caused by microan- Periarteriitis Nodosa giopathic and to a lesser extent macroangiopathic vessel Periarteriitis nodosa involves the small and mediumchanges (Figs. 5.4.49, 5.4.50). Therefore, peripheral MRA sized arteries of almost all organs. Characteristic findings to detect vascular lesions is often combined with mor- are small aneurysms in the visceral, renal, or distal limb phologic sequences (STIR and T1-weighted TSE pre and vessels, which are prone to rupture. Due to their small post-contrast) to assess the soft tissue infection. In these size, detection of these microaneurysms with MRA is ofpatients single phase MRA frequently exhibits venous ten not possible. A perfusion inhomogeneity of the kidand soft tissue enhancement due to inflammatory shunts. neys on fast time-resolved images may be seen, however. Time-resolved studies are essential in these patients to As discussed above, T1-weighted dark-blood fat-satucorrectly identify the vascular anatomy and stenotic le- rated sequences after the administration of contrast agent sions. If no time-resolved study can be performed it is are indicated to detect vessel wall enhancement.

843

844

5  Thorax and Vasculature Fig. 5.4.49  a Time-resolved (TREAT) MRA at 1.5 T of a 62-year-old male diabetic patient with chronic infection of the right foot. In this temporal series with each image 3 s apart there is early enhancement of the right arteries (arrow) and of a diffuse enhancing area at the level of the ankle which represents inflammatory tissue. No arterial vessels below the ankle are seen, which is a sign of microangiopathy. As no target vessel for a bypass operation is visible, an amputation of the forefoot may be the consequence. Please note that the flow of the healthy side is markedly delayed. b Precontrast T1-weighted TSE image of the same patient demonstrating destruction of the talus and extensive tissue edema

Fig. 5.4.50  Time-resolved (TREAT) MRA of a patient with a diabetic left foot. Similar to Fig. 5.4.46, it demonstrates very early enhancement of the vessels of the left leg and of the inflammatory vessel shunts at the level of the ankle. The vessels of the left leg are already enhanced in the first frame, which is indicative of an increased reactive blood flow to the left leg. All arterial vessels of the left leg are patent and do not show macroangiopathic changes. The healthy side demonstrates delayed flow of six frames

5.4  MR Angiography

Behçet’s Disease Behçet’s disease is a vasculitis that again affects small and large arteries and veins. All vessels can be affected by either thrombosis or formation of saccular aneurysms. Patients with Behçet’s disease are prone to false aneurysm formation after arterial or venous puncture (O’Duffy 1990a, b; Ko et al. 2000).

5.4.6.10  Popliteal Artery Entrapment

Similar to the arcuate ligament syndrome in the abdomen (Lee et al. 2003), the popliteal artery can be entrapped. This entrapment results from an anatomic variant in which the popliteal artery passes medial to and underneath the medial head of the gastrocnemius muscle. Popliteal entrapment is a disease of young men (male to female ratio: 9 : 1), and presents with calf or foot claudication that is bilateral in up to 25% of all cases. 5.4.6.9  Hereditary Disorders Angiographic findings include medial deviation of the with Peripheral Arterial Involvement proximal popliteal artery (P1 segment) in combination There are numerous hereditary (connective tissue) disor- with segmental occlusion or post-stenotic dilatation. If ders that may affect the vascular system. As they are rare there is high clinical suspicion but a negative MRA an diseases with similar imaging findings in the peripheral additional MRA with dorsiflexion of the gastrocnemius arteries, only the most important diseases will be briefly can be obtained. Review of the source images may greatly facilitate the detection of the anatomical relationship of discussed. arteries and muscles (Leiner 2005; Wright et al. 2004). 5.4.6.9.1  Marfan Syndrome Marfan syndrome is caused by an autosomal dominant defect in the cross-linking of collagen, with about 95% of patients having cardiovascular involvement—mainly aneurysms of the ascending aorta (compare Sect. 5.4.5.1). Peripheral arterial disease symptoms may be mimicked by descending aortic aneurysms (Leiner 2005; Crivello et al. 1986). In case of dissecting aneurysms, MRA is particularly helpful in assessing the blood-flow hemodynamics with time-resolved MRA techniques. 5.4.6.9.2  Ehlers-Danlos Syndrome Ehlers-Danlos syndrome (more than 10 subtypes), an autosomal dominant disease with failure of correct to collagen formation, leads to diffuse thinning of the arterial media. In terms of arterial involvement, type IV (arterial-ecchymotic type) is the most important in which the patients may present with spontaneous arterial rupture, dissections, and arteriovenous fistulae (Leiner 2005; Wilcken 2003; Germain 2002). 5.4.6.9.3  Neurofibromatosis Neurofibromatosis is an autosomal dominant disease with dysplasia of mesodermal and neuroectodermal tissues. Patients in their second and third decades may present with hypertension due to renal artery stenoses and sometimes claudication from peripheral artery involvement. The vascular lesions associated with neurofibromatosis include tapering stenoses, occlusion, aneurysm, pseudoaneurysm, and rupture or fistula formation of small-, medium-, and large-sized arteries (Leiner 2005; Ilgit et al. 1999; Haust 1987).

5.4.6.11  Imaging of Hemodialysis Shunts Hemodialysis shunts (HS) are artificially created between upper extremity arteries and veins and serve as a prerequisite for hemodialysis. In the direct arteriovenous bypass an increased flow of 350–400 ml/min in AV shunts and 800–1,000 ml/min in PTFE grafts (Konner et al. 2003) can be reached. HS are prone to thrombosis, and stenosis formation particularly at the anastomosis site where a turbulent flow is present. Occlusions and aneurysms can occur as well (Konner et al. 2003). The incidence of thrombosis is 0.5–2.5% of all HS. The inflow volume and the inflow velocity were found to be a good predictor of impending HS failure (Smits et al. 2002). 3D CE MRA has been used in some studies to image HS at the forearm with equally good interobserver agreement as the conventional angiography (Waldman t al. 1996; Planken et al. 2003; Cavagna et al. 2000; Han et al. 2003). Because of the 3D volumetric nature, the 3D MRA can be beneficial in the workup of suspected hemodialysis shunt pathology. The lack of inherent motion at the forearm allows prolongation of the acquisition times. However, the high flow in HS may lead to flow-related artifacts and false positive results. To avoid flow-related artifacts, a blood pressure cuff can be inflated to decrease inflow into the HS. This allows for prolonged imaging time (Han et al. 2003). Using a combined approach of flow measurements and MRA a sensitivity of 100% and a specificity of 94% were reported in a recent study (Han et al. 2003). Timeresolved MRA studies can be of particular value when a HS-induced steal syndrome is suspected. In HS-induced steal syndrome the extremity distal to the HS is hypoperfused and may develop ischemia and rest pain. A high temporal resolution of not more than a 2-s acquisition time per 3D volume should be chosen. An additional advantage of MRA in HS imaging is the ability to assess the

845

846

5  Thorax and Vasculature

entire inflow and outflow vascular system in order to de- part exclusively. Apart from coronary artery disease tect proximal stenosis or thrombi in the subclavian and (CAD) atherosclerotic variances occur in the carotid arbrachial vessel if the field of view is chosen large enough teries in a relatively high percentage of patients with pe(Michaely et al. 2005). ripheral artery occlusive disease (PAOD) (Dormandy et al. 1999; Goyen et al. 2003). In this respect a clarification of the complete arterial situation from the head down to the feet seems to be reasonable. Until now the limiting 5.4.7  Whole-Body MRA factor was the absence of a modality to image the comH. Kramer and H. Schlemmer plete arterial system in only one exam within reasonable time and without limiting factors like invasiveness and 5.4.7.1  Introduction ionizing radiation. Cardiovascular disease still ranks number one in mortality and morbidity statistics in most developed countries and the incidence is further rising (Reddy and Yusuf 1998). As mentioned in Section 5.4.6 atherosclerosis may be understood as systemic in nature, and it influences the entire macro- and microvascular system. Seventy percent of patients suffering from an atherosclerotic aneurysm of the peripheral arteries also show an aneurysm of the abdominal aorta. Moreover 50% of all patients with atherosclerotic changes of the peripheral arteries also suffer from significant coronary artery disease (Dormandy et al. 1999). Despite these facts the diagnostic approach to atherosclerosis has remained locally focused on the symptomatic clinic problem for a long time. Today it is a standard procedure to additionally display the abdominal aorta when asked for angiography of the lower extremity. This procedure is performed regardless of which angiography modality is used. In DSA the catheter is positioned proximal to the renal arteries and a first overview series is acquired (Fig. 5.4.51). In CTA and MRA, CA arrival time is calculated at the same position and the datasets are acquired consecutively down to the feet. This is essential and suggestive firstly because the above-mentioned incidence of findings not only in one vascular territory and secondly for optimal treatment planning. Certainly atherosclerotic changes are not found in the lower body

Fig. 5.4.51  Overview of the abdominal aorta as first series of a DSA of the lower body part. Catheter tip is positioned proximal to the renal arteries. At the same level the measurement region of interest (ROI) is positioned for CTA or MRA to calculate delay between CA application and start of measurement

5.4.7.2  Angiographic Modalities Whole-body DSA is beyond any discussion because of its invasiveness, the ionizing radiation as well as the amount of potentially nephrotoxic CA. Doppler ultrasound is a very cheap and accessible diagnostic approach to arterial vessels of the extremities and the neck (Howard et al. 1996; Yamamoto et al. 1996). But there are also a few limitations like superimposed air from lungs or intestine and bones like the skull when looking for the intracranial vessels. CTA again suffers from nearly the same limitations as DSA. Certainly it is possible to image the entire arterial system by CT in a life threatening situation in multiple-trauma patients with the suspicion of arterial bleeding due to the trauma (Matsubara et al. 1990; Laghi et al. 2001; Rubin 1997) (Fig. 5.4.52). But it is not justifiable to perform whole-body CTA only because of the suspicion of atherosclerotic changes in the arterial vasculature. 5.4.7.3  Whole-Body MRA For the above reason the only available method for imaging the complete arterial vasculature is MRA. Here the problem arises, that MRI has been limited to single stations for a long time. First approaches to cover larger anatomic areas suffered from reduced spatial resolution. The implementation of parallel imaging made it possible to expand the imaged area at adequate spatial resolution. However, the limiting factor was hardware with a restricted range of table movement. Furthermore special coils for imaging of large body parts at adequate image quality were not available. Today nearly all vendors provide special equipment for MRA and MR systems offer more flexibility in terms of table movement and coil systems as well as receiver channels (Ekelund et al. 1996; Fellner et al. 2003; Goyen et al. 2001b, c; Ho et al. 1998; Janka et al. 2000) (Fig. 5.4.53). Therefore today there are different approaches to whole-body MRA. 1 When using a standard MR system with still-limited table movement it is impossible to image the entire arterial vascular system in consecutive steps. The exam

5.4  MR Angiography

Fig. 5.4.52  Patient presenting with an acute onset of strongest abdominal pain. CTA of the abdominal aorta with detection of a ruptured infrarenal aortic aneurysm. CTA was performed from the aortic arch down to the knees

Fig. 5.4.53  Matrix-coil setup for wholebody imaging on a dedicated wholebody MR system. Matrix-coil system (Tim [total imaging matrix], Siemens Medical Solutions, Erlangen, Germany) consists of integrated spine-matrix and posteriorly positioned head and neck coil elements (all left side) and anteriorly positioned parts of head and neck coil as well as two or three body matrix coils (depending on patient’s height) and a dedicated peripheral angio array matrix coil (all right side)

then is divided into two steps. Step one only includes MRA of the cervical and cranial vessels; step 2 consists of MRA of the abdominal aorta and the arteries of the lower extremity. In this case two CA injections are necessary; a divided CA bolus with different flow velocities is most suitable for the second injection in part two (Kramer et al. 2005) (Fig. 5.4.54). 2 An option to overcome the limitation of restricted table movement in standard MR systems is to use dedicated rolling table platforms (AngioSURF®). Here the patient is pulled manually through the magnet and abdominal aorta and the thigh arteries are acquired five consecutive MRA steps are acquired. This amount (Fenchel et al. 2006b; Kramer et al. 2005). of CA is injected in a biphasic manner; the second half 4 A recently developed method not yet clinically routine of the CA is injected at reduced flow velocity followed is the so-called moving-during-scan technique. Here by a saline flush (Goyen et al. 2002) (Fig. 5.4.55). the patient is automatically and continuously moved through the magnet. The most promising approach 3 An expanded range of table movement of dedicated whole-body MR systems allow whole-body MRA to to whole-body MRA with this technique is combined be performed in four to five consecutive steps, withwith a special sequence technique. During the first out repositioning the patient. There are different CA passage of the patient through the magnet and the injection protocols when using this kind of MR sysarterial phase of the CA bolus only the centre of ktem. The most promising application scheme consists space is acquired; this is directly followed by a second of two different CA injections. With the first bolus passage in which the periphery of k-space is acquired the carotid arteries are imaged. Then the table moves (Barkhausen 2006). down to the calves, and with the same CA bolus the Methods numbers 1 and 2 were the first methods atarteries from the knee down to the feet are imaged. tempted for performing whole-body MRA and still sufWith a second CA injection MRA datasets of the fer from some technical limitations. These limitations

847

848

5  Thorax and Vasculature Fig. 5.4.54  Coil setup when using a standard MR system without dedicated rolling platforms. Data acquisition is divided into two parts. Part one consists of imaging of the heart, the lungs, as well as the brain and an MRA of the neck and head vessels. Coils used for that are the integrated spine array, two body array coils positioned next to each other to allow implementation of parallel imaging (PI) as well as the head coil (left side). Part two consists of imaging of the abdomen as well as MRA of the lower body part. Here again the integrated spine array, two body array coils and a dedicated peripheral angio array coil are used (right side)

Fig. 5.4.55  Whole-body imaging in five consecutive steps. A dedicated rolling platform (AngioSURF®, MR-Innovation GmbH, Essen, Germany) is pulled manually through a standard MR system. To increase image quality in terms of SNR an integrated spine array coil and a body array coil are used. The patient moves through the MR system and between both coils

affect image quality mostly in terms of venous enhance- techniques like parallel imaging, data acquisition for the ment, resulting in disturbing venous overlay. Even in the proximal MRA stations requires too much time to get a second part of method number one, when only MRA perfect arterial CA bolus in the most distal MRA station from the diaphragm down to the feet is performed, it is without venous overlay. Method number 2 suffers even nearly impossible to avoid venous enhancement in the more from this problem because one additional station calves. Despite the implementation of faster acquisition is acquired. To get sufficiently long CA bolus a dosage of

5.4  MR Angiography Fig. 5.4.56  To avoid venous enhancement in the lower extremity venous compression of the thigh can be applied. Therefore cuffs are positioned proximal to the knee and inflated to subdiastolic values. Cushions are placed directly below the hollow of the knee and the ankle to avoid compression of the calf

0.1 mmol/kg body weight is diluted to 60 ml with saline. This amount then is injected in a biphasic matter, the second part with a decreased flow-rate to get an expanded arterial CA bolus. Nevertheless, in this imaging protocol venous enhancement may already occur in MRA of the abdominal aorta. CA is injected and first the carotid arteries are scanned. While this data acquisition runs CA is already distributed in the lower body part. Renal veins show a fast back current and may already restrict the delineation of the renal arteries. An alternative to further accelerate data acquisition to avoid venous overlay in the more distal MRA stations is to reduce image quality in terms of spatial resolution. This is not acceptable when evaluating renal arteries or distal calf arteries for atherosclerotic changes. A procedure to reduce venous enhancement in the lower extremity, especially in the calves, is venous compression in the thigh. Here a cuff is positioned at the middle of the thigh and inflated to a subdiastolic pressure (60–70 mmHg). This avoids rapid venous backflow, especially of the superficial veins (Herborn et al. 2004a) (Fig. 5.4.56). Method number 3 is currently the most promising approach to whole-body MRA. Certainly this method needs dedicated hardware but it offers excellent image quality in all MRA stations, with good arterial signal compared to background and without disturbing venous enhancement (Fenchel et al. 2006b; Kramer et al. 2005). Because of the large range of table movement of dedicated whole-body MR systems between 185 and 205 cm (e.g., Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany), the patient need not be repositioned.

In combination with dedicated matrix-coil systems and multiple receiver channels these MR systems offer maximal flexibility, not only for MRA but also for detection of tumors or metastases as well as inflammatory changes as they occur in patients suffering from rheumatism or diabetes (Weckbach 2006). Before starting the exam all needed coils are placed on the patient and selectively chosen when they are needed during the exam. After injection of the first CA bolus the arteries of head and neck are imaged first. To find the perfect timing for starting the measurement a test-bolus technique or MR fluoroscopy can be used. The time needed for data acquisition of this first MRA station and table movement to the most distal MRA station (calves) is shorter than is the transit time of the CA bolus down to the feet. So it is possible to image the arteries of the calves in an arterial phase as well. Because in this most distal MRA station neither a test-bolus nor an MR fluoroscopic approach to CA arrival is possible, it is recommended that two acquisition phases of this region be performed because it is more likely that the scan will begin too early than too late and that arterial CA filling will be missed (Kramer et al. 2005). The second part of this approach to whole-body MRA works in nearly the same way; this time the abdominal aorta is imaged first. Again one of the common techniques for timing to start MRA data acquisition is used. After the measurement ends the table automatically moves to the next station and MRA of the thigh is done (Fig. 5.4.57). When using this technique it is not necessary to shorten scan times to avoid venous enhancement and therefore spatial resolution of 1 × 1 × 1 mm3 can be reached in the head and neck as well as in the lower-leg MRA station. Resolution of the abdominal MRA can be reduced to 1.6 × 1 × 1.5 mm3. Acquisition time for each station does not exceed a maximum of 26 s. The last-mentioned method is not yet clinically implemented, and only initial results are available. This method is also only possible with MR systems that offer the required table movement. It does not matter if a dedicated whole-body MR system or a rolling platform system is utilized as long as a continuous table movement over the needed range is provided. The used sequence consists of acquisition of the central parts of k-space more often than the periphery of k-space. This leads to a high spatial resolution without venous enhancement (Willinek et al. 2002). In Table 5.3.16 exemplary whole-body MRA examination protocols are given for different whole-body MRA techniques. The intracranial arterial system is assessed using 3D TOF sequences. Imaging of the arteries from the base of the skull to the foot is performed by CE MRA using 3D FLASH sequences in four to five subsequent stations. Using surface-coil technology the achieved spatial resolution of whole-body MRA is comparable to that

849

850

5  Thorax and Vasculature Table 5.4.16  Exemplary whole-body MRA examination protocols are given for different whole-body MRA techniques AngioSURF

a

Conventional b MRI system

Multiple receiver channel MRI system

RF-receiver method

Rolling table platform with one single body surface coil

Partial body coverage with few surface coils

Complete body coverage with multiple surface coils

Patient positioning

Feet first, no repositioning

(1) Head first and (2) feet first after repositioning

Head first, no repositioning

Examination performance

5 Subsequent stations: supraclavicular, thorax, abdomen/pelvis, upper legs, lower legs

5 Stations, 2 imaging blocks: I: carotids (1 station) II: abdomen, thigh, calf, foot (4 subsequent stations)

4 Station, 2 imaging blocks: I: lower thorax/abdomen, pelvis/upper leg, lower leg (3 subsequent stations) II: upper thorax/carotids (1 station)

Intravenous contrast

0.5 M Gd-based

0.5 M Gd-based

0.5 M Gd-based

Test bolus

1 ml at 1.3 ml/s (CM)

2 ml at 1.5 ml/s (CM) (each for [1], [2])

2 ml at 2 ml/s (CM) 18 ml at 2 ml/s (NaCl) 25 ml at 1 ml/s (NaCl)

Total volume

20 ml at 1.3 ml/s (CM) +20 ml at 0.7 ml/s (CM) +30 ml(s) at 0.7 ml/s (NaCl)

Carotid arteries: 15 ml at 1.5 ml/s (CM) +25 ml at 1.5 ml/s (NaCl) Abdomen-legs: 10 ml at 1.5 ml/s (CM) +15 ml at 0.7 ml/s (CM) +25 ml at 0.7 ml/s (NaCl)

Supra-aortic vessels: 10 ml at 2 ml/s (CM) +10 ml at 2 ml/s (NaCl) +25 ml at 1 ml/s (NaCl) Thorax-legs: 20 ml at 2 ml/s (CM) +8 ml at 1 ml/s (CM) +20 ml at 1 ml/s (NaCl)

Sequence

Brain: 3D TOF Body: CE 3D FLASH

Brain: 3D TOF Body: CE 3D FLASH (PI 2)

Brain: 3D TOF Body: CE 3D FLASH (PI 2)

TR(ms)/ TE(ms)/α

2.1/0.7/20°

3.4/1.14/25° (I) 3.4/1.14/20° (II)

I(1): 3.11/1.14/25° I(3–4): 3.46/1.21/25° II: 2.85/1.11/25°

Spatial resolution (CE 3D FLASH)

1.74 × 1.56 × 4 mm3

I: 0.9 × 1.7 × 1.3 mm3 II(1–3): ≤ 1.6 × 1 × 1.5 mm3

I(2+3): ≤ 1.6 × 1.0 × 1.5 mm3 I(4): 1.6 × 1 × 1.2 mm3 II: 1.3 × 1 × 1.5 mm3

MRA

CM contrast medium, CE contrast enhanced a AngioSURF technology (Goyen 2002) b Conventional MRI system with eight receiver channels (Kramer 2005) c Multiple receiver channel technology (total imaging matrix [TIM], Siemens, Erlangen, Germany) (Kramer 2005; Fenchel 2006). In Kramer (2005) whole-body MRA is performed in two separate blocks: first, carotids/calf/foot and second, abdomen/thigh

c

5.4  MR Angiography Fig. 5.4.57  Whole-body MRA acquired on a dedicated whole-body MR system in combination with matrix coils and parallel imaging (PI). Spatial resolution is increased to 1 mm3 in the head and neck vessels as well as 2.4 mm3 in the abdomen. Total acquisition time for all four stations is 88 s

vasculature has to be mentioned. But there are certainly some more systemic vascular diseases like Marfan syndrome (Sect. 5.4.5.2.8) and Takayasu’s arteritis (Sect. 5.4.5.2.9) that can easily be assessed by whole-body MRA. Another new indication for whole-body MRI in general and MRA in particular is disease-specific imaging. It is well known that long-standing diabetes mellitus leads to pathologic vascular changes that often are not diagnosed until they become symptomatic because of concomitant neuropathic changes. Whole-body MRA seems to be very helpful because of the systemic nature of many vessel threatening diseases. Only one circumscribed alteration of a vessel may become symptomatic. But in most cases there are multiple other pathologic changes at different vessel stations. Imaging of the complete arterial vasculature has an important impact on treatment planning. Therefore it is very important to do whole-body at an excellent quality in terms of spatial resolution, image quality, and reproducibility. References 1.

2.

3.

4.

achieved in conventional MRA of one dedicated body region. Venous overlap, however, may be problematic, particularly in the lower-leg region. For achieving optimal arterial enhancement by avoiding concurrent venous overlap, different imaging strategies with sophisticated contrast media injection protocols are applied after test bolus timing, whereas one single or two separate contrast media injections with generally biphasic injection protocols are applied depending on the particular whole-body MRA technique used. CE images are as usual subtracted from baseline images to increase the vessel-to-background contrast. With all these methods to perform whole-body 3D CE MRA having become available, several indications for this kind of examination have arisen. First of all screening for atherosclerotic changes in the complete arterial

5.

6.

7.

8.

Abolmaali ND, Hietschold V, Appold S, Ebert W, Vogl TJ (2002) Gadomer-17-enhanced 3D navigator-echo MR angiography of the pulmonary arteries in pigs. Eur Radiol 12:692–697 Adams WM, Laitt RD, Jackson A (2000) The role of MR angiography in the pretreatment assessment of intracranial aneurysms: a comparative study. AJNR Am J Neuroradiol 21:1618–1628 Amann M, Bock M, Floemer F, Schönberg SO, Schad LR (2002) Three-dimensional spiral MR imaging: application to renal multiphase contrast-enhanced angiography. Magn Reson Med 48:290–296 Anzalone N, Scomazzoni F, Castellano R et al. (2005) Carotid artery stenosis: intraindividual correlations of 3D time-of-flight MR angiography, contrast-enhanced MR angiography, conventional DSA, and rotational angiography for detection and grading. Radiology 236:204–213 Arnold SM, Strecker R, Scheffler K et al. (2003) Dynamic contrast enhancement of paragangliomas of the head and neck: evaluation with time-resolved 2D MR projection angiography. Eur Radiol 13:1608–1611 Auer A, Felber S, Schmidauer C, Waldenberger P, Aichner F (1998) Magnetic resonance angiographic and clinical features of extracranial vertebral artery dissection. J Neurol Neurosurg Psychiatry 64:474–481 Aumann S, Schönberg SO, Just A et al. (2003) Quantification of renal perfusion using an intravascular contrast agent (part 1): Results in a canine model. Magn Reson Med 49:276–287 Aviv RI, Benseler SM, Silverman ED et al. (2006) MR imaging and angiography of primary CNS vasculitis of childhood. AJNR Am J Neuroradiol 27:192–199

851

852

5  Thorax and Vasculature 9.

10. 11. 12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

Bachmann R, Nassenstein I, Kooijman H et al. (2006) Spontaneous acute dissection of the internal carotid artery: high-resolution magnetic resonance imaging at 3.0 Tesla with a dedicated surface coil. Invest Radiol 41:105–111 Barer D (1998) Endarterectomy for recently symptomatic carotid stenosis. Lancet 352:143–144 Barkhausen J (2006) New imaging protocols for peripheral MRA. In: ECR 2. Vienna, p 500 Barnett HJ, Taylor DW, Eliasziw M et al. (1998) Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 339:1415–1425 Baumgartner RW, Mattle HP, Aaslid R (1995) Transcranial color-coded duplex sonography, magnetic resonance angiography, and computed tomography angiography: methods, applications, advantages, and limitations. J Clin Ultrasound 23:89–111 Becker BN, Odorico JS, Becker YT et al. (1999) Peripheral vascular disease and renal transplant artery stenosis: a reappraisal of transplant renovascular disease. Clin Transplant 13:349–355 Berg R van den, Wasser MN, van Gils AP, van der Mey AG, Hermans J, van Buchem MA (2000) Vascularization of head and neck paragangliomas: comparison of three MR angiographic techniques with digital subtraction angiography. AJNR Am J Neuroradiol 21:162–170 Berg R van den, Schepers A, de Bruine FT et al. (2004a) The value of MR angiography techniques in the detection of head and neck paragangliomas. Eur J Radiol 52:240–245 Berg R van den, Verbist BM, Mertens BJ, van der Mey AG, van Buchem MA (2004b) Head and neck paragangliomas: improved tumor detection using contrast-enhanced 3D time-of-flight MR angiography as compared with fat-suppressed MR imaging techniques AJNR Am J Neuroradiol 25:863–870 Berg R van den, (2005) Imaging and management of head and neck paragangliomas. Eur Radiol 15:1310–1318 Bezooijen R, van den Bosch HC, Tielbeek AV et al. (2004) Peripheral arterial disease: sensitivity-encoded multiposition MR angiography compared with intraarterial angiography and conventional multiposition MR angiography. Radiology 231:263–271 Blatter DD, Parker DL, Robison RO (1991) Cerebral MR angiography with multiple overlapping thin slab acquisition. Part I, Quantitative analysis of vessel visibility. Radiology 179:805–811 Bley TA, Wieben O, Uhl M, Thiel J, Schmidt D, Langer M (2005a) High-resolution MRI in giant cell arteritis: imaging of the wall of the superficial temporal artery. AJR Am J Roentgenol 184:283–287 Bley TA, Wieben O, Uhl M et al. (2005b) Integrated headthoracic vascular MRI at 3 T: assessment of cranial, cervical and thoracic involvement of giant cell arteritis. Magma 18:193–200

23. Bock M, Schönberg SO, Floemer F, Schad LR (2000) Separation of arteries and veins in 3D MR angiography using correlation analysis. Magn Reson Med 43:481–487 24. Bogousslavsky J, Regli F (1987) Ischemic stroke in adults younger than 30 years of age. Cause and prognosis. Arch Neurol 44:479–482 25. Bongartz GM, Boos M, Winter K, Ott H, Scheffler K, Steinbrich W (1997) Clinical utility of contrast-enhanced MR angiography. Eur Radiol 7 Suppl 5:178–186 26. Born M, Willinek WA, Gieseke J, von Falkenhausen M, Schild H, Kuhl CK (2005) Sensitivity encoding (SENSE) for contrast-enhanced 3D MR angiography of the abdominal arteries. J Magn Reson Imaging 22:559–565 27. Bosmans H, Wilms G, Marchal G, Demaerel P, Baert AL (1995) Characterisation of intracranial aneurysms with MR angiography. Neuroradiology 37:262–266 28. Brierly J, Redington A, Anderson (2001) Coarctation of the aorta and interrupted aortic arch. In: RH A (ed) Paediatric cardiology, 2nd edn. Churchill Livingstone, Edinburgh, pp 1523–1551 29. Brugieres P, Blustajn J, Le Guerinel C, Meder JF, Thomas P, Gaston A (1998) Magnetic resonance angiography of giant intracranial aneurysms. Neuroradiology 40:96–102 30. Bruijn SF de, Stam J, Kappelle LJ (1996) Thunderclap headache as first symptom of cerebral venous sinus thrombosis. CVST Study Group. Lancet 348:1623–1625 31. Buturovic-Ponikvar J (2003) Renal transplant artery stenosis. Nephrol Dial Transplant 18 Suppl 5:v74–77 32. Carr JC, Shaibani A, Russell E, Finn JP (2001) Contrastenhanced magnetic resonance angiography of the carotid circulation. Top Magn Reson Imaging 12:349–357 33. Carr JC, Finn JP (2003) MR imaging of the thoracic aorta. Magn Reson Imaging Clin N Am 11:135–148 34. Carrell TW, Burnand KG, Wells GM, Clements JM, Smith A (2002) Stromelysin-1 (matrix metalloproteinase-3) and tissue inhibitor of metalloproteinase-3 are overexpressed in the wall of abdominal aortic aneurysms. Circulation 105:477–482 35. Carroll TJ, Korosec FR, Petermann GM, Grist TM, Turski PA (2001) Carotid bifurcation: evaluation of time-resolved three-dimensional contrast-enhanced MR angiography. Radiology 220:525–532 36. Cavagna E, D’Andrea P, Schiavon F, Tarroni G (2000) Failing hemodialysis arteriovenous fistula and percutaneous treatment: imaging with CT, MRI and digital subtraction angiography. Cardiovasc Intervent Radiol 23:262–265 37. Chan YL, Leung CB, Yu SC, Yeung DK, Li PK (2001) Comparison of non-breath-hold high resolution gadoliniumenhanced MRA with digital subtraction angiography in the evaluation on allograft renal artery stenosis. Clin Radiol 56:127–132 38. Charon JP, Milne W, Sheppard DG, Houston JG (2004) Evaluation of MR angiographic technique in the assessment of thoracic outlet syndrome. Clin Radiol 59:588–595

5.4  MR Angiography 39. Cinat ME, Pham H, Vo D, Gordon I, Wilson SE (1999) Improved imaging of carotid artery bifurcation using helical computed tomographic angiography. Ann Vasc Surg 13:178–183 40. Clouse WD, Hallett JW Jr, Schaff HV, Gayari MM, Ilstrup DM, Melton LJ III (1998) Improved prognosis of thoracic aortic aneurysms: a population-based study. JAMA 280:1926–1929 41. Coady MA, Rizzo JA, Goldstein LJ, Elefteriades JA (1999) Natural history, pathogenesis, and etiology of thoracic aortic aneurysms and dissections. Cardiol Clin 17:615–635; vii 42. Connolly JO, Higgins RM, Walters HL et al. (1994) Presentation, clinical features and outcome in different patterns of atherosclerotic renovascular disease. Qjm 87:413–421 43. Crivello MS, Porter DH, Kim D, Critchlow JF, Scoutt L (1986) Isolated external iliac artery aneurysm secondary to cystic medial necrosis. Cardiovasc Intervent Radiol 9:139–141 44. Dagirmanjian A, Ross JS, Obuchowski N et al. (1995) High resolution, magnetization transfer saturation, variable flip angle, time-of-flight MRA in the detection of intracranial vascular stenoses. J Comput Assist Tomogr 19:700–706 45. Derdeyn CP, Powers WJ, Moran CJ, Cross DT III, Allen BT (1995) Role of Doppler US in screening for carotid atherosclerotic disease. Radiology 197:635–643 46. Derdeyn CP, Graves VB, Turski PA, Masaryk AM, Strother CM (1997) MR angiography of saccular aneurysms after treatment with Guglielmi detachable coils: preliminary experience. AJNR Am J Neuroradiol 18:279–286 47. Derex L, Nighoghossian N, Hermier M, Adeleine P, Froment JC, Trouillas P (2002) Early detection of cerebral arterial occlusion on magnetic resonance angiography: predictive value of the baseline NIHSS score and impact on neurological outcome. Cerebrovasc Dis 13:225–229 48. Dormandy J, Heeck L, Vig S (1999) Lower-extremity arteriosclerosis as a reflection of a systemic process: implications for concomitant coronary and carotid disease. Semin Vasc Surg 12:118–122 49. Du YP, Parker DL, Davis WL, Cao G (1994) Reduction of partial-volume artifacts with zero-filled interpolation in three-dimensional MR angiography. J Magn Reson Imaging 4:733–741 50. Dumoulin CL, Hart HR Jr (1986) Magnetic resonance angiography. Radiology 161:717–720 51. Dumoulin CL, Souza SP, Walker MF, Wagle W (1989) Three-dimensional phase contrast angiography. Magn Reson Med 9:139–149 52. Dumoulin CL, Yucel EK, Vock P et al. (1990) Two- and three-dimensional phase contrast MR angiography of the abdomen. J Comput Assist Tomogr 14:779–784 53. Duna GF, Calabrese LH (1995) Limitations of invasive modalities in the diagnosis of primary angiitis of the central nervous system. J Rheumatol 22:662–667

54. Duran M, Schönberg SO, Yuh WT, Knopp MV, van Kaick G, Essig M (2002) Cerebral arteriovenous malformations: morphologic evaluation by ultrashort 3D gadolinium-enhanced MR angiography. Eur Radiol 12:2957–2964 55. Dymarkowski S, Bosmans H, Marchal G, Bogaert J (1999) Three-dimensional MR angiography in the evaluation of thoracic outlet syndrome. AJR Am J Roentgenol 173:1005–1008 56. Edelman RR, Siewert B, Adamis M, Gaa J, Laub G, Wielopolski P (1994) Signal targeting with alternating radiofrequency (STAR) sequences: application to MR angiography. Magn Reson Med 31:233–238 57. Ekelund L, Sjoqvist L, Thuomas KA, Asberg B (1996) MR angiography of abdominal and peripheral arteries. Techniques and clinical applications. Acta Radiol 37:3–13 58. Essig M, Engenhart R, Knopp MV et al. (1996) Cerebral arteriovenous malformations: improved nidus demarcation by means of dynamic tagging MR-angiography. Magn Reson Imaging 14:227–233 59. Essig M, Reichenbach JR, Schad LR, Schönberg SO, Debus J, Kaiser WA (1999) High-resolution MR venography of cerebral arteriovenous malformations. Magn Reson Imaging 17:1417–1425 60. Evans AL, Coley SC, Wilkinson ID, Griffiths PD (2005) First-line investigation of acute intracerebral hemorrhage using dynamic magnetic resonance angiography. Acta Radiol 46:625–630 61. Farb RI, Scott JN, Willinsky RA, Montanera WJ, Wright GA, terBrugge KG (2003) Intracranial venous system: gadolinium-enhanced three-dimensional MR venography with auto-triggered elliptic centric-ordered sequence—initial experience. Radiology 226:203–209 62. Farb RI, Nag S, Scott JN et al. (2005) Surveillance of intracranial aneurysms treated with detachable coils: a comparison of MRA techniques. Neuroradiology 47:507–515 63. Fattori R, Celletti F, Descovich B et al. (1998) Evolution of post-traumatic aortic aneurysm in the subacute phase: magnetic resonance imaging follow-up as a support of the surgical timing. Eur J Cardiothorac Surg 13:582–586; discussion 586–587 64. Fattori R, Napoli G, Lovato L et al. (2003) Descending thoracic aortic diseases: stent-graft repair. Radiology 229:176–183 65. Fattori R, Nienaber CA (1999) MRI of acute and chronic aortic pathology: pre-operative and postoperative evaluation. J Magn Reson Imaging 10:741–750 66. Fellner C, Lang W, Janka R, Wutke R, Bautz W, Fellner FA (2005) Magnetic resonance angiography of the carotid arteries using three different techniques: accuracy compared with intraarterial X-ray angiography and endarterectomy specimens. J Magn Reson Imaging 21:424–431 67. Fellner FA, Requardt M, Lang W, Fellner C, Bautz W, Cavallaro A (2003) Peripheral vessels: MR angiography with dedicated phased-array coil with large-field-of-view adapter feasibility study. Radiology 228:284–289

853

854

5  Thorax and Vasculature 68. Fenchel M, Nael K, Ruehm S, Finn JP, Miller S, Laub G (2006a) Isotropic high spatial resolution magnetic resonance angiography of the supra-aortic arteries using twodimensional parallel imaging (iPAT2) at 3 Tesla: a feasibility study. Invest Radiol 41:545–552 69. Fenchel M, Scheule AM, Stauder NI et al. (2006b) Atherosclerotic disease: whole-body cardiovascular imaging with MR system with 32 receiver channels and total-body surface coil technology—initial clinical results. Radiology 238:280–291 70. Ferrari VA, Scott CH, Holland GA, Axel L, Sutton MS (2001) Ultrafast three-dimensional contrast-enhanced magnetic resonance angiography and imaging in the diagnosis of partial anomalous pulmonary venous drainage. J Am Coll Cardiol 37:1120–1128 71. Festa P, Ait-Ali L, Cerillo AG, De Marchi D, Murzi B. Magnetic resonance imaging is the diagnostic tool of choice in the preoperative evaluation of patients with partial anomalous pulmonary venous return. Int J Cardiovasc Imaging 2006 72. Fink C, Bock M, Puderbach M, Schmahl A, Delorme S (2003a) Partially parallel three-dimensional magnetic resonance imaging for the assessment of lung perfusion—initial results. Invest Radiol 38:482–488 73. Fink C, Schmaehl A, Bock M, Tuengerthal S, Delorme S (2003b) Images in cardiovascular medicine. Pulmonary vein stenosis after radiofrequency ablation for atrial fibrillation: image findings with multiphasic pulmonary magnetic resonance angiography. Circulation 107:e129–130 74. Fink C, Risse F, Buhmann R et al. (2004a) Quantitative analysis of pulmonary perfusion using time-resolved parallel 3D MRI—initial results. RoFo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 176:170–174 75. Fink C, Ley S, Puderbach M, Plathow C, Bock M, Kauczor HU (2004c) 3D pulmonary perfusion MRI and MR angiography of pulmonary embolism in pigs after a single injection of a blood pool MR contrast agent. Eur Radiol 14:1291–1296 76. Fink C, Puderbach M, Bock M et al. (2004b) Regional lung perfusion: assessment with partially parallel three-dimensional MR imaging. Radiology 231:175–184 77. Fink C, Puderbach M, Ley S et al. (2005a) Intraindividual comparison of 1.0 M gadobutrol and 0.5 M gadopentetate dimeglumine for time-resolved contrast-enhanced threedimensional magnetic resonance angiography of the upper torso. J Magn Reson Imaging 22:286–290 78. Fink C, Ley S, Risse F et al. (2005b) Effect of inspiratory and expiratory breathhold on pulmonary perfusion: assessment by pulmonary perfusion magnetic resonance imaging. Invest Radiol 40:72–79 79. Fink C, Kauczor HU (2006) Clinical applications of MRI in respiratory diseases. In: Beckmann N (ed) In vivo MR techniques in drug discovery and development. CRC Press, Boca Raton, pp 378–389

80. Fishman AP (1063) Dynamic of the pulmonary circulation. In: Hamilton WF (ed) Handbook of physiology. American Physiological Society, Washington D.C., p 1708 81. Forsting M (2005) CTA of the ICA bifurcation and intracranial vessels. Eur Radiol 15 Suppl 4:D25–27 82. Fukui M (1997) Guidelines for the diagnosis and treatment of spontaneous occlusion of the circle of Willis (“moyamoya” disease). Research Committee on Spontaneous Occlusion of the Circle of Willis (Moyamoya Disease) of the Ministry of Health and Welfare, Japan. Clin Neurol Neurosurg 99 Suppl 2:S238–240 83. Fushimi Y, Miki Y, Kikuta K et al. (2006) Comparison of 3.0- and 1.5-T three-dimensional time-of-flight MR angiography in moyamoya disease: preliminary experience. Radiology 239:232–237 84. Gandy SJ, Almahri A, Armoogum K, Sudarshan TA, Milne W, Houston JG (2004) Perfusion parameters of MR renography are associated with cardiovascular disease risk factors and clinical indices of kidney function. In: Annual Conference of the International Society for Magnetic Resonance in Medicine. Kyoto 85. Gauvrit JY, Leclerc X, Caron S, Taschner CA, Lejeune JP, Pruvo JP (2006) Intracranial aneurysms treated with Guglielmi detachable coils: imaging follow-up with contrastenhanced MR angiography. Stroke 37:1033–1037 86. Germain DP (2002) Clinical and genetic features of vascular Ehlers-Danlos syndrome. Ann Vasc Surg 16:391–397 87. Gibbs GF, Huston J III, Bernstein MA, Riederer SJ, Brown RD Jr (2004) Improved image quality of intracranial aneurysms: 3.0-T versus 1.5-T time-of-flight MR angiography. AJNR Am J Neuroradiol 25:84–87 88. Golay X, Brown SJ, Itoh R, Melhem ER (2001) Time-resolved contrast-enhanced carotid MR angiography using sensitivity encoding (SENSE). AJNR Am J Neuroradiol 22:1615–1619 89. Goldfarb JW (2004) The SENSE ghost: field-of-view restrictions for SENSE imaging. J Magn Reson Imaging 20:1046–1051 90. Gonner F, Heid O, Remonda L et al. (1998) MR angiography with ultrashort echo time in cerebral aneurysms treated with Guglielmi detachable coils. AJNR Am J Neuroradiol 19:1324–1328 91. Gottschalk S, Gaebel C, Haendler G et al. (2002) [Contrast-enhanced intracranial 3 D MR angiography (CE MRA) in assessing arterial stenoses and aneurysms]. RoFo 174:704–713 92. Goyen M, Laub G, Ladd ME et al. (2001a) Dynamic 3D MR angiography of the pulmonary arteries in under four s. J Magn Reson Imaging 13:372–377 93. Goyen M, Debatin JF, Ruehm SG (2001b) Peripheral magnetic resonance angiography. Top Magn Reson Imaging 12:327–335 94. Goyen M, Ruehm SG, Barkhausen J et al. (2001c) Improved multi-station peripheral MR angiography with a dedicated vascular coil. J Magn Reson Imaging 13:475–480

5.4  MR Angiography 95. Goyen M, Quick HH, Debatin JF et al. (2002) Wholebody three-dimensional MR angiography with a rolling table platform: initial clinical experience. Radiology 224:270–277 96. Goyen M, Herborn CU, Kroger K, Lauenstein TC, Debatin JF, Ruehm SG (2003) Detection of atherosclerosis: systemic imaging for systemic disease with whole-body three-dimensional MR angiography—initial experience. Radiology 227:277–282 97. Greil GF, Powell AJ, Gildein HP, Geva T (2002) Gadolinium-enhanced three-dimensional magnetic resonance angiography of pulmonary and systemic venous anomalies. J Am Coll Cardiol 39:335–341 98. Griffiths PD, Hoggard N, Warren DJ, Wilkinson ID, Anderson B, Romanowski CA (2000) Brain arteriovenous malformations: assessment with dynamic MR digital subtraction angiography. AJNR Am J Neuroradiol 21:1892–1899 99. Griswold MA, Jakob PM, Heidemann RM et al. (2002) Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47:1202–1210 100. Griswold MA, Kannengiesser S, Heidemann RM, Wang J, Jakob PM (2004) Field-of-view limitations in parallel imaging. Magn Reson Med 52:1118–1126 101. Gupta A, Frazer CK, Ferguson JM et al. (1999) Acute pulmonary embolism: diagnosis with MR angiography. Radiology 210:353–359 102. Hagspiel KD, Spinosa DJ, Angle JF, Matsumoto AH (2000) Diagnosis of vascular compression at the thoracic outlet using gadolinium-enhanced high-resolution ultrafast MR angiography in abduction and adduction. Cardiovasc Intervent Radiol 23:152–154 103. Han KM, Duijm LE, Thelissen GR et al. (2003) Failing hemodialysis access grafts: evaluation of complete vascular tree with 3D contrast-enhanced MR angiography with high spatial resolution: initial results in 10 patients. Radiology 227:601–605 104. Hasuo K, Mihara F, Matsushima T (1998) MRI and MR angiography in moyamoya disease. J Magn Reson Imaging 8:762–766 105. Haust MD (1987) Arterial involvement in genetic diseases. Am J Cardiovasc Pathol 1:231–285 106. Heidemann RM, Griswold MA, Muller M et al. (2004) Feasibilities and limitations of high field parallel MRI. Radiologe 44:49–55 107. Heiserman JE, Drayer BP, Fram EK et al. (1992) Carotid artery stenosis: clinical efficacy of two-dimensional timeof-flight MR angiography. Radiology 182:761–768 108. Henderson RD, Eliasziw M, Fox AJ, Rothwell PM, Barnett HJ (2000) Angiographically defined collateral circulation and risk of stroke in patients with severe carotid artery stenosis. North American Symptomatic Carotid Endarterectomy Trial (NASCET) Group. Stroke 31:128–132 109. Herborn CU, Goyen M, Lauenstein TC, Debatin JF, Ruehm SG, Kroger K (2003) Comprehensive time-resolved MRI of peripheral vascular malformations. AJR Am J Roentgenol 181:729–735

110. Herborn CU, Ajaj W, Goyen M, Massing S, Ruehm SG, Debatin JF (2004a) Peripheral vasculature: whole-body MR angiography with midfemoral venous compression—initial experience. Radiology 230:872–878 111. Herborn CU, Goyen M, Quick HH et al. (2004b) Wholebody 3D MR angiography of patients with peripheral arterial occlusive disease. AJR Am J Roentgenol 182:1427–1434 112. Ho KY, de Haan MW, Kessels AG, Kitslaar PJ, van Engelshoven JM (1998) Peripheral vascular tree stenoses: detection with subtracted and nonsubtracted MR angiography. Radiology 206:673–681 113. Hoksbergen AW, Majoie CB, Hulsmans FJ, Legemate DA (2003) Assessment of the collateral function of the circle of Willis: three-dimensional time-of-flight MR angiography compared with transcranial color-coded duplex sonography. AJNR Am J Neuroradiol 24:456–462 114. Howard G, Baker WH, Chambless LE, Howard VJ, Jones AM, Toole JF (1996) An approach for the use of Doppler ultrasound as a screening tool for hemodynamically significant stenosis (despite heterogeneity of Doppler performance). A multicenter experience. Asymptomatic Carotid Atherosclerosis Study Investigators. Stroke 27:1951–1957 115. Hu HH, Madhuranthakam AJ, Kruger DG, Huston J III, Riederer SJ (2004) Improved venous suppression and spatial resolution with SENSE in elliptical centric 3D contrastenhanced MR angiography. Magn Reson Med 52:761–765 116. Ilgit ET, Vural M, Oguz A, Ozdogan ME (1999) Peripheral arterial involvement in neurofibromatosis type 1—a case report. Angiology 50:955–958 117. Janka R, Fellner F, Fellner C et al. (2000) Dedicated phasedarray coil for peripheral MRA. Eur Radiol 10:1745–1749 118. Jeong JY, Kim SH, Lee HJ, Sim JS (2002) Atypical low-signal-intensity renal parenchyma: causes and patterns. Radiographics 22:833–846 119. Jeremias A, Huegel H, Lee DP et al. (2000) Spatial orientation of atherosclerotic plaque in non-branching coronary artery segments. Atherosclerosis 152:209–215 120. Johnson MB, Wilkinson ID, Wattam J, Venables GS, Griffiths PD (2000) Comparison of Doppler ultrasound, magnetic resonance angiographic techniques and catheter angiography in evaluation of carotid stenosis. Clin Radiol 55:912–920 121. Julsrud PR, Breen JF, Felmlee JP, Warnes CA, Connolly HM, Schaff HV (1997) Coarctation of the aorta: collateral flow assessment with phase-contrast MR angiography. AJR Am J Roentgenol 169:1735–1742 122. Kahara VJ, Seppanen SK, Ryymin PS, Mattila P, Kuurne T, Laasonen EM (1999) MR angiography with three-dimensional time-of-flight and targeted maximum-intensity-projection reconstructions in the follow-up of intracranial aneurysms embolized with Guglielmi detachable coils. AJNR Am J Neuroradiol 20:1470–1475 123. Keller E, Flacke S, Gieseke J et al. (1997) Kraniozervikale Dissektionen: Untersuchungsstrategien in der MR-Tomographie und MR-Angiographie. RoFo 167:565–571

855

856

5  Thorax and Vasculature 124. Khurshid I, Downie GH (2002) Pulmonary arteriovenous malformation. Postgrad Med J 78:191–197 125. Kirchhof K, Welzel T, Jansen O, Sartor K (2002) More reliable noninvasive visualization of the cerebral veins and dural sinuses: comparison of three MR angiographic techniques. Radiology 224:804–810 126. Kitanaka C, Tanaka J, Kuwahara M, Teraoka A (1994) Magnetic resonance imaging study of intracranial vertebrobasilar artery dissections. Stroke 25:571–575 127. Klufas RA, Hsu L, Barnes PD, Patel MR, Schwartz RB (1995) Dissection of the carotid and vertebral arteries: imaging with MR angiography. AJR Am J Roentgenol 164:673–677 128. Kluge A, Dill T, Ekinci O et al. (2004) Decreased pulmonary perfusion in pulmonary vein stenosis after radiofrequency ablation: assessment with dynamic magnetic resonance perfusion imaging. Chest 126:428–437 129. Ko GY, Byun JY, Choi BG, Cho SH (2000) The vascular manifestations of Behçet’s disease: angiographic and CT findings. Br J Radiol 73:1270–1274 130. Koelemay MJ, Nederkoorn PJ, Reitsma JB, Majoie CB (2004) Systematic review of computed tomographic angiography for assessment of carotid artery disease. Stroke 35:2306–2312 131. Konner K, Nonnast-Daniel B, Ritz E (2003) The arteriovenous fistula. J Am Soc Nephrol 14:1669–1680 132. Korosec FR, Frayne R, Grist TM, Mistretta CA (1996) Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 36:345–351 133. Kramer H, Schönberg SO, Nikolaou K et al. (2005) Cardiovascular screening with parallel imaging techniques and a whole-body MR imager. Radiology 236:300–310 134. Kransdorf MJ, Turner-Stepahin S, Merritt WH (1998) Magnetic resonance angiography of the hand and wrist: evaluation of patients with severe ischemic disease. J Reconstr Microsurg 14:77–81 135. Kreitner KF, Ley S, Kauczor HU et al. (2004) Chronic thromboembolic pulmonary hypertension: pre- and postoperative assessment with breath-hold MR imaging techniques. Radiology 232:535–543 136. Krings T, Hans FJ, Moller-Hartmann W et al. (2002) Timeof-flight-, phase contrast and contrast enhanced magnetic resonance angiography for pre-interventional determination of aneurysm size, configuration, and neck morphology in an aneurysm model in rabbits. Neurosci Lett 326:46–50 137. Krinsky G, Rofsky NM (1998) MR angiography of the aortic arch vessels and upper extremities. Magn Reson Imaging Clin N Am 6:269–292 138. Kuntz KM, Skillman JJ, Whittemore AD, Kent KC (1995) Carotid endarterectomy in asymptomatic patients—is contrast angiography necessary? A morbidity analysis. J Vasc Surg 22:706–714; discussion 714–706 139. Laghi A, Iannaccone R, Catalano C, Passariello R (2001) Multislice spiral computed tomography angiography of mesenteric arteries. Lancet 358:638–639

140. Leclerc X, Lucas C, Godefroy O et al. (1999) Preliminary experience using contrast-enhanced MR angiography to assess vertebral artery structure for the follow-up of suspected dissection. AJNR Am J Neuroradiol 20:1482–1490 141. Lee VS, Morgan JN, Tan AG et al. (2003) Celiac artery compression by the median arcuate ligament: a pitfall of end-expiratory MR imaging. Radiology 228:437–442 142. Leiner T (2005) Magnetic resonance angiography of abdominal and lower extremity vasculature. Top Magn Reson Imaging 16:21–66 143. Leiner T, Gerretsen S, Botnar R et al. (2005a) Magnetic resonance imaging of atherosclerosis. Eur Radiol 15:1087–1099 144. Leiner T, Kessels AG, Nelemans PJ et al. (2005b) Peripheral arterial disease: comparison of color duplex US and contrast-enhanced MR angiography for diagnosis. Radi­ olo­gy 235:699–708 145. Leseche G, Schaetz A, Arrive L, Nussaume O, Andreassian B (1992) Diagnosis and management of 17 consecutive patients with inflammatory abdominal aortic aneurysm. Am J Surg 164:39–44 146. Levy C, Laissy JP, Raveau V et al. (1994) Carotid and vertebral artery dissections: three-dimensional time-of-flight MR angiography and MR imaging versus conventional angiography. Radiology 190:97–103 147. Ley S, Fink C, Zaporozhan J et al. Value of high spatial and high temporal resolution magnetic resonance angiography for differentiation between idiopathic and thromboembolic pulmonary hypertension: initial results. Eur Radiol 2005 148. Ley S, Kauczor HU, Heussel CP et al. (2003) Value of contrast-enhanced MR angiography and helical CT angiography in chronic thromboembolic pulmonary hypertension. Eur Radiol 13:2365–2371 149. Ley S, Kreitner KF, Fink C, Heussel CP, Borst MM, Kauczor HU (2004) Assessment of pulmonary hypertension by CT and MR imaging. Eur Radiol 14:359–368 150. Litt AW, Eidelman EM, Pinto RS et al. (1991) Diagnosis of carotid artery stenosis: comparison of 2DFT time-of-flight MR angiography with contrast angiography in 50 patients. AJNR Am J Neuroradiol 12:149–154 151. Lovblad KO, Schneider J, Bassetti C et al. (2002) Fast contrast-enhanced MR whole-brain venography. Neuroradiology 44:681–688 152. Ma SK, Kim SW, Kim NH, Choi KC (2002) Renal vein and inferior vena cava thrombosis associated with acute pancreatitis. Nephron 92:475–477 153. Maki DD, Siegelman ES, Roberts DA, Baum RA, Gefter WB (2001) Pulmonary arteriovenous malformations: three-dimensional gadolinium-enhanced MR angiography-initial experience. Radiology 219:243–246 154. Markl M, Uhl M, Wieben O et al. (2006) High resolution 3 T MRI for the assessment of cervical and superficial cranial arteries in giant cell arteritis. J Magn Reson Imaging

5.4  MR Angiography 155. Marks MP, Napel S, Jordan JE, Enzmann DR (1993) Diagnosis of carotid artery disease: preliminary experience with maximum-intensity-projection spiral CT angiography. AJR Am J Roentgenol 160:1267–1271 156. Mathews VP, Elster AD, King JC, Ulmer JL, Hamilton CA, Strottmann JM (1995) Combined effects of magnetization transfer and gadolinium in cranial MR imaging and MR angiography. AJR Am J Roentgenol 164:169–172 157. Matsubara TK, Fong HM, Burns CM (1990) Computed tomography of abdomen (CTA) in management of blunt abdominal trauma. J Trauma 30:410–414 158. Meaney JF, Weg JG, Chenevert TL, Stafford-Johnson D, Hamilton BH, Prince MR (1997) Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 336:1422–1427 159. Meissner OA, Rieger J, Weber C et al. (2005) Critical limb ischemia: hybrid MR angiography compared with DSA. Radiology 235:308–318 160. Mermuys KP, Vanhoenacker PK, Chappel P, Van Hoe L (2005) Three-dimensional venography of the brain with a volumetric interpolated sequence. Radiology 234:901–908 161. Michaely HJ, Herrmann KA, Kramer H, Laub G, Reiser MF, Schönberg SO (2004a) The significance of MR angiography for the diagnosis of carotid stenoses. Radiologe 44:975–984 162. Michaely HJ, Schönberg SO, Ittrich C, Dikow R, Bock M, Guenther M (2004b) Renal disease: value of functional magnetic resonance imaging with flow and perfusion measurements. Invest Radiol 39:698–705 163. Michaely HJ, Schönberg SO, Oesingmann N, Reiser MF (2004c) Assessment of renal transplants using functional magnetic resonance. In: MRA Workshop. London, Ont., p111 164. Michaely HJ, Nael K, Schönberg SO et al. (2005a) The feasibility of spatial high-resolution magnetic resonance angiography (MRA) of the renal arteries at 3.0 T. RoFo 177:800–804 165. Michaely HJ, Schönberg SO, Rieger JR, Reiser MF (2005b) MR angiography in patients with renal disease. Magn Reson Imaging Clin N Am 13:131–151 166. Michaely HJ, Schönberg SO, Oesingmann N et al. (2006a) Renal artery stenosis: functional assessment with dynamic MR perfusion measurements—feasibility study. Radiology 238:586–596 167. Michaely HJ, Nael K, Schönberg SO et al. (2006b) Renal perfusion: Comparison of SR-TurboFLASH measurements at 1.5T with SR-TurboFLASH and TREAT (time-resolved echo-shared angiographic technique) measurements at 3 T. J Magn Reson Imaging 24:1413–1419 168. Michaely HJ, Schönberg SO, Oesingmann N et al. (2006c) Functional assessment of renal artery stenosis using dynamic MR perfusion measurements - feasibility. Radiology 238:586–596

169. Michaely HJ, Herrmann KA, Kramer H, Dietrich O, Reiser MF, Schönberg SO (2006d) High-resolution renal MRA: comparison of image quality and vessel depiction with different parallel imaging acceleration factors. J Magn Reson Imaging 24(1):95–100 170. Michaely HJ, Herrmann KA, Dietrich O, Reiser MF, Schönberg SO (2007) Quantitative and qualitative characterization of vascularization and hemodynamics in head and neck tumors with a 3D magnetic resonance time-resolved echo-shared angiographic technique (TREAT) - Initial Results. European Radiology 17(4):1101–1110 171. Missouris CG, Buckenham T, Cappuccio FP, MacGregor GA (1994) Renal artery stenosis: a common and important problem in patients with peripheral vascular disease. Am J Med 96:10–14 172. Mitra D, Connolly D, Jenkins S et al. (2006) Comparison of image quality, diagnostic confidence and interobserver variability in contrast enhanced MR angiography and 2D time of flight angiography in evaluation of carotid stenosis. Br J Radiol 79:201–207 173. Modaresi KB, Cox TC, Summers PE et al. (1999) Comparison of intra-arterial digital subtraction angiography, magnetic resonance angiography and duplex ultrasonography for measuring carotid artery stenosis. Br J Surg 86:1422–1426 174. Mohiaddin RH, Schoser K, Amanuma M, Burman ED, Longmore DB (1990) MR imaging of age-related dimensional changes of thoracic aorta. J Comput Assist Tomogr 14:748–752 175. Molinari F, Fink C, Risse F, Tuengerthal S, Bonomo L, Kauczor HU (2006) Assessment of differential pulmonary blood flow using perfusion magnetic resonance imaging: comparison with radionuclide perfusion scintigraphy. Invest Radiol 41:624–630 176. Nael K, Laub G, Finn JP (2005) Three-dimensional contrast-enhanced MR angiography of the thoraco-abdominal vessels. Magn Reson Imaging Clin N Am 13:359–380 177. Nael K, Ruehm SG, Michaely HJ et al. (2006a) High spatialresolution CE MRA of the carotid circulation with parallel imaging: comparison of image quality between 2 different acceleration factors at 3.0. Tesla Invest Radiol 41:391–399 178. Nael K, Michaely HJ, Villablanca P, Salamon N, Laub G, Finn JP (2006b) Time-resolved contrast enhanced magnetic resonance angiography of the head and neck at 3.0 Tesla: initial results Invest Radiol 41:116–124 179. Nagaraja S, Capener D, Coley SC et al. (2005) Brain arteriovenous malformations: measurement of nidal volume using a combination of static and dynamic magnetic resonance angiography techniques. Neuroradiology 47:387–392 180. Nagele T, Klose U, Grodd W, Petersen D, Tintera J (1994) The effects of linearly increasing flip angles on 3D inflow MR angiography. Magn Reson Med 31:561–566 181. Nederkoorn PJ, Elgersma OE, van der Graaf Y, Eikelboom BC, Kappelle LJ, Mali WP (2003a) Carotid artery stenosis: accuracy of contrast-enhanced MR angiography for diagnosis Radiology 228:677–682

857

858

5  Thorax and Vasculature 182. Nederkoorn PJ, van der Graaf Y, Hunink MG (2003b) Duplex ultrasound and magnetic resonance angiography compared with digital subtraction angiography in carotid artery stenosis: a systematic review Stroke 34:1324–1332 183. Nikolaou K, Schönberg SO, Attenberger U et al. (2005) Pulmonary Arterial Hypertension: Diagnosis with Fast Perfusion MR Imaging and High-Spatial-Resolution MR Angiography—Preliminary Experience. Radiology 236:694–703 184. Nikolaou K, Schönberg SO, Brix G et al. (2004) Quantification of pulmonary blood flow and volume in healthy volunteers by dynamic contrast-enhanced magnetic resonance imaging using a parallel imaging technique. Invest Radiol 39:537–545 185. Nollen GJ, van Schijndel KE, Timmermans J et al. (2003) Magnetic resonance imaging of the main pulmonary artery: reliable assessment of dimensions in Marfan patients on a simple axial spin echo image. Int J Cardiovasc Imaging 19:141–147; discussion 149–150 186. O’Duffy JD (1990a) Vasculitis in Behçet’s disease Rheum Dis Clin North Am 16:423–431 187. O’Duffy JD (1990b) Behçet’s syndrome N Engl J Med 322:326–328 188. Oelerich M, Lentschig MG, Zunker P, Reimer P, Rummeny EJ, Schuierer G (1998) Intracranial vascular stenosis and occlusion: comparison of 3D time-of-flight and 3D phasecontrast MR angiography. Neuroradiology 40:567–573 189. Ohno Y, Hatabu H, Takenaka D, Adachi S, Hirota S, Sugimura K (2002) Contrast-enhanced MR perfusion imaging and MR angiography: utility for management of pulmonary arteriovenous malformations for embolotherapy. Eur J Radiol 41:136–146 190. Ohno Y, Kawamitsu H, Higashino T et al. (2003) Time-resolved contrast-enhanced pulmonary MR angiography using sensitivity encoding (SENSE). J Magn Reson Imaging 17:330–336 191. Ohno Y, Hatabu H, Murase K et al. (2004a) Quantitative assessment of regional pulmonary perfusion in the entire lung using three-dimensional ultrafast dynamic contrastenhanced magnetic resonance imaging: Preliminary experience in 40 subjects. J Magn Reson Imaging 20:353–365 192. Ohno Y, Higashino T, Takenaka D et al. (2004b) MR angiography with sensitivity encoding (SENSE) for suspected pulmonary embolism: comparison with MDCT and ventilation-perfusion scintigraphy. AJR Am J Roentgenol 183:91–98 193. Olin JW (1994) Thromboangiitis obliterans. Curr Opin Rheumatol 6:44–49 194. Olin JW (2000) Thromboangiitis obliterans (Buerger’s disease). N Engl J Med 343:864–869 195. Olin JW (2004) Renal artery disease: diagnosis and management. Mt Sinai J Med 71:73–85 196. Olin JW, Young JR, Graor RA, Ruschhaupt WF, Bartholomew JR (1990) The changing clinical spectrum of thromboangiitis obliterans (Buerger’s disease). Circulation 82:IV3–8

197. Oudkerk M, van Beek EJ, Wielopolski P et al. (2002) Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 359:1643–1647 198. Ozsarlak O, Van Goethem JW, Maes M, Parizel PM (2004) MR angiography of the intracranial vessels: technical aspects and clinical applications. Neuroradiology 46:955–972 199. Ozsvath RR, Casey SO, Lustrin ES, Alberico RA, Hassankhani A, Patel M (1997) Cerebral venography: comparison of CT and MR projection venography. AJR Am J Roentgenol 169:1699–1707 200. Paddon AJ, Nicholson AA, Ettles DF, Travis SJ, Dyet JF (2000) Long-term follow-Up of percutaneous balloon angioplasty in adult aortic coarctation. Cardiovasc Intervent Radiol 23:364–367 201. Patel MR, Kuntz KM, Klufas RA et al. (1995) Preoperative assessment of the carotid bifurcation. Can magnetic resonance angiography and duplex ultrasonography replace contrast arteriography? Stroke 26:1753–1758 202. Pereles FS, McCarthy RM, Baskaran V et al. (2002) Thoracic aortic dissection and aneurysm: evaluation with nonenhanced true FISP MR angiography in less than 4 minutes. Radiology 223:270–274 203. Pierot L, Delcourt C, Bouquigny F et al. (2006) Follow-up of intracranial aneurysms selectively treated with coils: Prospective evaluation of contrast-enhanced MR angiography. AJNR Am J Neuroradiol 27:744–749 204. Planken RN, Tordoir JH, Dammers R et al. (2003) Stenosis detection in forearm hemodialysis arteriovenous fistulae by multiphase contrast-enhanced magnetic resonance angiography: preliminary experience. J Magn Reson Imaging 17:54–64 205. Polak JF, Bajakian RL, O’Leary DH, Anderson MR, Donaldson MC, Jolesz FA (1992) Detection of internal carotid artery stenosis: comparison of MR angiography, color Doppler sonography, and arteriography. Radiology 182:35–40 206. Polak JF, Kalina P, Donaldson MC, O’Leary DH, Whittemore AD, Mannick JA (1993) Carotid endarterectomy: preoperative evaluation of candidates with combined Doppler sonography and MR angiography. Work in progress. Radiology 186:333–338 207. Potchen EJ (1992) Magnetic resonance angiography (MRA). Semin Ultrasound CT MR 13:225–226 208. Pozniak MA, Balison DJ, Lee FT Jr, Tambeaux RH, Uehling DT, Moon TD (1998) CT angiography of potential renal transplant donors. Radiographics 18:565–587 209. Prince MR, Yucel EK, Kaufman JA, Harrison DC, Geller SC (1993) Dynamic gadolinium-enhanced three-dimensional abdominal MR arteriography. J Magn Reson Imaging 3:877–881 210. Prince MR, Narasimham DL, Stanley JC et al. (1995) Breath-hold gadolinium-enhanced MR angiography of the abdominal aorta and its major branches. Radiology 197:785–792

5.4  MR Angiography 211. Prince MR, Schönberg SO, Ward JS, Londy FJ, Wakefield TW, Stanley JC (1997) Hemodynamically significant atherosclerotic renal artery stenosis: MR angiographic features. Radiology 205:128–136 212. Prince MR, Grist TM, Debatin JF (2003) 3D contrast MR angiography. Berlin; New York: Springer-Verlag 213. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42:952–962 214. Pui MH (2004) Cerebral MR venography. Clin Imaging 28:85–89 215. Radermacher J, Chavan A, Bleck J et al. (2001) Use of Doppler ultrasonography to predict the outcome of therapy for renal-artery stenosis. N Engl J Med 344:410–417 216. Randoux B, Marro B, Koskas F, Chiras J, Dormont D, Marsault C (2003) Proximal great vessels of aortic arch: comparison of three-dimensional gadolinium-enhanced MR angiography and digital subtraction angiography. Radiology 229:697–702 217. Reddy KS, Yusuf S (1998) Emerging epidemic of cardiovascular disease in developing countries. Circulation 97:596–601 218. Renowden S (2004) Cerebral venous sinus thrombosis. Eur Radiol 14:215–226 219. Rubin GD (1997) Helical CT angiography of the thoracic aorta. J Thorac Imaging 12:128–149 220. Safian RD, Textor SC (2001) Renal-artery stenosis. N Engl J Med 344:431–442 221. Sankhla SK, Gunawardena WJ, Coutinho CM, Jones AP, Keogh AJ (1996) Magnetic resonance angiography in the management of aneurysmal subarachnoid haemorrhage: a study of 51 cases. Neuroradiology 38:724–729 222. Sawicki PT, Kaiser S, Heinemann L, Frenzel H, Berger M (1991) Prevalence of renal artery stenosis in diabetes mellitus—an autopsy study. J Intern Med 229:489–492 223. Schellinger PD, Jansen O, Fiebach JB et al. (2000) Feasibility and practicality of MR imaging of stroke in the management of hyperacute cerebral ischemia. AJNR Am J Neuroradiol 21:1184–1189 224. Schönberg SO, Knopp MV, Bock M et al. (1997) Renal artery stenosis: grading of hemodynamic changes with cine phase-contrast MR blood flow measurements. Radiology 203:45–53 225. Schönberg SO, Bock M, Floemer F et al. (1999a) High-resolution pulmonary arterio- and venography using multiple-bolus multiphase 3D-Gd-mRA. J Magn Reson Imaging 10:339–346 226. Schönberg SO, Bock M, Knopp MV et al. (1999b) Renal arteries: optimization of three-dimensional gadoliniumenhanced MR angiography with bolus-timing-independent fast multiphase acquisition in a single breath hold. Radiology 211:667–679 227. Schönberg SO, Wunsch C, Knopp MV et al. (1999c) Abdominal aortic aneurysm. Detection of multilevel vascular pathology by time-resolved multiphase 3D gadolinium MR angiography: initial report. Invest Radiol 34:648–659

228. Schönberg SO, Bock M, Kallinowski F, Just A (2000) Correlation of hemodynamic impact and morphologic degree of renal artery stenosis in a canine model. J Am Soc Nephrol 11:2190–2198 229. Schönberg SO, Knopp MV, Londy F et al. (2002) Morphologic and functional magnetic resonance imaging of renal artery stenosis: a multireader tricenter study. J Am Soc Nephrol 13:158–169 230. Schönberg SO, Rieger J, Nittka M, Dietrich O, Johannson LO, Reiser MF (2003a) Renal MR angiography: current debates and developments in imaging of renal artery stenosis. Semin Ultrasound CT MR 24:255–267 231. Schönberg SO, Aumann S, Just A et al. (2003b) Quantification of renal perfusion abnormalities using an intravascular contrast agent (part 2): Results in animals and humans with renal artery stenosis. Magn Reson Med 49:288–298 232. Schönberg SO, Rieger J, Weber CH et al. (2005) High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 235:687–698 233. Schoepf UJ, Costello P (2004) CT angiography for diagnosis of pulmonary embolism: state of the art. Radiology 230:329–337 234. Schuster DP, Kaplan JD, Gauvain K, Welch MJ, Markham J (1995) Measurement of regional pulmonary blood flow with PET. J Nucl Med 36:371–377 235. Scoble JE, Hamilton G (1990) Atherosclerotic renovascular disease. BMJ 300:1670–1671 236. Sebastia C, Quiroga S, Boye R, Cantarell C, Fernandez-Planas M, Alvarez A (2001) Helical CT in renal transplantation: normal findings and early and late complications. Radiographics 21:1103–1117 237. Shariat Razavi I, Stacul F, Cova M et al. (1998) Morphofunctional study of the kidney in patients with kidney disease and liver disease with magnetic resonance. Radiol Med (Torino) 95:72–81 238. Sheehy N, MacNally S, Smith CS, Boyle G, Madhavan P, Meaney JF (2005) Contrast-enhanced MR angiography of subclavian steal syndrome: value of the 2D time-of-flight “localizer” sign. AJR Am J Roentgenol 185:1069–1073 239. Slovut DP, Olin JW (2004) Fibromuscular dysplasia. N Engl J Med 350:1862–1871 240. Smits JH, Bos C, Elgersma OE et al. (2002) Hemodialysis access imaging: comparison of flow-interrupted contrastenhanced MR angiography and digital subtraction angiography. Radiology 225:829–834 241. Stella A, Gargiulo M, Faggioli GL et al. (1993) Postoperative course of inflammatory abdominal aortic aneurysms. Ann Vasc Surg 7:229–238 242. Strecker R, Scheffler K, Klisch J et al. (2000) Fast functional MRA using time-resolved projection MR angiography with correlation analysis. Magn Reson Med 43:303–309 243. Subramaniam M, Mizzi A, Roditi G (2004) Magnetic resonance angiography in potential live renal donors: a joint radiological and surgical evaluation. Clin Radiol 59:335–341

859

860

5  Thorax and Vasculature 244. Summers PE, Jarosz JM, Markus H (2001) Mr angiography in cerebrovascular disease. Clin Radiol 56:437–456 245. Suzuki J, Kodama N (1983) Moyamoya disease—a review. Stroke 14:104–109 246. Swan JS, Weber DM, Grist TM, Wojtowycz MM, Korosec FR, Mistretta CA (1992) Peripheral MR angiography with variable velocity encoding. Work in progress. Radiology 184:813–817 247. Tan KT, van Beek EJ, Brown PW, van Delden OM, Tijssen J, Ramsay LE (2002) Magnetic resonance angiography for the diagnosis of renal artery stenosis: a meta-analysis. Clin Radiol 57:617–624 248. Textor SC (2000) Epidemiology and clinical presentation. Semin Nephrol 20:426–431 249. Thiene G, Frescura C (1999) Etiology and pathology of aortic arch malformations. In: R NCF (ed) Diagnosis and treatment of aortic diseases. Kluwer, Dordrecht, pp 225–269 250. Tsuchiya K, Aoki C, Fujikawa A, Hachiya J (2004) Threedimensional MR digital subtraction angiography using parallel imaging and keyhole data sampling in cerebrovascular diseases: initial experience. Eur Radiol 14:1494–1497 251. Uehara T, Tabuchi M, Kozawa S, Mori E (2001) MR angiographic evaluation of carotid and intracranial arteries in Japanese patients scheduled for coronary artery bypass grafting. Cerebrovasc Dis 11:341–345 252. Unlu E, Cakir B, Gocer B, Tuncbilek N, Gedikoglu M (2005) The role of contrast-enhanced MR angiography in the assessment of recently ruptured intracranial aneurysms: a comparative study. Neuroradiology 47:780–791 253. Van Grimberge F, Dymarkowski S, Budts W, Bogaert J (2000) Role of magnetic resonance in the diagnosis of subclavian steal syndrome. J Magn Reson Imaging 12:339–342 254. Vasbinder GB, Nelemans PJ, Kessels AG et al. (2004) Accuracy of computed tomographic angiography and magnetic resonance angiography for diagnosing renal artery stenosis. Ann Intern Med 141:674–682 255. Vasbinder GB, Nelemans PJ, Kessels AG, Kroon AA, de Leeuw PW, van Engelshoven JM (2001) Diagnostic tests for renal artery stenosis in patients suspected of having renovascular hypertension: a meta-analysis. Ann Intern Med 135:401–411 256. Vasbinder GBC, Maki JH, Nijenhuis RJ et al. (2002) Motion of the distal renal artery during three-dimensional contrast-enhanced breath-hold MRA. J Magn Reson Imaging 16:685–696 257. Vogl TJ, Juergens M, Balzer JO et al. (1994) Glomus tumors of the skull base: combined use of MR angiography and spin-echo imaging (2002) Radiology 192:103–110 258. Vosshenrich R, Fischer U Contrast-enhanced MR angiography of abdominal vessels: is there still a role for angiography? Eur Radiol 12:218–230 259. Waldman GJ, Pattynama PM, Chang PC, Verburgh C, Reiber JH, de Roos A (1996) Magnetic resonance angiography of dialysis access shunts: initial results. Magn Reson Imaging 14:197–200

260. Wardlaw JM, Chappell FM, Best JJ, Wartolowska K, Berry E (2006) Non-invasive imaging compared with intra-arterial angiography in the diagnosis of symptomatic carotid stenosis: a meta-analysis. Lancet 367:1503–1512 261. Weber MA, Kroll A, Gunther M et al. (2004) [Noninvasive measurement of relative cerebral blood flow with the blood bolus MRI arterial spin labeling: basic physics and clinical applications]. Radiologe 44:164–173 262. Weckbach S (2006) Comprehensive diabetes imaging with whole body MR imaging at 1.5 and 3.0 T in patients with longstanding diabetes. In:ECR 2Vienna, 259 263. Wehrli FW, Shimakawa A, Gullberg GT, MacFall JR (1986) Time-of-flight MR flow imaging: selective saturation recovery with gradient refocusing. Radiology 160:781–785 264. Westphal M, Grzyska U (2000) Clinical significance of pedicle aneurysms on feeding vessels, especially those located in infratentorial arteriovenous malformations. J Neurosurg 92:995–1001 265. Wetzel S, Bongartz G (1999) MR angiography: supra-aortic vessels. Eur Radiol 9:1277–1284 266. Wetzel SG, Law M, Lee VS, Cha S, Johnson G, Nelson K (2003) Imaging of the intracranial venous system with a contrast-enhanced volumetric interpolated examination. Eur Radiol 13:1010–1018 267. White CS, Baffa JM, Haney PJ, Pace ME, Campbell AB (1997) MR imaging of congenital anomalies of the thoracic veins (2003) Radiographics 17:595–608 268. White PM, Wardlaw JM, Lindsay KW, Sloss S, Patel DK, Teasdale EM. The non-invasive detection of intracranial aneurysms: are neuroradiologists any better than other observers? Eur Radiol 13:389–396 269. Wilcken DE (2003) Overview of inherited metabolic disorders causing cardiovascular disease. J Inherit Metab Dis 26:245–257 270. Willig DS, Turski PA, Frayne R et al. (1998) Contrast-enhanced 3D MR DSA of the carotid artery bifurcation: preliminary study of comparison with unenhanced 2D and 3D time-of-flight MR angiography. Radiology 208:447–451 271. Willinek WA, Gieseke J, Conrad R et al. (2002) Randomly segmented central k-space ordering in high-spatial-resolution contrast-enhanced MR angiography of the supra-aortic arteries: initial experience. Radiology 225:583–588 272. Willinek WA, Born M, Simon B et al. (2003a) Time-offlight MR angiography: comparison of 3.0-T imaging and 1.5-T imaging—initial experience. Radiology 229:913–920 273. Willinek WA, Gieseke J, von Falkenhausen M, Neuen B, Schild HH, Kuhl CK (2003b) Sensitivity encoding for fast MR imaging of the brain in patients with stroke. Radiology 228:669–675 274. Wilman AH, Riederer SJ, Huston J III, Wald JT, Debbins JP (1998) Arterial phase carotid and vertebral artery imaging in 3D contrast-enhanced MR angiography by combining fluoroscopic triggering with an elliptical centric acquisition order. Magn Reson Med 40:24–35

5.4  MR Angiography 275. Winterer JT, Strey C, Wolffram C et al. (2000) Preoperative examination of potential kidney transplantation donors: value of gadolinium-enhanced 3D MR angiography in comparison with DSA and urography. RoFo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 172:449–457 276. Wong-You-Cheong JJ, Grumbach K, Krebs TL et al. (1998) Torsion of intraperitoneal renal transplants: imaging appearances. AJR Am J Roentgenol 171:1355–1359 277. Wright LB, Matchett WJ, Cruz CP et al. (2004) Popliteal artery disease: diagnosis and treatment. Radiographics 24:467–479 278. Xu C, Zarins CK, Glagov S (2001) Aneurysmal and occlusive atherosclerosis of the human abdominal aorta. J Vasc Surg 33:91–96 279. Yamada I, Suzuki S, Matsushima Y (1995) Moyamoya disease: comparison of assessment with MR angiography and MR imaging versus conventional angiography. Radiology 196:211–218 280. Yamada I, Nakagawa T, Matsushima Y, Shibuya H (2001) High-resolution turbo magnetic resonance angiography for diagnosis of Moyamoya disease. Stroke 32:1825–1831 281. Yamada N, Takamiya M, Kuribayashi S, Okita Y, Minatoya K, Tanaka R (2000) MRA of the Adamkiewicz artery: a preoperative study for thoracic aortic aneurysm. J Comput Assist Tomogr 24:362–368

282. Yamamoto T, Ogasawara Y, Kimura A et al. (1996) Blood velocity profiles in the human renal artery by Doppler ultrasound and their relationship to atherosclerosis. Arterioscler Thromb Vasc Biol 16:172–177 283. Yanaka K, Asakawa H, Matsumaru Y, Kujiraoka Y, Nose T (2004) Diagnosis of vascular compression at the thoracic outlet using magnetic resonance angiography. Eur Neurol 51:122–123 284. Yang CW, Lee SH, Choo SW et al. (1996) Early graft dysfunction due to renal vein compression. Nephron 73:480–481 285. Yang JJ, Hill MD, Morrish WF et al. (2002) Comparison of pre- and postcontrast 3D time-of-flight MR angiography for the evaluation of distal intracranial branch occlusions in acute ischemic stroke. AJNR Am J Neuroradiol 23:557–567 286. Zaporozhan J, Ley S, Eichinger M, Fink C (2005) Unklarer Befund im Thoraxröntgen. Radiologe 45:644–648 287. Zeller T, Frank U, Spath M, Roskamm H (2001) Color duplex ultrasound imaging of renal arteries and detection of hemodynamically relevant renal artery stenoses. Ultraschall Med 22:116–121 288. Zeller T, Frank U, Muller C et al. (2002) Duplex ultrasound for follow-up examination after stent-angioplasty of ostial renal artery stenoses. Ultraschall Med 23:315–319

861

Chapter 6

Abdomen and Retroperitoneum

6.1 

Abdominal MRI  .. . . . . . . . . . . . . . . . . . . . .   864 N.C. Balci, E. Altun, K. Hermann, and R.C. Semelka

6.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   864

6.1.2 

Imaging Technique  .. . . . . . . . . . . . . . . . . . .   864

6.1.3 

Liver  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   864

6

6. 1.8.5  Inflammation  .. . . . . . . . . . . . . . . . . . . . . . . .   905 6.1.9 

6.1.9.1  Gastric Carcinoma  . . . . . . . . . . . . . . . . . . . .   906 6.1.9.2  Crohn’s Disease  .. . . . . . . . . . . . . . . . . . . . . .   907 6.1.9.3  Colorectal Carcinoma   .. . . . . . . . . . . . . . . .   909 References  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   909

6.1.3.1  Normal Variations  . . . . . . . . . . . . . . . . . . . .   864 6.1.3.2  Diseases of the Hepatic Parenchyma  .. . .   864

6.2 

Kidneys, Adrenals, and Retroperitoneum  .. . . . . . . . . . . . . . . .   912 H.J. Michaely, M. Laniado, and S.O. Schönberg

6.2.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . . .   912

6.2.2 

General Examination Techniques  .. . . . . .   912

6.1.3.3  Diffuse Liver Parenchymal Diseases  . . . .   877 6.1.4 

Pancreas  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   886

6.1.4.1  Developmental Anomalies  .. . . . . . . . . . . .   886 6.1.4.2  Genetic Diseases  .. . . . . . . . . . . . . . . . . . . . .   886 6.1.4.3  Mass Lesions  . . . . . . . . . . . . . . . . . . . . . . . . .   886 6.1.4.4  Lymphoma  . . . . . . . . . . . . . . . . . . . . . . . . . . .   891 6.1.4.5  Metastases  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   891 6.1.4.6  Inflammatory Disease  . . . . . . . . . . . . . . . . .   891 6.1.5 

Gallbladder and Bile Ducts  . . . . . . . . . . . .   893

6.1.5.1  Gallbladder  .. . . . . . . . . . . . . . . . . . . . . . . . . .   893 6.1.6 

Bile Ducts  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   895

6.1.6.1  Benign Diseases  .. . . . . . . . . . . . . . . . . . . . . .   895 6.1.6.2  Mass Lesions  . . . . . . . . . . . . . . . . . . . . . . . . .   896 6.1.7 

Spleen  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   897

6.1.7.1  Normal Variants   and Congenital Diseases  .. . . . . . . . . . . . . .   897 6.1.7.2  Mass Lesions  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   6.1.7.2  Mass Lesions  . . . . . . . . . . . . . . . . . . . . . . . . .   899 6.1.8 

Peritoneum  .. . . . . . . . . . . . . . . . . . . . . . . . . .   903

6.1.8.1  Hernias  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   903 6.1.8.2  Benign Masses  .. . . . . . . . . . . . . . . . . . . . . . .   903 6.1.8.3  Malignant Masses  .. . . . . . . . . . . . . . . . . . . .   903 6.1.8.4  Intraperitoneal Fluid  . . . . . . . . . . . . . . . . . .   904

Gastrointestinal Tract  . . . . . . . . . . . . . . . . .   906

6.2.2.1  Patient Preparation and Positioning  . . . .   912 6.2.2.2  Choice of Coils  . . . . . . . . . . . . . . . . . . . . . . .   912 6.2.2.3  Imaging Planes  . . . . . . . . . . . . . . . . . . . . . . .   912 6.2.2.4  Sequences   .. . . . . . . . . . . . . . . . . . . . . . . . . . .   912 6.2.3 

Kidney  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   913

6.2.3.1  Dedicated Examination Techniques  .. . . .   913 6.2.3.2  Normal Anatomy and Developmental Anomalies  .. . . . . . . .   914 6.2.3.3  Pathologies  .. . . . . . . . . . . . . . . . . . . . . . . . . .   916 6.2.3.4  Differential Diagnosis  . . . . . . . . . . . . . . . . .   941 6.2.3.5  Clinical Value MRI in Comparison with Other Diagnostic Modalities  . . . . . .   941 6.2.3.6  Indications for Imaging  .. . . . . . . . . . . . . . .   943 6.2.4 

Adrenal Gland  .. . . . . . . . . . . . . . . . . . . . . . .   943

6.2.4.1  Dedicated Examination Technique  . . . . .   943 6.2.4.2  Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   944 6.2.4.3  Pathophysiology  . . . . . . . . . . . . . . . . . . . . . .   944 6.2.4.4  Pathology  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   944 6.2.4.5  Differential Diagnosis  . . . . . . . . . . . . . . . . .   951

864

6  Abdomen and Retroperitoneum 6.2.4.6  Value of MRI in Comparison with Other Imaging Modalities   . . . . . . . .   951 6.2.5 

Lymph Nodes and Retroperitoneal Tumors  . . . . . . . . . . .   953

6.2.5.1  Dedicated Examination Techniques  .. . . .   953 6.2.5.2  Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   953 6.2.5.3  Pathology  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   954 6.2.5.4  Value of MRI and Comparison with Other Imaging Modalities  .. . . . . . . .   958 6.2.5.5  Diagnostic Procedure  . . . . . . . . . . . . . . . . .   958

6.1  Abdominal MRI N.C. Balci, E. Altun, K. Hermann, and R.C. Semelka 6.1.1  Introduction MR imaging has achieved a level of maturation that has rendered it the most accurate imaging modality for investigating diseases of the abdomen. With new imaging techniques and correlative studies, MRI findings have come close to matching histopathological findings regarding abdominal disease. In this section, we present the current MRI techniques and findings for the most common abdominal disease entities. 6.1.2  Imaging Technique The standard imaging technique for the abdomen includes breath-hold sequences. T1-weighted images are obtained using a non-fat-suppressed spoiled gradientecho (SGE) sequence with dual-echo acquisition in phase and out of phase, and a fat-suppressed SGE sequence or three-dimensional gradient-echo (3D GE) sequence. T2weighted images are obtained using single-shot breathing-independent T2-weighted sequences (e.g., SS ETSE) with and without fat suppression. Dynamic contrast-enhanced images including hepatic arterial dominant, portal venous, and interstitial phases are acquired with the use of T1-weighted SGE or 3D GE sequences after injection of gadolinium. A phased-array body coil is used to obtain an optimal signal-to-noise ratio. In uncooperative patients, SGE may be modified as a single-shot technique using the minimum TR to achieve breathing-independent images. Such sequences have included so-called magnetization-prepared rapid-acquisition gradient-echo (MP RAGE) and turbo-fast low-angle shot (turboFLASH) sequences (Semelka et al. 1999a). Parallel imaging techniques can be used to further reduce the scanning time.

6.2.5.6  Indications  . . . . . . . . . . . . . . . . . . . . . . . . . . .   958 6.2.6 

Psoas Muscle  . . . . . . . . . . . . . . . . . . . . . . . . .   958

6.2.6.1  Dedicated Examination Techniques  .. . . .   958 6.2.6.2  Normal Anatomy  . . . . . . . . . . . . . . . . . . . . .   959 6.2.6.3  Pathology  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   959 6.2.6.4  Value of MRI in Comparison with Other Imaging Modalities  .. . . . . . . .   959 6.2.6.5  Diagnostic Procedure  . . . . . . . . . . . . . . . . .   959 Suggested Reading  . . . . . . . . . . . . . . . . . . . .   959

6.1.3  Liver 6.1.3.1  Normal Variations Common variations include horizontal elongation of the lateral segment of the left lobe, hypoplasia of the left lobe, and vertical elongation of the right lobe, termed the Riedel lobe. Hypoplasia of the left lobe does not generally result in diagnostic difficulties, although it may simulate a left hepatectomy, which clinical history readily establishes. Diaphragmatic insertions are not an uncommon finding along the lateral aspect of the liver. They tend to be multiple and closely related to overlying ribs, having wedge-shaped margins with the capsular surface of the liver. Insertions are low in signal on T2- and T1-weighted images. These features help to distinguish diaphragmatic insertions from peripheral mass lesions (Semelka et al. 1993a). 6.1.3.2  Diseases of the Hepatic Parenchyma 6.1.3.2.1  Benign Masses Solitary (Non-Parasitic) Cysts Hepatic cysts are common lesions and are usually divided into unilocular (95%) or multilocular varieties. On MR imaging, cysts are homogeneous, well-defined lesions that possess a sharp boundary with the liver. Occasionally, cysts are so closely grouped that they resemble a multicystic mass. Simple cysts are low in signal intensity on T1-weighted images, markedly high in signal intensity on T2-weighted images. Single-shot breathing-independent T2-weighted sequences (e.g., SS-ETSE) are especially effective at showing small (≤ 5 mm) cysts. Because cysts do not enhance with gadolinium on MR images, delayed post-gadolinium images (up to 5 min) may be useful to ensure that lesions are cysts and not poorly vascularized metastases that show gradual enhancement (Semelka et al. 1993a) (Fig 6.1.1).

6.1  Abdominal MRI

Fig. 6.1.1a–d  Simple hepatic cyst (arrows). Axial T2-weighted fat-suppressed single-shot echo-train spin-echo (a) image reveals hyperintense lesion in the right liver lobe. Post-contrast

axial T1-weighted spoiled gradient-echo hepatic arterial dominant- (a), portal venous– (b), and interstitial (c), phase images reveal no enhancement of the lesion

Ciliated hepatic foregut cysts are an uncommon type of solitary unilocular cyst. These congenital lesions are believed to arise from the embryonic foregut and to differentiate toward bronchial tissue in the liver. These cysts are most frequently located at the anterosuperior margin of the liver, superficially. On MRI, foregut cysts characteristically create a bulge in the liver contour and show marked hyperintensity on T2-weighted images; on T1weighted images, they range from hypo- to hyperintense. On post-contrast images, a subtly enhancing cyst wall is present (Shoenut et al. 1994a).

Biliary Hamartoma Biliary hamartomas (or von Meyenburg complexes) are benign biliary malformations, which are currently considered part of the spectrum of fibropolycystic diseases of the liver due to ductal plate malformation. On MR images, lesions are small (usually <1 cm) and well defined. The high fluid content gives these lesions markedly high signal intensity on T2-weighted images, low signal intensity on T1-weighted images, and faint rim enhancement on early and late post-gadolinium images (Semelka et al. 1999b) (Fig 6.1.3).

Autosomal Dominant Polycystic Kidney Disease In autosomal dominant polycystic kidney disease (ADPKD), the liver is the most common extrarenal organ in which cysts occur. It is estimated that up to 75% of patients with ADPKD have hepatic involvement. Although these cysts vary in number and size, they tend to be multiple and smaller than the renal cysts, measuring up to 4 cm. However, extensive hepatic replacement with large cysts has been described (Fig 6.1.2). Liver cysts in the setting of ADPKD exhibit the same MR features as simple cysts. Occasionally, hemorrhage may be observed in cysts (Mosetti et al. 2003).

Biliary Cystadenoma/Cystadenocarcinoma Although rare, benign and malignant cystic tumors of biliary origin may arise in the liver. There is a peak incidence of these lesions in the fifth decade, with a great predominance in women. Biliary cystadenomas are typically large and multiloculated and filled with clear or mucinous fluid. Mural nodules may be a component of some cysts. On imaging, these tumors frequently have solid enhancing nodules associated with cystic components (Palacios et al. 1990). Occasionally, mucin content renders these tumors high in signal intensity on T1-weighted images. Enhancing thick septa and mural nodules on post-con-

865

866

6  Abdomen and Retroperitoneum

Fig. 6.1.2a–d  Autosomal dominant polycystic kidney disease— liver involvement. Axial T2-weighted single-shot echo-train spin-echo image (a) demonstrates innumerable hyperintense simple cysts in the liver and the kidneys The cysts reveal no enhancement on the axial T1-weighted spoiled gradient-echo ar-

terial (b) and late-venous-phase (c) contrast-enhanced images. Coronal fat-suppressed three dimensional gradient-echo image (d) demonstrates the residual liver and renal parenchyma and the renal vessels (arrows)

trast images may be associated with cytadenocarcinoma (Palacios et al. 1990).

signal intensity on T1-weighted images, and exhibit a diffuse heterogeneous enhancement on immediate postgadolinium SGE images (Worawattanakul et al. 1996).

Angiomyolipomas Angiomyolipomas of the liver are uncommon benign mesenchymal tumors. The tumor is composed of mature fat, blood vessels, and smooth muscle. Angiomyolipomas are well-defined, sharply marginated masses, which frequently have a high fat content, and therefore are high in signal intensity on T1-weighted images and low in signal intensity on fat-suppressed images. Angiomyolipomas may also have a low fat content and appear moderately high in signal intensity on T2-weighted images, low in

Hemangiomas Hemangiomas are the most common benign hepatic neoplasm, with an autopsy incidence between 0.4 and 20%. Pathologically, hemangiomas are characterized grossly as well-circumscribed, sponge-like, blood-filled mesenchymal tumors. On MRI, hemangiomas have long T2 and T1 values, so they are markedly high in signal intensity on T2-weighted images and low in signal intensity on T1-weighted images,

6.1  Abdominal MRI

Fig. 6.1.3a–c  Biliary hamartomas. On axial T2-weighted fat suppressed single-shot echo-train spin-echo image (a), small multiple hyperintense lesions are scattered throughout the liver parenchyma. These small lesions reveal subtle peripheral enhancement on arterial (b) and late venous (c) phase contrast enhanced axial T1-weighted spoiled gradient-echo images

maintaining signal intensity on longer echo times (e.g., >120 ms). On dynamic serial post-contrast images, the most distinctive imaging feature of hemangiomas is the demonstration of a discontinuous ring of nodules immediately after contrast administration. Hemangiomas may fade away in signal intensity toward parenchyma isointensity over time, but they will fade in a homogeneous fashion with no evidence of peripheral or heterogeneous lesional washout (Semelka and Sofka 1997). Three types of enhancement patterns are observed in hemangiomas: Uniform high signal intensity immediately after contrast (type 1); peripheral nodular enhancement with centripetal progression to uniform high signal intensity (type 2) (Fig 6.1.4); and peripheral nodular enhancement with centripetal progression and a persistent central scar (type 3). Type 1 enhancement is observed only in small tumors (<1.5 cm); type 2 and type 3 enhancements are observed in all size categories. Type 3 enhancement is observed especially in larger tumors (>5.0 cm) (Fig 6.1.5) (Semelka and Sofka 1997). Hepatocellular Adenoma Hepatocellular adenomas (HCA) are benign epithelial neoplasms. Approximately 90% of HCAs occur in young

women. These lesions are associated with the use of oral contraceptive steroids or, less frequently, the use of anabolic steroids or abnormal carbohydrate metabolism, such as familial diabetes mellitus, galactosemia, and glycogen storage disease type Ia. Malignant transformation is sporadic. The typical MR appearance of HCA is homogeneously mild hyperintensity on T2-weighted images, homogeneously mild hypointensity or isointensity on T1-weighted images, and transient homogeneous blush immediately after contrast that uniformly fades to isointensity with liver parenchyma by 1 min. The intensity and heterogeneity of signal on T2- and T1-weighted images may vary and reflects the quantity of fat, hemorrhage, and necrosis within the tumor (Paulson et al. 1994). The enhancement pattern of HCA in arterial dominant phase images may resemble hepatocellular carcinoma (HCC). However, the presence of lesional washout and complete capsule enhancement on late-phase imaging suggests the diagnosis of HCC. HCA may decrease homogeneously in signal intensity on out-of-phase or fat-suppressed images due to their fat content, which is commonly uniform (Fig 6.1.6). HCA and focal nodular hyperplasia (FNH) have similar signal intensities on T2- and T1-weighted pre-contrast images and reveal similar enhancement pat-

867

868

6  Abdomen and Retroperitoneum

6.1  Abdominal MRI 9 Fig. 6.1.4a–g  Hepatic hemangioma. Axial T1-weighted SGE image (a) reveals hypointense lesion in the left hepatic lobe that is hyperintense on axial T2-weighted fat-saturated fast spinecho (b) and single-shot echo-train spin-echo (c) images. The lesion shows peripheral nodular enhancement and coalescence of the nodules in centripetal fashion in serial contrast-enhanced axial spoiled gradient-echo images (d–g)

terns on arterial dominant phase images. Moreover, FNH rarely contains fat or blood and is more likely to exist in the setting of fatty liver (Paulson et al. 1994). Focal Nodular Hyperplasia Focal nodular hyperplasia (FNH) is an uncommon lesion defined by a localized region of hyperplasia within otherwise normal liver and found predominantly in women during the third to fifth decades of life. In contrast to hepatocellular adenomas (HCA), FNH does not appear to have a clear-cut association with oral-contraceptive use. FNH is mildly hyperintense/isointense on T2-weighted images and mildly hypointense/isointense on T1-weighted images. FNH enhances with an intense uniform blush on immediate post-gadolinium images and fades rapidly to near isointensity. Small (<1.5 cm) FNHs are commonly isointense on all pre-contrast images and may be appreciated only on the immediate post-gadolinium images (Fig

Fig. 6.1.5a–f  Giant hepatic hemangioma. Two hypointense lesions are visualized on axial T1-weighted spoiled gradient-echo image (a), which are hyperintense on axial T2-weighted fat- saturated fast spin-echo (b) and single-shot echo-train spin-echo (c)

6.1.7). FNH may contain a central scar that is hyperintense on T2-weighted images. The central scar exhibits lack of enhancement on immediate post-gadolinium images, and enhances on late-phase images. In small FNH (<1.5 cm), the central scar is often not apparent. On imaging, background fatty liver is common in FNH. Fatty infiltration of FNH is very rare. There are several imaging features that help to distinguish between FNA and HCA. A central scar, which shows delayed enhancement is typical for FNH, whereas pseudocapsule, internal hemorrhage, focal necrosis, and intralesional fat are features more commonly noted in HCA (Mortelé et al. 2000). Hepatocyte-selective contrast agents may distinguish between HCA and atypical FNH, and hypervascular metastases and atypical FNHs. On Gd-EOB-DTPA and GdBOPTA enhanced images, FNH and HCA will show an early capillary blush but late hepatocellular uptake is only observed with FNH (Fig 6.1.8). 6.1.3.2.2 Malignant Masses Liver Metastases Metastases are moderately high in signal intensity on T2weighted images and moderately low in signal intensity on T1-weighted images in general. The degree of enhancement of the metastases depends on their vascularity. In order to analyze the degree of

images. On serial contrast-enhanced axial T1-weighted spoiled gradient-echo images, bigger lesion reveals type 3 enhancement; smaller lesion reveals complete centripetal fill in (d–f)

869

870

6  Abdomen and Retroperitoneum

Fig. 6.1.6a–e  Hepatocellular adenoma. On axial T2-weighted fat-suppressed single-shot echo-train spin-echo image (a), no focal liver lesion is identified. On axial T1-weighted outof-phase spoiled gradient-echo (SGE) image (b) focal signal loss (arrow) is observed compared to in-phase (SGE) image (c). This focal lesion reveals intense enhancement on SGE arterial phase image (d) and washout on portal venous-phase image (e)

6.1  Abdominal MRI

Fig. 6.1.7a–e  Focal nodular hyperplasia. Axial T2-weighted single-shot echo-train spin-echo image (a), T1-weighted fatsuppressed three dimensional gradient-echo (3D GE) images acquired on hepatic arterial dominant- and portal venousphase after intravenous application of gadolinium (b,c) and T2*-weighted GE images in axial (d) and coronal (e) planes after administration of super-paramagnetic iron-oxide particles

(SPIO): The well-defined mass is barely visible and isointense to liver on T2-weighted image (a). On immediate post-gadolinium image (b), there is a strong blush. On delayed image (c), the lesion fades away and persists as a slightly hyperintense lesion. There is clear uptake of SPIO within the lesion, inducing a signal loss almost to the same extent as in the surrounding normal liver tissue (d,e)

Fig. 6.1.8a–d  Atypical focal nodular hyperplasia. Dynamic post-contrast axial T1-weighted spoiled gradient-echo images 30 s (a), 100 s (b), 5 min (c), and 20 min (d) after the administration of a liver-specific contrast agent with inherent hepatocytespecific properties (hepatobiliary contrast agent gadoxetic acid, Primovist®, Schering, Germany). Two intrahepatic lesions are present, one of which involves almost the entire left liver lobe;

the second is situated in the right liver lobe. After pronounced contrast enhancement in the arterial dominant phase (a), these lesions preserve hyperintense signal until late phases after contrast application (b–d). This behavior is indicative of the presence of hepatocytes within the lesion, in this case atypical focal nodular hyperplasia

871

872

6  Abdomen and Retroperitoneum

enhancement on arterial dominant-phase images, it is crucial to determine if the sequence was acquired with the “perfect enhancement timing.” Perfect enhancement timing is considered to occur when contrast is seen in hepatic arteries and portal veins but not in hepatic veins (Semelka et al. 1994). Avascular metastases appear as completely cystic or necrotic metastases and include ovarian cancer and posttreatment (chemotherapy, chemoembolization, or ablation) metastases. They are characterized by high signal intensity on T2-weighted images, low signal intensity on T1-weighed images, and lack of enhancement on arterial dominant and interstitial phase images. A thin lesional or perilesional enhancement in the margin of the metastases is often demonstrated in one of the phases post-contrast. Hypovascular metastases are characterized by near isointensity or high signal intensity on T2-weighted images and low signal on T1-weighted images. After contrast, hypovascular metastases demonstrate minimal enhancement on arterial dominant-phase images, which tends to be more conspicuous on interstitial phase images (Semelka et al. 1994) (Fig 6.1.9). Primary tumors that commonly result in hypovascular metastases include colorectal carcinoma, transitional-cell carcinoma, pancreatic ductal adenocarcinoma, small bowel adenocar-

cinoma, pulmonary carcinoma, bladder carcinoma, and prostate carcinoma (Semelka et al. 1994). Isovascular metastases are characterized by lesional enhancement similar to background parenchyma on arterial dominant-phase images. On interstitial-phase images, isovascular metastases often, but not always, show a decrease in the degree of enhancement (washout), becoming more conspicuous. Isovascular metastases are generally well demonstrated on pre-contrast images, with high signal intensity on T2-weighted images or low signal intensity on T1-weighted images or both. This appearance is most often observed in metastases after chemotherapy, presumably reflecting an antiangiogenic effect. Most commonly, metastases from colon, thyroid, and endometrium may demonstrate isovascularity (Semelka et al. 1994). Hypervascular metastases are generally high in signal intensity on T2-weighted images, low in signal intensity on T1-weighted images, and possess a moderate or intense peripheral ring of enhancement on early-phase images, comparable with the extent of enhancement of the pancreas and/or renal cortex. On interstitial phase images these metastases are the most likely to show centripetal enhancement and peripheral washout (Semelka et al. 1994). Hypervascular metastases are more conspicuous on arterial dominant phase images due to the great

Fig. 6.1.9a–d  Hepatic metastasis from colon cancer. Axial nonenhanced T1-weighted spoiled gradient-echo image (a) and T1-weighted fat-suppressed three dimensional gradient-echo images after administration of a hepatocyte-specific contrast agent gadoxetic acid (Primovist®, Schering, Germany) in the arterial dominant phase (b), portal venous-phase (c) and delayed phase (d) (20 min p.i.): Three hypointense metastatic lesions are

appreciated before and after application of the hepatocyte-specific contrast agent. On the hepatic arterial dominant phase, the lesions show peripheral rim type enhancement. On the delayed phase, the lack of contrast uptake within the lesions is proof of the absence of hepatic tissue. Note the high signal intensity of the bile within the common hepatic duct due to the biliary excretion of the contrast agent

6.1  Abdominal MRI

Fig. 6.1.10a–g  Hepatic metastases from neuroendocrine tumor. Axial in-phase and out-of-phase T1-weighted spoiled gradient-echo (SGE) images (a,b), T2-weighted fast spin-echo imaging with fat saturation (c), pre-contrast T1-weighted SGE image with fat-saturation (d), contrast-enhanced T1-weighted fat-saturated three dimensional GE images in arterial dominant and late venous phase (e,f) and T2*-weighted GE image after administration of super-paramagnetic iron-oxide (SPIO) (g). Four irregularly shaped lesions are identified in both liver lobes exhibiting low signal intensity on T1-weighted pre-contrast images

(a,b,d) and moderately high signal intensity on T2-weighted image (c). All lesions show increased and pronounced contrast uptake in the arterial dominant phase (e) after gadolinium application indicating hypervascularity. On late venous phase (f), the lesions show fading and washout of the contrast medium and almost isointense signal behavior relative to the surrounding liver parenchyma. The lack of iron uptake after SPIO (g) rules out the presence of hepatocytes within the lesions and suggests metastatic disease

signal difference between intensely enhanced lesions and minimal enhancement of the background parenchyma. Small (<1.5 cm) hypervascular metastases are commonly homogeneously high in signal intensity on T2-weighted images, homogeneously low in signal intensity on T1weighted images, and show either fading to background or washout. Often, small hypervascular metastases (especially those <1.0 cm) are only evident on hepatic arterial dominant phase images, i.e., the lesion is isointense on T2- and T1-weighed images and interstitial phase postcontrast images (Semelka et al. 1994) (Fig 6.1.10). The malignancies that most commonly result in hypervascular liver metastases include breast cancer, renal cell carcinoma, carcinoid tumor, islet cell tumor, thyroid carcinoma, adenocarcinoma of unknown primary site, leiomyosarcoma, and malignant melanoma.

Hepatocellular Carcinoma Hepatocellular carcinoma (HCC) is the most common primary malignancy of the liver and usually develops in patients with cirrhosis. HCC does occur in the non-cirrhotic liver as well. Incidence of HCC is particularly high in patients with cirrhosis from chronic hepatitis C infection, chronic hepatitis B infection, and alcoholic liver disease (Kelekis et al. 1998). Small HCCs (<2 cm) are frequently isointense on T2-weighted images. Isointensity on T2-weighted images may correlate with well-differentiated HCC. Signal intensity on T1-weighted image varies from moderately low to moderately high. High signal intensity on T1weighted images may reflect the presence of fat or protein (Hussain et al. 2002). The most sensitive sequence for detecting small HCCs is hepatic arterial dominant-

873

874

6  Abdomen and Retroperitoneum

phase imaging, in which the majority of small tumors will enhance moderately; not uncommonly these tumors may only be apparent on this set of images (Hussain et al. 2002) (Fig 6.1.11). Intense enhancement revealed on early phase images is not specific to HCC, as high-grade dysplastic nodules (DNs) may show similar findings. Rarely, small HCCs may be hypovascular, shown by a minimal extent of enhancement on arterial dominantphase images (Hussain et al. 2002). Isovascular HCCs on arterial-phase images have been reported, and attention should be paid to interstitial-phase images that may show washout of the tumors with capsule enhancement (Hussain et al. 2002). Large HCC (>2 cm) may range from hypo- to hyperintense on T2- and T1-weighted images. The most frequent appearance is mildly high signal intensity on T2-weighted images and minimally low signal intensity on T1-weighted images (Hussain et al. 2002). The hyperintensity on T2-weighted images and hypointensity on T1-weigthed images are highly suggestive of moderately differentiated HCC (Hussain et al. 2002). In the interstitial phase, large HCCs tend to demonstrate washout below the signal intensity of background parenchyma, and capsule enhancement, which may only be apparent around some portions of the tumor (Hussain et al. 2002). In histopathological analysis, 60–87% of large HCCs have a fibrotic tumor capsule (Kelekis et al. 1998; Hussain et al. 2002). The typical signal intensity of a capsule is mildly hyperintense on T2-weighted images, hypointense on T1weighted images; negligible mild enhancement occurs on immediate post-gadolinium images, which becomes more intense on interstitial-phase images. Tumor extension occurs most commonly into portal veins; however, hepatic venous extension also occurs. Although tumor thrombus is observed in fewer than 50% of cases, it is common in the setting of large and advanced tumors. Higher doses of gadolinium or higher flow rates may improve visualization of HCCs that may possess minimal increased vascularity (Kelekis et al. 1998; Hussain et al. 2002). Newer contrast agents with higher T1 relaxivity may aid in lesion detection of hypovascular, isovascular, or minimally hypervascular tumors. Also, new techniques such as parallel imaging may improve detection of HCCs in debilitated patients unable to suspend breathing for 20 s with conventional spoiled gradient-echo sequences (Kelekis et al. 1998; Hussain et al. 2002). Diffuse HCC The most common appearance of diffuse infiltrative HCC is extensive hepatic parenchymal involvement with mottled punctate mildly to moderately high signal intensity on T2-weighted images and mildly to moderately low signal on T1-weighted images. A patchy or miliary pattern of enhancement is often observed on immediate post-gadolinium images with tumor washout and segments of late capsule enhancement on late images (Fig.

6.1.12). Diffuse HCC may also appear as irregular linear strands that are iso- to moderately hyperintense on T2weighted images and hypo- to isointense on T1-weighted images. On immediate post-gadolinium images these tumor strands tend to enhance variably. Late increased enhancement of the tumor strands may reflect a high fibrous composition. Portal vein thrombus is associated with diffuse HCC (Hussain et al. 2002). Tumor thrombus is commonly of high signal intensity on T2-weighted images and enhances with gadolinium; meanwhile, bland thrombus is low in signal intensity on T2-weighted images and does not enhance after gadolinium (Hussain et al. 2002). Fibrolamellar Carcinoma Fibrolamellar carcinoma is a distinct morphologic subtype of liver cell carcinoma. This tumor occurs in younger patients, frequently females, without underlying cirrhosis or chronic liver disease. On imaging, fibrolamellar carcinomas are generally large, solitary tumors that are heterogeneous and moderately high in signal intensity on T2weighted images, and heterogeneous and moderately low in signal intensity on T1-weighted images. Enhancement of the tumor is diffuse heterogeneous and moderately intense on immediate post-gadolinium images. A huge central scar with a radiating appearance is present. The central scar is variable in signal and has large low-signal components on T2-weighted images that enhance negligibly on delayed gadolinium enhanced images (Corrigan and Semelka 1995). On MR imaging, the scar is characterized by a complex arborizing pattern, radiating from a central focus and extending out to the tumor periphery. This profile is distinctly different from the appearance of FNH, in which the scar occupies a small central portion of the tumor and exhibits more uniform signal enhancement characteristics on late-phase images (Corrigan and Semelka 1995). Lymphoma Secondary involvement of the liver by Hodgkin’s and non-Hodgkin’s lymphoma is common in stage IV disease (Kelekis et al. 1997). On imaging, non-Hodgkin’s lymphoma more frequently results in focal hepatic lesions than does Hodgkin’s disease. Lesions vary in signal intensity from low to moderately high on T2-weighted images and are typically low in signal intensity on T1-weighted images. On post-gadolinium images, lesions that are low in signal intensity on T2-weighted images tend to enhance minimally, whereas lesions that are high in signal intensity tend to enhance in a substantial fashion (Kelekis et al. 1997). Enhancement on immediate post-gadolinium images usually is predominantly peripheral. Lesions of malignant lymphoma may possess transient, ill-defined perilesional enhancement on immediate post-gadolinium images independent of the degree of enhancement of the lesions themselves.

6.1  Abdominal MRI

Fig. 6.1.11a–e  Hepatocellular carcinoma (arrows). On axial T2-weighted fat-suppressed single-shot echo-train spin-echo image (a) a subcapsular nodule is identified. On axial T1weighted out-of-phase (b) and in-phase (c) spoiled gradientecho (SGE) images, the lesion is iso-hyperintense compared to the liver parenchyma. Arterial dominant-phase SGE image (d) reveals intense enhancement of the lesion, with washout on portal venous-phase image (e)

Primary hepatic lymphoma is considerably rarer than secondary involvement and histologically the majority is non-Hodgkin’s lymphomas. Most tumors are characterized grossly as a large solitary mass, but they may vary in appearance from multiple nodules to diffuse involve-

ment. Tumors are mild to moderately high in signal on T2-weighted images and moderately low in signal intensity on T1-weighted images, and show relatively diffuse heterogeneous enhancement on immediate post-gadolinium gradient-echo images.

875

876

6  Abdomen and Retroperitoneum

Fig. 6.1.12a–d  Diffuse hepatocellular carcinoma. On axial T2weighted fat-suppressed single-shot echo-train spin-echo (a) and T1-weighted spoiled gradient-echo (SGE) (b) images, most of the liver parenchyma is replaced by diffuse tumor involvement, which is isointense to hyperintense on T2-weighted image and hypointense on T1-weighted image. On axial T1-weighted

arterial phase SGE image (c), diffuse tumor involvement throughout the liver shows heterogeneous increased enhancement. On late-venous-phase axial SGE image (d), the diffuse tumor involvement demonstrates washout throughout the liver. On late-venous-phase SGE, thrombosis of the right portal vein (arrows) was detected

Intrahepatic or Peripheral Bile-Duct Carcinoma (Cholangiocarcinoma) Intrahepatic or peripheral cholangiocarcinoma are terms applied to lesions that originate in the ducts proximal to (i.e., above) the hilum of the liver. The tumor is frequently large at presentation. Cholangiocarcinoma resembles HCC with moderate high signal intensity on T2-weighted images and low signal intensity on T1-weighted images. Enhancement with gadolinium varies from minimal to intense diffuse heterogeneous enhancement immediately after contrast administration. Minimal enhancement is most commonly observed. Persistent enhancement on delayed images is relatively common (Hamrick et al. 1992) (Fig 6.1.13). Malignant tumors of mixed liver cell and bile duct differentiation are rare. Mixed HCC–cholangiocarcinoma may occur,

and the imaging appearance is generally indistinguishable from that of HCC. Angiosarcoma Angiosarcoma is the most common sarcoma arising in the liver, and accounts for 1.8% of all liver cancers. On MRI, angiosarcoma may have high signal intensity on T2weighted images and low signal intensity on T1-weighted images. The frequent presence of hemorrhage results in focal areas of low signal intensity on T2-weighted images and high signal intensity on T1-weighted images. After contrast, angiosarcoma may demonstrate peripheral nodular enhancement with centripetal progression, mimicking the appearance of hemangioma (Worawattanakul et al. 1997). Presence of hemorrhage may help to distinguish angiosarcoma form hemangioma.

6.1  Abdominal MRI

Fig. 6.1.13a–d  Peripheral cholangiocarcinoma. A big heterogenous mass is present in the left lobe of the liver on T2-weighted fat-suppressed single-shot echo-train spin-echo image (a) and on T1-weighted spoiled gradient-echo (SGE) image (b). The

mass enhances heterogeneously on arterial dominant (c) and late venous phases (d). The lesion demonstrates more enhancement on the late venous phase

Epithelioid Hemangioendothelioma Epithelioid hemangioendothelioma (EHE) is a malignant, slow-growing vascular tumor, usually occurring in middle-aged patients. Females predominate over males in a 2 : 1 ratio, and oral contraceptives have been implicated as possible causative agents in younger patients (Leonardou et al. 2002). Pathologically, lesions tend to be multiple, tough, fibrous masses distributed throughout the liver. EHE have moderately high signal intensity on T2-weighted images, moderately low signal intensity on T1-weighted images, and show a mild heterogeneous or moderately intense diffuse heterogeneous enhancement on early phase images (Leonardou et al. 2002).

casionally lobulated mass surrounded by a pseudocapsule. Although it is usually solitary, multiple lesions can be seen in less than 20% of cases. Areas of necrosis and calcifications are frequently present (Helmberger et al. 1999). On MR imaging, hepatoblastoma resembles hepatocellular carcinoma in that tumor shows diffuse heterogeneous enhancement on immediate post-gadolinium images.

Hepatoblastoma Hepatoblastoma is the most common primary malignant tumor of the liver in children and may occur from the newborn to adolescent period and rarely later. On gross inspection, hepatoblastoma is a solid, well-defined, oc-

6.1.3.3  Diffuse Liver Parenchymal Diseases 6.1.3.3.1  Chronic Liver Diseases Primary Biliary Cirrhosis Primary biliary cirrhosis (PBC) is a chronic progressive autoimmune liver disorder that causes the obliteration of the intrahepatic bile ducts, portal inflammation, fibrosis, and cirrhosis. A few imaging studies describe the imag-

877

878

6  Abdomen and Retroperitoneum

ing appearance of PBC. A recent report described the “periportal halo sign” in 43% (9 of 21) of patients with PBC, characterized by a hypointense rounded area surrounding the portal vein branches on both T2- and T1weighted images. The authors attributed this finding to the presence of stellate, periportal hepatocellular parenchyma extension, surrounded by regenerative nodules. This finding was better appreciated on portal venous and interstitial phase. It may be that this appearance reflects a prominent pattern of fibrosis as also observed in autoimmune hepatitis (Wenzel et al. 2001). Viral Hepatitis Acute hepatitis is diagnosed by clinical and serologic studies. MRI findings of acute hepatitis include heterogeneous hepatic signal intensity, which is most apparent on T2-weighted images and immediate post-gadolinium images. Periportal edema may be identified. Chronic hepatitis may be defined as symptomatic, biochemical or serologic evidence of continuing inflammation of the liver without improvement for at least 6 months. Chronic inflammation followed by progressive fibrosis may lead to fully developed cirrhosis (Semelka et al. 2001). In patients with chronic viral hepatitis, imaging studies are more commonly obtained, usually to detect the presence of cirrhosis or HCC. Focal inflammatory changes or fibrosis may develop in chronic active hepatitis, resulting in diffuse or regional areas of high signal intensity on T2-weighted images and heterogeneous enhancement after contrast administration on gradientecho images, most often appreciated as linear stromal enhancement on late fat-suppressed images (Semelka et al. 2001). On T2-weighted images, chronic active hepatitis often has periportal high signal intensity, corresponding to inflamed or enlarged lymph nodes, or both (Semelka et al. 2001). Porta hepatis lymph nodes measuring 2 cm or more are common in HCV hepatitis. 6.1.3.3.2  Cirrhosis Cirrhosis is a stage in the evolution of many chronic liver diseases including viral infections, alcohol abuse, hemochromatosis, autoimmune disease, Wilson’s disease, and primary sclerosing cholangitis. Pathologic gross inspection of cirrhotic livers generally shows two types of patterns, (1) micronodular in which parenchymal nodules are small (< 3 mm diameter) and separated by thin fibrous septa; and (2) macronodular in which parenchymal nodules are large (>3 mm) and separated by fibrous septa sometimes reaching the proportions of large scars. Cirrhotic nodules may be characterized based on the histological features into three major categories, namely: (1) regenerative, representing a benign proliferation of hepatocytes surrounded by fibrous septa; (2) dysplastic, representing regenerative nodules (RNs) with cellular atypia,

an intermediate step in the pathogenesis of HCC; and (3) malignant or HCC (Ito et al. 2002). On MRI a variety of morphological findings are observed. Atrophy of the right lobe and the medial segment of the left lobe are common in cirrhotic livers. The caudate lobe and lateral segment of the left lobe may undergo hypertrophy. In cirrhotic livers, enlargement of the hilar periportal space is commonly observed in patients with atrophy of the medial segment of the left lobe (Ito et al. 2002). Expansion of the major interlobar fissure may be seen in the late stage of disease, causing extrahepatic fat to fill the space between the left medial and lateral segments (Ito et al. 2002). These findings are accompanied by enlargement of the pericholecystic space (gallbladder fossa), which is subsequently filled with fat, in what is known as the “expanded gallbladder fossa sign.” The presence of regenerative nodules and confluent or diffuse parenchymal fibrosis causes irregularities and distortion in the liver surface and parenchyma (Ito et al. 2002). The most consistent morphological feature of cirrhosis is the demonstration of focal or diffuse fibrous tissue, that have low signal intensity on T1-weighted images and high or low signal intensity on T2-weighted images, depending on chronicity, with acute fibrous tissue having a higher fluid content and therefore higher signal intensity. Fibrous tissue enhances negligibly on hepatic arterial dominant-phase images and demonstrates late enhancement on hepatic venous phase images. Fibrous tissue is most consistently shown on TE = 2 ms out-of-phase imaging at 1.5 T, appearing as low-signal-intensity reticular tissue. Fibrosis is also well shown as late-enhancing stroma on 2-min post-gadolinium fat-suppressed gradient-echo images (Ito et al. 2002). On MRI, the majority of RNs are isointense on T2and T1-weighted images. Occasionally, RNs may appear low in signal intensity on T2-weighted images relative to high-signal-intensity inflammatory fibrous septa or damaged liver (Ito et al. 2002). RNs containing iron have low signal intensity on T2-weighted and T2*-weighted gradient-echo images. Approximately 16% of RNs are hyperintense on T1-weighted images. RNs demonstrate negligible enhancement on both hepatic arterial dominant phase and interstitial phase images (Fig 6.1.14). Dysplastic nodules (DNs) are defined as neoplastic, clonal lesions that represent an intermediate step in the pathway of carcinogenesis of hepatocytes in cirrhotic livers. On MR imaging, DNs are most commonly recognized as isointense or hypointense on T2-weighted images and hyperintense on T1-weighted images. Like RNs, DNs may also contain iron, which then results in low signal intensity on both T2- and T1-weighted images. Unlike RNs, DNs have been found to contain isolated arteries unaccompanied by bile ducts. Correlations exist between extent of enhancement on arterial dominant images and the grade of DNs. Increase in arterial blood supply and decrease of portal blood supply of hepatic nodules is closely related to

6.1  Abdominal MRI

Fig. 6.1.14a–e  Cirrhotic liver with regenerative nodules. On T2-weighted fat-suppressed single-shot echo-train spin-echo image (a), the nodules are isointense. On T1-weighted inphase (b) and out-of-phase (c) spoiled gradient-echo images (SGE), multiple nodules are observed that are hyperintense in signal. The nodules do not reveal enhancement on arterial dominant (d) and late-venous (e) phase contrast enhanced SGE images

the process of malignant transformation to HCC. On MR imaging, low-grade DNs show negligible enhancement, or similar enhancement to the background parenchyma (i.e., isointense) on arterial dominant-phase images; and high-grade DNs may demonstrate enhancement ranging from mild to intense on arterial dominant-phase images. DNs tend to fade toward background signal of the liver in the interstitial phase of enhancement; whereas small HCCs are more likely to exhibit lesion washout with late capsule enhancement (Fig 6.1.15). A focus of small HCC that develops in a high-grade DN appears as a high sig-

nal intensity focus within a low signal intensity nodule on T2-weighted images—a nodule within a nodule. This reflects the development of a high T2 signal malignancy within a low T2 signal dysplastic nodule. On T1-weighted images, the high-grade DN exhibits low signal intensity, and the foci of small HCC may appear isointense with the liver parenchyma. Portal hypertension results from obstruction at presinusoidal (e.g., portal vein), sinusoidal (e.g., cirrhosis), postsinusoidal (e.g., hepatic vein), or multiple levels. The most common cause of portal hypertension is cirrhosis.

879

880

6  Abdomen and Retroperitoneum

Fig. 6.1.15a–d  Dysplastic nodule in a cirrhotic liver. On T2weighted fat-suppressed single-shot echo-train spin-echo image (a), and T1-weighted spoiled gradient-echo (SGE) image (b), no focal liver lesion is identified. On arterial dominant-phase

SGE image (c), an enhancing lesion is present (arrow). The lesion fades and becomes isointense with the parenchyma on late venous phase SGE image (d)

Portal hypertension causes or exacerbates complications of cirrhosis such as variceal bleeding, ascites, and splenomegaly. Portosystemic shunts may be identified with gadolinium-enhanced imaging. In the early stages of portal hypertension, the portal venous system dilates, but flow is maintained. Later, substantial portosystemic shunting develops, reducing the volume of flow to the liver and decreasing the size of the portal vein. With advanced portal hypertension, portal flow may reverse and become hepatofugal. Thrombosis of the portal veins may develop with development of collaterals referred to as cavernous transformation. Other associated findings of portal hypertension includes: mesenteric, omental and retroperitoneal edema, gastrointestinal wall thickening. Portal varices arise from increased portal pressure, and portal blood is shunted into systemic veins, bypassing hepatic parenchyma. This may play a role in the development of hepatic atrophy in advanced cirrhosis. Major sites of portosystemic collateralization include gastroesophageal junction, paraumbilical veins, retroperitoneal

regions, perigastric, splenorenal, omentum, peritoneum, and hemorrhoidal veins. Esophageal varices are a serious complication because they may rupture and produce lifethreatening hemorrhage. Varices are particularly conspicuous using fat suppression or water excitation with gadolinium enhancement on gradient-echo images (Ito et al. 2002). 6.1.3.3.3  Iron Overload Primary (Idiopathic) Hemochromatosis Genetic hemochromatosis (GH) results from excessive gastrointestinal absorption and deposition of iron in tissues such as liver, heart, pancreas, anterior pituitary, joints, and skin. Early in the disease process, iron accumulation is restricted to the liver. Over time, iron deposition progresses to involve other organs, primarily the pancreas and heart. A diagnostic feature of idiopathic hemochromatosis is that signal intensity of the spleen is not

6.1  Abdominal MRI

Fig. 6.1.16a–c  Primary hemochromatosis. Axial T2-weighted single-shot echo-train spin-echo image (a) and T1-weighted spoiled gradient-echo images before (b) and after gadolinium application (c): On pre-contrast images (a,b), the liver has markedly low signal intensity relative to the spleen or muscle due to

increased iron deposition within the hepatic parenchyma. Note that the spleen has normal intermediate signal intensity. The pancreas (c small arrows) is also low in signal intensity, comparable to liver parenchyma, a characteristic feature in primary hemochromatosis

substantially decreased on T2-weighted or T2*-weighted images, due to accumulation of iron within the parenchyma of the liver and pancreas and lack of selective uptake by the RES in the spleen (Fig 6.1.16). The presence of iron deposition in the pancreas correlates with irreversible changes of cirrhosis in the liver. In patients with GH, cirrhosis and HCC develop over time. HCC reveals high signal intensity on a low-signalintensity background of liver parenchyma on T2-weighted images (Siegelman et al. 1996).

malnutrition, and exposure to toxins. Fatty degeneration may present as diffuse uniform or patchy, focal, or with spared foci of normal liver. At times focal fatty infiltration or geographic regions of normal liver within fatty liver (fat sparing) may mimic the appearance of mass lesions. Out-of-phase gradient-echo (TE = 2.1) imaging is a highly accurate MRI technique to examine for fatty liver and to distinguish focal fat from neoplastic masses. Fat in substantial amounts has high signal intensity on in-phase T1-weighted images due to its short T1.Comparing outof-phase (TE = 2.1 ms) to in-phase (TE = 4.2 ms) gradientecho images, the presence of fatty metamorphosis results in signal loss (Fig. 6.1.17). An area that is isointense or hyperintense to surrounding background parenchyma on in-phase images and loses signal homogeneously on out-of-phase images is highly diagnostic for focal fatty infiltration. Focal fat usually has angular, wedge-shaped margins that are relatively well-defined. Masses that contain fat usually have a rounded configuration. Common locations for focal fat are adjacent to the ligamentum teres, the central tip of segment IV; less commonly it appears along the gallbladder. On out-of-phase images focal normal liver in the setting of diffuse fatty infiltration (focal sparing) appears as a focus of high signal intensity on a background of liver of diminished signal intensity. The central tip of segment IV is a common location for focal fatty sparing. Liver metastases in the setting of fatty liver may demonstrate peritumoral fat sparing due to compression of this circumferential liver (Mitchell 1992). Nonalcoholic fatty liver disease (NAFLD) is related to obesity and type 2 diabetes. NAFLD is one of the causes of cryptogenic cirrhosis. Cirrhosis related to obesity and NAFLD are risk factors for HCC. Nonalcoholic steatohepatitis (NASH) occupies a middle position in the range of NAFLD and represents an intermediate stage of fatty liver damage.

Secondary Hemochromatosis Transfusional iron overload can cause excess iron deposition. Fibrosis is usually mild despite even heavy iron stores, and cirrhosis is rare. Iron deposition in the RES results in low signal intensity of the spleen, liver, and bone marrow on MR images, best shown on T2- or T2*weighted images. Transfusional hemochromatosis can be differentiated from genetic hemochromatosis by the signal intensity of the spleen, which is usually normal in genetic hemochromatosis, whereas signal intensity of the pancreas is normal with most cases of transfusional overload. In massive iron overload (e.g., >100 units) direct tissue deposition may occur in other cells and tissues, notably the pancreas (Siegelman et al. 1996). In moderate to severe forms of iron deposition the T2-shortening effect of iron results in low signal on T1weighted images as well. If liver and spleen are gray on in-phase SGE (TE = 4.2 ms) iron deposition is considered moderate, and if liver and spleen are near to being signal voids iron deposition is severe. 6.1.3.3.4  Fatty Liver Fatty liver or steatosis is defined as accumulation of triglycerides within hepatocytes. The causes of hepatic steatosis include alcohol abuse, diabetes mellitus, obesity,

881

882

6  Abdomen and Retroperitoneum

Fig. 6.1.17a,b  Fatty liver. Axial in-phase (a) and out-of-phase (b) T1-weighted spoiled gradient-echo images. The homogeneous signal loss of the entire liver parenchyma on opposed-phase im-

age (b) as compared with in-phase image (a) is caused by the chemical shift effects of intra-voxel fat and water indicating the increased fat content within the hepatocytes

6.1.3.3.5  Portal Venous Obstruction/Thrombosis

outflow. Obstruction of venous outflow from the liver results in portal hypertension, ascites, and progressive hepatic failure. Causes of intraluminal venous obstruction include polycythemia vera, pregnancy, postpartum state and intra-abdominal cancer, especially HCC. BuddChiari syndrome most often results in atrophy of peripheral liver, which experiences severe venous obstruction, and hypertrophy of the caudate lobe and central liver, which are relatively spared. Absence of hepatic veins may be demonstrated by time-of-flight techniques or portalphase gadolinium-enhanced gradient-echo sequences. Generally, a combination of both approaches results in the highest diagnostic accuracy. However, bright-blood technique is the most accurate and usually suffices. On immediate post-gadolinium MR images, the peripheral atrophic liver in Budd-Chiari syndrome may enhance to a greater or lesser extent than normal or hypertrophied liver, which provides insight into the chronicity of the disease process. In acute-onset Budd-Chiari syndrome, the peripheral liver enhances less than central liver. This is associated with moderately high signal intensity on T2-weighted images and low signal intensity on T1-weighted images reflecting associated edema (Fig 6.1.18). In subacute Budd-Chiari syndrome, reversal of flow in portal veins and development of small intra- and extrahepatic venovenous collaterals occurs. Many of the collaterals are capsule-based. Signal of the peripheral liver is mildly increased on T2-weighted images and mildly decreased on T1-weighted images, similar to acute BuddChiari syndrome. On dynamic gadolinium-enhanced MR images, mildly increased and heterogeneous enhancement is apparent in the peripheral liver relative to central liver on hepatic arterial dominant-phase images that, over time, becomes more homogeneous with the remainder of the liver. Caudate lobe hypertrophy is mild to

Thrombosis of the portal vein is generally associated with the presence of a hypercoagulable state, vascular injury, or stasis. Blockage of the portal vein may be extrahepatic or intrahepatic. On imaging, portal vein thrombosis may be demonstrated by using black-blood techniques (e.g., spin-echo techniques with superior and inferior saturation pulses) and bright-blood techniques (e.g., time-offlight gradient echo or gadolinium-enhanced gradient echo). A combination of both approaches is often useful to increase diagnostic confidence. Portal veins may be occluded by tumor thrombus, bland thrombus, or extrinsic compression. MRI usually is able to distinguish between these entities. Tumor and bland thrombus may be distinguished from each other by the observation that tumor thrombus is higher in signal intensity on T2-weighted images, has soft-tissue signal intensity on time-of-flight gradient-echo images, and enhances with gadolinium. Bland thrombus is low in signal intensity on T2-weighted and time-of-flight gradient-echo images, and does not enhance with gadolinium. Tumor thrombus is most often observed with hepatocellular carcinoma, most commonly in the diffuse type (Fig. 6.1.12), although it may also occur with metastases. After administration of intravenous gadolinium, transient increased enhancement of hepatic parenchyma may be apparent in areas with decreased portal perfusion during the hepatic arterial dominant phase of enhancement (Ito et al. 2002). 6.1.3.3.6  Hepatic Venous Thrombosis Budd-Chiari Syndrome Budd-Chiari syndrome is a disorder with numerous causes, resulting from obstruction to hepatic venous

6.1  Abdominal MRI

Fig. 6.1.18a–g  Budd-Chiari syndrome. Axial T2-weighted fatsaturated fast spin-echo (a,d), pre-contrast T1-weighted spoiled gradient-echo (b) and contrast-enhanced T1-weighted fat-saturated three dimensional gradient-echo images (c,e). The diffusely enlarged liver is inhomogeneous in signal intensity before and after contrast application, with patchy and geographical signal changes circumferentially around the vena cava. The vena

cava is compressed and the hepatic veins are occluded as best seen on coronal MR-angiography (f). Some small intrahepatic veins are thrombosed (e small arrow) and the portal vein is patent (f) in this patient. Perihepatic ascites as seen on coronal and sagittal steady state free precision sequences (g) and other signs of portal hypertension may be associated with Budd-Chiari syndrome

moderate, and collateral vessels are not prominent in the subacute setting (Noone et al. 2000). In chronic Budd-Chiari syndrome, hepatic edema is not a prominent feature and fibrosis develops. Fibrosis results in decreased signal of peripheral liver on T2- and T1-weighted images. Enhancement differences between peripheral and central liver on serial post-gadolinium images become more subtle. Venous thrombosis, appreciated in acute and subacute disease, is usually not observed in chronic disease. Massive caudate lobe hypertrophy, massive enlarged bridging intrahepatic collaterals, extrahepatic collaterals, and regenerative nodules are all features observed in chronic Budd-Chiari syndrome. Curvilinear intrahepatic collaterals and capsule-based collaterals are characteristic of chronic Budd-Chiari syndrome. Varices are usually prominent in chronic BuddChiari syndrome and are well shown on interstitial-phase fat-suppressed images. Extensive portosystemic varices, as observed in other chronic liver diseases, are also present.

The development of nodular regenerative hyperplasia in the chronic setting is the result of hepatic ischemia caused by hepatic venous obstruction. The nodules are isointense or of low signal intensity on T2-weighted images, and have high signal intensity on T1-weighted images similar to macroregenerative nodules. These nodules reveal moderately intense enhancement on immediate postgadolinium gradient-echo images (Noone et al. 2000). 6.1.3.3.7  Infectious Parenchymal Diseases Pyogenic Abscess Pyogenic abscesses are the most frequent form of focal hepatic infections resulting from an infectious process of bacterial origin. Pathologically, pyogenic liver abscesses may occur as solitary or multiple lesions, ranging from millimeters to massive lesions. Portal vein thrombosis is frequently associated with bacterial abscesses. The in-

883

884

6  Abdomen and Retroperitoneum

fected bland thrombus is characterized by low signal intensity on T2- and T1-weighted images and time-of-flight gradient-echo images. The thrombus does not enhance after contrast; however the vein wall shows a moderateto-intense enhancement after contrast administration, best seen on late-phase fat-suppressed images caused by the inflammatory reaction. Characteristic MRI findings of pyogenic abscesses are high signal intensity on T2-weighted images, low signal intensity on T1-weighted images, moderate enhancement of stroma on immediate post-gadolinium images with persistent enhancement on interstitial phase images, and no enhancement of additional stroma or progressive fill in of the lesion over time (Balci and Sirvanci 2002). Pyogenic abscesses also possess markedly thick walls and internal septations, which enhance moderately to intensely on early-phase images and demonstrate persistent enhancement on late-phase images that often appears more intense (Balci and Sirvanci 2002). Abscesses typically have a moderate perilesional enhancement with indistinct outer margins on immediate post-gadolinium images because of a surrounding rim of granulation tis-

sues and a hyperemic inflammatory response in adjacent liver. The perilesional enhancement rapidly diminishes, and is often nearly resolved by 1 min postinjection. Layering of debris and gas within the abscess cavity, mainly after biliary drainage, is also commonly appreciated (Fig 6.1.19). Gas is identified as signal void on both T2- and T1-weighted images and debris are identified as low signal intensity on T2- and high signal intensity on T1weighted images, since debris are usually composed of protein (Balci and Sirvanci 2002).

Fig. 6.1.19a–d  Pyogenic liver abscess. T2-weighted fat-suppressed single-shot echo-train spin-echo image (a) and T1weighted spoiled gradient-echo (SGE) image (b) reveal hyperintense lesion and hypointense lesion with air-fluid levels (arrows),

respectively. Arterial phase SGE image (c) reveals peripheral rim (arrow) and perilesional enhancement. On portal venous–phase SGE image (d), peripheral rim enhancement (arrow) persists

Amebic Abscess Amebic liver abscesses are caused by a protozoan parasite, Entamoeba histolytica, and are not uncommon in developing tropical countries (Balci and Sirvanci 2002). Lesions are usually solitary, affect the right lobe more often than the left lobe, and are prone to invade the diaphragm with development of pulmonary consolidation and empyema. Lesions are encapsulated, thick walled (5–0 mm), and demonstrate substantial enhancement of the capsule on gadolinium-enhanced images, which permits differentiation from liver cysts (Balci and Sirvanci 2002).

6.1  Abdominal MRI

Echinococcal Disease Echinococcal disease is a worldwide zoonosis produced by two main types of larval forms of Echinococcus tapeworms: E. granulosus and E. alveolaris. E. granulosus is the causative organism for hydatid cysts. The typical imaging feature is an intrahepatic encapsulated multicystic lesion with daughter cysts arranged peripherally within the larger cyst. Satellite cysts located exterior to the fibrinous membrane of the main hepatic cyst are not uncommon. Lesions are frequently complex, with mixed high signal intensity on T2-weighted images and mixed low signal intensity on T1-weighted images due to the presence of proteinaceous and cellular debris. The fibrous capsule and internal septations are well shown on T2-weighted images and gadolinium-enhanced T1-weighted images (Fig 6.1.20). E. alveolaris is the causative organism for hepatic alveolar echinococcosis (HAE), a rare parasitic disease for which the fox is the main host of the adult parasite, with dogs and cats being less frequently reported hosts. Pathologically, HAE is characterized grossly by multilocular or confluent cystic, necrotic cavities. A fibrous rim is not present. Calcification is common in HAE and appears as clusters of microcalcifications or large calcified foci. HAE tends to involve extensive regions in the liver in an infiltrative pattern because it does not form membranes or capsules. HAE is more likely to involve the porta hepatitis causing stenoses of portal veins, intrahepatic bile ducts and hepatic veins, which commonly result in portal hypertension (Balci and Sirvanci 2002). Mycobacterial Infection Hepatic tuberculosis is the most frequent form of infectious hepatic granulomas. The most common setting for the bacilli of Mycobacterium tuberculosis to reach the liver is through the bloodstream. Although abdominal tuberculosis preferentially affects lymph nodes and the ileo-

Fig. 6.1.20a–c  Hydatid cyst. Axial T1-weighted spoiled gradient-echo image (a), T2-weighted single-shot echo-train spinecho image (b) and T1-weighted contrast-enhanced fat-suppressed three-dimensional gradient-echo image in portal venous phase: The well-defined lesion in segment IVa is encapsulated with a small rim of granulation tissue. The lesion is mainly hyperintense and heterogeneous on T2-weighted images (b) and

cecal junction, the liver is also commonly involved. The incidence of hepatic tuberculosis is increasing, reflecting, in part, an increase in numbers of patients who are immunocompromised, such as patients with HIV infection. Focal hepatic lesions are typically small and multiple with an appearance similar to that of fungal lesions. Infection has a propensity to involve the portal triads and spread in a superficial infiltrating fashion. This can be visualized as periportal high signal intensity on T2weighted fat-suppressed images and moderate-to-intense enhancement on late-phase fat-suppressed post-gadolinium images. Associated porta hepatis nodes are common (Balci and Sirvanci 2002). Fungal Infection Hepatosplenic or visceral candidiasis is a form of invasive fungal infection that has emerged as a serious complication of the immunocompromised state, especially in AIDS patients, patients on medical therapy for acute myelogenous leukemia (AML), and patients with bone marrow transplantation. The most common infecting organism is Candida albicans, but other fungi may be found. Acute hepatosplenic candidiasis involves the liver and spleen, with renal involvement occurring in less than 50% of patients. Multifocal microabscesses or granulomas are observed in the liver. Liver lesions are frequently smaller than 1 cm and subcapsular in location. T2-weighted fatsuppressed spin-echo sequences are effective at demonstrating these lesions because of the high conspicuity of this sequence for small lesions and the absence of chemical shift artifact that may mask small peripheral lesions. The fungal abscesses are high in signal intensity on T2weighted images. They also may be seen on gadoliniumenhanced T1-weighted images as signal-void foci with no appreciable abscess wall enhancement. It has been observed that patients with hepatosplenic candidiasis

low in signal intensity on T1-weighted images both before (a) and after (c) intravenous application of gadolinium with a mild contrast enhancement of the surrounding rim. Multiple hypointense linear structures inside the lesion on T2-weighted images suggesting cyst wall remnants are indicative of the infectious origin of this lesion from Echinococcus granulosus

885

886

6  Abdomen and Retroperitoneum

who are immunocompetent possess abscesses that demonstrate mural enhancement. The absence of abscess wall enhancement may reflect the patient’s neutropenic state. Overall sensitivity of MRI is 100%, and specificity is 9%. After institution of antifungal antibiotics, successful response may be demonstrated. Central high signal intensity develops within lesions on T2-weighted and T1-weighted images that enhances with gadolinium, representing granuloma formation. In addition, a distinctive dark perilesional ring is observed on all sequences, representing collections of iron-laden macrophages throughout granulation tissue at the periphery of lesions. This represents the subacute treated phase, which may be consistent with a good prognostic finding, reflecting the patient’s ability to mount an immune response. MRI also demonstrates chronic healed lesions that have responded to antifungal therapy. Chronic healed lesions are irregularly shaped, isointense and poorly shown on T2-weighted images, hypointense on T1-weighted images and demonstrate negligible enhancement after contrast. The lesions are most conspicuous as low-signal-intensity defects with angular margins on immediate post-gadolinium SGE images. Capsular retraction may also be observed adjacent to the lesions. This constellation of imaging features is consistent with chronic scar formation (Balci and Sirvanci 2002). 6.1.4  Pancreas 6.1.4.1  Developmental Anomalies 6.1.4.1.1  Pancreas Divisum In this congenital abnormality, portions of the pancreas have separate ductal systems: a very short ventral duct of Wirsung drains only the lower portion of the head while the dorsal duct of Santorini drains the tail, body, neck, and upper aspect of the head. The incidence of this anomaly varies between 1.3 to 6.7% of the population (Fulcher and Turner 1999). On MRCP images, separate entries of the ducts of Santorini and Wirsung into the duodenum are consistently demonstrated due to the good conspicuity of the linear high-signal-intensity tubular structures. Pancreas divisum may be a predisposing factor in recurrent pancreatitis (Fulcher and Turner 1999). 6.1.4.1.2  Annular Pancreas Annular pancreas is an uncommon congenital anomaly in which glandular pancreatic tissue, in continuity with the head of the pancreas, encircles the second part of the duodenum. Patients may present with duodenal obstruction. On MR images, pancreatic tissue is identified encasing the duodenum. Non-contrast T1-weighted fat-suppressed

and/or immediate post-gadolinium gradient-echo images are particularly effective at demonstrating this entity due to the high signal intensity of pancreatic tissue, which is readily distinguished from the lower signal intensity of adjacent tissue and duodenum (Fulcher and Turner 1999). 6.1.4.2  Genetic Diseases 6.1.4.2.1  Cystic Fibrosis Cystic fibrosis is characterized by a dysfunction of the secretory process of all exocrine glands and reduced mucociliar transport, which results in mucous plugging of the exocrine glands (Ferroi et al. 1996). MRI findings of cystic fibrosis include pancreatic enlargement with complete fatty replacement with or without loss of the lobulated contour, atrophic pancreas with partial fatty replacement, and diffuse atrophy of the pancreas without fatty replacement (Ferroi et al. 1996). Pancreatic enlargement with complete fatty replacement is the most common pattern observed in cystic fibrosis. Fatty replacement is high in signal intensity on T1-weighted images and demonstrates loss of signal intensity on T1-weighted fat-suppressed images. Pancreatic cysts secondary to duct obstruction by secretion are another manifestation of cystic fibrosis (Ferroi et al. 1996). 6.1.4.2.2  von Hippel-Lindau Disease von Hippel-Lindau disease is an autosomal dominant condition with variable penetration. This condition is characterized by tumors in the cerebellum and retina. Patients may have cysts of the liver and kidney, with a strong propensity to develop renal cell carcinoma. Patients with von Hippel-Lindau may develop pancreatic cysts, islet cell tumors, or microcystic cystadenoma (Friedman et al. 1983). 6.1.4.3  Mass Lesions 6.1.4.3.1  Adenocarcinoma Pancreatic ductal adenocarcinoma accounts for 95% of the malignant tumors of the pancreas. Pancreatic cancer arising in the head of the pancreas may cause obstruction of the common bile duct (CBD) and pancreatic duct (Semelka and Ascher 1993). This appearance on MRCP studies results in the “double-duct sign,” which was originally described on ERCP. A characteristic imaging appearance of pancreatic carcinoma consists of enlargement of the head of the pancreas with dilatation of the pancreatic and common bile duct and atrophy of the body and tail of the pancreas. Other MRI findings of pancreatic

6.1  Abdominal MRI

Fig. 6.1.21a–f  Adenocarcinoma of the pancreas. Coronal T2weighted single-shot echo-train spin-echo image (a) reveals abrupt termination of the common bile duct. Same finding is present on the MRCP image (b). On axial T2-weighted singleshot echo-train spin-echo image (c), heterogenous hyperintensity is present (arrow). Axial T1-weighted fat-saturated spoiled

gradient-echo (SGE) image (d) of the same patient reveals better delineation of the mass (arrow). On arterial-phase axial T1weighted SGE image (e), the mass does not enhance (arrow). Gradual enhancement of the mass is observed on late-venousphase axial SGE image (f)

cancer include the presence of lymphadenopathy, encasement of the celiac axis or superior mesenteric artery, and liver metastases (Semelka and Ascher 1993). Detection of carcinoma is best performed using immediate post-gadolinium T1-weighted gradient-echo images, with the demonstration of a focal hypovascular mass within pancreatic parenchyma (Semelka and Ascher 1993; Semelka et al. 1996) (Fig 6.1.21). Tumors are usually minimally hypointense relative to pancreas on T2-weighted images and are therefore difficult to visualize. Pancreatic cancers appear as low-signal-intensity masses on non-contrast T1weighted fat-suppressed images and are clearly separated from normal pancreatic tissue, which is high in signal intensity (Semelka and Ascher 1993). Pancreatic tissue distal to pancreatic cancer is often lower in signal intensity than normal pancreatic tissue due to tumor-associated obstructive pancreatitis (Semelka and Ascher 1993; Semelka et al. 1996). Demonstration of a rim of increased

enhancement representing surrounding pancreas is helpful for the delineation of the tumor within the chronically inflamed pancreas. Staging of pancreatic cancer is performed with the use of a combination of sequences including non-suppressed T1-weighted images and interstitial-phase gadoliniumenhanced fat-suppressed T1-weighted images. Vascular encasement by tumor is best-shown using thin-section 3D-gradient-echo images, which can be analyzed as both source images in the traverse plane and reformatted images in the coronal plane. Immediate post-gadolinium gradient-echo images are useful for evaluating arterial patency, and portal venous and interstitial phase postgadolinium gradient-echo images are useful for evaluating venous patency. Lymph nodes are well shown on T2-weighted fat-suppressed images and interstitial-phase gadolinium-enhanced fat-suppressed T1-weighted images with high signal intensity (Semelka and Ascher 1993).

887

888

6  Abdomen and Retroperitoneum

Islet-cell tumors are a subgroup of gastrointestinal neuroendocrine tumors that occur within the endocrine pancreas. Insulinomas are most commonly benign tumors, gastrinomas are malignant in approximately 60% of cases, and almost all other types, including nonfunctioning tumors are malignant in the great majority of cases. The liver is the most common organ for metastatic spread (Martin and Semelka 2000).

cystic. Gastrinomas may occur outside the pancreas, and fat-suppressed T2-weighted images are particularly effective at detecting these high-signal-intensity tumors in a background of suppressed fat. Gastrinoma metastases to liver are very high in signal intensity on T2-weighted fat-suppressed images and have well-defined margins. Centripetal enhancement of gastrinoma metastases may occur on serial post-gadolinium images (Fig 6.1.22). Peripheral washout is commonly observed for hypervascular gastrinoma metastases.

Gastrinomas (G-Cell Tumors) Zollinger-Ellison syndrome is defined by the clinical triad of pancreatic islet cell gastrinoma, gastric hypersecretion, and recalcitrant peptic ulcer disease. Gastrinomas occur most frequently in the region of the head of the pancreas including the pancreatic head, duodenum, stomach, and lymph nodes in a territory termed the gastrinoma triangle (Martin and Semelka 2000). Gastrinomas are low in signal intensity on T1-weighted fat-suppressed images and high in signal intensity on T2weighted fat-suppressed images, demonstrating peripheral ring-like enhancement on immediate post-gadolinium gradient-echo images (Martin and Semelka 2000). These imaging features are observed in the primary lesion and in hepatic metastases. Occasionally, lesions will be

Insulinomas Insulinomas are one of the most common of the islet cell tumors and are frequently functionally active. Tumors frequently come to clinical attention when they are small (<2 cm) due to the severity of the symptomatology (Martin and Semelka 2000). Insulinomas are usually richly vascular. Angiography has been reported as superior to CT imaging in detecting these tumors due to their small size and increased vascularity (Martin and Semelka 2000). Insulinomas are low in signal intensity on T1-weighted images and high in signal intensity on T2-weighted images. Insulinomas are well shown on T1-weighted fat-suppressed images (Martin and Semelka 2000). Small insulinomas typically enhance homogeneously on immediate post-gadolinium gradient-echo images. Larger tumors,

6.1.4.3.2  Islet Cell Tumors

Fig. 6.1.22a–c  Pancreatic gastrinoma. Axial T1-weighted fatsaturated three-dimensional gradient-echo (3D GE) image in arterial phase (a) reveals nodular mass with marked enhancement (arrow) in the pancreatic head. The same mass retains contrast on axial and coronal 3D GE late venous phase images (b,c)

6.1  Abdominal MRI

which measure greater than 2 cm in diameter, often show ring enhancement. Liver metastases from insulinomas are hypervascular and typically have peripheral ring-like or homogenous enhancement. Glucagonoma, Somatostatinoma, VIPoma, and ACTHoma These islet-cell tumors are considerably rarer than are insulinomas or gastrinomas. They are usually malignant, with liver metastases present at the time of diagnosis (Martin and Semelka 2000). The primary pancreatic tumors of glucagonoma and somatostatinoma are large and heterogeneous on MR images. They are usually moderately low in signal intensity on T1-weighted fat-suppressed images and moderately high in signal intensity on T2-weighted fat-suppressed images, enhancing heterogeneously on immediate post-gadolinium images. Liver metastases are generally heterogeneous in size and shape, unlike gastrinoma metastases, which are typically uniform. VIPomas may have a characteristic appearance of a small primary tumor despite large and extensive liver metastases (Martin and Semelka 2000). 6.1.4.3.3  Cystic Pancreatic Neoplasms Serous Cystadenoma (Microcystic/Macrocystic Serous Cystadenoma) Serous cystadenoma is a benign neoplasm characterized by numerous tiny serous fluid-filled cysts (Friedman et al.

Fig. 6.1.23a–f  Microcystic serous cystadenoma of the pancreas. Axial T2-weighted single-shot echo-train spin-echo images in axial (a) and coronal (b) planes reveal multiple grape-like, small cysts in the head of the pancreas. On axial T1-weighted fat-saturated spoiled gradient-echo (SGE) image (c), the lesion shows

1983; Martin and Semelka 2000). This tumor frequently occurs in older patients and has an increased association with von Hippel-Lindau disease (Friedman et al. 1983). Tumors range in size from 1 to 12 cm, with an average diameter at presentation of 5 cm. T2-weighted images reveal small cysts and intervening septations as a cluster of small grape-like high-signal-intensity cysts with a central scar (Fig 6.1.23). Cystic pancreatic masses that contain cysts measuring less than 1 cm in diameter may represent microcystic cystadenoma or side-branch type intraductal papillary mucinous tumor (IPMT), which can be difficult to distinguish. Definition of communication with the pancreatic duct on MRCP images establishes the diagnosis of side branch IPMT. Uncommonly serous cystadenomas may be microcystic (cysts measuring from 1–8 cm) oligo- or unilocular (Friedman et al. 1983; Martin and Semelka 2000) (Fig 6.1.24). Microcystic and macrocystic serous tumors represent morphologic variants of the same benign pancreatic neoplasm, namely serous cystadenoma. Relatively thin uniform septations and absence of infiltration of adjacent organs and structures are features that distinguish serous cystadenoma from serous cystadenocarcinoma. Tumor septations usually enhance minimally with gadolinium on early and late post-contrast images, although moderate enhancement on early post-contrast images may occur. Delayed enhancement of the central scar may occasionally be observed, and is more typical of large tumors.

homogeneous hypointense signal. Coronal MRCP image (d) reveals the same mass with hyperintense signal. On post-contrast T1-weighted SGE images on axial (e) and coronal (f) planes, the cystic mass reveals septal enhancement

889

890

6  Abdomen and Retroperitoneum

Fig. 6.1.24a–e  Macrocystic serous cystadenoma of the pancreas. T2-weighted single-shot echo-train spin-echo images on axial (a) and coronal (b) planes reveal cystic mass on the tail of the pancreas with multiple septae. On axial T1-weighted fat-sat-

urated spoiled gradient-echo (SGE) image (c), the lesion is hypointense without visible septae. On post-contrast T1-weighted SGE images on axial (d) and coronal (e) planes, the septae are visible

Serous Cystadenocarcinoma (Microcystic Serous Cystadenocarcinoma) This malignant pancreatic tumor is extremely rare. Distinction from benign serous cystadenoma is difficult on histologic grounds alone and may only be established by the presence of metastatic disease or local invasion (Friedman et al. 1983; Martin and Semelka 2000).

produced by these tumors may result in high signal intensity on T1- and T2-weighted images of the primary tumor and liver metastases (Martin and Semelka 2000).

Mucinous Cystadenoma/Cystadenocarcinoma Mucinous cystic neoplasms of the pancreas are characterized by the formation of large unilocular or multilocular cysts filled with abundant, thick, gelatinous mucin. Mucinous cystic neoplasms should be interpreted as mucinous cystadenocarcinomas of low-grade malignant potential in order to reinforce the need for complete surgical resection and close clinical follow up. Mucinous cystic neoplasms occur more frequently in females (6 to 1) and approximately 50% occur in patients between the ages of 40 and 60 years. These tumors usually are located in the body and tail of the pancreas. They may be large (mean diameter of 10 cm), multiloculated, and encapsulated. There is a great propensity for invasion of local organs and tissues (Martin and Semelka 2000). On gadolinium-enhanced T1-weighted fat-suppressed images, large, irregular cystic spaces separated by thick septa are demonstrated. Mucinous cystadenomas are well circumscribed and they show no evidence of metastases or invasion of adjacent tissues. Mucinous cystadenomas described pathologically as having borderline malignant potential, may be very large, but may not show imaging or gross evidence of metastases or local invasion. Mucin

Main pancreatic duct involvement presents as diffuse ductal dilatation, copious mucin production, and papillary growth. These tumors are rare and typically malignant (Martin and Semelka 2000). A greatly expanded main pancreatic duct is demonstrated on T2-weighted images or MRCP images. Irregular enhancing tissue along the ductal epithelium is appreciated on post gadolinium images, confirming that underlying tumor is the cause of the ductal dilatation. Intraductal papillary mucinous tumors involving predominantly side branch ducts appear as oval-shaped cystic masses in proximity to the main pancreatic duct. Septations are generally present creating a cluster-of-grapes appearance.

6.1.4.3.4  Intraductal Papillary Mucinous Tumors (Duct Ectatic Mucin-Producing Tumor)

6.1.4.3.5  Solid and Papillary Epithelial Neoplasm These tumors are generally considered benign neoplasms with occasional examples exhibiting low-grade malignant potential. Solid and papillary epithelial neoplasms occur most frequently in females between 20 and 30 years of age. MRI findings of solid and papillary epithelial neoplasms are virtually diagnostic in the appropriate clinical

6.1  Abdominal MRI

setting. The MR appearance is a large, well-encapsulated mass, which demonstrates focal signal void calcification and regions of hemorrhagic degeneration, (as evidenced by fluid-debris levels, or signal intensities consistent with blood products). MRI findings of solid and papillary epithelial neoplasms include well-demarcated lesions that contain central high signal intensity on T1-weighted images. This central high signal intensity corresponds to hemorrhagic necrosis (Martin and Semelka 2000). 6.1.4.4  Lymphoma Non-Hodgkin’s lymphoma may involve peripancreatic lymph nodes or may directly invade the pancreas (Zieman et al. 1985). Intermediate-signal-intensity peripancreatic lymph nodes are distinguished from high-signalintensity normal pancreas on T1-weighted fat-suppressed images. Invasion of the pancreas is shown by loss of the normal high signal intensity of the pancreas on T1weighted fat-suppressed images. 6.1.4.5  Metastases Involvement of the pancreas by metastatic tumor may be the result of spread by direct extension or hematogenous metastases. Hematogenous metastases may occur with carcinomas of the lung, breast and kidney and malignant melanoma. Metastases from renal cell carcinoma are hypointense on T1- and hyperintense on T2-weighted images and reveal increased enhancement due to their hypervascular nature (Martin and Semelka 2000). Metastases from other primary tumors generally appear as focal pancreatic masses that are mildly hypointense on T1-weighted images, moderately hypointense on T1weighted fat suppressed images and mildly hyperintense on T2-weighted images. Metastases to the pancreas often enhance in a ring-like fashion. 6.1.4.6  Inflammatory Disease 6.1.4.6.1  Acute Pancreatitis Acute pancreatitis arises in the majority of cases secondary to alcoholism or cholelithiasis. Alcohol-related acute pancreatitis most frequently results in acute recurrent pancreatitis, whereas gallstone-related pancreatitis typically results in a single attack. The signal-intensity features of the pancreas in uncomplicated mild acute pancreatitis resemble those of normal pancreatic tissue. The acutely inflamed pancreas shows either focal or diffuse enlargement, which may be subtle. Peripancreatic fluid is well shown on non-contrast or immediate post-gadolinium gradient-echo images and

appears as low-signal-intensity strands of fluid or fluid collections in a background of high-signal-intensity fat. T2-weighted single-shot echo train spin-echo imaging employing fat suppression is the most sensitive technique for showing small volume peripancreatic fluid, which appears as high signal intensity in a background of intermediate- to low-signal intensity pancreas and low-signalintensity fat. As the extent of pancreatitis becomes more severe, the pancreas develops a heterogeneous appearance on pre-contrast T1-weighted fat-suppressed images and enhances in a more heterogeneous, diminished fashion on immediate post-gadolinium images (Fig 6.1.25). Percentage of pancreatic necrosis has been considered an important prognostic indicator in patients with acute pancreatitis. Dynamic gadolinium-enhanced gradientecho images may be useful for this determination. Complications of acute pancreatitis such as hemorrhage, pseudocyst formation, or abscess are clearly shown on MRI. Hemorrhagic fluid collections are high in signal intensity on T1-weighted fat-suppressed images, and depiction of hemorrhage is superior on MR images compared to CT images. Simple pseudocysts are low in signal intensity or represented by a signal void in a background of normal signal intensity pancreatic tissue on both non-contrast gradient-echo and T1-weighted fat-suppressed images. Pseudocyst walls enhance minimally on early post-gadolinium images and show progressive intense enhancement on 5 min post-contrast images, consistent with the appearance of fibrous tissue. Simple pseudocysts are relatively homogeneous and high in signal intensity on T2weighted images. Pseudocysts complicated by necrotic debris, hemorrhage, or infection, are heterogeneous in signal intensity on T2-weighted images (Saifuddin et al. 1993). 6.1.4.6.2  Chronic Pancreatitis Chronic pancreatitis is defined pathologically by continuous or relapsing inflammation of the organ leading to irreversible morphologic injury and, typically, impairment of function. Chronic pancreatitis is caused by complication of repeated attacks of acute pancreatitis. Alcoholism and obstruction of the pancreatic duct from various causes, including pancreatic ductal cancer, result in chronic pancreatitis. MRI may perform better than CT imaging at detecting changes of chronic pancreatitis, since MRI detects not only morphological findings but also the presence of fibrosis. Fibrosis is shown by diminished signal intensity on T1-weighted fat-suppressed images and diminished heterogeneous enhancement on immediate post-gadolinium gradient-echo images. Most cases of chronic pancreatitis show progressive parenchymal enhancement on 5-min post-contrast images, reflecting the pattern of enhancement of fibrous tissue.

891

892

6  Abdomen and Retroperitoneum

Fig 6.1.25a–e  Acute pancreatitis. On axial T2-weighted single-shot echo-train spin-echo image (a), minimal peripancreatic fluid is observed with hyperintense signal (arrows). T1-weighted axial fat-saturated spoiled gradient-echo (SGE) image (b) reveals diminished parenchymal signal of the pancreas. On arterial dominant-phase (c) and late-venous-phase (d) axial SGE images, decreased enhancement of the pancreas is observed

Focal enlargement of the head of the pancreas with chronic pancreatitis may be difficult to distinguish from cancer on CT images. MR images permit the distinction between these two entities with greater reliability. Both chronic pancreatitis and carcinoma show similar signal intensity changes of the enlarged region of pancreas on non-contrast T1-weighted fat-suppressed and T2-weighted images; generally mildly hypointense on T1weighted images and heterogenous and mildly hyperintense on T2-weighted images. On immediate post-gado-

linium images, focal pancreatitis shows heterogeneous enhancement with the presence of signal-void cysts and calcifications, without evidence of a marginated definable, minimally enhancing mass lesion. Demonstration of a definable, circumscribed mass lesion is most often diagnostic for tumor. In chronic pancreatitis, the focally enlarged portion of the pancreas usually shows preservation of a glandular, feathery, or marbled texture similar to that of the remaining pancreas. In contrast, in pancreatic cancer, the focally enlarged portion of the pancreas loses

6.1  Abdominal MRI

its usual anatomic detail. Tumor disrupts the underlying architecture and generally exhibits irregular, heterogeneous, diminished enhancement. Diffuse low signal intensity of the entire pancreas, similar to and including the area of focal enlargement, on T1-weighted fat-suppressed and immediate post-gadolinium SGE images is typical for chronic pancreatitis. In the setting of pancreatic cancer, the enhancement of the tumor is less than adjacent pancreatic parenchyma. Acute or chronic pancreatitis is well shown on MR images. Pancreatic pseudocysts observed in patients with chronic pancreatitis often arise as a sequel of episodes of acute inflammation (Semelka et al. 1993b). Small pseudocysts and cysts are present and are well shown on gadolinium-enhanced T1-weighted fat suppressed images as nearly signal-void oval structures. 6.1.5  Gallbladder and Bile Ducts 6.1.5.1  Gallbladder 6.1.5.1.1  Non-Neoplastic Disease Gallstone Disease MRCP sequences are highly sensitive and accurate in depicting cholecystolithiasis and can outperform ultrasound and computed tomography (Kelekis and Semelka 1996). Gallstones generally present as intraluminal, signal-void, round or facetted structures on both T1- and T2-weighted images (Fig 6.1.26). High signal intensity of the gallstones on T1- and T2-weighted sequences is not uncommon. A gallstone can be differentiated from a gallbladder polyp by the lack of enhancement on T1-weighted post-gadolinium images.

Fig. 6.1.26a–c  Choledocholithiasis with cholecystolithiasis. On coronal thick-section projection MRCP image (a) and T2weighted coronal thin section fat-saturated single-shot echotrain spin-echo image (b,c), the biliary system including the gall

Acute Cholecystitis Acute inflammation of the gallbladder is caused by obstruction of the cystic duct (e.g., by cystic duct stones) in 80–95% of patients. Findings that are indicative of acute cholecystitis on post-gadolinium T1-weighted images are (1) increased wall enhancement, (2) transient increased enhancement of adjacent liver parenchyma on immediate post-gadolinium images, and (3) increased thickness of the gallbladder wall (>3 mm) (Kelekis and Semelka 1996; Motohara et al. 2003) (Fig 6.1.27). Findings on T2weighted images that are helpful to establish the diagnosis are (1) presence of gallstones, (2) presence of pericholecystic fluid, (3) presence of intramural abscess as hyperintense focus in the gallbladder wall, and (4) increased wall thickness. Periportal high signal intensity may be observed but is a non-specific finding. Acute acalculous cholecystitis makes up about 5–15% of all acute cholecystitis cases. It can be caused by depressed motility (e.g., in patients with severe trauma/ surgery, burns, shock, anesthesia, diabetes mellitus), by decreased blood flow in the cystic artery due to extrinsic obstruction or embolization, or by bacterial infection (Kelekis and Semelka 1996; Motohara et al. 2003). Chronic Cholecystitis Chronic cholecystitis is more common than acute cholecystitis. In chronic cholecystitis mural gadolinium enhancement is mild and most prominent on delayed post-gadolinium images. Pericholecystic enhancement is minimal or absent (Kelekis and Semelka 1996). The wall of the gallbladder may calcify, resulting in porcelain gallbladder. On MR images, calcifications may appear as signal void foci.

bladder presents as high signal intensity structures equivalent to fluid. Irregularly shaped signal voids within the bile ducts and the gall bladder (b arrows) represent gallstones

893

894

6  Abdomen and Retroperitoneum

Fig. 6.1.27a–d  Acute cholecystitis. Axial T2-weighted fat suppressed single-shot echo-train spin-echo image (a) reveals thickened gallbladder wall and pericholecystic fluid (arrows). On axial T1-weighted arterial dominant-phase, spoiled gradient-echo image (b), there is transient pericholecystic increased enhancement (arrows) in the liver parenchyma. On portal venous-phase three-dimensional gradient-echo (3D GE) image (c), transient increased enhancement detected on the arterial dominant phase fades away and becomes isointense with the remaining liver parenchyma. On late venous phase 3D GE image (d), enhancing thickened gallbladder wall (arrow) is demonstrated

Diffuse Gallbladder Wall Thickening Diffuse gallbladder wall thickening may be present in a number of hepatic, biliary, and pancreatic diseases. Among non-tumorous causes are: hepatitis, liver cirrhosis, hypoalbuminemia, renal failure, systemic or hepatic venous hypertension, AIDS cholangiopathy, and graftversus-host disease. Important features to discriminate these conditions from cholecystitis are minimal enhancement of the gallbladder wall, and lack of increased enhancement of adjacent structures on post-gadolinium images, in particular the lack of transient increased enhancement of adjacent liver parenchyma (Kelekis and Semelka 1996). 6.1.5.1.2  Neoplastic Disease Gallbladder Polyps Gallbladder polyps are often incidentally identified arising from the gallbladder wall and are either sessile or pedunculated. The majority are cholesterol polyps, which do not have malignant potential. Approximately 10% of gallbladder polyps, however, are adenomas, which may have malignant potential. Polyps typically have homogeneously low to intermediate signal intensity on T1- and T2-weighted MR images. On T1-weighted post-gadolinium images, they show moderate homogeneous en-

hancement, which is most pronounced on delayed images. Polyps can be readily distinguished from calculi regarding their gadolinium enhancement, or by location, if the polyp is located on the nondependent surface of the gallbladder wall. Symptomatic lesions, polyps larger than 1 cm, or interval increase in size are suggestive of malignancy, and in such cases, cholecystectomy is required (Kelekis and Semelka 1996). Gallbladder Adenomyomatosis This disease entity is characterized by hyperplasia of epithelial and muscular elements with mucosal outpouching of epithelium-lined cystic spaces into a thickened muscularis layer. On MR images, these fluid-filled sinuses appear as small intramural foci of low signal intensity on T1-weighted images and of high signal intensity on T2weighted images. After gadolinium administration, early mucosal enhancement and late homogeneous enhancement are observed (Kelekis and Semelka 1996). Gallbladder Carcinoma Gallbladder carcinoma is the most common biliary malignancy and occurs predominantly in the sixth and seventh decade, with a slight female predominance. Findings on MRI, which are suggestive of gallbladder carcinoma, are (1) a mass either protruding into the gallbladder lumen or replacing the lumen completely; (2) focal or diffuse

6.1  Abdominal MRI

thickening of the gallbladder wall greater than 1 cm; and (3) soft tissue (tumor) invasion of adjacent organs such as the liver, duodenum, and pancreas, which occurs frequently. On T1-weighted MR images, the tumor is hypoor isointense compared to adjacent liver. On T2-weighted sequences, it is usually hyperintense relative to the liver and poorly delineated. The tumor usually enhances on T1-weighted immediate post-gadolinium images in a heterogeneous fashion, which facilitates differentiation from chronic cholecystitis (Kelekis and Semelka 1996). 6.1.6  Bile Ducts 6.1.6.1  Benign Diseases 6.1.6.1.1  Choledocholithiasis Calculi in the biliary ducts, although less frequent than in the gallbladder, are the most common cause of extrahepatic obstructive jaundice. MRCP is a non-invasive technique that is ideally suited for detecting bile duct stones owing to the high contrast of calculi as intraluminal low signal intensity or signal-void structures against high signal intensity bile. At MRI, ductal biliary stones typically have a rounded or oval-shaped configuration with a meniscus of fluid above their proximal edge. On thin-section source images, stones consistently appear as signal void foci, and can be detected at sizes as small as 2 mm in dilated and non-dilated ducts (Fig 6.1.26) (Motohara et al. 2003; Holzknecht et al. 1998). 6.1.6.1.2  Primary Sclerosing Cholangitis Approximately 71% of patients with primary sclerosing cholangitis (PSC) also have inflammatory bowel disease. Approximately 87% of these patients have ulcerative colitis and 13% have Crohn’s disease. PSC results in cholestasis with progression to secondary biliary cirrhosis and hepatic failure. The imaging appearance of PSC is characterized by multifocal, irregular strictures and dilatations of segments of the intra- and extrahepatic biliary tree. MRCP has shown to be an adequate method for the diagnosis and follow-up of PSC (Fig 6.1.28). Factors that can lead to difficulties in interpreting the MR images and to false-positive and false-negative diagnoses are (1) the presence of liver cirrhosis and (2) PSC limited to the peripheral intrahepatic ducts. Cirrhosis may lead to distortion of the biliary tree, which may mimic PSC even on ERCP images. The MRI findings of PSC include: peripheral, wedgeshaped zones of hyperintense signal on T2-weighted images measuring 1–5 cm in diameter, periportal edema, or inflammation, seen as high signal intensity along the porta hepatis on T2-weighted images. On pre-gado-

linium T1-weighted images, areas of increased signal intensity are observed without fatty infiltration. On immediate post-gadolinium images, increased parenchymal enhancement is present with patchy, peripheral, segmental, or a combination of these patterns. On delayed phase post-gadolinium images, thickening of bile duct walls and wall enhancement may be seen. Other findings occasionally associated with PSC are atrophy of liver segments, periportal lymphadenopathy, and findings attributable to liver cirrhosis and portal hypertension, such as hypertrophy of the caudate lobe, regenerative nodules, and abdominal varices. PSC is associated with an increased malignant potential, and the most important and common malignant entity that may occur in these patients is cholangiocarcinoma (Bader et al. 2003). 6.1.6.1.3  Infectious Cholangitis Infectious, bacterial, or ascending cholangitis is a clinically defined syndrome caused by complete or partial biliary obstruction with associated ascending infection from the intestine. The distribution of inflammatory changes may be diffuse or segmental. The most consistent imaging finding in infectious cholangitis is generalized or segmental biliary dilatation, which can be mild or severe, but does not correlate well with the severity or stage of the disease. Bile duct wall is commonly mildly to moderately thickened and shows increased enhancement, which can be best appreciated on T1-weighted fat-suppressed 2-min post-gadolinium images. Imaging findings on T2-weighted images are streaky increased signal in the periportal area and wedge-shaped hyperintense regions in the liver parenchyma. Liver abscesses may complicate infectious cholangitis and are best visualized on T2-weighted and T1-weighted dynamic post-gadolinium images. Thrombosis of the portal vein can be associated with infective cholangitis (Bader et al. 2001). 6.1.6.1.4  AIDS Cholangiopathy In HIV-positive patients, involvement of the pancreaticobiliary tract may be an early feature of AIDS. Inflammation and edema of the biliary mucosa resulting in mucosal thickening and irregularity is the hallmark of AIDS cholangiopathy. This may lead to strictures, dilatations, and pruning resembling sclerosing cholangitis. When the papilla of Vater is involved, ampullary stenosis with common bile duct dilatation may result. The gallbladder may also be involved and show acalculous cholecystitis with similar imaging features to acute cholecystitis. Furthermore, patients have a predisposition for superimposed infectious cholangitis often by unusual pathogens (e.g., cytomegalovirus, cryptosporidium, mycobacteriae, Candida albicans) (Miller et al. 1996).

895

896

6  Abdomen and Retroperitoneum

Fig. 6.1.28a–e  Primary sclerosing cholangitis (PSC). Maximum-intensity projection (MIP) from high-resolution T2weighted three-dimensional fast spin-echo imaging in two patients with primary sclerosing cholangitis. Focal bile-duct irregularities, stenosis, and interruptions can be appreciated

in all parts of the intrahepatic and extrahepatic biliary system (a arrows). High- and low-grade stenoses of peripheral bile ducts (b arrows) induce signs of obstruction and dilatation of the bile ducts until far into the liver periphery

6.1.6.1.5  Cystic Diseases of Bile Ducts

can demonstrate Caroli’s disease, with rounded cystic dilatations of equivalent signal intensity compared to bile communicating with bile ducts (Matos et al. 1998).

Congenital biliary cysts are classified by Todani as type I, choledochal cyst; type II, diverticulum of extrahepatic ducts; type III choledochocele; type IV, multiple segmental cysts; type V, Caroli’s disease. The most common cystic dilatations are choledochal cysts (77–87%), which are segmental aneurysmal dilatations of the CBD alone or the CBD and common hepatic duct (CHD). Choledochal cysts are associated with an increased incidence of other biliary anomalies, gallstone disease, pancreatitis, and cholangiocarcinoma. Choledochal cysts may also be coexistent with intrahepatic bile duct cysts (multiple segmental cysts). Cystic expansion of the common bile duct may also be short in length, the etiology of these are at present uncertain. Choledochoceles are cystic dilatations of the distal CBD that herniate into the lumen of the duodenum and create a “cobra-head” appearance on cholangiographic images. Caroli’s disease is an uncommon form of congenital dilatations of intrahepatic bile ducts with normal extrahepatic ducts. Thin-section T2- or T1-weighted images

6.1.6.2  Mass Lesions 6.1.6.2.1  Papillary and Ampullary Adenoma Papillary and ampullary adenomas of the biliary tract are rare benign epithelial tumors. They are depicted by MRCP and reveal homogenous enhancement on postgadolinium images (Motohara et al. 2003). 6.1.6.2.2  Cholangiocarcinoma Cholangiocarcinoma is a biliary malignancy of older patients (>50 years). Patients usually present with jaundice and weight loss. Three types of cholangiocarcinomas can be differentiated based on the anatomical distribution: the peripheral (or intrahepatic) type arising from

6.1  Abdominal MRI

peripheral bile ducts in the liver, the hilar type (Klatskin tumor) with its origin at the confluence of the right and left hepatic ducts, and the extrahepatic type arising from the main hepatic ducts, CHD or CBD. Peripheral cholangiocarcinomas usually present as mass-like lesions that do not obstruct the central bile ducts. Their typical MR imaging appearance is a mass lesion that is mildly heterogeneous with moderately low signal intensity on T1weighted images and mildly to moderately hyperintense signal on T2-weighted images (Worawattanakul et al. 1998). On immediate post-gadolinium images, they usually show mild to moderate enhancement that is usually diffuse heterogeneous in pattern. Progressive enhancement may be observed on late fat-suppressed images reflecting a high content of fibrous tissue. Klatskin tumors are usually small-volume, superficial, spreading tumors that result in early biliary obstruction and dilatation of proximal ducts. They show circumferential growth and spread along bile ducts with poor conspicuity on non-contrast MR images. Lobar atrophy of the liver combined with marked biliary dilatation is an associated MRI finding. Extrahepatic cholangiocarcinomas usually grow in a circumferential pattern similar to Klatskin tumors. They arise in the CBD and result in biliary obstruction in the vast majority of patients. On T1-weighted MR images with or without fat-suppression, cholangiocarcinomas appear mildly to moderately hypointense but may also be isointense relative to liver parenchyma. On T2-weighted images, they are isointense or mildly hyperintense. Thickening of bile duct walls greater than 5 mm is highly suggestive of cholangiocarcinoma. On immediate post-gadolinium images, cholangiocarcinomas are usually hypovascular showing minimal or moderate enhancement that intensifies on delayed images (Fig 6.1.29). A combination of early and late fat-suppressed gadolinium-enhanced images is very helpful to identify these tumors. Fat-suppression also reduces the signal of fatty tissue in the porta hepatis, which improves the conspicuity of cholangiocarcinomas and facilitates the evaluation of the extent of tumor and infiltration into adjacent tissues and organs (Worawattanakul et al. 1998). 6.1.6.2.3  Periampullary and Ampullary Carcinoma Carcinomas arising from the ampulla of Vater, periampullary duodenum or distal CBD are grouped together and termed periampullary carcinomas. Their presentation is similar to that of pancreatic head ductal adenocarcinoma including obstruction of both the CBD and pancreatic duct. The prognosis of periampullary carcinoma is significantly better than that of pancreatic carcinoma with a 5-year survival rate up to 85%. Periampullary carcinomas can cause ampullary obstruction and become clinically symptomatic even when they are only

a few millimeters in size. MRCP is very effective for the visualization of biliary and pancreatic ductal dilatation and the determination of the level of obstruction. On T1-weighted fat-suppressed images, periampullary carcinomas typically appear as low-signal-intensity masses. On immediate post-gadolinium T1-weighted images, pancreatic parenchyma enhances more than tumor, even in the presence of chronic pancreatitis that results from pancreatic outlet obstruction. Periampullary carcinomas enhance minimally on early post-gadolinium images due to their hypovascular character. On 2-min post-gadolinium fat-suppressed images, delayed enhancement is a typical finding. A thin rim of enhancement is commonly observed along the periphery of these tumors, and may also be a relatively specific finding (Semelka et al. 1997). 6.1.7  Spleen Normal splenic parenchyma is invariably low in signal intensity on T1-weighted images and usually high in signal intensity on T2-weighted images. Low signal intensity on T2-weighted images of the spleen is associated with prior blood transfusions, which result in iron deposition in the reticuloendothelial system (RES) of the spleen. Immediate post-gadolinium T1-weighted gradient-echo sequences demonstrate the different circulations in the normal spleen as regions of transient higher- and lower-contrast enhancement, usually in an arciform or serpiginous pattern. In the neonate and until the infant is approximately 8 months old, the spleen signal intensity is isointense to the liver on T1-weighted images and varies from iso to hypointense relative to the liver on T2-weighted images. As the reticuloendothelial system matures, the spleen displays a hypointense signal relative to the liver on T1-weighted images, with a gradual increase in the signal relative to the liver on T2-weighted images, approaching the normal appearance of the adult spleen (Semelka et al. 1992). 6.1.7.1  Normal Variants and Congenital Diseases An accessory spleen is a congenital ectopic focus of splenic tissue that fails to fuse with the spleen. This anatomic variant is found in 10% of the population. The majority of accessory spleens are located in close proximity to the splenic hilum. Asplenia syndrome (right isomerism or Ivemark’s Syndrome) is a congenital syndrome characterized by absence of the spleen associated with thoracoabdominal abnormalities. The majority of patients die in infancy, with few surviving longer than 1 year. The mortality in the first year of life approaches 80% due to complex and severe cardiovascular anomalies and a compromised immune system. Polysplenia syndrome is a congenital syndrome characterized by multiple small splenic masses and left isomerism. The splenic

897

898

6  Abdomen and Retroperitoneum

Fig. 6.1.29a–h  Central cholangiocarcinoma: Klatskin tumor axial T2-weighted fast spin-echo (a) and single-shot echo-train spin-echo (b,c) images, pre-contrast T1-weighted spoiled gradient-echo (SGE) (d) and post-gadolinium T1-weighted fatsaturated three dimensional gradient-echo images in arterial dominant, portal venous– and late-venous phase (e–g). On T2-weighted images, slightly hyperintense tissue can be identified encasing the bifurcation of the hepatic bile ducts (a,b small arrows). The proximal portion of the common bile duct is in-

volved in this process, which causes marked dilatation intrahepatic bile ducts that is best seen on coronal MRCP image (h); the site of compression is located between the two arrowheads. The enclosing tissue is hypointense relative to the adjacent hepatic tissue on T1-weighted image before and early after contrast application (e,f arrows) and shows discrete contrast uptake in late phase (g). Note that the portal vein is also stenosed at the left branch and the bifurcation due to the encasement in this patient with central cholangiocarcinoma

masses vary from 2 to 16 in number and are distributed along the greater curvature of the stomach. Other associated abnormalities include cardiopulmonary anomalies, malrotation of the intestinal tract, absence of the hepatic

segment of the inferior vena cava with azygous or hemiazygous continuation and a short pancreas. Polysplenia has also been associated with polycystic kidney disease (Semelka et al. 1992).

6.1  Abdominal MRI

masses vary from 2 to 16 in number and are distributed along the greater curvature of the stomach. Other associated abnormalities include cardiopulmonary anomalies, malrotation of the intestinal tract, absence of the hepatic segment of the inferior vena cava with azygous or hemiazygous continuation and a short pancreas. Polysplenia has also been associated with polycystic kidney disease (Semelka et al. 1992). 6.1.7.1.1  Sickle-Cell Disease The manifestations of sickle-cell anemia vary and depend on whether the patient is homozygous or heterozygous for the hemoglobinopathy. In patients with homozygous disease, the spleen is nearly a signal void due to the sequela of iron deposition from blood transfusions coupled with microscopic perivascular and parenchymal calcifications (Worawattanakul et al. 1997). This decrease in signal intensity was found to be diffuse in most patients with signal-void foci due to calcifications and/or foci of greater iron deposition. Hyperintense focal lesions on proton-density images may occur and are believed to represent infarcts (Adler et al. 1986). 6.1.7.2  Mass Lesions 6.1.7.2.1  Benign Masses Cysts Cysts are the most common of the benign splenic lesions. Three types of non-neoplastic cysts exist: posttraumatic or pseudocysts, epidermoid cysts, and hydatid cysts. Most splenic cysts are posttraumatic in origin. Epidermoid cysts are true cysts discovered in childhood or early adulthood that may have trabeculations or septations in their walls with occasional peripheral calcification. Hydatids, or echinococcal cysts, are rare. They are characterized by extensive wall calcification. The MRI features of cysts include sharp lesion margination, low signal intensity on T1-weighted images, and very high signal intensity on T2-weighted images. Complicated cysts are high signal intensity on T1-weighted images, regions of mixed signal intensity on T2-weighted images, or both. Cysts do not enhance on post-gadolinium images. Pseudocysts may be complicated by hemorrhage particularly early in their evolution and thus may contain foci of high signal intensity on pre-contrast T1-weighted images (Urrutia et al. 1996). Hemangiomas Hemangiomas are the most common of the benign splenic neoplasms. Lesions may be single or multiple. Splenic hemangiomas are mildly low to isointense on T1weighted images and mildly to moderately hyperintense on T2-weighted images, similar to hepatic hemangiomas.

Three patterns of contrast enhancement are observed: (1) immediate homogeneous enhancement with persistent enhancement on delayed images, (2) peripheral enhancement with progression to uniform enhancement on delayed images (Fig 6.1.30), and (3) peripheral enhancement with centripetal progression but persistent lack of enhancement of central scar. Unlike hepatic hemangiomas, splenic hemangiomas generally do not demonstrate welldefined nodules on early post-gadolinium images. Uniform high signal intensity on immediate post-gadolinium SGE images is a common appearance for small (<1.5 cm) hemangiomas, as it is with hepatic hemangiomas. Rarely hemangiomas with a very large central scar can appear hypointense on T2-weighted images, reflecting the lower fluid content of the central scar (Disler and Chew 1991). Hamartomas Hamartomas are rare and composed of structurally disorganized mature splenic red pulp elements. The lesions tend to be single, spherical, and predominantly solid. Hamartomas are mildly low to isointense on T1weighted images and moderately high in signal intensity on T2-weighted images. They frequently are moderately heterogeneous and may have regions of low signal intensity on T2-weighted images. Hamartomas enhance on immediate post-gadolinium SGE images in an intense diffuse heterogeneous fashion. Enhancement becomes homogeneous on more delayed images with signal intensity slightly greater than in background spleen. The early diffuse heterogeneous enhancement permits distinction from hemangioma (Ramani et al. 1997) (Fig. 6.1.31). 6.1.7.2.2  Malignant Masses Lymphoma and Other Hematologic Malignancies Hodgkin’s and non-Hodgkin’s lymphomas often involve the spleen. Lymphomatous deposits in the spleen frequently parallel the signal intensity of splenic parenchyma on T1- and T2-weighted images. Diffuse involvement of lymphoma may appear as large, irregularly enhancing regions of high and low signal intensity, in contrast to the uniform bands that characterize normal arciform enhancement. Multifocal disease appears as focal low-signal-intensity mass lesions scattered throughout the spleen (Fig. 6.1.32). Focal involvement appears as spherical lesions in distinction to the wavy tubular pattern of arciform enhancement of uninvolved spleen. Focal lymphomatous deposits may be low in signal intensity compared to background spleen on T2-weighted images, which is a feature distinguishing lymphomas from metastases. It is critical to acquire gradient-echo images within the first 30 s after contrast administration because foci of lymphoma equilibrate early, becoming isointense with normal splenic tissue within 2 min and frequently earlier. Superparamagnetic iron-oxide particles also improve the accuracy of diagnosing splenic lymphoma. These parti-

899

900

6  Abdomen and Retroperitoneum

Fig. 6.1.30a–d  Splenic hemangioma. Axial T2-weighted fatsuppressed single-shot echo-train spin-echo image (a) reveals hyperintense lesion in the spleen (arrow) that is not appreciated on axial T1-weighted spoiled gradient-echo (SGE) image (b).

Axial arterial dominant-phase SGE contrast-enhanced image reveals peripheral enhancement of the lesion (c arrow). On late phase axial contrast-enhanced SGE image (d), contrast fills in and the lesion reveals persistent enhancement (arrow)

cles are selectively taken up by the RES cells, and cause a decrease in signal intensity. By contrast, malignant cells do not take up superparamagnetic iron-oxide particles. Therefore, splenic lymphoma remains hyperintense compared to the normal spleen, improving tumor-spleen contrast. Chronic lymphocytic leukemia frequently involves the spleen and may result in massive splenomegaly. Focal deposits are more infiltrative and less well defined than lymphoma. Deposits are well shown after gadolinium administration and appear as irregular hypointense masses on early post-contrast images. Lymphadenopathy is frequently present (Hahn et al. 1988; Rabushka et al. 1994).

Lesion detection is best performed, by acquiring immediate post-gadolinium SGE images. Metastases are lower in signal intensity than normal splenic tissue on these images. Images must be acquired within the first 30 s after gadolinium administration because metastases rapidly equilibrate with splenic parenchyma. Image acquisition with superparamagnetic iron oxide particles renders metastases higher in signal intensity than normal spleen. An attractive feature of iron oxide particles is that the imaging window is longer (60 min) than for gadolinium (< 1 min) (Hahn et al. 1988; Rabushka et al. 1994).

Metastases Although tumors may invade the spleen from contiguous viscera, true tumor metastasis to the spleen is rare, usually occurring only in the setting of disseminated disease in the terminal stage. Breast cancer, lung cancer, and melanoma are the most common primary tumors.

6.1.7.2.3  Miscellaneous Splenomegaly Splenomegaly may be observed in a number of disease states including venous congestion (portal hypertension), leukemia, lymphoma, metastases, and various infections. On immediate post-gadolinium images demonstration of

6.1  Abdominal MRI

Fig. 6.1.31a–d  Splenic hamartoma. Axial T2-weighted fat-suppressed single-shot echo-train spin-echo image (a) reveals heterogenous hyperintense mass in the spleen (arrows). On axial T1-weighted spoiled gradient-echo (SGE) image (b), the mass is minimally and homogenously hypointense relative the normal

splenic parenchyma. Arterial phase (c) axial contrast-enhanced image reveals heterogenous enhancement of the lesion (arrows). On portal venous-phase axial SGE image (d), the lesion tends to be isointense with the spleen, although the lesion shows some heterogeneity (arrows)

arciform or uniform high-signal-intensity enhancement is consistent with portal hypertension and excludes the presence of malignant disease.

visualized on pre-contrast gradient-echo images. Bacterial and fungal abscesses are rare in the spleen. Abscesses appear slightly hypointense to isointense on T1-weighted images and heterogeneous and mildly to moderately hyperintense on T2-weighted images. These lesions show intense mural enhancement on early gadolinium-enhanced images. This pattern persists on later post-gadolinium images, accompanied by the presence of periabscess-increased enhancement of surrounding tissue on immediate post-gadolinium images (Rabushka et al. 1994).

Infection Viral infection may result in splenomegaly. The three most common viruses to involve the spleen are EpsteinBarr, varicella, and cytomegalovirus. Nonviral infectious agents that involve the spleen in patients with normal immune status include histoplasmosis, tuberculosis, and echinococcosis. These infectious agents are observed in immunocompromised patients with an even greater frequency. In the immunocompromised patient, the most common hepatosplenic infection is fungal infection with Candida albicans and Cryptococcus. In the acute phase, hepatosplenic candidiasis results in small (<1 cm) welldefined abscesses in the spleen and liver. They are well shown on T2-weighted fat-suppressed images as highsignal-intensity rounded foci. Lesions also may be visible on post-gadolinium images, but they usually are not

Sarcoidosis Lesions of sarcoidosis are small (<1 cm) and hypovascular. Due to their hypovascularity, the lesions are low in signal intensity on T1- and T2-weighted images and enhance on gadolinium-enhanced images in a minimal and delayed fashion. Low signal intensity on T2-weighted images is a feature that distinguishes these lesions from acute infective lesions (Warshauer et al. 1994).

901

902

6  Abdomen and Retroperitoneum

Fig. 6.1.32a–d  Splenic involvement in Hodgkin’s lymphoma. Axial T2-weighted single-shot echo- train spin-echo image (a) shows multiple hyperintense lesions in the spleen. On axial T1weighted fat-saturated spoiled gradient-echo image (SGE) (b), no definite lesion is appreciated. On axial arterial phase post-

contrast SGE image (c) and portal venous-phase SGE image (d), the splenic lesions are hypointense without contrast enhancement. Adjacent to the portal hilus, a conglomerate lymphadenopathy is noted

Gamna-Gandy Bodies Foci of iron deposition occur commonly in patients with cirrhosis and portal hypertension due to micro hemorrhages in the splenic parenchyma. On occasion, such foci are observed in patients receiving blood transfusions. Lesions vary in size but are generally smaller than 1 cm. Lesions demonstrate signal void on all pulse sequences (Rabushka et al. 1994). Susceptibility artifact is demonstrated on gradient-echo images as blooming artifact, and this artifact is pathognomonic for this entity.

Traumatic injury of the spleen, especially devascularization, is well shown on immediate post-gadolinium gradient-echo images. Areas of devascularization are nearly signal voids compared to the high signal intensity of vascularized tissue (Rabushka et al. 1994).

Trauma The spleen is the most commonly ruptured abdominal organ in the setting of trauma. Injury to the spleen may take several forms: subcapsular hematoma, contusion, laceration, and devascularization/infarct. Subcapsular or intraparenchymal hematoma secondary to contusion or laceration demonstrates a time course of changes in signal intensity due to the paramagnetic properties of the degradation products of hemoglobin. Subacute hemorrhage is particularly conspicuous because of its distinctive high signal intensity on T1- and T2-weighted images.

Infarcts Splenic infarcts are a common occurrence in the setting of obstruction of the splenic artery or one of its branches. The most common cause is cardiac emboli, but local thrombosis, vasculitis, and splenic torsion are also described. Infarcts appear as peripheral wedge-shaped, round, or linear defects that are most clearly defined on 1 to 5 min post-gadolinium images as low-signal-intensity wedge-shaped regions. The splenic capsule is commonly observed as a thin peripheral, enhancing linear structure. Massive splenic infarcts may appear as diffuse low signal intensity on T1-weighted images and inhomogeneous high signal intensity on T2-weighted images. Lack of enhancement on early and late post-gadolinium images of wedge-shaped regions is the most diagnostic feature (Rabushka et al. 1994).

6.1  Abdominal MRI

6.1.8  Peritoneum 6.1.8.1  Hernias Bochdalek’s hernia is characterized by a posterolateral defect of the diaphragm. Defective formation and/or fusion of the pleuroperitoneal membrane result in herniation of abdominal contents into the thoracic cavity at the posterior aspect of the diaphragm. Esophageal hiatal hernias are acquired lesions occurring during adult life and considered as sliding (axial) and paraesophageal. Abdominal wall hernias are classified as spigelian, paraumbilical and inguinal hernias and MR images acquired with breath-hold or single-shot techniques can identify abdominal wall hernias (Lee and Cohen 1993). 6.1.8.2  Benign Masses 6.1.8.2.1  Cysts Mesenteric cysts most commonly occur in the small bowel mesentery. The cysts tend to be single and thin-walled, and may contain septae. Different types of mesenteric cysts may be lined by a diversity of cell types including endothelium, mesothelium, and fallopian tube-like epithelium. The MRI appearance reflects the cyst’s contents. Simple cysts have round, well-marginated contours, with low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Complex cysts have higher signal intensity on T1-weighted images and/or heterogeneous signal intensity on T2-weighted images. Following contrast administration, the cyst wall and septae, if present, enhance (Vanek and Phillips 1984). 6.1.8.2.2  Lipomas and Mesenteric Lipomatosis Lipomas are benign tumors that rarely involve the peritoneal cavity. These lesions are high in signal intensity on non-suppressed T1- and T2-weighted images (Lewis et al. 1982). 6.1.8.2.3  Endometriosis Endometriosis is defined as the presence of endometrial glands or stroma in abnormal locations outside of the uterus. The most common peritoneal sites of involvement are, in decreasing order of frequency, the ovaries, uterine ligaments, cul-de-sac, and pelvic peritoneum reflected over the uterus, fallopian tubes, rectosigmoid region, and bladder. Endometriomas have variable signal intensity but are commonly high in signal intensity on T1-weighted images and heterogeneously high in signal intensity on

T2-weighted images (Wenzel et al. 2001. Protein and blood breakdown products tend to demonstrate a gradation of signal intensity on T2-weighted images, which has been termed shading. Non-contrast T1-weighted fat-suppressed imaging is the most sensitive MRI technique for identifying endometriomas (Ascher et al. 1995). 6.1.8.2.4  Desmoid Tumor (Aggressive Fibromatosis) Desmoid tumor is a rare gastrointestinal mesenchymal tumor. Intra-abdominal desmoid tumors occur in the mesentery or the pelvic wall. Grossly, desmoid tumors vary in size, from 1 to 15 cm in greater diameter. In general, they are unicentric, infiltrative lesions with poorly defined borders. Longstanding tumors are low in signal intensity on T1- and T2-weighted images and enhance only minimally following intravenous gadolinium chelate. In the acute phase, tumors may have regions of high signal intensity on T2-weighted images that also show heterogeneous increased enhancement (Ascher et al. 1995). 6.1.8.3  Malignant Masses 6.1.8.3.1  Diffuse Malignant Mesothelioma The term mesothelioma describes a malignant tumor derived of peritoneum and pleura. MRI findings include nodules and plaques along the surface of the peritoneum that become confluent over time. Malignant mesothelioma may spread along serosal surfaces, invading underlying tissue, especially the wall of the intestine, and adjacent organs, such as the liver. Intraperitoneal adhesions are not uncommon. A distinctive feature of mesotheliomas may be the cystic foci interspaced through solid tumor, which may be unusual in many other forms of peritoneal disease (with the exception of ovarian cancer) (Hamrick et al. 1992b). 6.1.8.3.2  Metastases Metastatic tumors that involve the peritoneum most commonly arise from the female genital tract, particularly the ovary, followed by the colon, stomach, and pancreas. The gross appearance of metastases ranges from single, well-defined nodules to diffuse peritoneal thickening. Peritoneal metastases occasionally appear as large cystic masses with multiple septations and layering low signal proteinaceous material in a multi-tier fashion along the septations. This may be a distinctive appearance for cystic metastasis. Septations are best seen on T2-weighted single-shot echo-train spin-echo images.

903

904

6  Abdomen and Retroperitoneum

Dissemination occurs by several routes: contiguous spread, intraperitoneal seeding, hematogenous spread, and lymphatic dissemination (Hamrick et al. 1992b) (Fig 6.1.33). 6.1.8.4  Intraperitoneal Fluid 6.1.8.4.1  Ascites Ascites results from either overproduction or impaired resorption or leakage of fluid. It is a common manifestation of many diseases: cirrhosis, pancreatitis, obstruction (venous or lymphatic), inflammation, low albumin states, malignancy, and trauma. Simple transudates are low in signal intensity on T1-weighted sequences and very high in signal intensity on T2-weighted images, whereas exudates, blood, and enteric contents will have higher signal intensity on T1-weighted images and more variable signal intensity on T2-weighted images. Benign processes favor the greater sac, whereas malignant fluid tends to involve the greater and lesser sac proportionally (Walls et al. 1986).

Fig. 6.1.33a–d  Peritoneal metastasis from endometrial carcinoma. Axial T2-weighted fat-suppressed single-shot echo-train spin-echo image (a) reveals peritoneal mass with hyperintense signal (arrow). The mass is hypointense on axial T1-weighted

6.1.8.4.2  Intraperitoneal Blood Intraperitoneal blood most frequently occurs in the setting of trauma. MRI can readily distinguish blood from ascites. The age of hemorrhage is determined by MRI. Acute blood (<48 h), in the form of deoxyhemoglobin, is low in signal intensity on both T1- and T2-weighted images. From 48 h to 7 days intracellular methemoglobin may be observed, which has high signal intensity on T1and low signal intensity on T2-weighted images. Subacute blood, in the form of extracellular methemoglobin, is high in signal intensity on T1- and T2-weighted images. A structure with a high-signal-intensity rim on non-contrast T1-weighted images is characteristic of subacute hematoma. This distinctive imaging feature represents extracellular methemoglobin encircling the retracting clot (Walls et al. 1986). As hematomas age, a low-signalintensity rim develops around the hematoma on both T1- and T2-weighted sequences. This rim corresponds to hemosiderin and/or fibrosis.

spoiled gradient-echo (SGE) image (b) (arrow). The mass reveals enhancement on arterial phase (c) and late venous phase (d) contrast-enhanced axial SGE images (arrows)

6.1  Abdominal MRI

6.1.8.4.3  Intraperitoneal Bile Free intraperitoneal bile is usually the result of surgery (Walls et al. 1986). Free bile preferentially collects in the right upper quadrant where it incites an inflammatory reaction. A biloma results if the bile is walled off by a pseudocapsule and adhesions. The signal intensity of a biloma is variable and mimics that of the gallbladder. Bilomas may be low, intermediate, or high in signal intensity on T1-weighted images. They are high in signal intensity on T2-weighted images. Enhancement of peritoneum on gadolinium-enhanced T1-weighted images reflects the inflammation associated with bile leak. 6. 1.8.5  Inflammation 6.1.8.5.1  Mesenteric Panniculitis (Isolated Lipodystrophy of the Mesentery, Retractile Mesenteritis, Sclerosing Mesenteritis) Mesenteric panniculitis is a rare disorder characterized grossly by a diffuse, localized, or multinodular fibrofatty thickening of the mesentery of the small and/or large bowel. The etiology of mesenteric panniculitis is unclear, and infection, trauma, ischemia, autoimmune disorders, and a history of previous abdominal surgery have been suggested as causative factors (Walls et al. 1986). Focal or diffuse changes are observed in the mesentery. When diffuse, the mesenteric fat is traversed by low-signal-intensity strands on T1-weighted images. In the focal form, heterogeneous nodular masses of fat necrosis are noted (Katz et al. 1985). 6.1.8.5.2  Peritonitis Peritonitis results from a variety of infectious or noninfectious causes, many of which are related to bowel perforation. Trauma, complications of surgery and inflammatory bowel disease are common underlying causes. Peritonitis appears as diffuse increased enhancement of the peritoneum and mesentery, and is most clearly defined on interstitial-phase gadolinium-enhanced fat-suppressed gradient-echo images. Pseudocysts may develop in the setting of peritonitis as walled-off collections of fluid. In uncomplicated cases they are low in signal intensity on T1-weighted images and very high in signal intensity on T2-weighted images. Complex fluid is characterized by either increased intensity signal on T1-weighted images, decreased or heterogeneous signal intensity on T2-weighted images, or a combination of both. Gadolinium-enhanced T1-weighted images reveal increased enhancement and occasionally

Fig. 6.1.34a,b  Peritonitis. Axial T2-weighted single-shot echotrain spin-echo image (a) reveals moderate amount of ascites. On post-contrast T1-weighted axial spoiled gradient-echo image (b), increased peritoneal enhancement is observed (arrows)

increased thickness of the inflamed peritoneum, which is more conspicuous in combination with fat-suppression (Fig. 6.1.34) (Kanematsu et al. 1997). 6.1.8.5.3  Abscess Intra-abdominal abscesses are most often the sequelae of gastrointestinal or biliary surgery, diverticulitis, and Crohn’s disease. In the appropriate clinical setting, a focal fluid collection that demonstrates rim enhancement on gadolinium-enhanced images suggests the correct diagnosis. The addition of fat suppression and image acquisition 2–10 min after injection (interstitial phase) can highlight the enhancement of the abscess wall and surrounding tissues. Layering of lower-signal-intensity debris in the dependent portion of the cystic lesion on T2-weighted images is a common finding in abscesses reflecting the layering of high protein content dependently in abscesses. This is a very specific finding for abscess.

905

906

6  Abdomen and Retroperitoneum

When air is identified within a fluid collection, active infection, and/or fistula to the bowel is present. The combination of breathing independent T2-weighted echo-train spin-echo, gadolinium-enhanced capillary-phase T1weighted gradient-echo, interstitial-phase fat-suppressed gradient-echo, and multiplanar imaging, render MRI a very accurate technique for detecting intraperitoneal abscesses. MRI may be the technique of choice in patients who have dense intraluminal barium contrast, renal failure, or allergy to iodine (Noone et al. 1998). 6.1.9  Gastrointestinal Tract MRI is able to identify virtually all abnormalities of the gastrointestinal tract. In this section, the most common indications will be described: gastric carcinoma, Crohn’s disease, and colorectal carcinoma.

Fig. 6.1.35a–d  Gastric carcinoma. There is an exophytic, polypoid gastric carcinoma at the lesser curvature of the stomach projecting into the lumen. The tumor (arrows) is heterogeneous in signal intensity on coronal T2-weighted single-shot echotrain spin-echo image (a) and hypointense on axial T1-weighted

6.1.9.1  Gastric Carcinoma Carcinoma is the most important and the most common tumor of the stomach. Most gastric carcinomas are adenocarcinomas. Predisposing conditions include atrophic gastritis, pernicious anemia, adenomatous polyps, dietary nitrates, and Japanese heritage. The tumors show a predilection for the lesser curvature of the antropyloric region. Grossly, adenocarcinomas of the stomach can be divided generally into three forms: (1) exophytic or polypoid, projecting into the lumen; (2) ulcerated, with a shallow or deeply erosive crater; and (3) diffusely infiltrative (linitis plastica). Gastric cancer may spread hematogenously to the liver and lung, contiguously to adjacent organs, lymphatically to regional and remote lymph nodes, and/or intraperitoneally to the abdominal lining, mesentery, and serosa. The goals of MRI in patients with gastric cancer (Fig. 6.1.35) are to demonstrate the primary tumor, as-

spoiled gradient-echo (SGE) image (b). The tumor demonstrates mild enhancement in the arterial phase (c) and prominent enhancement in the portal venous phase (d) on axial post-gadolinium T1-weighted SGE images

6.1  Abdominal MRI

sess the depth of invasion, and detect extra gastric disease. Adequate distention is necessary for surveying the gastric wall. On T1-weighted sequences, gastric adenocarcinoma is isointense to normal stomach wall and may be apparent as focal wall thickening. On T2-weighted images, tumors usually are slightly higher in signal intensity than adjacent normal wall except diffusely infiltrative carcinoma (linitis plastica carcinoma), which tends to be lower in signal intensity than normal adjacent wall because of its desmoplastic nature (Auh et al. 1994). Tumors show heterogeneous enhancement that may be decreased or increased relative to the gastric wall on early, late, or both sets of images (Marcos and Semelka 1999). Linitis plastica carcinoma enhances only modestly after intravenous contrast. Gadolinium-enhanced fat-suppressed GE imaging aids in identification of transmural spread including peritoneal disease, tumor involvement of lymph nodes and metastases particularly if it is dynamic.

Crohn’s disease is the most common inflammatory condition affecting the small bowel in North America. Although any part of the gastrointestinal tract, from the mouth to the anus, may become involved with Crohn’s disease, it most commonly involves the terminal ileum. Involvement of the terminal ileum occurs in approximately 70% of patients often together with cecum. Twenty to 30% will have isolated colon involvement. Five percent of patients will manifest Crohn’s disease in the duodenum or jejunum. Patients with longstanding Crohn’s disease have a well-documented increased incidence of cancer (approximately 3% of patients) of the gastrointestinal tract usually involving the colon or ileum. Changes of Crohn’s disease are well shown on MRI (Fig. 6.1.36). T2-weighted single-shot echo-train spinecho and gadolinium-enhanced T1-weighted fat-suppressed SGE images demonstrate characteristic findings:

Fig. 6.1.36a–c  Crohn’s disease. In a patient with a known history of Crohn’s disease and past terminal ileum resection, the neoterminal ileum (thick arrows) shows wall thickening and prominent enhancement at the junction of the cecum (arrowhead) on axial T2-weighted single-shot echo-train spin-echo (a),

axial (b), and sagittal (c) T1-weighted post-gadolinium three dimensional gradient-echo (3D GE) images. Additionally, the wall thickening with prominent enhancement (thin arrow) is seen in the sigmoid colon on axial T1-weighted post-gadolinium 3D GE image (b)

6.1.9.2  Crohn’s Disease

907

908

6  Abdomen and Retroperitoneum

Fig. 6.1.37a–c  Rectal cancer. There is a tumor arising from the left aspect of the rectum and extending to the perirectal fascia on coronal T2-weighted single-shot echo-train spin-echo (a) and high-resolution axial T2weighted spin-echo (b) images. The tumor constricts the rectal lumen and there is surrounding perirectal fat stranding. On axial T1-weighted post-gadolinium spoiled gradient-echo image (c), heterogeneous and prominent enhancement of the tumor is noticed

transmural involvement, skip lesions, and mesenteric inflammatory changes. Single-shot echo-train spin-echo and gadolinium-enhanced T1 SGE images are also useful for the evaluation of complications of Crohn’s disease (Marcos and Semelka 2000). Single-shot echo-train spinecho image is a very effective technique to demonstrate dilated obstructed bowel developing secondary to strictures and edema, whereas gadolinium-enhanced fat-suppressed SGE is useful in demonstrating inflammatory changes in bowel and mesentery. Both techniques were effective in showing wall thickening, abscess, and fistulae formation. The MRI criteria of mild, moderate, and severe disease has been described and is a function of wall thickness, length of diseased segment, and extent of mural contrast enhancement. The extent of mural enhancement may also be determined by comparison of bowel enhancement on gadolinium-enhanced fat-suppressed SGE with that of

the renal parenchyma (Marcos and Semelka 2000). Bowel should not enhance to the same degree as renal cortex on either early capillary-phase images or > 1 min interstitialphase images. Enhancement equivalent or greater than renal cortex is abnormal and most often reflects the presence of inflammatory change. MRI also may have a role in the evaluation of acute exacerbations of Crohn’s disease. Specifically, in patients with longstanding disease, marked enhancement of the mucosa with substantially thickened wall and minimal enhancement of the outer layer is suggestive of acute-on-chronic involvement, and may have a role in the evaluation of acute exacerbations of Crohn’s disease. In patients with non-active chronic disease, there may remain persistent thickening and abnormal enhancement seen on delayed post-gadolinium images. Acute disease will result in edema that may be best visualized on fat-suppressed single shot T2-weighted images.

6.1  Abdominal MRI

MRI may play an important role in the determination of Crohn’s disease activity because of its high contrast resolution, high sensitivity to the contrast material and its safety compared to CT and barium studies. 6.1.9.3  Colorectal Carcinoma Adenocarcinoma of the colon is the most common gastrointestinal tract malignancy and the second most common visceral cancer in North America. The conditions that predispose to the development of colon cancer include heredity, familial adenomatous polyposis, Gardner syndrome, Lynch syndrome, ulcerative colitis, and Crohn’s colitis. Cancers occur most often in the rectosigmoid colon, but right-sided cancers tend to occur in increasing frequency. Tumors may be polypoid, circumferential (“apple core”), or plaque-like. Good correlation is observed between gadoliniumenhanced fat-suppressed MRI techniques and surgical specimens for tumor size, bowel wall involvement, peritumoral extension, and lymph node detection (Shoenut et al. 1993). Malignant lymph nodes are usually not enlarged in gastrointestinal adenocarcinoma. However, the presence of more than five lymph nodes that measure smaller than 1 cm in a regional distribution related to the tumor correlates well with tumor involvement. All segments of the colon and the appendix are well shown on MR images. The combination of T2-weighted single-shot echo-train spin-echo and gadolinium-enhanced fat-suppressed SGE images result in the most reproducible image quality for the colon above the rectum. Rectal and colon cancers benefit from the combined use of gadolinium-enhanced fat-suppressed SGE and high-resolution T2-weighted echo-train spin-echo images. MR colonography employing a bowel cleansing preparation and administration of rectal water enema has been shown effective in demonstrating small polyps and tumors (Ajaj et al. 2003). Gadolinium-enhanced fat-suppressed SGE imaging is valuable in demonstrating perirectal tumor extension (Fig. 6.1.37), regional lymph nodes, and seeding of peritoneal by tumor. This reflects the high-contrast resolution of this technique for detecting enhancing diseased tissue. Image acquisition of T2-weighted echo-train spin-echo or single-shot echo train-spin echo after the administration of gadolinium is commonly done when abdomen and pelvis studies are combined in one examination. As an additional benefit to a shortened MR examination, dependent, concentrated gadolinium in the bladder, which is low in signal intensity, may increase the conspicuity of high-signal intensity rectal tumor invasion of the bladder wall. Endorectal coil imaging permits differentiation of the anatomic layers of the rectal wall on T2-weighted fatsuppressed images. Local staging of rectal carcinoma also benefits from endorectal coil imaging.

References 1.

2.

3.

4.

5.

6.

7. 8.

9. 10.

11.

12.

13. 14.

15.

16.

17.

Adler DD, Glazer GM, Aisen AM (1986) MRI of the spleen: normal appearance and findings in sickle-cell anemia. AJR Am J Roentgenol 147:843–845 Ajaj W, Pelster G, Treichel U, Vogt FM, Debatin JF, Ruehm SG, Lauenstein TC (2003) Dark lumen magnetic resonance colonography: comparison with conventional colonoscopy for the detection of colorectal pathology. Gut 52:1738–1743 Ascher SM, Agrawal R, Bis KG, Brown E et al. (1995) Endometriosis: Appearance and detection with conventional, fat-suppressed, and contrast-enhanced fat-suppressed spin-echo techniques. J Magn Reson Imaging 5:251–257 Auh Yh, Lim T-H, Lee DH, Young YK et al. (1994) In vitro MR imaging of the resected stomach with a 4.7-T Super conducting magnet. Radiology 191:129–134 Bader TR, Beavers KL, Semelka RC (2003) MR imaging features of primary sclerosing cholangitis: patterns of cirrhosis in relationship to clinical severity of disease. Radiology 226:675–685 Bader TR, Braga L, Beavers KL, Semelka RC (2001) MR imaging findings of infectious cholangitis. Magn Reson Imaging 19:781788 Balci NC, Sirvanci M (2002) MR imaging of infective liver lesions. Magn Reson Imaging. 10:121–135 Corrigan K, Semelka RC (1995) Dynamic contrast-enhanced MR imaging of fibrolamellar hepatocellular carcinoma. Abdom Imaging 20:122–125 Disler DG, Chew FS (1991) Splenic hemangioma. AJR Am J Roentgenol 157:44 Ferroi F, Bova D, Campodonico F et al. (1996) Cystic fibrosis: MR assessment of pancreatic damage. Radiology 198:875–879 Friedman AC, Liechtenstein JE, Dachman AH (1983) Cystic neoplasms of the pancreas: radiological-pathological correlation. Radiology 149:45–50 Fulcher AS, Turner MA (1999) MR pancreatography: A useful tool for evaluating pancreatic disorders. Radiographics 19:5–24 Hahn PF, Weissleder R, Stark DD et al. (1988) MR imaging of focal splenic tumors. AJR Am J Roentgenol 150:823–827 Hamrick-Turner J, Abbitt PL, Ros PR (1992a) Intrahepatic cholangiocarcinoma: MR appearance. AJR Am J Roentgenol 158:77–79 Hamrick-Turner JE, Chiechi MV, Abbitt PL, Ros PR (1992b) Neoplastic and inflammatory processes of the peritoneum, omentum, and mesentery: diagnosis with CT. Radiographics 12:1051–1068 Helmberger TK, Ros PR, Mergo PJ, Tomczak R, Reiser MF (1999) Pediatric liver neoplasms: a radiologic-pathologic correlation Eur Radiol 9:1339–1347 Holzknecht N, Gauger J, Sackmann M, Thoeni RF, Schurig J, Holl J, Weinzierl M, Helmberger T, Paumgartner G, Reiser M (1998) Breath-hold MR cholangiography with snapshot techniques: prospective comparison with endoscopic retrograde cholangiography. Radiology 206:657–664

909

910

6  Abdomen and Retroperitoneum 18. Hussain SM, Semelka RC, Mitchell DG (2002) MR imaging of hepatocellular carcinoma. Magn Reson Imaging Clin N Am 10:31–52 19. Ito K, Mitchell DG, Siegelman ES (2002) Cirrhosis: MR imaging features. Magn Reson Imaging Clin N Am 10:75–92 20. Kanematsu M, Hoshi H, Murakami T, Tsuda K, Yokoyama R, Nakamura H (1997) Spontaneous bacterial peritonitis in cirrhosis: enhancement of ascites on delayed MR imaging.15:185–187 21. Katz ME, Heiken JP, Glazer HS, Lee JKT (1985) Intra-abdominal panniculitis: clinical, radiographic, and CT features. AJR Am J Roentgenol 145:293–296 22. Kelekis NL, Semelka RC (1996) MR imaging of the gallbladder. Top Magn Res Imaging 8:312–320 23. Kelekis NL, Semelka RC, Siegelman ES (1996) MRI of pancreatic metastases from renal cancer. J Comp Assist Tomogr 20:249–253 24. Kelekis NL, Semelka RC, Siegelman ES, Ascher SM, Outwater EK, Woosley TJ, Reinhold C, Mitchell DG (1997) Focal hepatic lymphoma: MR demonstration using current techniques including gadolinium enhancement. J Magn Reson Imaging 15:625–636 25. Kelekis NL, Semelka RC, Worawattanakul S, de Lange EE, Ascher SM, Ahn IO, Reinhold C, Remer EM, Brown JJ, Bis KG, Woosley JT, Mitchell DG (1998) Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. AJR Am J Roentgenol 170:1005–1013 26. Lee GHM, Cohen AJ (1993) CT imaging of abdominal hernias. AJR Am J Roentgenol 161:1209–1213 27. Leonardou P, Semelka RC, Mastropasqua M, Kanematsu M, Woosley JT (2002) Epithelioid hemangioendothelioma of the liver: MR imaging findings. Magn Reson Imaging 20:631–633 28. Lewis VL, Shaffer HA Jr, Williamson BRJ (1982) Pseudotumoral lipomatosis of the abdomen. J Comput Assist Tomogr 6:79–82 29. Marcos H, Semelka RC (1999) Stomach Diseases: MR evaluation using combined T2-weighted single-shot echo train spin echo and gadolinium–enhanced spoiled gradient-echo sequences. J Magn Reson Imaging 10:950–960 30. Marcos HB, Semelka RC (2000) Evaluation of Crohn’s disease using half-fourier RARE and gadolinium-enhanced SGE sequences inicial results. Mag Reson Imaging 18:263–268 31. Martin DR, Semelka RC (2000) MR imaging of pancreatic masses. Magn Reson Imaging Clin N Am 8:787–812 32. Matos C, Nicaise N, Deviere J, Cassart M, Metens T, Struyven J, Cremer M (1998) Choledochal cysts: comparison of findings at MR cholangiopancreatography and endoscopic retrograde cholangiopancreatography in eight patients. Radiology 209:443–448 33. Miller FH, Gore RM, Nemcek AA Jr, Fitzgerald SW (1996). Pancreaticobiliary manifestations of AIDS. AJR Am J Roentgenol 166:1269–1274

34. Mitchell DG (1992) Focal manifestations of diffuse liver disease at MR imaging. Radiology 185:1–11 35. Mortelé KJ, Praet M, Vlierberghe HV, Kunnen M, Ros PR (2000) CT and MR imaging findings in focal nodular hyperplasia of the liver: radiologic - pathologic correlation. AJR Am J Roentgenol 175:687–692 36. Mosetti MA, Leonardou P, Motohara T, Kanematsu M, Armao D, Semelka RC (2003) Autosomal dominant polycystic kidney disease: MR imaging evaluation using current techniques. J Magn Reson Imaging 18:210–215 37. Motohara T, Semelka RC, Bader TR (2003) MR cholangiopancreatography. Radiol Clin North Am 41:89–96 38. Noone TC, Semelka RC, Worawattanakul S, Marcos HB (1998) Intraperitoneal abscesses: diagnostic accuracy of and appearances at MR imaging. Radiology 208:525–528 39. Noone TC, Semelka RC, Siegelman ES, Balci NC, Hussain SM (2000) Budd-Chiari syndrome: spectrum of appearances of acute, subacute, and chronic disease with magnetic resonance imaging. J Magn Reson Imaging 11:44–50 40. Palacios E, Shannon M, Solomon C, Guzman M (1990) Biliary cystadenoma: ultrasound, CT, and MRI. Gastrointest Radiol 15:313–316 41. Paulson EK, McClellan JS, Washington K, Spritzer CE, Meyers WC, Baker ME (1994) Hepatic adenoma: MR characteristics and correlation with pathologic findings. AJR Am J Roentgenol 163:113–116 42. Rabushka LS, Kawashima A, Fishman EK (1994) Imaging of the spleen: CT with supplemental MR examination. Radiographics 14:307–332 43. Ramani M, Reinhold C, Semelka RC, Siegelman ES, Liang L, Ascher SM, Brown JJ, Eisen RN, Bret PM (1997) Splenic hemangiomas and hamartomas: MR imaging characteristics of 28 lesions. Radiology 202:166–172 44. Saifuddin A, Ward J, Ridgway J, Chalriners AG (1993) Comparison of MR and CT scanning in severe acute pancreatitis: initial experiences. Clin Radiol 48:111–116 45. Semelka RC, Ascher SM (1993) MRI of the pancreas—state of the art. Radiology 188:593–602 46. Semelka RC, Sofka CM (1997) Hepatic hemangiomas. Magn Reson Imaging Clin N Am 5:241–253 47. Semelka RC, Shoenut JP, Lawrence PH, Greenberg HM, Madden TP, Kroeker MA (1992) Spleen: Dynamic enhancement patterns on gradient-echo MR images enhanced with gadopentetate dimeglumine. Radiology, 185:479–482 48. Semelka RC, Shoenut JP, Greenberg HM, Mickflickier AB (1993a) The liver. In: Semelka RC, Shoenut JP (eds) MRI of the abdomen with CT correlation. Raven, New York, pp 13–41 49. Semelka RC, Shoenut JP, Kroeker MA, Micflikier AB (1993b) Chronic pancreatitis: MR imaging features before and after administration of gadopentetate dimeglumine. J Magn Reson Imaging 3:79–82

6.1  Abdominal MRI 50. Semelka RC, Shoenut JP, Ascher SM, Kroeker MA, Greenberg HM, Yaffe CS, Micflikier AB (1994) Solitary hepatic metastasis: comparison of dynamic contrast-enhanced CT and MR imaging with fat-suppressed T2-weighted, breath-hold T1-weighted FLASH, and dynamic gadolinium-enhanced FLASH sequences. J Magn Reson Imaging 4:319–323 51. Semelka RC, Kelekis NL, Molina PL, Scharp T, Calvo B (1996) Pancreatic masses with inconclusive findings on spiral CT. Is there a role for MRI? J Magn Reson Imaging 6:585–588 52. Semelka RC, Kelekis NL, John G, Ascher SM, Burdeny DA, Siegelman ES (1997) Ampullary carcinoma: demonstration by current MR techniques. J Magn Reson Imaging 7:153–156 53. Semelka RC, Balci NC, Op de Beeck B, and Reinhold C (1999a) Evaluation of a 10-min comprehensive MR imaging examination of the upper abdomen. Radiology 211:189–195 54. Semelka RC, Hussain SM, Marcos HB, Woosley JT (1999b) Biliary hamartomas: solitary and multiple lesions shown on current MR techniques including gadolinium enhancement. J Magn Reson Imaging 10:196–201 55. Semelka RC, Chung JJ, Hussain SM, Marcos HB, Woosley JT (2001) Chronic hepatitis: correlation of early patchy and late linear enhancement patters on gadolinium-enhanced MR images with histopathology-initial experience. J Magn Reson Imaging 13:385–391 56. Shoenut JP, Semelka RC, Silverman R, Yaffe CS, Mickflikier AB (1993) Magnetic resonance imaging evaluation of the local extent of colorectal mass lesions. J Clin Gastroenterol 17:248–253

57. Shoenut JP, Semelka RC, Levi C, Greenberg H (1994) Ciliated hepatic foregut cysts: US, CT, and contrast- enhanced MR imaging. Abdom Imaging 19:150–152 58. Siegelman ES, Mitchell DG, Semelka RC (1996) Abdominal iron deposition: metabolism, MR findings, and clinical importance. Radiology 199:13–22 59. Urrutia M, Mergo PJ, Ros LH, Torres GM, Ros PR (1996) Cystic masses of the spleen: radiologic-pathologic correlation. Radiographics 16:107–129 60. Vanek VW, Phillips AK (1984) Retroperitoneal, mesenteric and omental cysts. Arch Surg 119:838–842 61. Walls SD, Hricak H, Baily GD, Kerlan RK Jr et al. (1986) MR of pathologic abdominal fluid collections. J Comput Assist Tomogr 10:746–750 62. Warshauer DM, Semelka RC, Ascher SM (1994) Nodular sarcoidosis of the liver and spleen: Appearance on MR images. J Magn Reson Imaging 4:553–557 63. Wenzel JS, Donohoe A, Ford KL III, Glastad K, Watkins D, Molmenti E (2001) Primary biliary cirrhosis: MR imaging findings and description of MR imaging periportal halo sign. AJR Am J Roentgenol 176:885–889 64. Worawattanakul S, Kelekis NL, Semelka RC, Woosley JT (1996) Hepatic angiomyolipoma with minimal fat content: MR demonstration. Magn Reson Imaging 14:687–689 65. Worawattanakul S, Semelka RC, Kelekis NL, Woosley JT (1997) Angiosarcoma of the liver: MR imaging pre- and post-chemotherapy. Magn Reson Imaging 15:613–617 66. Worawattanakul S, Semelka RC, Noone TC, Calvo BF, Kelekis NL, Woosley JT (1998) Cholangiocarcinoma: spectrum of appearances on MR images using current techniques. Magn Reson Imaging 16:993–1003 67. Zeman RK, Schiebler M, Clark LR et al. (1985) The clinical and imaging spectrum of pancreaticoduodenal lymph node enlargement. AJR Am J Roentgenol 144:1223–1227

911

912

6  Abdomen and Retroperitoneum

6.2  Kidneys, Adrenals, and Retroperitoneum H.J. Michaely, M. Laniado, and S.O. Schönberg 6.2.1  Introduction For the examination of the retroperitoneum, ultrasound, computed tomography, and in an increasing number of situations, magnetic resonance imaging, are being used. In this section, magnetic resonance imaging of the kidneys, adrenals, lymph nodes, retroperitoneal tumors, and the large vessels, as well as of the psoas muscle is discussed. 6.2.2  General Examination Techniques 6.2.2.1  Patient Preparation and Positioning Typically, the patient is positioned in a headfirst, supine position with the arms parallel to the body. In shorter exams, the arms can also be positioned above the head, but this position is less comfortable for the patient and may lead to an increased number of motion artifacts. The administration of oral contrast agents has been advocated by some experts, but it is currently not considered standard. For many clinical indications, the administration of an intravenous contrast agent is required. To administer a short, compact contrast-agent bolus for a dynamic examination, the intravenous access should be chosen generously. Typically, an 18 or 20 ga needle is placed into an antecubital vein of the patient. If the patient exhibits severe artifacts due to bowel motility, N-butyl-scopolamine (BuscopanTM) can be employed using the same venous access. Children of 5 years or less typically require sedation for motion-free imaging. 6.2.2.2  Choice of Coils Due to the profound technical developments that have occurred in recent years, the examination of the retroperitoneum is usually performed using a phased-array surface coil. The body coil should only be used for imaging if surface coils cannot be applied. The surface coils have the advantages of yielding a higher signal-to-noise ratio than the body coil does, and of allowing the application of parallel imaging techniques to eventually speedup image acquisition or to increase the spatial resolution. If clinically warranted, a combination of several surface coils can be used to increase the scan volume. Newer scanner concepts such as Tim (total imaging matrix) from Siemens allow the combination of multiple different coils. Other vendors are expected to follow this approach.

6.2.2.3  Imaging Planes Conventionally, a transverse slice orientation is chosen for most MR exams of the retroperitoneum. Most of the time coronal slices are acquired as well. Coronal images are particularly useful for demonstrating the anatomic relationship between the adrenals and renal tumors. In addition, they demonstrate the psoas muscle in relation to the other retroperitoneal organs. Sagittal slices are not commonly used for imaging of the kidneys and the retroperitoneum. The slice thickness is dependent on the sequence technique and the field strength used. For 1.5 T scanners a slice thickness of 5–4 mm should be chosen. For dynamic examinations with 3D gradient echo sequences a thinner slice thickness of 3 mm at 1.5 T and 2 mm at 3 T can be chosen. 6.2.2.4  Sequences 6.2.2.4.1  T1-Weighted Sequences T1-weighted sequences depict the parenchymal organs with a good contrast to the retroperitoneal fat tissue. Gradient-recalled echo (GRE) sequences with a TE time where water and fat protons are in phase are recommended. To facilitate breath-hold examinations, a TR time of roughly 100 ms should be chosen. This allows covering a complete scan volume of the abdomen and retroperitoneum in less than 30 s. The flip angle for T1-weighted sequences should be larger than 70°, or at least as high as possible to stay within the specific absorption rate (SAR) limits. To minimize flow related phase artifacts (ghosting) flowcompensated sequences should be used. Alternatively, a parallel saturation band can be applied. Due to recent technical advances, prospective respiratory gating is now an option for T1-weighted sequences. When respiratory gating techniques are used, the position of the diaphragm is used as a reference for the prospective correction of the different slices in respect to the diaphragm. This allows T1-weighted multi-breath-hold exams to cover the entire abdomen. However, one has to keep in mind that the inand opposed-phase times are different at 1.5 and 3 T. All sequences therefore need to be adapted to the higher field strength. Spin-echo sequences are meanwhile considered obsolete for the morphologic T1-weighted imaging of the abdomen due to the longer total acquisition time. Newer sequence types such us VIBE (volume-interpolated breath-hold examination) or LAVA (liver acquisition with volume acceleration) allow imaging of the entire abdomen and retroperitoneum during the arterial or venous phase in less than 25 s. Due to their 3D volumetric nature they yield a higher signal-to-noise ratio than do conventional 2D GRE approaches and they allow for 3D reformatting. Hence, these sequences should be used

6.2  Kidneys, Adrenals, and Retroperitoneum

for dynamic imaging of the kidneys or the adrenals. At 3 T, the sequence parameters of these sequences can be altered in that even almost MR angiographic data sets can be acquired. If retroperitoneal bleeding or intrarenal hemorrhage is suspected, then T1-weighted sequences should be acquired with fat-sat pre-contrast. Beyond this indication, fat-saturated sequences should also be used for at least one acquisition post-contrast to clearly depict contrast enhancement, particularly in partially fatty or cystic lesions. 6.2.2.4.2  T2-Weighted Sequences T2-weighted sequences should be acquired as fast spinecho sequences (vendor-specific acronyms: fast SE, turbo SE). Compared to conventional spin-echo sequences, a 5 to 30 times faster acquisition can be achieved. By this means, acquisitions during a breath hold are feasible. Nevertheless, the echo train should be no longer than 25 in order to allow for sufficient T2 weighting. Due to the longer echo train and the multiple refocusing RF pulses, turbo spin-echo sequences yield a higher signal from the fat than do conventional SE sequences secondary to jjcoupling effects. In some cases, fatty tissue can be hard to be discriminate from fluid. If clinically warranted fat suppression techniques can be added to selectively cancel

out the fat signal. With the newer scanner techniques, the combination of prospective respiratory gating and T2weighted sequences allows for longer imaging times and thus yields a higher spatial resolution with fewer motion artifacts. Particularly single-shot sequences like HASTE (half-Fourier–acquired turbo spin-echo sequence) allow very fast imaging with little motion artifacts. These sequences are suitable for less cooperative patients or patients with a large amount of ascites, which normally degrades the image quality. However, one has to keep in mind that single shot sequences suffer from poor T2 contrast. The application of heavily T2-weighted sequences like RARE can be indicated in the work-up of hydronephrosis. 6.2.3  Kidney 6.2.3.1  Dedicated Examination Techniques T2/T1-weighted fast steady-state free-precession sequences (FIESTA, TrueFISP) in the coronal and axial orientations are mostly used as initial localizing sequences. T1-weighted gradient-echo images during a breath-hold period in the axial and coronal orientations are used to assess the renal borders, the corticomedullary differentiation, and intrarenal hemorrhage if present. Particularly for

Table 6.2.1  Sequences used for imaging of the kidneys and the retroperitoneum Sequence

Orientation

Comment

2D T1-weighted GRE inphase and opposed-phase

Axial/coronal

Depiction of the corticomedullary differentiation Signal cancellation in opposed-phase images for particular pathologies with fat content Depiction of perirenal infiltration Decreased SI in patients with iron overload

2D T2-weighted STIR

Coronal

Optional sequence to demonstrate and assess the retroperitoneum and the psoas muscle (inflammation, edema, abscess)

2D T1-weighted GRE fat-sat

Axial/coronal

Optional sequence to detect renal/adrenal/retroperitoneal hemorrhage

2D T2-weighted TSE fat-sat

Axial

Assessment of renal/adrenal lesions (tumor, cysts/pheochromocytoma) Depiction of pathologic lymph nodes Assessment of renal vein/inferior vena cava involvement

3D T1-weighted GRE pre-contrast, arterial and venous phase

Axial

Detection of lesions Assessment of lesion vascularity Assessment of contrast enhancement Assessment of renal vein/IVC involvement

2D T1-weighted GRE fat-sat

Coronal

Assessment of extent of enhancing lesions with respect to liver and retroperitoneum

913

914

6  Abdomen and Retroperitoneum

the detection of hemorrhage, T1-weighted sequences with fat-saturation are most suitable. As most commercially available gadolinium chelates are freely filtrated within the glomerulus, dynamic images first allow assessment of the arterial and venous enhancement of the kidney or intrarenal lesions. Second, the transit of the contrast agent from the renal cortex through the renal medulla into the renal pelvis and ureter can be monitored at the same time. In this way, the renal filtration and excretion function can be assessed. For dynamic examinations, images in coronal orientation as well as images in axial orientation have been used. A coronal orientation allows for a better overview of the anatomical structures (cortex, medulla, renal pelvis, and pyelon). On newer scanners, fast 3D T1weighted sequences such as VIBE and LAVA are the optimal sequences for dynamic imaging of the kidneys. If the renal function is the focus of the examination even faster, then heavily T1-weighted sequences with dedicated pulse preparation schemes can be used to monitor the first pass and the excretion of the contrast agent with high temporal and spatial resolution. In return, these dynamic saturation-recovery gradient echo sequences such as TurboFLASH are characterized by relatively low in-plane spatial resolution. Using dedicated software and dedicated post-processing methods the split renal function can be determined and the absolute renal perfusion can be quantified. Post-contrast administration is recommended to acquire two different T1-weighted sequences, one in the axial plane and one in the coronal plane. To one of these sequences spectral fat suppression should be applied. T2weighted sequences should be acquired with prospective respiratory gating (respiratory belt or dedicated sequence techniques) to minimize the motion artifacts. In addition, the application of a fat saturation pulse allows for good differentiation of the kidney and the adrenals from the retroperitoneal fat. T2-weighted sequences are of particular value in the assessment of (tumor) thrombus in the renal veins and the inferior vena cava.

Fig. 6.2.1a–c  Benign cystic dysplasia. T1-weighted axial (a left side), T2-weighted coronal (b middle image) and T1-weighted coronal post-contrast image (c right side) of a young patient with suspicious renal ultrasound. A hypointense spiculated mass can be seen on the T1-weighted image (a), which reveals a high SI on

6.2.3.2  Normal Anatomy and Developmental Anomalies T1-weighted sequences yield relatively hyperintense signal intensity (SI) in the renal cortex and hypointense SI in the renal medulla. This contrast allows clear differentiation between the renal medulla and the cortex, which is called corticomedullary differentiation (CMD). CMD can be even clearer in the first minute post-bolus injection of gadolinium chelates. Complete loss of CMD can be seen with chronic renal failure and acute tubular necrosis, although loss of CMD is a non-specific sign . Radiologists have to keep in mind that due to chemical shift artifacts there seems to be low SI border around the kidneys predominately on GRE images with out-of-phase conditions, which must not be confused with the renal capsule. The renal capsule is too thin to be demonstrated on MR images. T2-weighted images do not show the typical corticomedullary differentiation. The renal parenchyma shows substantially higher SI on T2-weighted-images than it does on comparable T1-weighted images. Typically, the renal vessels and the renal hilus have almost no SI due to flow voids. The pyelon cannot be marked off in a majority of patients and reveals hypointense SI on T1-weighted images but high SI on T2-weighted images. Gerota’s fascia often can be marked off as a low SI line between the perirenal and pararenal fat tissue. Developmental anomalies of the kidneys are relatively frequent. An estimated 10% of the population suffers from some kind of urogenital malformation. The most common developmental anomaly is the persistent fetal lobulation in which an undulating outer contour of the kidney can be demonstrated, particularly on coronal images. Other anomalies include renal fusions such as the horseshoe kidney with the lower poles of the kidneys fused. These patients are at a higher risk for ureteral obstruction as their ureters may be compressed by vessels.

T2 (b), where also thin septa can be appreciated. After contrast enhancement, the lesion shows no enhancement. The histologic work-up of the lesion revealed a benign cystic dysplasia of the kidney

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.2  Multicystic dysplastic kidney. Multicystic dysplastic right kidney in a young male patient. On the coronal steady-state image (upper left image) three cystic masses can be appreciated at the location of the right, missing kidney. The left kidney shows a compensatory increase in size. On the axial T2-weighted image

(lower left image) the same finding can be seen again. There are no cysts in the liver or the contralateral kidney. After contrast injection excretion of the contrast agent can be seen in the left hypertrophic kidney while there is no enhancement and hence no excretion on the right side (right image, coronal T1-weighted GRE)

Fig. 6.2.3  Duplicated kidney. On the T2-weighted coronal images (left and middle images) this patient reveals a duplicated left kidney, which is a relatively common anomaly. A duplicated

ureter can be appreciated (left and middle images). There are also two renal arteries (right image), which can be clearly depicted on this coronal thin T1-weighted MRA MIP

Fig. 6.2.4a–c  Malrotation. Coronal T2-weighted-HASTE image in a 25-year-old volunteer, revealing a lacking kidney on the left side (a left image). The right kidney appears normal. The second kidney is also located on the left side anteriorly and

caudally to the right kidney. The renal hilus is rotated ventrally (b middle image). In the volume-rendered image of the arterial phase (c right image) the origin of the malrotated renal artery can be appreciated at the level of the aortic bifurcation

915

916

6  Abdomen and Retroperitoneum

The kidneys can also be ectopic in their position, which may vary. Common locations for the kidneys include the lower abdomen and the pelvis; in rare cases, also intrathoracic kidneys have been described. A pelvic kidney due to renal migration has a reported incidence of 1 : 725 births. In addition to an aberrant location, there is often malrotation of the kidneys, which leads to ventrally located collecting structures. 6.2.3.3  Pathologies The main two pathologies that radiologists encounter when analyzing images of the kidneys are renal cysts and malignant renal tumors. Typical renal cysts are easy to identify and do not cause a problem even for an inexperienced radiologist. However, in the case of a complex cyst, differentiation between a benign complex cyst, a hemorrhagic cyst, and a cystic renal cell carcinoma is much harder to achieve. The challenge in assessing renal tumors is not to detect a simple cyst as such but to differentiate between a complex but probably benign lesion and a malignant lesion. Renal cell carcinoma is the most important malignant lesion. All the other malignant lesions such as primary tumors of the kidney or lymphomas of the kidney are rare (Table 6.2.2). These diagnoses should be kept in mind to serve as a potential differential diagnosis.

6.2.3.3.1  Renal Cysts Simple Renal Cysts The occurrence of simple renal cysts is dependent on the patient’s age. Simple renal cysts mostly originate from the renal cortex, are round, demonstrate a smooth thin wall and do not exhibit any septa. There are often multiple cysts in affected kidney parenchyma and cysts can be bilateral. On T1-weighted images, kidneys show homogeneous low SI. On T2-weighted images, there is homogenous high SI. The cyst wall should be smooth and almost not appreciable. After intravenous injection of contrast agent, the cyst must not enhance. Peripelvine renal cysts are mainly differentiated by their more variable shape. They have, however, the same signal characteristics. One can distinguish a peripelvine renal cyst from an ampullary renal pelvis simply by looking at post-contrast T1-weighted images, in which peripelvine renal cysts do not enhance. Simple renal cysts can turn into complex cysts by hemorrhage, infection, rupture, calcification, or the occurrence of an intracystic malignant tumor. Hemorrhage in renal cysts is mostly secondary to hemorrhagic diathesis and trauma but in most cases of unknown origin. Depending on the age of the hemorrhage, the SI of the cysts is very different. Even within a single cyst, there can be a signal gradient due to sedimentation of the degradation products of the hemoglobin. The dependent part of the cysts shows a higher SI in the T1-weighted sequences and a lower SI on the T2-weighted sequences, respectively.

Table 6.2.2  Epidemiology of renal lesions Tumor

Prevalence (%)

Peak age (years)

Comment

Cyst

Rare 50

<40 >50

Polycystic kidney disease

0.1

40–50

Acquired form

Renal cell carcinoma

0.2 Up to 45%

50–70

Incidence von Hippel-Lindau syndrome

Adenoma

7–22 35

Autopsy data von Hippel-Lindau syndrome

Angiomyolipoma

42–80 1

In patients with tuberous sclerosis Percentage of all renal tumors

Metastasis

8–13

Autopsy data

Lymphoma infiltration

Up to 25

In patients suffering from lymphomas

Transitional cell carcinoma of the renal pelvis

5–10

60–80

Percentage of all renal tumors

Wilms’ tumor

0.8

2–3

Incidence per 100,000 people

6.2  Kidneys, Adrenals, and Retroperitoneum Table 6.2.3  Bosniak classification of renal cysts Bosniak category

Name

Imaging features

Malignant lesions (%)

I

Simple cyst

Hairline-thin wall, no septa No contrast enhancement No calcifications

1.7

II

Probably benign cyst

Hairline-thin wall, few thin septa Fine calcifications No contrast enhancement

18.5 (II and IIF)

IIF

Indeterminate cyst

More septa Minimal contrast enhancement Minimal wall thickening Nodular to thick calcifications

18.5 (II and IIF)

III

Probably malignant cyst

Irregular walls Enhancing septa

33.0

IV

Malignant cyst

Enhancing soft tissue masses

92.5

Fig. 6.2.5a,b  Simple renal cyst. a (left side) Simple renal cyst at the upper pole of the kidney demonstrates low SI on a T1-weighted GRE-sequence.   b (right side) In the T2-weighted image the simple renal cyst demonstrates a homogenous high SI with an almost invisible cyst wall

During the further temporal course, a hemorrhagic cyst may behave like a protein rich fluid with high SI on the T1-weighted images. MR imaging does not allow safe differentiation between a hemorrhagic cyst and a hemorrhagic renal cell cancer based on the signal intensities on T1-weighted and T2-weighted images. Contrast-enhancement of the cyst wall, which can be appreciated much better with MRI than with CT, suggests potentially malignant tissue. To visualize even faint contrast uptake, subtraction techniques are suitable and make MRI superior to CT in the detection of malignant lesions. Cysts with partially enhancing wall or nodular components should be closely followed or even surgically removed.

Infection of renal cysts can occur spontaneously, during a bacteremia or after the puncture of the cyst. Morphologically, a thickened or nodular appearance of the cyst wall can be appreciated. Due to the higher protein content and the resultant increased relaxivity, infected cysts may also exhibit higher SI than normal cysts on T1weighted images. Calcification of the cyst wall may be due to hemorrhage or infection. In contrast to the irregularly distributed calcifications of renal cell carcinoma, calcifications of cyst walls are typically located in the periphery of the cyst. The calcifications cannot be directly visualized using MRI. Rarely, they can be indirectly seen as

917

918

6  Abdomen and Retroperitoneum

Fig. 6.2.6a,b  Complicated renal cyst. a (upper row) T2-weighted image demonstrating a multilobulated cyst with multiple septa at the dorsal aspect of the left kidney. b (lower row) In the corresponding CT image of the same patient in the arterial phase the cyst is characterized by water-dense attenuation values. However, CT does not allow visualization of the septa, which could be seen with MRI

Fig. 6.2.7a–d  Autosomal dominant polycystic kidney disease 1. a (upper row, left side) In the T2-weighted coronal image the right kidney demonstrates multiple cysts with different signal intensities ranging from hyperintense to isointense. b (upper row, right side) The T1-weighted coronal image with spectral fat saturation reveals hypointense simple cysts but also numerous cysts with high SI, which are due to either high protein content or intracystic hemorrhage. c (lower row, left side) The T2-weighted axial image underlines the above-mentioned findings. In this image particularly the thin septa of the cysts can be seen. Please note again the varying signal intensities of the cysts. d (lower row, right side) After contrast administration the T1-weighted fat-saturated GRE sequence demonstrates enhancement of the walls of the cysts but no contrast uptake in the cysts. The cysts with higher SI demonstrated high signal intensity in the non-enhanced sequences most likely due to hemorrhage (see b)

signal voids. Calcifications are a minor characteristic in the Bosniak cyst classification and are often used in CT. In studies comparing CT and MRI with regard to the assessment of cystic lesions according to the Bosniak classification (Table 6.2.3), a slightly higher sensitivity for MR has been reported. Seven out of 69 lesions were upgraded due to the higher soft tissue contrast of MRI, which helps in depicting a greater number of thin, hairline septa and allows better appreciation of enhancing wall structures.

nal cell carcinoma. In contrast, patients who suffer from acquired polycystic kidney disease are at a higher risk of developing renal cell carcinoma. These patients are often on hemodialysis. After 3 years on hemodialysis, on the other hand, up to 50% of all patients reveal polycystic kidney changes. With increasing time of hemodialysis, the incidence of polycystic kidneys changes is increasing steadily. Therefore, these patients need to be followed closely since in up to 8% a renal cell carcinoma will develop in these cystic changes (Table 6.2.4).

6.2.3.3.2  Polycystic Kidneys The autosomal dominant form of polycystic kidney disease is characterized by either unilaterally or bilaterally enlarged kidneys with cystic degeneration. At this stage, the kidneys often exhibit simple cysts and complex cysts at the same time. Patients with polycystic kidney disease often also reveal cystic changes of the pancreas, spleen, or the liver. In patients suffering from autosomal dominant polycystic kidney disease, there is no increased risk of re-

Table 6.2.4  Key features autosomal dominant polycystic kidney disease Mutation of chromosome 16, prevalence 1 : 1,000 Third most common cause of chronic renal failure Formation of cysts in first decade of life (54%) Additional cysts in liver (25–50%) and pancreas (9%) Saccular aneurysms of basal cerebral vessels (3–13%) Only slightly elevated risk of renal cancer

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.8a–c  Autosomal dominant polycystic kidney disease 2. a (left image) Axial T1-weighted image pre-contrast in a patient with ADPKD demonstrating innumerable cysts of various sizes in the liver and in both kidneys. Due to the increased volume of the liver and the kidneys, they can hardly be distinguished from each other. A single cyst in the left lobe of the liver demonstrates

high SI. b (middle image) The transverse T2-weighted image in the same patient underlines these findings. All cysts reveal hyperintense SI. c (right image)T1-weighted image with fat-saturation after the administration of contrast agent demonstrating no enhancement in the cysts

Fig. 6.2.9a,b  Autosomal recessive polycystic kidney disease. In the coronal T2-weighted-HASTE sequence with fat-saturation (a left image) and in the axial T2-weighted-HASTE sequence with fat-saturation (b right image), innumerable small cysts of the kidneys are depicted in this young patient. The liver is not

affected. The autosomal recessive polycystic kidney disease is characterized by an earlier onset and smaller cysts. The liver can also be affected. Most patients hardly achieve the adolescent age and die of renal failure

Fig. 6.2.10  Acquired multi-cystic disease with renal cell carcinoma and tumor thrombus. Axial T2-weighted-image showing a tumor thrombus of the left renal vein (arrow) in a patient with bilateral acquired multi-cystic disease

919

920

6  Abdomen and Retroperitoneum

6.2.3.3.3  Renal Cell Carcinomas and Adenomas Renal Cell Carcinomas More than 85% of solid renal tumors are renal cell carcinomas (RCC), which occur bilaterally in up to 5% of the cases. Every non-fatty solid renal tumor should therefore be considered RCC unless proven otherwise. Men are twice as often affected as women are. Due to the lack of sensitive innervation of the kidney and the renal capsule, renal cell cancer often exhibits a substantial size at the time of the diagnosis. One third of the patients already have metastases (lungs, mediastinum, bones, liver, brain, skin). Up to two thirds of the patients suffering from von Hippel-Lindau syndrome develop bilateral, often multiple renal cell carcinomas. Renal cell carcinomas exhibit calcifications in up to 50% and necrosis, while hemorrhage is less frequent. Therefore, RCC often exhibits an inhomogeneous SI on T1-weighted and T2-weighted images. In case of necrosis, mainly occurring centrally, these tumors exhibit central hyperintense signal on the T2-weighted images. Up to 15% of all renal cell tumors are iso-intense to the kidney in all pre-contrast sequences. Often a distortion of the renal architecture is the only hint for renal tumors in non-enhanced studies. Therefore, the administration of intravenous contrast agents is mandatory to reliably detect and grade renal cell carcinomas. Small RCC manifest as round, smooth lesions, which can be clearly distinguished from normal renal parenchyma. Larger RCC however, have a more polycyclic appearance. The exact border of larger renal cell carcinomas may be hard to detect due to the invasive growth pattern and require multiplanar imaging for delineation from adjacent organs. Post-contrast T1-weighted images allow for a good delineation of necrotic and viable tumor components. The thick and irregular tumor is readily detected on contrast-enhanced scans. The delineation of the tumor from normal renal parenchyma is dependent on the exact time of image acquisition following contrast injection and on the vascularity of the tumor. The acquisition of at least two sequences after intravenous contrast administration (arterial phase 15–20 s after bolus injection and venous phase 20–30 s after the arterial phase) greatly facilitates the characterization of renal tumors. Renal tumors often demonstrate an early washout of the contrast agent compared to the normal renal parenchyma. An enhancement of more than 15% compared with the precontrast-images has been found to be very sensitive (100%) and specific (94%) for detection of renal cell carcinomas. T2-weighted sequences with fat-suppression also allow for highly sensitive detection of renal tumors. Non-enhanced T1-weighted images are good at demonstrating the tumor infiltration into the perirenal space. This can be even better done on fat suppressed T1weighted images post-contrast. However, the differentiation between renal cell carcinoma, which is confined to the kidney (Robson stage I, T1, and T2), and renal cell

carcinoma with spread into the perirenal space (Robson stage II, T3a) is often hard to achieve with MRI. Tumor infiltration into the renal vein or the inferior vena cava (Robson stage IIIa, T3b) can be appreciated as a moderate SI in the affected vessels that should normally show no SI at all. T2-weighted respiratory-gated images are most suitable for the detection of tumor growth into the renal vein and inferior vena cava. The reported sensitivity in the detection of (tumor-) thrombus is 93–100% for MRI and 93% for CT, while the reported specificity is 75% for MRI and 80% for CT. To differentiate a tumor thrombus from an apposition thrombus, post-contrast images are required. While the latter must not show any contrast enhancement, a tumor thrombus exhibits substantial enhancement. The infiltration of renal cell carcinoma into adjacent organs (Robson stage IV, T4) can be seen by a depleted perirenal fat space or direct infiltration into the adjacent structures. Current approaches allow the acquisition of 3D volumes, which can be post-processed in any desired way. This enables the use of curved multiplanar reformats and of thin maximum intensity projections for advanced postprocessing. However, in some cases—even with the use of the newest techniques—it may not be possible to clearly demonstrate or rule out an infiltration into the adjacent organs. While the Robson classification has been in use for a long time, the TNM classification should be preferred for staging patients with renal cell carcinoma. A major change in the most current version of the TNM guidelines is that T1 carcinomas now include tumors of up to 7 cm in size as long, as they are confined to the kidney. The TNM system is superior to the Robson classification as it allows for a detailed assessment of the primary tumor (T), the lymph nodes (N), and distant metastases (M). To ease the further management of patients with renal cell carcinoma the American Joint Committee on Cancer has issued a staging system based on the TNM classification (Tables 6.2.5, 6.2.6, 6.2.7, 6.2.8).

Fig. 6.2.11a,b  Cystic renal cell carcinoma. T1-weighted axial image without fat-saturation post-contrast (a left image) demonstrates a hypointense mass of the right kidney with no obvious contrast enhancement. The subtraction image (b right image), however, reveals strong enhancement of the wall of this lesion. The patient subsequently underwent total nephrectomy where a cystic renal cell carcinoma was found

6.2  Kidneys, Adrenals, and Retroperitoneum Table 6.2.5  TNM staging of the renal cell carcinoma (International Union Against Cancer) Stage

Description

T0

No evidence of a primary tumor

T1a

Tumor <4 cm in diameter and is limited to the kidney

T1b

Tumor >4 cm but <7 cm and is limited to the kidney

T2

Tumor >7 cm but is still limited to the kidney.

T3a

Tumor has spread into the adrenal gland or into fatty tissue around the kidney, but not beyond Gerota’s fascia

T3b

Tumor spread into renal vein and/or intraabdominal part of the inferior vena cava (IVC)

T3c

Tumor spread into the thoracic part of the IVC

T4

Tumor spread beyond Gerota fascia

N0

No regional lymph node metastases

N1

One regional lymph node affected

N2

More than one regional lymph node affected

M0

No distant metastases

M1

Distant metastases

T/N/M X

Local tumor/lymph nodes/distant metastases not assessed

Table 6.2.6  AJCC staging of the renal cell carcinoma Stage I

T1a–T1b, N0, M0

Stage II

T2, N0, M0

Stage III

T1a–T2, N1, M0 or T3a–T3c, N0, M0

Stage IV

T4, N0–N1, M0 or any T, N2, M0 or any T, any N, M1

Table 6.2.7  Accuracy of CT/MRI for the staging of renal cell carcinoma (Hallscheidt et al. 2004) Modality

Sensitivity: T1/T2 vs. T3/T4 (%)

Specificity: T1/T2 vs. T3/T4 (%)

Accuracy (%)

CT

88

72

89

MRI

91

83

84

Table 6.2.8  Features of von Hippel-Lindau disease Autosomal dominant hereditary phakomatosis, familial in 20% of cases Age at onset: 2nd to 3rd decade, M : F = 1 : 1 Diagnostic criteria: • >1 hemangioblastoma of CNS • 1 hemangioblastoma + visceral manifestation • 1 manifestation + known family history CNS • Retinal angiomatosis = von Hippel tumor (>45%) earliest manifestation of disease • Hemangioblastomas of CNS = Lindau tumor (40%), cerebellum (65%) Kidneys • Polycythemia due to elevated erythropoietin level (in 15% with hemangioblastoma, in 10% with renal cell carcinoma) • Cortical renal cysts (75%), multiple + bilateral (may be confused with adult polycystic kidney disease) • Renal cell carcinoma (20–45%), age: 20–50 years, multicentric in 87%, bilateral in 10–75%, may arise from cyst wall, 50% metastatic at time of discovery • Renal adenoma • Renal hemangioma

921

922

6  Abdomen and Retroperitoneum Fig. 6.2.12a–d  Hemorrhagic renal cell carcinoma T2. Axial T2-weighted (a upper row left), T1-weighted pre-contrast (b upper row right), T1-weighted post-contrast arterial phase (c lower row left), and T1-weighted post-contrast late-phase (d lower row right) images demonstrating an encapsulated hemorrhagic renal cell carcinoma of less than 7 cm in diameter. Hence, this tumor has to be staged as a T2 tumor. On the T2-weighted image the SI is mixed, mainly hyperintense. After contrast enhancement, the SI of the tumor is not significantly changed without central enhancement. This signal behavior is indicative of hemorrhagic renal cell cancer, which was histologically confirmed

Fig. 6.2.13  RCC with renal vein invasion. T2-weighted image with respiratory gating (left image) demonstrating a tumor thrombus with direct extension into the inferior vena cava. The arterial phase T1-weighted-image Post-contrast (middle and

right images) show an intensely enhancing tumor thrombus (arrow, middle image) of a large, exophytic renal cell cancer of the left kidney (arrow right image)

Fig. 6.2.14  Renal cell carcinoma with renal vein invasion CT case. Venous phase axial 16-slice MDCT demonstrating a leftsided renal tumor with extension into the left renal vein and also presumably into the IVC. In contrast to MRI, CT does not allow for easy detection of thrombus and easy differentiation of tumor thrombus from apposition thrombus

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.15a–c  Renal cell carcinoma with apposition thrombus to the right atrium. Steady-state free-precession images in a tilted coronal (a upper left row) and tilted sagittal (b lower left row) orientation and contrast-enhanced T1-weighted coronal image (c left image). An apposition thrombus extending into the right

atrium is clearly visible (arrows), which showed motion with the cardiac cycle. Particularly the sagittal image demonstrates the entire extent of this thrombus. In the post-contrast T1-weighted image, the thrombus shows no enhancement, allowing differentiation of tumor thrombus from apposition thrombus

Fig. 6.2.16a,b  Renal cell carcinoma T1 versus T3a. T2-weighted respiratory triggered images with 512 matrix in a patient with renal cell cancer stage T1 (a left side) and stage T3a (b right side). The T1 tumor, which must not exceed 7 cm in diameter, demonstrates a thin but clearly visible capsule and is confined to the kidney (a). In contrast the T3a (b) is invading the perinephric fat tissue (arrow). Both tumors have mixed signal intensities due to partly necrotic tissue

923

924

6  Abdomen and Retroperitoneum

Fig. 6.2.17a–c  Renal cell carcinoma T3a. Axial (a left image) and coronal (b,c middle and right images) steady-state free precession images of a patient with T3a renal cell carcinoma. The tumor of the right kidney bulges into the liver without proof of

direct invasion of the liver, which would make it a T4 tumor. The SI of the tumor is mixed hyperintense and hypointense representing partially necrotic tissue. The contralateral kidney demonstrates a small cyst at the lower pole

Fig. 6.2.18  Renal cell carcinoma with invasion of the IVC. T2-weighted respiratory triggered images in the axial plane demonstrating a tumor thrombus in the inferior vena cava (arrows) making this a T3b tumor. The primary tumor in the right kidney can also be appreciated (arrowhead)

Fig. 6.2.19  Renal cell carcinoma with multiple relapses. Axial T1-weighted images post-contrast in a patient status post-multiple tumor enucleations of his right kidney. There are now several enhancing masses (arrows), which are highly suspicious for reoccurring multifocal renal cell carcinomas

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.20a–d  Renal cell carcinoma with renal vein invasion (T3b). Axial T2-weighted respiratory triggered image (a left side), T1-weighted coronal images post-contrast (b,c) demonstrating renal vein and inferior vena cava infiltration from renal cell carcinoma. Due to flow voids, the vessel should appear dark on the T2-weighted image. In this case, however, an isointense tumor mass can be appreciated whose tip is located in the IVC (a). The

first coronal T1-weighted image demonstrates a large partly necrotic tumor at the lower pole of the left kidney (b) from which a tumor thrombus extends into the renal vein and the IVC (c). In the arterial-phase MRA image (c) the tumor-thrombus already shows a strong enhancement due to hypervascularization of the tumor-thrombus

Fig. 6.2.21a,b  Reoccurring renal cell carcinoma. Axial T1weighted image pre-contrast (a left side) and post-contrast (b right side) in a patient status post-partial nephrectomy of his right kidney. The MRI reveals an enhancing recurrent tumor at

the anterior aspect of the kidney (arrows) as well as an enhancing mass paraspinally (arrowhead) representing another recurrent tumor manifestation

925

926

6  Abdomen and Retroperitoneum

Fig. 6.2.22  Whole-body staging. As the renal cell carcinoma tends to develop metastases early patients with metastatic disease are also treated with anti-angiogenetic substances. Wholebody MRI (composed whole body T1-weighted image at the left border) allows detection of soft tissue, bone, and brain metastases with high sensitivity as is suitable for staging and following these patients during therapy. Concerning bone and brain metastases MRI is even more sensitive than PET-CT. This ex-

ample demonstrates a large gluteal metastasis (upper left corner, arrowhead) and bilateral adrenal metastases (upper right image, arrowheads) on T1-weighted images post-contrast administration. On the T1-weighted image pre-contrast a metastasis to the femur (lower left image) can be clearly seen. MRI also allows the detection of pulmonary nodules of at least 7 mm as can be seen in this example of a patient with multiple pulmonary metastases (lower right image, T2-weighted HASTE)

Fig. 6.2.23a–c  Postoperative kidney. Axial T2-weighted HASTE image (a left image), T2-weighted fat-saturated image (b middle image) and T1-weighted GRE image post-contrast administration (c right image) in a patient post-partial nephrectomy. A large mass originating from the right kidney and extending anteriorly can be appreciated, which could mimic recurrent can-

cer. This mass, being centrally hyperintense on the T2-weighted image and not enhancing after contrast administration (star in a–c), most likely represents a seroma. The surrounding tissue demonstrates a signal drop in the T2-weighted fat-saturated image (b) and enhances only slightly post-contrast (c), which is consistent with a postoperative fat tissue plumb

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.24  von Hippel-Lindau syndrome. 32-year-old patient with known von Hippel-Lindau syndrome who presented with multiple renal cell carcinomas in both kidneys, which can be well appreciated the axial T1-weighted images post-contrast (left row upper and lower image). The tumors are labeled with an arrow-

head on the upper image while on the lower image an additional spinal tumor a hemangioblastoma can be seen (arrowhead). The right image demonstrates a color-coded parameter map on which the renal cell tumors can be differentiated due to their different exchange rates with the interstitial space (arrowheads)

Adenomas Renal adenomas are histologically not different from renal cell carcinomas. They are regarded as precursors of renal cell carcinomas by some authors. These solid tumors often show a small size, a slow growth pattern and do not cause any symptoms. Therefore, there are rarely detected. They may occur more often in patients suffering from von Hippel-Lindau syndrome. The diagnosis of renal adenoma is based on the visualization of a tumor with a maximum diameter of 3 cm that is located in the renal cortex and

does not have any necrosis or hemorrhage. Adenomas occurring in the same kidney as a renal cell carcinoma have been reported. The SI of renal adenomas is described as slightly hyperintense on T2-weighted images and slightly hypo-intense on T1-weighted images. They may enhance diffusely on the later phases of a multiphase renal exam. A clear differentiation between renal cell carcinoma and renal adenoma cannot be achieved based on MR imaging findings. Depending on the clinical picture, follow-up exams may be warranted to assess potential tumor growth.

Fig. 6.2.25a–c  Oncocytoma 1. Axial T1-weighted images postcontrast (a,b left and middle image) as well as 64-slice MDCT of an oncocytoma (c right image). This oncocytoma was controlled over a long period of time and did not show a change in size or contrast uptake. This lesion exhibits the characteristic features of

oncocytomas of being well circumscribed and of often having a central scar and a spoke-wheel enhancement pattern (just barely visible in this case). They show moderate to good enhancement after contrast administration

927

928

6  Abdomen and Retroperitoneum

Fig. 6.2.26a–c  Oncocytoma 2. T2-weighted fat-saturated (a left side), T1-weighted pre-contrast (b middle image) and T1-weighted post-contrast images in the axial orientation in a patient with bilateral oncocytomas. They reveal a rather homogenous SI on the

T2-weighted fat-saturated images and appear hypointense on the pre-contrast T1-weighted images. Post-contrast, the left oncocytoma reveals a homogenous contrast uptake, while the contralateral oncocytoma is slightly heterogeneous

6.2.3.3.4  Angiomyolipomas

vena cava has been described. With increasing size, angiomyolipomas have a tendency to rupture, with subsequent hemorrhage. MR images demonstrate angiomyolipomas often as round lesions that exhibit high signal intensity on T1-weighted images and a distinct signal drop on fat-saturated T1-weighted images or opposed-phased T1-weighted images. In opposed-phased images, the signal cancellation occurs particularly at the boundary between the angiomyolipomas and the adjacent renal parenchyma in case of macroscopic fatty tissue and within the lesion itself in case of mainly intracellular, microscopic fatty tissue. Rare cases in which the muscular or the vascular elements of the angiomyolipomas prevail may demonstrate moderate SI in all sequences with no distinct signal drop. Angiomyolipomas show little (<15%) enhancement. Intratumoral hemorrhage manifests with high SI on T1-weighted images.

Angiomyolipomas are hamartomas of the kidney, which means they are benign tumors containing fat tissue, smooth muscle cells, and atypical blood vessels. Angiomyolipomas reveal a variable appearance on MR images depending on the ratio of the abovementioned three different tissue types contained in them. Angiomyolipomas occur more often in patients suffering from tuberous sclerosis and autosomal dominantly inherited phacomatoses. Angiomyolipomas in these patients can be multiple in up to 20% of cases or even bilateral. In all other patients’ angiomyolipomas occur as solitary tumors most of the time. Women are more often affected than men are. Even though angiomyolipomas are benign tumors they have a tendency to grow and rare cases of tumor growth into the renal vein and inferior

Fig. 6.2.27a–d  Angiomyolipoma pre-contrast. This figure demonstrates a small angiomyolipoma of the dorsal aspect of the left kidney on T1-weighted in-phase image (a upper left image), T1-weighted opposed-phase image (b lower left image), T2-weighted non-fatsaturated image (FS) (c upper right image) and T2-weighted FS image (d lower right image). On T1-weighted in-phase and on T2-weighted non-FS images the angiomyolipoma exhibits a hyperintense signal. With fat-saturation by either means of opposed-phase imaging (c) or by means of spectral fat-saturation (d) the signal of the angiomyolipoma drops significantly

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.28a–d  Angiomyolipoma post-contrast. Typical low enhancement of angiomyolipomas after contrast administration. These T1-weighted-GRE sequences in the axial slice orientation were acquired during the arterial-phase 20-s postcontrast injection (a upper left corner), the venous-phase 40-s

post-contrast injection (b upper right corner) and a late-phase 1min post-contrast agent administration (c lower right corner). In all these images, virtually no enhancement of the lesion can be appreciated. In the delayed axial T1-weighted GRE, only a faint enhancement can be seen (d lower right corner)

Fig. 6.2.29  Angiomyolipoma CT and MR correlation. Axial CT images (upper row from left to right: unenhanced, venous phase, excretory phase) and axial T1-weighted MR in-phase (lower row, left image) and opposed-phase (lower row, right image) images demonstrating an angiomyolipoma. In CT the angiomyolipoma

is isodense or hypodense to the kidney and shows little to no contrast uptake. MR reveals a signal drop with the transition from in-phase to opposed-phase, which is characteristic and almost pathognomonic of angiomyolipomas

929

930

6  Abdomen and Retroperitoneum

Fig. 6.2.30a,b  Angiomyolipoma CT. Axial (a left) and coronal (b right) 64-MDCT venous-phase images of a patient undergoing a CT scan to rule out lymphoma. The right kidney demonstrates

two well-circumscribed lesions with attenuation values equivalent to fat. No significant enhancement can be appreciated. This finding is consistent with angiomyolipoma of the kidneys

6.2.3.3.5  Metastases

lateral involvement is frequent. Growth of lymphoma tissue from outside of the kidney into the kidney represents another manifestation of lymphoma spread to kidney. Primary lymphomas of the kidneys are rare, and there is ongoing discussion in the literature about whether the kidney has lymphatic tissue from which a primary lymphoma could arise. Therefore, one theory is that renal lymphomas arise from the retroperitoneum and invade the kidney secondarily. Rarely there is infiltration of the kidneys in patients with leukemia, mostly in children suffering from acute myelogenous leukemia. These children are likely to reveal a bilateral manifestation. Based on MRI, lymphoma cannot regularly be differentiated from other solid renal tumors. Lymphomas tend to grow along anatomic structures such as the renal vessels or the ureter. The SI of lymphoma is generally heterogeneous and slightly hypointense to isointense on T2-weighted images. Compared with the renal cortex in T1-weighted images lymphomas are hypointense. Postcontrast these tumors exhibit a heterogeneous enhancement. Generally, there is no central necrosis seen with lymphoma spread to the kidney. Direct invasion from adjacent tissue is best visualized on T2-weighted images in which the tumor is hypointense compared to the kidney. T1-weighted images without fat-suppression show hypointense tumor masses surrounded by the perirenal

Metastases to the kidneys may be detected at autopsy of patients suffering from malignant diseases. Only in the liver, the lungs, in bone and the adrenal glands metastases are found more often than in the kidneys. If present, metastases to the kidneys are often multiple and bilateral. These metastases most frequently originate from lung cancer, and less commonly originate from colorectal or breast cancer or malignant melanoma. At the time of the diagnosis of a metastasis to the kidney, other organ metastases are usually already known. The SI of metastases to the kidneys is variable and non-specific. It is mainly dependent on the size of the lesion (necrosis) and can be similar to that of the primary tumor. A dynamic scan may be required to differentiate highly vascularized lesions or poorly vascularized lesions from the renal parenchyma. 6.2.3.3.6  Lymphoma In up to 25%, lymphoma spread to the kidneys is found in non-Hodgkin’s lymphoma and Hodgkin’s disease. The kidney is rarely the only organ with manifestation of the lymphoma. There may be focal and multiple lesions or even diffuse parenchymal infiltration of the kidney. Bi-

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.31a–c  Renal lymphoma. Coronal HASTE image (a left image), axial T1-weighted VIBE image 3 min after contrast administration (b middle image) and coronal T1-weighted VIBE image 7 min after contrast administration (c right image) of 68year-old patient with biopsy proven lymphoma of the kidney.

There is an ill-defined moderately enhancing mass, which originates from the right kidney extending medially. As the right ureter is also involved there is now hydronephrosis of the directly involved part of the kidney best seen in a (Images courtesy of Prof. Elmar Merkle, Duke University, Durham, N.C.)

Fig. 6.2.32  Perirenal lymphoma. Steady-state free precession images of a patient with perirenal lymphoma, which can be clearly distinguished on the axial (left) and sagittal (middle) image. There is tumor spread along the renal capsule without invasion into the kidney. The lymphoma reveals high SI while the kidney is characterized by an intermediate SI. In another patient

(right image) infiltration into the lower pole of the kidney can be appreciated (arrowhead) on this coronal T1-weighted image post-contrast. At the site of the infiltration the enhancement of the kidney is slightly inferior to the enhancement of the healthy renal tissue

931

932

6  Abdomen and Retroperitoneum

Fig. 6.2.33  Renal hilum lymphoma. Axial T2-weighted fatsaturated (upper left image), T1-weighted pre-contrast image (upper right image), T1-weighted post-contrast image (lower left image), and close-up of the T1-weighted post-contrast image (upper right image) in a patient with lymphoma growth in the

renal hilum. While the changes are quite subtle in the pre-contrast image, the tumor can be well appreciated on the post-contrast images. This growth pattern along anatomic structures is quite typical of lymphomas

fat. A special type of renal lymphoma is the so-called Rosai-Dorfman disease, a sinus histiocytosis with massive lymphadenopathy, which may extend to internal organs, such as liver, kidney, and pancreas. This disease of unknown origin is often self-limiting. Other sites of involvement include bone (asymptomatic osteolytic lesions) respiratory tract, salivary glands, orbits, eyelids, and testes.

ureter or in the bladder. The transitional cell cancer of the renal pelvis infiltrates early into the renal parenchyma and into the peripelvine tissue. The SI of transitional cell carcinomas is mainly isointense to the renal parenchyma on T1-weighted images and T2-weighted images. Contrast-enhanced T1-weighted images often show a narrowing of the ureter at the side of the tumor or a bulging of the tumor into the renal pelvis. Heavily T2weighted images after the administration of furosemide may also demonstrate narrowing of the ureters in case of transitional cell carcinoma involvement. Transitional cell carcinomas of the renal pelvis have to be differentiated from stones of the renal pelvis, which do not reveal any signal on MR images, and from blood clots in the renal pelvis, which are rather hyperintense. Transitional cell carcinomas are mainly hypovascularized and hence appear hypointense compared to the kidney on post-

6.2.3.3.7  Carcinomas of the Renal Pelvis More than 40% of cases of carcinoma of the transitional cell epithelium arise in the renal pelvis. They account for 8% of all renal tumors. Men are three times more frequently affected than women are. Due to their tendency to grow along the upper urinary tract further tumors are often seen at the time of the diagnosis in the ipsilateral

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.34  Rosai-Dorfman disease. Axial T2-weighted images (upper row) and axial T1-weighted post-contrast images (lower row) demonstrating Rosai-Dorfman disease of the right kidney with growth into the liver. The renal lesion and the hepatic manifestation demonstrate rim-enhancement after contrast injection (arrows)

contrast T1-weighted images. Focal involvement of the ureter might be better appreciated on high-resolution multi-slice CT urograms. 6.2.3.3.8  Wilms’ Tumor (Nephroblastoma) Nephroblastoma is the most common malignant renal tumor in childhood. It accounts for approximately 20% of all abdominal tumors in this age group and for up to 6% of all renal tumors. The tumor is diagnosed most frequently in children aged 2–5 years. Nephroblastoma is rare in newborns and children of more than 7 years. After a symptom-free initial stage a large palpable mass (mean: 12 cm) is found in 90% of all patients. At the time of diagnosis, there are often metastases in the inguinal lymph nodes, the lungs and the liver. Infiltration into the

renal vein is present in 4–10%. In many cases, a tumor thrombus extending to the right atrium is detected. Often a tumor thrombus that reaches the right atrium can be appreciated. Five to 10% of the tumors are bilateral or even multifocal. Necrosis and hemorrhage (80%) as well as cyst formation and calcifications (15%) may result in an inhomogeneous SI pattern on T1-weighted images. In general, nephroblastoma is slightly hyperintense on T2-weighted images and slightly hypointense on T1-weighted images. In large tumors, central hyperintense signal is present on T1-weighted images, which is due to hemorrhage. Postgadolinium the tumor enhances heterogeneously. To clearly demonstrate the infiltration of the tumor into the renal vein and the inferior vena cava T2-weighted images with prospective respiratory gating should be acquired. Nephroblastoma has to be differentiated from hydro-

933

934

6  Abdomen and Retroperitoneum

Fig. 6.2.35  Transitional cell carcinoma. On the coronal T2weighted images (upper row), a hypointense structure can be appreciated, which originates from the renal pelvis and extends into the renal parenchyma of the right kidney. On the axial respiratory triggered T2-weighted images (lower row, left image), the tumor can be seen as an ill-defined hypointense structure

between the bright renal pelvis and the renal parenchyma. On the T1-weighted post-contrast injection image (lower row, right image), the entire extent of the tumor infiltration into the lower pole of the left kidney is presented. Like most transitional cell carcinomas, this tumor is hypointense to the kidney

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.36a,b  Carcinoid of the ureter. The coronal T1-weighted post-contrast image (a left side) and T2-weighted pre-contrast image (b right side) demonstrate an enhancing mass extending

along the right ureter from the renal hilus almost all the way down to the bladder. Due to the ureteral involvement urinary obstruction is present in the left kidney (b)

Fig. 6.2.37  Adult Wilms’ tumor. 38-year-old patient with newly diagnosed tumor of the right kidney, which was a histology proven Wilms’ tumor. On the T2-weighted image (upper left image) the tumor reveals a high SI with partially necrotic areas. On the T1-weighted images after the administration of a hepato-

cyte-specific contrast agent (Gd-DTPA-EOB) the tumor reveals enhancing areas in the periphery (axial image, lower left image; coronal image, right side). Due to the displacement/infiltration of the caudal segments of the liver, the adjacent segments show an altered perfusion status with decreased contrast uptake

935

936

6  Abdomen and Retroperitoneum

nephrosis and multicystic kidney disease, which can be achieved easily using MRI due to the higher SI of the latter two diseases on T2-weighted images. Nephroblastoma has to be differentiated from neuroblastoma. This can be easily achieved by determining the origin of the tumor, which is the kidney in case of nephroblastoma. 6.2.3.3.9  Rhabdoid Tumor of the Kidney This malignant neoplasm of childhood is a rare entity. It preferentially metastasizes to the brain. So far, there are no systematic descriptions of the MR characteristics of this tumor. 6.2.3.3.10  Liposarcoma of the Kidney Liposarcoma is also rare and does not manifest with specific clinical signs and symptoms. It is usually diagnosed in the fifth to seventh decade of life. The two most common sites of liposarcoma manifestation are the extremities and the retroperitoneum. Imaging may reveal tumors with or without high signal on T1-weighted non-fat-saturated images depending on the fat content and the tumor grade. Usually liposarcoma exhibits strong contrast enhancement. Myxoid type liposarcoma may appear cystic

with little fat content. Recurrent lesions are not unusual after tumor resection. High-grade tumors with metastatic potential spread hematogenously or grow directly into the pelvis. 6.2.3.3.11  Inflammatory and Infectious Renal Diseases The prevalence of pyelonephritis is about 3–9%. Acute pyelonephritis most often affects women between the ages of 15 and 40 years. T2-weighted images may demonstrate enlargement of the kidneys, a decreased SI of the renal parenchyma, and a loss in corticomedullary contrast. Gerota’s fascia might be thickened. Chronic pyelonephritis can result from the acute form or develop insidiously. It often manifests with secondary symptoms, such as hypertension and uremia. The morphologic features of chronic pyelonephritis such as shrinkage, irregular borders, and thin renal parenchyma are easily detected with MRI. Pyelonephritis is defined as an upper urinary tract infection with involvement of the renal pelvis. Risk factors include reflux disease in children and urinary obstruction or stones in adult patients. According to the course of the disease, acute and chronic pyelonephritis have to be differentiated. Acute pyelonephritis affects most often Fig. 6.2.38  Liposarcoma kidney. T2weighted HASTE images in axial (left column, upper image) and coronal (right row upper image) orientation show a large irregularly bordered mass of the right kidney with a distinct internal septation. The corresponding axial T1-weighted image (left row, middle image) shows a hypointense mass, which enhances irregularly after contrast injection (T1-weighted image post-contrast, left and right row lower image). Please note that this biopsy proven liposarcoma shows no fat-isointense signal characteristics

6.2  Kidneys, Adrenals, and Retroperitoneum

women aged 15–40 years. T1-weighted images may show swelling of the kidneys, with a decrease in corticomedullary differentiation. A thickening of the Gerota’s fascia can be seen in some cases. Contrast-enhanced T1-weighted images or T2-weighted images may show-wedge shaped foci of persistently increased SI. A third and rare form of pyelonephritis is xanthogranulomatous pyelonephritis (Table 6.2.9). Xanthogranulomatous pyelonephritis may develop in patients with abnormal immune response to bacterial infection. Particularly Proteus or Escherichia coli species are causative organisms. Often, a history of (partial) urinary obstruction (stone, ureteropelvic junction stenosis, tumor) can be found. Histology is characterized by a diffuse infiltration of the affected tissue with plasma cells and lipid-laden macrophages. This disease mainly affects women aged 45–65, but 10% of all patients are younger than 17 years. Xanthogranulomatous pyelonephritis can occur as a diffuse disease or less often as a segmental or focal disease. MRI findings include fatty masses replacing the renal parenchyma, which may show strong enhancement on postcontrast T1-weighted images. One percent of all cases of xanthogranulomatous pyelonephritis are bilateral, and up to 90% show a complete loss of renal function. The bearclaw sign—a characteristic appearance of kidneys with xanthogranulomatous pyelonephritis, with a thinned-out cortex and dilated renal pelvis—has been described as a

diagnostic finding suggestive of the presence of xanthogranulomatous pyelonephritis. Xanthogranulomatous pyelonephritis may pose a diagnostic dilemma as it may mimic malignant disease by involving adjacent structures. Due to this involvement, sinus tracts may arise. The different forms of glomerulonephritis have a loss of corticomedullary differentiation in common, which is a non-specific sign. The various forms of glomerulonephritis cannot be differentiated using MRI. Initial results suggest that USPIO contrast agents, which are being taken up by macrophages, can demonstrate inflammatory changes in kidneys and hence help to establish the diagnosis of glomerulonephritis. Table 6.2.9  Features of xanthogranulomatous pyelonephritis Xanthogranulomatous Pyelonephritis • Chronic granulomatous bacterial infection • Associated with urinary obstruction (stones, ureteropelvic junction obstruction, ureteral strictures, ureteral tumors) • 10% of all patients < 17 years • Bilateral in 1% • 73–90% diffuse spread with loss of renal function • Infiltration of the perirenal space and formation of fistula tracts

Fig. 6.2.39  USPIO kidney. Axial T2*weighted images before (upper left image) and 3 days after (upper right image) injection of Sinerem® demonstrating a diffuse signal intensity drop in a transplanted kidney with an acute rejection. In another patient with glomerulosclerosis before (lower left image) and 3 days after (lower right image) injection of Sinerem® no signal drop can be appreciated (Images courtesy of Prof. Grenier, Bordeaux, France, from: Hauger O, Grenier N, Deminière C, Lasseur C, Delmas Y, Merville P, Combe C USPIO-enhanced MR imaging of macrophage infiltration in native and transplanted kidneys: preliminary results in humans. Eur Radiol, in press)

937

938

6  Abdomen and Retroperitoneum

Malacoplakia is a rare inflammatory condition presenting as a plaque or a nodule that usually affects the genitourinary tract and the skin in immunocompromised patients and those with diabetes mellitus, status post–renal transplantation, and lymphoma. MRI depicts enlargement of the kidneys and inhomogeneous contrast enhancement. In T2-weighted images, inhomogeneous SI of the renal parenchyma is present. It is important to include this disease in the list of potential differential diagnoses as it may be cured with the use of antibiotics.

In most cases, acute and short-term obstruction of urinary outflow tract is caused by stones. Dilatation of the renal pelvis and of the ureters is only mild in these cases. The complete picture of a hydronephrosis develops only with chronic renal outflow obstruction (tumor, apparent crossing vessel at the renal pelvis, ureteropelvic junction stenosis). In chronic hydronephrosis, T1-weighted images demonstrate the low SI dilated renal calyceal system and a loss of the corticomedullary differentiation. In addition, the renal parenchyma may be thinned out as a sign of chronic parenchymal damage. On T2-weighted images, the renal calyceal system is hyperintense. Coronal images demonstrate the dilated ureter. At this, particularly heavily T2-weighted images such as RARE (rapid acquisition with relaxation enhancement) and HASTE sequences are suitable to acquire fast images with high T2-weighted SI. These sequences can also be applied after the administration of furosemide to increase the volume load on the urinary tract and hence achieve a better distension of the urinary tract. Before the administration of furosemide, however, a fast heavily T2-weighted sequence such as RARE should be acquired to rule out a massive dilatation

of the calyceal system in patients with acute obstruction. In these cases, the administration of furosemide is contraindicated as it may lead to a fornix rupture. The exact site of the renal tract obstruction can be safely diagnosed using T2-weighted sequences. A different approach for MR-imaging of hydronephrosis or suspected urinary outflow obstruction is to use Gadolinium chelates and postcontrast T1-weighted images. Typically, the excretion of the contrast agents starts immediately after the contrast injection. After as little as 1–2 min, initial enhancement can be seen in the renal pelvis and ureters. Heavily T1weighted sequences, such as 3D-GRE sequences that are mainly used for MRA, can also be applied to demonstrate the ureters. Due to the high concentration of gadolinium contrast agent in the bladder T2* effects may arise, frequently rendering some parts of the bladder non-diagnostic. Sequential acquisition of heavily T1-weighted post-contrast images allows the generation of a dynamic urography series. For T1-weighted techniques, the administration of furosemide is also advantageous. MRI also allows identification of the cause of hydronephrosis, such as retroperitoneal tumors compressing or invading the ureter. Carcinoma of the ureter cannot regularly be depicted. Stones of the renal pelvis and the ureter generally have a low SI on T1-weighted and T2weighted images. Therefore, only larger stones may be well marked off on T2-weighted images where they appear as black holes in the hyperintense urine. Ureteral obstruction can also be caused by ectopic endometrial tissue in patients with endometriosis. In these typically younger patients, MR demonstrates hyperintense lesions on fat-saturated T1-weighted sequences pre-contrast caused by hemorrhage into the endometrial lesions. Retroperitoneal fibrosis is another rare cause of hydronephrosis and is due to encasing and compressing the ureters. This disease entity is discussed in detail in the retroperitoneum section.

Fig. 6.2.40  Endometriosis. Coronal T2-weighted-RARE single shot urography (left image) and coronal T1-weighted image (middle image) post-contrast demonstrating a dilated left ureter and a dilated renal pelvis on the left side in a 28-year-old female

patient with intermittent pelvic pain. On the morphologic T1weighted axial image post-contrast (right image), a moderately enhancing mass encasing the left ureter was seen, which was resected and found to be endometriosis

6.2.3.3.12  Urinary Obstruction

6.2  Kidneys, Adrenals, and Retroperitoneum

The role of MRI in the assessment of renal transplants is twofold. First, vascular complications such as renal artery stenosis, clamp damage, of the iliac artery or renal vein thrombosis can be depicted with high sensitivity and specificity. Second, MRI allows comprehensive assessment of renal transplants by morphologic and functional imaging. Due to its superior soft tissue contrast, complications of renal transplantation such as granulomatous disease, lymphoceles, or posttransplant lymphoproliferative disease (PTLD) can be demonstrated. Vascular problems after renal transplantation occur in up to 10% of all patients. Renal artery stenosis is most often a short segment stenosis at the anastomosis site secondary to the vessel anastomosis. Clamp damage or ischemic changes of the donor vessel versus recipient iliac artery can also lead to renal transplant artery stenosis. Renal vein thrombosis may occur as an acute event and potentially cause early renal allograft dysfunction. For assessment of the transplant arteries and veins, the acquisition of a biphasic contrast-enhanced renal MR angiography is warranted. A high special resolution of roughly 1 mm isotropic is currently considered the gold standard at 1.5 T. At 3 T smaller voxel sizes can be used.

The sequences should be acquired consecutively to allow for strong arterial and strong venous signal. In imaging of renal transplants, fluid collections are of particular interest. These include urinomas, lymphoceles abscesses, and hematomas. As many of these fluid collections represent postoperative changes, they are quite often encountered in the early postoperative phase. Their imaging characteristics are quite distinct. Urinomas reveal a T2-weighted hyperintense signal but should also show some signal enhancement on postcontrast delayed T1-weighted images. In contrast, lymphoceles do not enhance on post-contrast images. In abscesses, a rim enhancement is frequently found on contrast-enhanced scans. A hematoma does not show any enhancement after contrast injection but may reveal a hyperintense signal on pre-contrast T1-weighted sequences. Acute transplant rejection does not have a morphologic correlation on MR images. In patients suffering from acute tubular necrosis (ATN), an entrapment of the contrast agent in the medulla can be appreciated. The use of functional imaging techniques such as perfusion measurements and renal oxygenation measurements (BOLD) may allow for the detection of acute or chronic rejection or ATN. In a recent publication the mean R2* values in

Fig. 6.2.41  Renal transplant urinoma. Dynamic coronal T1weighted images in a patient status post renal transplant with suspected urinoma. The images show the enhancing renal pel-

vis, the ureter, and the urinary bladder. In the late phases of the urography study contrast leakage can be seen (arrows), which is indicative of urinary leakage

6.2.3.3.13  Renal Transplantation

939

940

6  Abdomen and Retroperitoneum

the medulla/cortex of functioning renal transplants were found to be 23.9/12.7 (1/s), which was not significantly altered in ATN where the corresponding values were 21.3/14.2 (1/s). In patients with acute rejection, however, the values dropped to 15.8/12.4 (1/s), pointing at the medulla as the main site of pathology. Older publications using dynamic T1-weighted sequences reported on a 96% agreement between histopathology and this distinct per-

fusion pattern. However, their ultimate clinical value has not been evaluated in depth. The normal SI of the renal transplant is similar to that of the native kidney. Acute renal transplant rejection and acute tubular necrosis often result in an increased kidney size and a loss of the corticomedullary differentiation. Kidneys with cyclosporine-induced damage may appear with normal SI on T1weighted images.

Fig. 6.2.42  Renal transplant lymphocele. In the coronal SSFP image (left side) two hyperintense fluid collections are shown, of which the larger structure represents the urinary bladder. To differentiate between a postoperative lymphocele and an urinoma

a dynamic coronal T1-weighted urography sequence (right three images) was acquired in which no contrast leakage could be found. Therefore, the cranial fluid collection was deemed to be a lymphocele

Fig. 6.2.43  Renal transplant RAS with perfusion deficit. 30 mmthin MIP of a high-resolution MRA at 3 T in a young patient after renal transplantation demonstrating a proximal high-grade stenosis of the transplant renal artery (left image). In the perfusion measurement before the dilatation of the stenosis (middle image) there is hardly any corticomedullary differentiation to be

seen on the color-coded parameter map of the plasma flow. The mean plasma flow to the transplant was 90 ml/100 g/min. The color-coded parameter map of the mean plasma flow after therapy (right image) demonstrates a clear corticomedullary differentiation with a restored mean plasma flow of 200 ml/100 g/min

6.2  Kidneys, Adrenals, and Retroperitoneum

6.2.3.3.14  Renal Trauma As CT is used in most instances for imaging trauma patients there are no systematic studies on the role of MRI in the detection of renal trauma. In the posttraumatic state, loss of corticomedullary differentiation can be visualized even though this is a very non-specific finding. Small subcapsular or perirenal fluid collections and/or intraparenchymal areas of increased SI (hematoma) or of decreased SI (edema, urinoma) can be well appreciated with MRI. This is important as in the post-traumatic state renal hypertension may develop when a perirenal hematoma leads to decreased blood inflow into the kidney, a status also known as Page kidney. Due to the secondary activation of the renin–angiotensin pathway, the blood pressure is elevated. 6.2.3.3.15  Paroxysmal Nocturnal Hemoglobinuria In patients suffering from paroxysmal nocturnal hemoglobinuria (PNH), lower SI on T1-weighted and T2weighted images is found, which is most pronounced in the cortex. In these patients, mild nocturnal hypoxia and mild acidosis result in hemolysis of the pathologically altered red blood cells, which have a decreased acid resistance. The hereof-resulting hemoglobin and its split products are trapped in tubules and shorten the T2 times of surrounding tissue by inducing local field susceptibilities. On T2-weighted images, an inverted corticomedullary contrast can then be seen with lower medullary signal and higher cortical signal.

Table 6.2.10  Differential diagnosis of renal lesions Lesion

Characteristics

Simple cyst

Well-defined, thin wall Homogeneous SI T1-weighted hypointense T2-weighted hyperintense No contrast enhancement

Renal cell cancer

Irregular border, ill defined Inhomogeneous SI Variable SI in T1-weighted and T2-weighted Contrast enhancement (often peripheral)

Adenoma

Clearly defined, <3cm

Metastasis

Irregular shape and size Variable signal characteristics

Lymphoma

Increased renal size Diffuse reduction of corticomedullary contrast Diffuse growth into adjacent structures Growth following anatomic structures

Wilms’ tumor

Quite large at initial diagnosis Often with tumor thrombus in renal vein and/or IVC Good differentiation from renal parenchyma in arterial phase Post-contrast

Transitional cell cancer (renal pelvis)

Often spread into the renal parenchyma Irregular configuration of renal pelvis Metastases in the ipsilateral ureter

6.2.3.4  Differential Diagnosis Typical morphologic characteristics of various renal tumors as detected in MRI are displayed in Table 6.2.10. Due to the overlapping character of many renal tumor entities, MRI cannot always exactly define the character of the lesion. 6.2.3.5  Clinical Value MRI in Comparison with Other Diagnostic Modalities 6.2.3.5.1  Intravenous Pyelogram The intravenous pyelogram (IVP) including conventional renal tomography is infrequently used nowadays with some remaining relevance particularly in smaller hospitals. For assessment of morphology, it has been replaced by CT urography to a large extent. If functional assessment of the urinary outflow tract is required, then

IVP still plays a role in case dynamic MRI techniques are not available. The IVP consists of two parts. The first part is the pre-contrast image of the abdomen, which shows opacifications such as concrements in the renal pelvis or calcifications of the renal parenchyma. After contrast injection, opacification of the renal pelvis and ureters takes place. Non-calcified renal stones can then be visualized due to a filling defect of the ureter. Masses of the renal pelvis and the ureters are directly depicted as contours bulging into either the renal pelvis or the ureter. The position of the kidney and the excretory function in comparison to the contralateral side can be assessed. The information derived from the IVP is often sufficient to direct further clinical workup. Typical diagnoses that can be seen with IVP are urolithiasis, delayed opacification of the renal pelvis (urinary obstruction). In most cases however, the information derived from IVP is used for the assessment of the most appropriate next imaging modality.

941

942

6  Abdomen and Retroperitoneum

6.2.3.5.2  Ultrasound Ultrasound plays a pivotal role for the imaging of the kidneys, particularly in children. In comparison with CT and MRI, ultrasound can be easily performed even in children of 5 years or less without sedation. Ultrasound is the easiest way to differentiate between solid and cystic tumors. The approximate accuracy of ultrasound for this indication is about 87%. However, ultrasound does not allow further differentiation of solid tumors. Depending on the body habitus of the patient, the renal veins and the inferior vena cava can be nicely depicted using ultrasound. When assessing the operability of renal tumors ultrasound is inferior to CT and MRI because the exact borders of renal tumors and their relation to Gerota’s fascia cannot always be determined. With optimal ultrasound conditions, renal tumors of about 1 cm in size are the smallest that can be detected. On ultrasound, angiomyolipomas are typically hyper echoic. Yet, one has to be aware that up to 30% of all renal cell carcinomas are also hyperechoic. In patients with urinary obstruction, inflammatory renal diseases, and complications after renal transplantation, ultrasound is the method of choice. This is mainly due to the non-invasiveness and the widespread availability of this method. Ultrasound allows visualizing the flow patterns in the arteries and veins. Theoretically, ultrasound can accurately show posttraumatic changes after a blunt abdominal trauma. Yet the use of the ultrasound probe can be painful in these patients when rip fractures are present. It is also well known that obese patients and patients with abdominal bloating are not suitable for ultrasound exams. 6.2.3.5.3  Computed Tomography For imaging of the kidneys, computed tomography (CT) is the imaging modality that is most often used and that has good diagnostic accuracy. This is especially true for the diagnosis of renal cell carcinomas, which are accurately detected in up to 96% of cases. One of the few disadvantages of CT is impaired detection of infiltration of the perirenal fat tissue in early stage T3a of renal cancer as well as the impaired detection of the invasion of the renal vein (T3b). This is particularly true in cases of infiltration of the intrarenal parts of the renal vein. As in both cases, the operation method is identical (total tumor nephrectomy), these disadvantages from an imaging prospective are not clinically relevant. Differentiation of hyperdense from solid and potentially hemorrhagic renal cell cancers is hard to perform with CT. In addition, the differentiation between adenomas, lymphomas, metastases, and renal cell carcinomas compared to complex cysts and cystic malignancies is not always feasible with CT. Different CT criteria including the detection of calcifications and their morphology

as well as their pattern of distribution can be used for differential diagnosis. A common classification for cystic renal masses is the Bosniak classification. This classification has been designed for use in CT. It differentiates between non-surgical and surgical renal cysts. Due to the dependence of this classification on the presence of calcifications, which cannot be seen on MRI, this classification is not suitable to be applied to MRI images. As CT allows direct measurement of the density of tissues examined, angiomyolipomas can often be reliably diagnosed when they contain mainly lipid tissue. If angiomyolipomas contain a higher content of muscle tissue, then they may not be picked up readily on non-enhanced CT images. CT has developed into the main clinical tool for the assessment of trauma patients. Lacerations of the renal parenchyma or vessel disruptions after blunt abdominal trauma can be very well diagnosed using CT. In these cases, CT demonstrates delayed or lacking renal enhancement after contrast injection. Whether additional invasive angiography is necessary can then be decided based on the CT images. In patients status post–renal transplantation in early stages CT is superior compared to ultrasound because foreign material does not prevent imaging. However, one has to keep in mind that the administration of iodine-based contrast agents can potentially lead to contrast induced nephropathy. Children of 5 years or less should be sedated for a CT exam. Oral contrast agent administration is desirable in those patients as children often have little retroperitoneal fat tissue and thus the demarcation of renal masses can be challenging. In the case of huge renal masses, CT does not always allow exact determination of the organ of origination. Another disadvantage of CT is the administration of iodine-based contrast agents, which as mentioned above can lead to contrast-induced nephropathy. The most important single measure to prevent patients from contrastinduced nephropathy is to hydrate them well before and after the CT exam. 6.2.3.5.4  Invasive Angiography Invasive angiography has evolved from a diagnostic modality to a predominately therapeutic modality. The previous indications for angiography such as detection of renal masses and characterization of the blood supply of renal masses as well as detection of renal vein infiltration can now easily be taken care of with CT or MRI. At this time, invasive angiography only seems to be warranted for the interventional treatment of renal artery stenosis or the embolization of renal tumors preoperatively to minimize intraoperative hemorrhage.

6.2  Kidneys, Adrenals, and Retroperitoneum

6.2.3.5.5  Scintigraphy Scintigraphy using 99Tc-DTPA and 131I or 123I-hippuran and 99Tc-MAG 3 yields quantitative and reproducible functional data (perfusion and secretion). The technique is applied in patients post–renal transplant, in whom the renal function as well as the presence of ureteral obstructions has to be evaluated. The methods of scintigraphy have been considered superior to IVP and CT as well as standard morphologic MRI in terms of determination of the renal function. They are, however, characterized by a much lower spatial resolution and a lower signal-to-noise ratio as well as long examination times. In the past 10 years new MRI techniques have arisen, which allow exact quantification of renal blood flow, split renal function and renal excretion. Therefore, the use of scintigraphy seems to be of decreasing importance. 6.2.3.5.6  MRI MRI has accuracy similar to that of CT for the detection and grading of renal cell carcinomas. As with CT, metastases in normal sized lymph nodes cannot be reliably detected. Unlike CT, MRI allows exact determination of infiltration of the renal vein and the inferior vena cava with tumor thrombus and apposition thrombus. For this purpose, respiratory gated T2-weighted sequences or T1weighted MRA sequences in the venous phase are used. Coronal slice orientation facilitates detection of larger renal masses and better assessment of the extent of a tumor thrombus in relation to the diaphragm. MRI, CT, and ultrasound have similar accuracy levels in the diagnosis of cystic renal masses. Due to the higher soft-tissue contrast, MRI reveals septa that cannot be seen on CT examinations. The SI for hemorrhagic cysts is quite characteristic but not pathognomonic. Therefore, hemorrhagic renal cell carcinomas cannot be reliably differentiated from hemorrhagic cysts. Apart from angiomyolipoma, which shows characteristic signal cancellation in opposed-phase images, MRI does not allow characterization of solid renal masses. MRI is superior to ultrasound and CT in the detection and grading of Wilms’ tumors and for the characterization of the relationship of the tumor to the neighboring organs and the large vessels. As with CT, children of 5 years or less should be sedated for the MR examination. MRI is particularly suitable for the assessment of renal transplant dysfunction. Due to the high soft tissue contrast complications such as uroceles, lymphoceles as well as vascular complications such as renal artery stenosis, clamp damage or renal vein thrombosis can be reliably detected with MRI. With arising new imaging techniques, even a functional characterization of renal transplants can be achieved to yield additional information on function and focal rejection.

The disadvantages of MRI are the long measurement times and the susceptibility of MRI to motion. Another disadvantage is the inability to depict calcifications. 6.2.3.6  Indications for Imaging Due to the comprehensive morphologic and functional information as well as the lack of iodinated contrast media, MRI is the method of choice for evaluation of renal donors and transplant patients. Like ultrasound and CT, MRI has high sensitivity for the detection of pathologic renal changes. MRI often cannot improve upon CT or ultrasound in the characterization of pathologic findings. Therefore, renal pathologies do not represent an immediate indication for MRI. MRI is warranted in those cases where CT and ultrasound are either non-diagnostic or equivocal. Also for patients with Wilms’ tumor or patients post-transplant where additional functional images can be obtained, MRI seems to be superior to the other two imaging modalities. In addition, in patients with a known allergy to iodine-based contrast agents or impaired renal function MRI is the method of choice. 6.2.4  Adrenal Gland 6.2.4.1  Dedicated Examination Technique The examination of the adrenal gland region should be started with axial T1-weighted gradient-recalled echo-sequences in breath-hold mode. These sequences allow for a good orientation over the adrenal gland region. With current hardware, a slice thickness of 4 mm should be chosen. This allows for sufficient signal-to-noise ratio and permits the detection of smaller lesions of the adrenal glands. To differentiate between masses originating from the adrenal gland and those originating from the kidney, coronal, or sagittal plane imaging needs to be done as well. Additionally, T1-weighted fat-saturated sequences should be acquired as the adrenal gland can then be detected with high SI and high contrast to the surrounding fat tissue. For further characterization of adrenal gland lesions, in-phase and opposed-phase T1weighted sequences should be acquired. Adrenal gland adenomas demonstrate a characteristic signal drop in the opposed-phase images compared to the in-phase images. This signal drop should be at least 40%. With the chemical shift technique, the sensitivity and specificity for differentiating adenomas from metastases range from 81 to 100% and 94 to 100%, respectively. T2-weighted sequences allow for further characterization of adrenal gland masses. Cystic changes, carcinomas, and pheochromocytomas will reveal high SI on T2-weighted images. To minimize partial volume effects with the surrounding fat tissue, the

943

944

6  Abdomen and Retroperitoneum Table 6.2.11  Adrenal disease entities with hormone abnormalities Syndrome

Hyperplasia (%)

Adenoma (%)

Carcinoma (%)

Cushing’s syndrome   (incidence 0.01%, peak age: 25–35 years)

75–85

10–15

5–10

Conn’s syndrome   (prevalence in patients with hypertension 0.5–1.0%,   peak age: 30–50 years)

25–40

60–75

Extremely rare

Adrenogenital syndrome (1 in 1,000 live births)

Common

Rare

Extremely rare

application of fat suppression techniques is helpful, also for T2-weighted images. Contrast administration should be performed in a dynamic fashion to demonstrate enhancement of the lesion and to derive information on the vascularity of the lesion. Definite MRI criteria for the typical enhancement of adenomas do not exist at this time. In CT, lesions exhibiting an increase of less than 30 HU after contrast injection and a washout of at least 50% after 10 min can be safely diagnosed as adenomas. T1weighted images post-contrast should also be acquired with fat-saturation. This facilitates the demonstration of contrast-enhancing lesions. 6.2.4.2  Normal Anatomy The craniocaudal diameter of the adrenal glands located in the perirenal space is about 2–4 cm. The right adrenal gland, which is located slightly more cranially than the left adrenal gland, can be localized between the right crux of the diaphragm and the right lobe of the liver and has a thickness similar to that of the right crux of the diaphragm. The morphology is often described as an inverted W. The left adrenal gland also has thickness similar to that of the ipsilateral crux of the diaphragm. It is located lateral of the aorta and slightly superior or anterior to the left kidney. The configuration of the left adrenal gland has most often been described as an inverted Y. A normal adrenal gland has intermediate SI on T1weighted and T2-weighted sequences. The contrast to the surrounding perirenal fat tissue is higher on T1-weighted images than on T2-weighted images. The adrenal glands show higher signal intensities than the crux of the diaphragm on all pulse sequences. 6.2.4.3  Pathophysiology Cushing’s disease is caused by a pituitary tumor with an excess ACTH excretion, leading to bilateral adrenal gland hyperplasia. Hyperfunctioning tumors of the adrenal cortex can produce Cushing’s syndrome from cortisol

Table 6.2.12  Epidemiology of adrenal gland diseases Entity

Prevalence (%)

Peak age (years)

Adenoma

2–9

Metastasis

27 (in patients with malignant disease)

Carcinoma

0.0002

40–50

Pheochromocytoma

0.005–0.1

40–50

Neuroblastoma

0.01 (incidence)

2

Myelolipoma

0.2–0.4

hyperproduction, Conn’s syndrome from production of aldosterone, or hyperandrogenism from overproduction of androgens. Conn’s syndrome is often caused by microadenomas of the adrenal gland, which cannot be detected with imaging studies. In adrenogenital syndrome (AGS), an inherited disease of the adrenal gland enzyme deficiency leads to decreased/lacking secretion of cortisol or aldosterone. Through the lack of feedback regulation, the central release of ACTH is increased, leading to bilateral adrenal gland hyperplasia with release of 17-ketosteroids with androgenic effect. 6.2.4.4  Pathology 6.2.4.4.1  Hyperplasia of the Adrenal Gland Cortex Hyperplasia of the adrenal gland cortex can occur in patients suffering from Cushing’s syndrome and Cushing’s disease, Conn’s syndrome and in patients suffering from adrenogenital syndrome (AGS) (Tables 6.2.11, 6.2.12). Also, hyperplasia can be found in patients who exercise frequently. On MRI, a bilateral broadening of the crux of the adrenal gland can be seen, which often reaches

6.2  Kidneys, Adrenals, and Retroperitoneum

10 mm or more. At this, the adrenal glands can be configured in a nodular fashion bilaterally. In patients suffering from Cushing’s syndrome or Cushing’s disease the increased amount of intra-abdominal fat tissue facilitates the delineation of the adrenal glands from the perirenal fat tissue. In patients suffering from adrenogenital syndrome a nodular hyperplasia of the adrenal glands can be seen. The SI of hyperplastic adrenal glands is not different from the SI of normal adrenal glands.

Adrenal gland adenomas can be found in patients suffering from Cushing’s syndrome, and Conn’s syndrome. They are extremely rare in patients suffering from adrenogenital syndrome. Hormone-producing adenomas often lead to atrophy of the remaining parenchyma of the affected adrenal gland as well as of the contralateral

adrenal gland. Non- or only slightly hormone-producing adrenal gland adenomas occur far more often than hormone-producing adrenal gland adenomas. These non-hormone-producing adenomas are most often detected incidentally and occur bilaterally. Even in cancer patients, about 50% of all incidentally detected adrenal masses are adenomas despite the high frequency of adrenal metastases. Most of the time they measure less than 3 cm in diameter. Adrenal gland adenomas causing Cushing’s syndrome are usually greater than 2 cm in diameter and often readily visualized on CT scans while adenomas causing Conn’s syndrome are less than 2.5 cm in size in 90% of all cases, and in 20% of all cases even smaller than 1 cm in size. Adrenal gland adenomas are isointense or slightly hyperintense compared to the liver on T2weighted images. On fat-saturated T2-weighted images, a slightly higher SI of the tumor can be found. Adrenal gland adenomas most often show a round configuration

Fig. 6.2.44  Adrenal adenoma. Axial T1-weighted image in phase (upper left image), T1-weighted opposed-phase image (upper right image) and T1-weighted image post-contrast injection (lower left image) of a patient with adrenal adenoma. This example demonstrates typical imaging features of adrenal adenomas: they are round and well defined in size, of moderate size

(2–4 cm), show an at least 30–40% signal drop in the opposedphase image and do not enhance significantly. CT correlation of the same patient (lower right image) again demonstrating the well-defined small adenoma. On CT the masses are isodense to hypodense pre-contrast and show only little uptake after contrast injection

6.2.4.4.2  Adrenal Gland Adenomas

945

946

6  Abdomen and Retroperitoneum

with smooth borders. Their SI is homogenous over the entire tumor. An almost perfectly pathognomonic finding is a signal drop of 30–40% on opposed-phase images as compared to in-phase images. This signal drop is due to the presence of cytoplasmatic fat in adenomas. After contrast-bolus injection, a homogeneous slightly low enhancement with early washout can be seen in adenomas. Metastases tend to show a more heterogeneous enhancement with no early washout. 6.2.4.4.3  Adrenal Gland Metastases Metastases are the most common malignant tumor of the adrenal glands. In up to 40% of cases, both adrenal glands are affected. However an insufficiency of the adrenal glands due to metastatic spread to the adrenal glands is very rare. Probably due to the higher blood supply, the adrenal gland is the fourth most common organ to which other cancers metastasize. In patients with underlying malignant disease, an incidentally detected adrenal gland tumor is a metastasis in up to 50% of cases. Small metastases reveal a homogeneous signal on T1weighted and T2-weighted sequences, while larger metastases reveal a heterogeneous structure and a poorly defined border. In T1-weighted images, adrenal gland metastases are most of the time hypointense compared to the liver and hyperintense on T2-weighted images. They can have the same SI as the primary tumor. After

Fig. 6.2.45  Adrenal metastasis. Adrenal metastasis in a patient with bronchial carcinoma. The pre-contrast T1-weighted image (left image) and the T2-weighted image (right image) demonstrate slightly heterogeneous SI. In the T2-weighted image, the metastasis is only slightly hyperintense, which allows for dif-

bolus contrast injection of paramagnetic contrast agents, adrenal gland metastases demonstrate a significant and long lasting (approximately 50 minute) SI increase on dynamic studies. Even though adrenal gland metastases reveal a characteristic SI and can often be differentiated from adenomas, lesions of more than 2 cm in size often require biopsy as the ultimate means of establishing a diagnosis. Lymphomas of the adrenal gland most often occur with non-Hodgkin’s lymphomas and are rarely seen in patients suffering from Hodgkin’s disease. They do not reveal a characteristic SI pattern. 6.2.4.4.5  Adrenal Gland Carcinomas Adrenal gland carcinomas are counted among the rare malignant diseases and can be hereditary. From 50 to 75% of adrenal gland carcinomas are hormone-producing tumors. Due to the relatively small hormone production, characteristic symptoms such as Cushing’s syndrome occur in delayed phases of the disease. In up to 90% of cases, adrenal gland carcinomas therefore measure more than 6 cm at the time of diagnosis. In one third of patients, metastases to the lung, the liver and the locoregional lymph nodes have occurred by the time of diagnosis. Direct spread into the kidney, the inferior vena cava, and the retroperitoneum are often seen. Therefore, in patients with suspected adrenal gland carcinoma examination of the liver and the entire retroperitoneum is

ferentiation to pheochromocytomas, which are strongly hyperintense. After contrast enhancement, metastases show a wide spectrum of enhancement patterns depending on the primary tumor

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.46  Adrenal carcinoma. Large adrenal carcinoma of the right adrenal gland with displacement of the liver and of the right kidney, which can be appreciated on the pre-contrast axial T1-weighted image (left image), the T2-weighted fat-saturated image (middle image), and the axial T1-weighted image

in the arterial phase after contrast injection (right image). The tumor reveals heterogeneous SI on the T1-weighted and the T2weighted images with moderate enhancement in the peripheral vital tumor parts after contrast injection

strongly recommended. Adrenal gland carcinomas are hyperintense compared to the liver on T2-weighted images. Large tumors often demonstrate inhomogeneous SI values on T2-weightedeigted images due to necrosis and calcifications. They are ill defined. The SI after bolus injection of a paramagnetic contrast agent can compare to that of metastases. On contrast-enhanced T1-weighted images, there is much better demarcation of necrotic areas within the tumor. Infiltration of the inferior vena cava is not uncommon.

of the sympathetic chain outside of the adrenal gland medulla. Therefore, in patients having a clinical suspicion of pheochromocytoma the examination should include the entire retroperitoneum as well as the pelvis including the organ of Zuckerkandl. In 5 to 10% of all cases, pheochromocytoma is associated with other neuroectodermal diseases such as neurofibromatosis, von Hippel-Lindau syndrome, tuberous sclerosis, and Sturge-Weber syndrome. Three percent of all pheochromocytomas are associated with multiple endocrine neoplasms (MEN II). In patients suffering from MEN syndromes, pheochromocytomas are virtually always bilateral and almost never extra-adrenally located. Malignant pheochromocytomas seem to be more often located extra-adrenally than in the adrenal medulla. They metastasize preferably into the lungs, the liver, and the skeleton (Table 6.2.13). At the time of diagnosis, pheochromocytomas are most often larger than 3 cm and measure on average 5 cm in diameter. Large tumors are irregularly bordered. A typical characteristic of pheochromocytomas is the high SI on T2-weighted images. Even compared to the retroperitoneal fat tissue they are hyperintense. Large pheochromocytomas can reveal necrosis, hemorrhagic transformed areas, and cysts. From these features, inhomogeneous SI can result on T2-weighted images. Calcifications of the tumor, which can be seen in up to 7% of the cases on CT images, are another source of inhomogeneous SI. After bolus injection of paramagnetic contrast agents, pheochromocytomas reveal a contrast pattern similar to that of metastases and carcinomas, with substantial signal increase and signal increase lasting as long as 50 min. Malignant pheochromocytomas cannot be differentiated from benign pheochromocytomas based on imaging findings. The sole criterion for the definition of malignant disease is the invasive nature of the tumor growth and the detection of distant metastases.

6.2.4.4.6  Pheochromocytoma The pheochromocytoma is the most common tumor of the adrenal gland medulla. The “rule of tens” applies to pheochromocytomas stating that bilateral affection is present in 10% of cases, malignant transformation occurs in 10% of cases, 10% are familial, and 10% of all pheochromocytomas may occur in the ganglions along the course

Table 6.2.13  Multiple endocrine neoplasia (MEN) syndromes MEN I (Wermer’s syndrome) Primary hyperparathyroidism Hormone-producing pituitary tumors Endocrine pancreatic tumors MEN Iia (Sipple’s syndrome) Primary hyperparathyroidism Medullary carcinoma of the thyroid (Bilateral) pheochromocytomas MEN IIb Similar to MEN IIa with additional phenotypic changes (habitus, skeleton, teeth)

947

948

6  Abdomen and Retroperitoneum

Fig. 6.2.47  Adrenal pheochromocytoma. Axial images of a patient with adrenal pheochromocytoma: The tumor reveals no signal drop from in-phase T1 (upper left image) to T1 opposedphase (upper right image) imaging. The T2-weighted image (lower left image) demonstrates typical areas of high SI, which

correspond to necrotic tumor areas. After contrast injection (lower right image), the tumor shows moderate enhancement of the non-necrotic tumor parts. As in this case, pheochromocytomas demonstrate a non-spherical growth pattern in contrast to adrenal adenomas, which may be helpful in differentiating both

Fig. 6.2.48  Extra-adrenal malignant pheochromocytoma. The sagittal T2-weighted HASTE sequence demonstrates a soft tissue enlargement paravertebrally (arrow, left image) in this 37-yearold female patient in whom a malignant pheochromocytoma has already been resected. The 15-mm-thin MIP images (middle

and right image) from the arterial phase of the MRA exam in the same patient demonstrate an enhancing mass, which was histologically proven to be a reoccurring extra-adrenal malignant pheochromocytoma

6.2  Kidneys, Adrenals, and Retroperitoneum Fig. 6.2.49  Adrenal myelolipoma. The medium-sized myelolipoma of the left adrenal gland shows high SI on the presented T2-weighted TSE image (upper image) and T1-weighted in phase image (middle image,) and only slight enhancement at the periphery after contrast enhancement on the coronal T1-weighted image (lower image). In contrast to adenomas, the myelolipoma does not show signal cancellation in opposed-phase images. In contrast, a spectral fat sat as in the post-contrast T1-weighted image will suppress the signal. This image characteristic is caused by the presence of fat tissue only in the lesion. Adenomas contain fat and muscle tissue and therefore reveal the characteristic signal drop in opposed-phase images

949

950

6  Abdomen and Retroperitoneum

6.2.4.4.7  Neuroblastoma

6.2.4.4.8  Myelolipomas

Neuroblastomas are the most common abdominal tumors of early childhood. In up to 50% of cases neuroblastomas are located intra-abdominally, and two thirds of all cases originate from the adrenal medulla. They can however originate from every ganglion of the sympathetic chain. Therefore these tumors can also be localized in the head and neck area (2–5%), in the thorax (10–20%) or in the pelvis (5%). At the time of diagnosis 50–75% of the children already have metastatic disease with metastases to the bone marrow, the skeleton, the liver, the skin and in later stages of this disease to the lungs. Neuroblastomas are often characterized by a polycyclic border and demonstrate growth over the midline. They can infiltrate the kidneys and encase the large vessel; in up to 50% of all cases they infiltrate the spinal canal. As the disease almost exclusively occurs in children, staging exams without ionizing radiation are desirable. Therefore, initial reports on whole-body MRI for the local staging and the detection of distant metastasis in patients with neuroblastoma have been published. The SI of neuroblastoma on T1-weighted images is almost identical to that of the renal medulla. Compared to the renal cortex the tumors often appear to be slightly hypointense. On T2-weighted images, neuroblastomas have an isointense SI compared to a normal kidney. In up to 84% of all cases, calcifications are present in neuroblastomas and account for an inhomogeneous SI distribution. Post-chemotherapy neuroblastomas also reveal inhomogeneous SI. The assessment of the relation of the tumor to the large vessels and to the spinal canal is of crucial importance for the planning of an operation. For this purpose, coronal images and transfers images are most suitable. Identification of the potential organ of origin of neuroblastomas can most easily be done on coronal slices.

Myelolipomas are rare tumors of the adrenal glands. They most often occur unilaterally and are asymptomatic. They have a typically high-fat-tissue component and hematopoeitic tissue and therefore reveal high SI on T1-weighted and T2-weighted images. On fat-suppressed images, they are hypointense. Due to their high fat content and the small number of voxels having both fat and water, they typically do not reveal a signal drop on opposed-phase images. Non-fatty components of these tumors demonstrate moderate SI, so inhomogeneous SI can occur. Up to 20% of myelolipomas reveal calcifications, which can also account for inhomogeneous SI.

Fig. 6.2.50  Adrenal myelolipoma CT companion case. The myelolipoma of the right adrenal gland shows high SI on T1weighted image (right image) and only slight enhancement in

6.2.4.4.9  Cysts Cysts of the adrenal glands are rare and comprise endothelial, hemorrhagic (pseudocyst), epithelial and parasitic cysts. Endothelial cysts can be multi-septated with fluid-fluid levels from old hemorrhage. Thickened irregular septa with strong enhancement should warrant further diagnostic work-up to rule out malignancy. Pseudocysts (39%) occur after hemorrhage into the adrenal gland. They can be large and can cause pain to the patient. Rarely parasitic cysts occur (7%). Cysts of the adrenal glands are calcified in 50% of cases, and 80% of cases occur unilaterally. Therefore, an adrenal gland insufficiency is extremely rare. The imaging morphology of adrenal gland cysts is the same as that of common renal cysts. They reveal low SI on T1-weighted images, high SI on T2-weighted images and should not reveal contrast enhancement after gadolinium injection.

the periphery after contrast administration (middle image). The corresponding CT image demonstrates a centrally hypodense adrenal gland representing the lipid center of the lesion

6.2  Kidneys, Adrenals, and Retroperitoneum Fig. 6.2.51  Adrenal cyst. Axial (left image) and coronal (right image) T2-weighted HASTE image of a 27-year-old female patient with a benign, multiloculated cyst of the left adrenal gland. There are multiple fluid levels

6.2.4.5  Differential Diagnosis Table 6.2.14 gives an overview of the morphologic and MR imaging findings of different adrenal gland tumors. One has to keep in mind that MRI is not specific enough for a differential diagnosis of adrenal gland tumors. MRI is only highly sensitive for the diagnosis of adrenal gland adenomas, which can be identified by means of the signal drop on opposed-phase images. There is a large overlap in the imaging findings of other tumor entities.

6.2.4.6  Value of MRI in Comparison with Other Imaging Modalities 6.2.4.6.1  Conventional X-Ray Calcification of the adrenal glands as seen with neuroblastomas and tuberculosis of the adrenal glands can be well seen on kidney, ureters, and bladder (KUB) images. For non-calcified methods, conventional X-ray techniques do not yield any additional valuable information.

Table 6.2.14  Differential diagnosis of adrenal lesions: diagnostic hints Adenoma

Metastasis

Carcinoma

Pheochromo­ cytoma

Neuroblastoma

Myelolipoma

Cyst

Size

<3 cm

Often >3 cm

>6 cm

>3 cm

>6 cm

Contour

Smooth

Variable

Irregular

Variable

Polycyclic

Smooth

Smooth

T2weighted pattern

Homogeneous

Variable

Inhomogeneous

Inhomogeneous

Inhomogeneous

Inhomogeneous

Homogeneous

T2weighted SI (reference: liver)

Variable

Isointense–hyperintense

Isointense–hyperintense

Hyperintense

Isointense–  hyperintense

Hyperintense on non-fat-sat T2-weighted

Hyperintense

T1weighted opp. phase

Homogenous signal drop of 30–40%

Unchanged

Unchanged

Unchanged

Unchanged

Inhomogeneous signal drop

Unchanged

Gd enhancement

<100%, early washout

>200%

>200%

>200%

No reliable data

Moderate

Non

951

952

6  Abdomen and Retroperitoneum

6.2.4.6.2  Ultrasound Ultrasound allows for depiction of the right adrenal gland in up to 80% of cases. The left adrenal gland is often obscured by abdominal gas in the stomach, the duodenum, and the colon. The depiction of the left adrenal gland can only be achieved in 45% of all cases. In obese patients such as those suffering from Cushing’s disease, performance of ultrasound is even less successful. In contrast, on CT the differentiation of the adrenal glands from retroperitoneal fat can be problematic in slender patients. These patients are well suited for ultrasound exams. Ultrasound cannot be considered as a first-line diagnostic modality for imaging of adrenal gland pathologies except in pediatric patients. Adrenal cysts 2 cm in size or larger can be differentiated from adrenal gland tumors. 6.2.4.6.3  Computed Tomography Due to its relatively low costs and broad availability, computed tomography is the number one imaging modality for the adrenal glands. The adrenal glands can be imaged successfully in almost all cases. Adrenal gland tumors 1 cm in size or larger can be detected with high sensitivity. The administration of contrast agent facilitates the detection of tumors. Also calcifications of the tumors can be reliably diagnosed with CT. A pre-contrast density of 10 HU and less with an increase of less than 30 HE after contrast injection and with a washout of at least 50% after 10 min virtually proves the presence of an adrenal adenoma. In cachectic patients and in very slender patients, the demarcation of the right adrenal gland from the inferior vena cava can be hard to achieve. While the CT allows for a very sensitive detection of adrenal gland masses, a differential diagnosis is hard to achieve apart from cysts and myelolipomas. In addition, one has to keep in mind that many adrenal gland tumors are found incidentally and patients are often re-examined to rule out malignant adrenal disease. In a typical low-pretest probability population, most adrenal tumors are benign. For these patients it seems questionable whether the use of ionizing radiation for a triphasic CT exam is warranted. 6.2.4.6.4  Scintigraphy Scintigraphy of the adrenal gland cortex with 131I-marked cholesterol or 75Se-marked norcholesterol demonstrates increased bilateral uptake in adrenal gland hyperplasia. Hormone-producing adenomas also demonstrate an increased uptake whereas the contralateral adrenal gland does not show any uptake. Adrenocortical carcinomas demonstrate low tracer uptake, whereas the contralateral adrenal gland most often shows normal tracer uptake. The disadvantages of scintigraphy include a long exami-

nation time with relatively low spatial resolution, and a signal-to-noise ratio lower than those of other imaging modalities. In addition, scintigraphy exposes the patient to 5–10 times more ionizing radiation than does CT. The adrenal gland medulla can be examined using 131I or 123Imeta-benzylguanidine (MIBG). Adrenal and extra-adrenal pheochromocytomas as well as neuroblastomas can be diagnosed. The very low uptake in the normal adrenal gland cannot be demonstrated. The sensitivity of this method for the detection of pheochromocytomas is 87% and specificity is almost 100%. Also because of the uptake in liver, lung and bone metastases MIBG scintigraphy can be a very helpful imaging modality in patients with suspected malignant pheochromocytoma or neuroblastoma. 6.2.4.6.5  Invasive Angiography Invasive angiography theoretically allows demonstration of the blood supply of adrenal gland masses. Due to the variable blood supply of the adrenal glands invasive angiography is, however, time consuming and technically demanding. For diagnostic purposes computed tomography, angiography and MR angiography seem to be completely sufficient. Angiography can, however, be used to sample blood from the renal veins to determine the amount of aldosterone. Although this is an invasive approach, a high sensitivity of 96% for the detection of aldosterone-producing adenomas can be achieved. 6.2.4.6.6  MRI MRI allows depiction of both adrenal glands in almost all cases. To determine the origin of large tumors in the retroperitoneal space coronal or sagittal slice orientation is mandatory. MRI allows detection of adrenal gland tumors as small as 1 cm in size. For the detection of extraadrenal, intra-abdominal pheochromocytomas, MRI is more sensitive than CT is, but probably less sensitive than scintigraphy using MIBG. For the diagnosis of neuroblastoma, MRI is more accurate in defining the relationship of the tumor to the large vessels and in respect to the spinal canal. The infiltration of the vena cava and renal vein can easily be detected using MRI. In this respect, MRI is also superior to CT. The differentiation of different adrenal gland tumors is limited using MRI. Solely adenomas and myelolipomas reveal characteristic signal intensities. Reliable differentiation between pheochromocytoma, metastasis, and primary carcinoma of the adrenal glands cannot be achieved using MRI. MRI has the well-known disadvantage of longer examination times compared with CT. Also MR examinations are more expensive. As with CT, children of 5 years or less should be sedated to allow for a motion-free imaging.

6.2  Kidneys, Adrenals, and Retroperitoneum

In summary, it seems appropriate to examine younger patients with MRI while CT may be an alternative in elderly patients. For children suffering from neuroblastoma whole-body MRI seems to be a very promising, accurate, and non-invasive modality for staging of local disease and metastases. In theory, whole-body MRI can also be applied to monitor response to therapy. For tumors of the adrenal glands, scintigraphy with 131I or 123I-meta-benzylguanidine (MIBG) seems to be the most appropriate examination. 6.2.5  Lymph Nodes and Retroperitoneal Tumors 6.2.5.1  Dedicated Examination Techniques MRI does not serve as a routine diagnostic modality for the diagnosis of retroperitoneal lymph nodes. T2-weighted fat-saturated sequences such as turbo- pin-echo fat-saturated sequences, and mildly T2-weighted inversion-recovery fat-saturation sequences such as STIR allow for a sensitive detection of increased water content in lymph nodes. In these sequences, lymph nodes often reveal a higher SI than surrounding fat tissue does. High SI is indicative of malignant transformation in a lymph node. In patients with Whipple’s disease or lymphangioleiomyomatosis and tuberous sclerosis, the lymph nodes may also reveal a high SI. However, these are rare diseases after all. In the last few years, diffusion-weighted echo planar sequences are increasingly used for detection of lymph nodes. With the use of dedicated software, these images can be fused with STIR images to yield a PET-like image. So far, this approach has only been shown to improve the display of lymph nodes; reliable differentiation of benign and malignant lymph nodes based on the different apparent diffusion coefficients has not been demonstrated in larger studies. Post-therapeutic states are often characterized by hypointense SI of the lymph nodes or of post-therapeutic tumor remnants. To allow for an exact planning of radiation therapy in patients with malignant lymphomas a coronal slice orientation is very helpful in determining the size of the radiation field. For most other applications, an axial slice orientation with 4–6 mm thin slices should be chosen. T1-weighted dynamic images post-contrast for the exact determination of the borders and vascularization of the lymph nodes are mandatory. The recent market approval of a USPIO contrast agent (Sinerem®, Guerbet) with high uptake in lymph nodes in one European country may be a first step towards a lymph node-specific imaging. The value of these contrast agents for the detection of lymph nodes metastases so far has only been convincingly demonstrated in large studies of patients with prostate cancer but needs to be determined for other malignant diseases in future studies.

Fig. 6.2.52  Tuberous sclerosis with multiple para-aortic lymph nodes. Axial T2-weighted-HASTE images (upper two images) in a patient suffering from tuberous sclerosis demonstrating enlarged and hyperintense lymph nodes para-aortally (arrows). In the T1-weighted post-contrast image, small hypoenhancing areas can be appreciated in the right kidney (arrow), which represent angiomyolipomas. On the left kidney a huge angiomyolipoma (arrowhead) can be appreciated, which led to a total disruption of the renal anatomy

6.2.5.2  Normal Anatomy As with CT, retroperitoneal lymph nodes of up to 15 mm in the longest diameter can be considered normal. Yet there is ongoing discussion about whether the long or the

953

954

6  Abdomen and Retroperitoneum

short axis of lymph nodes should be analyzed. Some authors are in favor of a 10 mm size limit for normal lymph nodes. On T1-weighted images, normal lymph nodes should have intermediate SI, which should be almost similar to that of the psoas muscle. Therefore, differentiation of normal lymph nodes and large vessels, which should be hypointense, as well as retroperitoneal fat, which should be hyperintense, can be performed without difficulty. On T2-weighted images, lymph nodes only have a slight contrast to the fat tissue surrounding them. Fatsuppressed T2-weighted sequences are therefore superior in depicting the lymph nodes as hyperintense structures within the dark, fat-saturated surrounding tissue. 6.2.5.3  Pathology 6.2.5.3.1  Lymphoma The main indication for imaging patients with malignant lymphomas, either Hodgkin’s disease or non-Hodgkin’s lymphoma, is to assess the extent of the disease. The results from these imaging studies have profound implications for the patient as they may alter the therapeutic approach. At this point, mainly morphologic parameters such as te size and shape of lymph nodes are included in the assessment of MRI and CT studies. Therefore, these modalities have low sensitivity for pathologic lymph nodes of normal size. PET-CT is much more sensitive in depicting small tumor-affected lymph nodes. The only advantage of MRI compared to the other cross-sectional imaging modalities is that on T2-weighted images tumor tissue is mainly hyperintense whereas post-therapeutic tumor tissue or scar tissue reveals low SI on T2-weighted images. Therefore, T2-weighted images can be used for orienting differentiation between scar tissue and lymph node metastases on images after adrenalectomy. MRI has lower sensitivity than PET-CT for the detection of the extent of disease in the soft tissue. However, MRI is superior to PET-CT in detecting involvement of the bone marrow and the brain by lymphoma. 6.2.5.3.2  Lymph Node Metastases Retroperitoneal lymph node metastases occur in a multitude of different malignant tumors such as renal cell carcinoma, testicular cancer, ovarian cancer, cancer of the uterus, bladder cancer or prostate cancer, gastric cancer, colon cancer, and pancreatic cancer as well as with malignant systemic diseases. Retroperitoneal lymphadenopathy can occur either as single lymph node enlargement or as a lymph node conglomerate tumor in which single lymph nodes cannot be delineated from each other. On MRI diagnosis of lymph node metastases is based on increased lymph node size and not on a specific SI change. Para-

aortic or paracaval lymph nodes more than 15 mm in size have to be considered pathologic. Retrocrural lymph nodes are considered pathologic starting from 6 mm in size. Grouped retroperitoneal lymph nodes between 10 and 15 mm size are also considered pathologic. Enlarged metastatic lymph nodes have homogeneous intermediate SI on T1-weighted images. On T2-weighted images, especially with fat suppression, these lymph nodes have high SI. Necrotic areas or calcifications, as are sometime seen with colorectal cancer, lead to an inhomogeneous SI. 6.2.5.3.3  Retroperitoneal Tumors The group of primary retroperitoneal tumors comprises mesenchymal tumors, myxoid tumors, or non-differentiated sarcomas such as liposarcoma, fibrosarcoma, or hemangiopericytoma. Apart from these tumors extra-nodal lymphomas, neuroectodermal tumors such as paragangliomas, extra-adrenal pheochromocytomas, and germ cell tumors such as teratomas can be found in a retroperitoneal space. Approximately 80–85% of these tumors are malignant with liposarcomas being the most common tumors. Primary retroperitoneal tumors often have a large extent at the time of diagnosis. Most of the time they originate from the perirenal ventral part of spine or the psoas muscle and grow in a ventral direction infiltrating the surrounding tissues. Less commonly, they originate from the paravertebral or dorsal pararenal space. They can grow to a relatively large size without infiltration of adjacent structures, thus potentially mimicking a benign lesion. Lipomas are homogeneously hyperintense on T1weighted images and slightly to strongly hyperintense on T2-weighted images, depending on the sequence technique used. Liposarcomas have different signal characteristics based on the content of lipid tissue and the tumor grading. Highly differentiated liposarcomas tend to mimic the SI characteristics of benign lipomas. An important diagnostic hint is the presence of nodular enhancing structures, which should direct the reader’s attention to the potential presence of a malignant liposarcoma. With increasing de-differentiation, liposarcomas reveal decreasing SI on T1-weighted images. Conversely, their SI becomes more and more hyperintense on T2weighted images, reflecting the inhomogeneous structure of these tissues including necrotic areas. The SI of teratomas is heavily dependent on the tissue they express. 6.2.5.3.4  Retroperitoneal Fibrosis Retroperitoneal fibrosis (Ormond’s disease) is a rare disease. Men are affected twice as often as women are, with a median age between 40 and 60 years at the time of diagnosis. 70% of the cases are considered idiopathic,

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.53  Retroperitoneal liposarcoma. Arterial-phase T1weighted axial image of a 45-year-old patient with reoccurring liposarcoma of the retroperitoneum infiltrating the right kidney (left image). The tumor shows inhomogeneous, peripheral enhancement in the arterial phase. In the late phase (middle im-

age), weak enhancement can be seen centrally in the tumor. The infiltrated kidney shows no contrast excretion in contrast to the contralateral kidney. The coronal T1-weighted post-contrast image (right side) demonstrates the entire extent of the tumor

and 10% are considered drug-induced. The other cases are currently seen to be associated with the presence of lymphomas, metastases, aortic aneurysms, retroperitoneal bleeding as well as with status post radiation. Fifty to 75% of all cases reveal bilateral affection. The SI of retroperitoneal fibrosis is in between that of fat tissue and muscle, the contrast enhancement is only moderate. To differentiate benign retroperitoneal fibrosis from malignant retroperitoneal fibrosis T2-weighted images with fat suppression are warranted. Malignant cases of retroperitoneal fibrosis, which can show a transition into lymphoma, are mainly characterized by hyperintense signal on T2-weighted images. A dilated ureter may be unilateral or bilateral and can be seen as an indirect suggestion of retroperitoneal fibrosis with encasement of the ureters. The large vessels (aorta inferior, vena cava) are often encased by retroperitoneal fibrosis as well; however, they are not occluded by this disease. The contour of retro-

peritoneal fibrosis is not lobular, and this distinction can help in distinguishing this disease from adenopathy. Malignant retroperitoneal fibrosis in addition reveals a slight enhancement post-gadolinium. As the differentiation between benign and malignant retroperitoneal fibrosis cannot be achieved with a high level of diagnostic confidence by imaging, clinical markers such as the age of the patient should be taken into account. Benign retroperitoneal fibrosis is often observed in younger patients of up to 40 years of age, with no preexisting malignant disease while malignant retroperitoneal fibrosis is a disease of the elderly of more than 40 years, who often have a history of malignant disease. Imaging criteria for differentiation of different forms of RF and lymphoma are shown in Table 6.2.15. Please note that these criteria are only weak and do not allow a definite differentiation (Table 6.2.16) among these entities.

Table 6.2.15  Diagnostic criteria for differential diagnosis of retroperitoneal masses Benign mass

Malignant mass

Lymphoma

Age

<40

>40

Variable

Signal on T2

Intermediate

High

High

Contour

Round

irregular

Lobular

Enhancement post-Gd

Variable

Strong

Moderate

Ureter involvement

Bilateral encasement with retraction towards midline

Unilateral involvement with obstruction

Unilateral involvement with obstruction

Vena cava

Anterior displacement by mass effect

Anterior displacement by mass effect

Encasement

955

956

6  Abdomen and Retroperitoneum Table 6.2.16  Differential diagnosis of perirenal masses Perirenal lymphoma Malignant melanoma Perirenal Ormond’s disease Hematoma (biopsy) Xantogranulomatous pyelonephritis Retroperitoneal sarcoma

Fig. 6.2.54a–c  Retroperitoneal fibrosis 1. Coronal (a upper row, left) and axial (b lower row, left) steady-state free precession images of a patient suffering from retroperitoneal fibrosis. A perirenal mass can be appreciated on both images (b arrows). As this proliferating tissue also included the proximal parts of the ureter

there was bilateral low-grade hydronephrosis, which can be best seen on the nephrographic T1-weighted post-contrast images (c upper row, right). The perirenal tissue has a typical brush-like appearance and moderately hyperintense signal on T2-weighted images (d lower row, right side)

6.2  Kidneys, Adrenals, and Retroperitoneum

Fig. 6.2.55a–e  Retroperitoneal fibrosis 2. Axial T2-weighted fat-saturated (a upper row, left), T2/T1-weighted (b upper row, right), T1-weighted pre-contrast (c lower row, left), T1-weighted fat-saturated post-contrast (d lower row, right) and T1-weighted fat-saturated coronal MR-urography image demonstrating a

periaortic tissue proliferation consistent with retroperitoneal fibrosis. The mass has a slightly hyperintense signal in T2, is hypointense in T1 and shows strong enhancement after contrast administration. Due to the involvement of the left ureter, a third-degree urinary obstruction is present on the left side

Fig. 6.2.56  Dynamic urography with retroperitoneal fibrosis. Dynamic urography of a patient suffering from retroperitoneal fibrosis with bilateral affection of the ureters. On these coronal T1-

weighted MIP images from a dynamic urography study, a dilatation of the renal pelvis can be seen bilaterally with delayed filling of the ureters. The ureters themselves appear only slightly dilated

957

958

6  Abdomen and Retroperitoneum

6.2.5.4  Value of MRI and Comparison with Other Imaging Modalities 6.2.5.4.1  Conventional X-Ray For the assessment of retroperitoneal lymph nodes or malignant diseases, conventional X-ray is obsolete at this time. Conventional X-ray may demonstrate calcification in projections to the retroperitoneum, which need further work-up with either ultrasound, CT or MRI. 6.2.5.4.2  Ultrasound Depending on the body habitus, retroperitoneal lymph node metastases and tumors can be seen with ultrasound. Due to the often-impaired ultrasound conditions, a sufficient assessment of the entire retroperitoneum cannot always be achieved. 6.2.5.4.3  Computed Tomography Contrast-enhanced CT is the method of choice for the detection of retroperitoneal lymph nodes metastases and retroperitoneal tumors. CT is also a suitable modality for the detection of retroperitoneal fibrosis. The cause of the retroperitoneal fibrosis (e.g., aortic aneurysm) and its consequences (e.g., urinary obstruction) can be imaged at the same time. For pathologic lymph node metastases in Hodgkin’s disease and non-Hodgkin’s disease CT demonstrates diagnostic accuracy of 87 and 82%, respectively. On the other hand, differentiated liposarcomas are difficult to see on CT due to the fat equivalent Hounsfield units. As noted earlier, CT may produce false negative results when pathologic but non-enlarged lymph nodes are present. Enlarged lymph nodes as seen with infection or status post-infection cannot be differentiated from pathologic lymph nodes either. Typically, hypodense lymph nodes are encountered in patients with lymphangioleiomyomatosis and Whipple’s disease. Calcified lymph nodes are typically seen with tuberculosis and metastatic ovarian cancer. Evaluation of cachectic patients and patients with poor contrast of the bowel loops by CT is challenging. 6.2.5.4.4  MRI MRI has sensitivity and specificity similar to those of CT for the detection of retroperitoneal lymphomas and lymph nodes. As with CT, with MRI, it is not possible to differentiate between pathologic but non-enlarged lymph nodes and normal lymph nodes, or between reactively enlarged lymph nodes and pathologic lymph nodes. However, USPIO contrast agents that have just been ap-

proved in one European country may result in higher diagnostic accuracy. MRI and CT perform equally well in depicting the extent of retroperitoneal fibrosis. MRI has the advantage of yielding additional information on the SI, which may provide some suggestion of whether the disease is malignant or benign in character. Due to the non-invasiveness of MRI, dynamic MR urography studies can be performed. These studies allow depiction of the filling of the renal pelvis, the ureter, and the bladder. Of course, MRI is more expensive and time-consuming than CT. Even though motion artifacts are not as pronounced in the retroperitoneum as in the abdomen, they can occur and decrease image quality significantly. 6.2.5.5  Diagnostic Procedure In case of clinical suspicion for a primary retroperitoneal tumor, ultrasound and mainly CT are the imaging modalities of choice. MRI can be seen as a problem-solving modality when the origin of the disease and the exact extent of disease as well as the relation of the tumor mass to the other organs is of interest. In addition, dynamic imaging studies can only be obtained with MRI. MRI can be used instead of CT when there are CT contraindications such as allergy to iodine-based contrast agents or issues with the ionizing radiation, as in children. 6.2.5.6  Indications As laid out above MRI is mainly indicated as a problem-solving tool. MRI has the advantages of being more precise in determining the organ or site of origin of the tumor and more precise in defining the outer margins of the tumor, particularly in relation to the adjacent organs. For some diseases such as retroperitoneal fibrosis, MRI may guide the biopsy towards those areas that show a higher SI on T2-weighted images as a marker for active disease or malignant transformation. 6.2.6  Psoas Muscle 6.2.6.1  Dedicated Examination Techniques Examination of the psoas muscle is usually included in the examination of the kidneys or the retroperitoneum. Therefore, the same sequences are used. The sequences should be acquired in transverse orientation with 4–6 mm slice thickness. Axial images will allow clear delineation of the psoas muscle and analysis of its borders in relationship to the neural foramina. To visualize the entire extent of a disease process often coronal images are helpful. To detect edema in the muscle as in the case of inflammatory disease, fat-saturated mildly T2-weighted

6.2  Kidneys, Adrenals, and Retroperitoneum

images such as STIR images are required. Tumors and inflammatory processes such as abscesses have to be examined with T1-weighted sequences post intravenous contrast agent administration. For these clinical questions, T1-weighted sequences post-contrast with fat saturation are warranted. 6.2.6.2  Normal Anatomy The psoas muscle has its origin at the 12th thoracic vertebral body or the 1st lumbar vertebral body and ends at the minor trochanter of the femur. At the level of L5–S2, the psoas muscle has its largest diameter. At this level, the psoas muscle and iliac muscle unite to form the iliopsoas muscle. The muscle is capsulated by a fascia forming a common compartment with the intraspinal space, the pelvis, and the upper thigh. Independent of the pulse sequence a healthy psoas muscle has a relatively low SI. Due to the high contrast between the low SI psoas muscle and surrounding fat tissue, it is easy to delineate the psoas muscle. 6.2.6.3  Pathology The psoas compartment is mainly affected by disease processes that originate elsewhere. The muscle can be shifted by processes such as paraspinal abscesses or it can be directly infiltrated in the case of lymphoma or in the case of metastases, which often originate from testicular cancer. Lymphoma infiltration and metastases to the psoas muscle have significantly higher SI on T2-weighted images than non-affected psoas muscle. Similar SI can be found in abscesses of the psoas, which, in addition, show a slightly hyperintense signal surrounding the abscess. Air inclusion or calcifications cannot be seen on MRI or can only be seen as signal void. The SI of retroperitoneal bleeding including the psoas muscle depends on the age of the bleeding. Anticoagulation therapy and hemophilia are the two main predisposing conditions for retroperitoneal bleeding. Hematomas more than 3 weeks of age often show a characteristic configuration with low SI in the periphery. In the center of the hematoma, T1- and T2weighted images reveal higher SI, which is thought to be due to paramagnetic hemoglobin split products. 6.2.6.4  Value of MRI in Comparison with Other Imaging Modalities At this time, CT is considered the method of choice for examination of pathologic changes of the psoas muscle. Particularly abscesses, infections and metastases as well as acute bleeding as in the case of aortic aneurysms can be detected with high sensitivity. Due to the higher con-

trast resolution and the better anatomical depiction of the psoas muscle and paraspinal tissue, MRI is increasingly often applied as a diagnostic modality for the depiction of the psoas muscle. MRI is of particular value when disease of the psoas muscle with involvement of the spinal canal is suspected and when only subtle changes are expected. This is often the case in patients with suspected spondylodiskitis, where the infection originates from the vertebral discs and can spread into the psoas muscle as well as into the spinal canal. 6.2.6.5  Diagnostic Procedure In case of pathologic changes of the psoas muscle CT seems to be warranted first as a screening modality. As discussed above, if the disease is thought to arise from the spine, MRI seems to be a better imaging modality. Suggested Reading 1.

Agrons GA, Lonergan GJ, Dickey GE, Perez-Monte JE (1999) Adrenocortical neoplasms in children: radiologicpathologic correlation. Radiographics 19:989–1008 2. Ahmadi T, Itai Y, Takahashi M, Onaya H, Kobayashi T, Tanaka YO, Matsuzaki Y, Tanaka N, Okada Y (1998) Angiomyolipoma of the liver: significance of CT and MR dynamic study. Abdom Imaging 23:520–526 3. Arslan A, Ciftci E, Cetin A, Selcuk H, Demirci A (1998) Tuberous sclerosis: ultrasound, CT and MRI features of two cases with multiple organ involvement. Australas Radiol 42:379–382 4. Baumgartner BR, Chezmar JL (1989) Magnetic resonance imaging of the kidneys and adrenal glands. Semin Ultrasound CT MR 10:43–62 5. Birnbaum BA, Weinreb JC, Megibow AJ, Sanger JJ, Lubat E, Kanamuller H, Noz ME, Bosniak MA (1990) Definitive diagnosis of hepatic hemangiomas: MR imaging versus Tc99m-labeled red blood cell SPECT. Radiology 176:95–101 6. Blake MA, Kalra MK, Maher MM, Sahani DV, Sweeney AT, Mueller PR, Hahn PF, Boland GW (2004) Pheochromocytoma: an imaging chameleon. Radiographics 24(Suppl 1): S87–S99 7. Bosniak MA (1993) Problems in the radiologic diagnosis of renal parenchymal tumors. Urol Clin North Am 20:217–230 8. Bosniak MA, Rofsky NM (1996) Problems in the detection and characterization of small renal masses. Radiology 200:286–287 9. Brown ED, Semelka RC (1995) Magnetic resonance imaging of the adrenal gland and kidney. Top Magn Reson Imaging 7:90–101 10. Chung JJ, Semelka RC, Martin DR (2001) Acute renal failure: common occurrence of preservation of corticomedullary differentiation on MR images. Magn Reson Imaging 19:789–793

959

960

6  Abdomen and Retroperitoneum 11. Cowie A (2002) Renal imaging in patients requiring renal replacement therapy. Semin Dial 15:237–249 12. Dann P, Thakur R, Chin D, Krinsky G, Israel GM (2006) Are T2-weighted images necessary in renal mass characterization? Eur J Radiol 59:112–116 13. Dunnick NR, Korobkin M, Francis I (1996) Adrenal radiology: distinguishing benign from malignant adrenal masses. AJR Am J Roentgenol 167:861–867 14. Eilber FC, Eilber KS, Eilber FR (2000) Retroperitoneal sarcomas. Curr Treat Options Oncol 1:274–278 15. Elsayes KM, Mukundan G, Narra VR, Lewis JS Jr, Shirkhoda A, Farooki A, Brown JJ (2004) Adrenal masses: mr imaging features with pathologic correlation. Radiographics 24(Suppl 1):S73–S86 16. Forman HP, Middleton WD, Melson GL, McClennan BL (1993) Hyperechoic renal cell carcinomas: increase in detection at US. Radiology 188:431–434 17. Grenier N, Hauger O, Delmas Y, Combe C (2006) MR imaging of nephropathies. Abdom Imaging 31:213–223 18. Grenier N, Pedersen M, Hauger O (2006) Contrast agents for functional and cellular MRI of the kidney. Eur J Radiol 60:341–352 19. Hallscheidt P, Weber MA, Schenk JP, Riedasch G (2002) [Magnetic resonance tomography of xanthogranulomatous pyelonephritis. Epidemiology, pathogenesis and symptoms]. Urologe A 41:577–582 20. Hallscheidt PJ, Thorn M, Radeleff BA, Noeldge G, Shab A, Meinzer P, Boese JM (2003) Comparison of spatial resolution in high-resolution multislice computed tomography and digital subtraction angiography using renal specimens. J Comput Assist Tomogr 27:864–868 21. Hallscheidt PJ, Bock M, Riedasch G, Zuna I, Schönberg SO, Autschbach F, Soder M, Noeldge G (2004) 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 28:333–339 22. Hallscheidt PJ, Fink C, Haferkamp A, Bock M, Luburic A, Zuna I, Noeldge G, Kauffmann G (2005) 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 29:64–68 23. Hallscheidt P, Wagener N, Gholipour F, Aghabozorgi N, Dreyhaupt J, Hohenfellner M, Haferkamp A, Pfitzen­maier J (2006) Multislice computed tomography in planning nephron-sparing surgery in a prospective study with 76 patients: comparison of radiological and histopathological findings in the infiltration of renal structures. J Comput Assist Tomogr 30:869–874 24. Hauger O, Delalande C, Deminiere C, Fouqueray B, Ohayon C, Garcia S, Trillaud H, Combe C, Grenier N (2000) Nephrotoxic nephritis and obstructive nephropathy: evaluation with MR imaging enhanced with ultrasmall superparamagnetic iron oxide-preliminary findings in a rat model. Radiology 217:819–826

25. Hecht EM, Israel GM, Krinsky GA, Hahn WY, Kim DC, Belitskaya-Levy I, Lee VS (2004) Renal masses: quantitative analysis of enhancement with signal intensity measurements versus qualitative analysis of enhancement with image subtraction for diagnosing malignancy at MR imaging. Radiology 232:373–378 26. Heesakkers RA, Futterer JJ, Hovels AM, van den Bosch HC, Scheenen TW, Hoogeveen YL, Barentsz JO (2006) Prostate cancer evaluated with ferumoxtran-10-enhanced T2*-weighted MR Imaging at 1.5 and 3.0 T: early experience. Radiology 239:481–487 27. Heidenreich A, Ravery V (2004) Preoperative imaging in renal cell cancer. World J Urol 22:307–315 28. Heiss SG, Shifrin RY, Sommer FG (2000) Contrast-enhanced three-dimensional fast spoiled gradient-echo renal MR imaging: evaluation of vascular and nonvascular disease. Radiographics 20:1341–1352; discussion 1353–1344 29. Helenon O, Andre M, Correas JM, Khairoune A, Merran S, Balleyguier C (2002) [Characterization of renal masses]. J Radiol 83:787–804 30. Hes FJ, Feldberg MA (1999) Von Hippel-Lindau disease: strategies in early detection (renal-, adrenal-, pancreatic masses). Eur Radiol 9:598–610 31. Hoffer FA (2005) Magnetic resonance imaging of abdominal masses in the pediatric patient. Semin Ultrasound CT MR 26:212–223 32. Hricak H, Terrier F, Demas BE (1986) Renal allografts: evaluation by MR imaging. Radiology 159:435–441 33. Huang AJ, Lee VS, Rusinek H (2003) MR imaging of renal function. Radiol Clin North Am 41:1001–1017 34. Israel GM, Bosniak MA (2003) Renal imaging for diagnosis and staging of renal cell carcinoma. Urol Clin North Am 30:499–514 35. Israel GM, Hecht E, Bosniak MA (2006) CT and MR imaging of complications of partial nephrectomy. Radiographics 26:1419–1429 36. Kawashima A, Sandler CM, Corl FM, West OC, Tamm EP, Fishman EK, Goldman SM (2001) Imaging of renal trauma: a comprehensive review. Radiographics 21:557–574 37. Kettritz U, Semelka RC, Brown ED, Sharp TJ, Lawing WL, Colindres RE (1996) MR findings in diffuse renal parenchymal disease. J Magn Reson Imaging 6:136–144 38. Kim J (2006) Imaging findings of renal cell carcinoma. Expert Rev Anticancer Ther 6:895–904 39. Kim KA, Choi JW, Park CM, Lee CH, Lee JH, Yoon DK, Seol HY (2006) Unusual renal cell carcinomas: a pictorial essay. Abdom Imaging 31:154–163 40. Kondo T, Nakazawa H, Sakai F, Kuwata T, Onitsuka S, Hashimoto Y, Toma H (2004) Spoke-wheel-like enhancement as an important imaging finding of chromophobe cell renal carcinoma: a retrospective analysis on computed tomography and magnetic resonance imaging studies. Int J Urol 11:817–824 41. Korobkin M, Francis IR (1995) Adrenal imaging. Semin Ultrasound CT MR 16:317–330

6.2  Kidneys, Adrenals, and Retroperitoneum 42. Korobkin M, Giordano TJ, Brodeur FJ, Francis IR, Siegelman ES, Quint LE, Dunnick NR, Heiken JP, Wang HH (1996) Adrenal adenomas: relationship between histologic lipid and CT and MR findings. Radiology 200:743–747 43. Kreft B, Schild HH (2003) [Cystic renal lesions]. Rofo 175:892–903 44. Lang EK, Macchia RJ, Gayle B, Richter F, Watson RA, Thomas R, Myers L (2002) CT-guided biopsy of indeterminate renal cystic masses (Bosniak 3 and 2F): accuracy and impact on clinical management. Eur Radiol 12:2518–2524 45. Lawrentschuk N, Gani J, Riordan R, Esler S, Bolton DM (2005) Multidetector computed tomography vs magnetic resonance imaging for defining the upper limit of tumour thrombus in renal cell carcinoma: a study and review. BJU Int 96:291–295 46. Levy P, Helenon O, Merran S, Paraf F, Mejean A, Cornud F, Moreau JF (1999) [Cystic tumors of the kidney in adults: radio-histopathologic correlations]. J Radiol 80:121–133 47. Lockhart ME, Smith JK, Kenney PJ (2002) Imaging of adrenal masses. Eur J Radiol 41:95–112 48. Lowe LH, Isuani BH, Heller RM, Stein SM, Johnson JE, Navarro OM, Hernanz-Schulman M (2000) Pediatric renal masses: Wilms’ tumor and beyond. Radiographics 20:1585–1603 49. Matsuura H, Hayashi N, Arima K, Yanagawa M, Kawa­ mura J, Takeda H (1997) [Evaluation of renal cystic mass on ultrasonogram and computed tomogram: usefulness of magnetic resonance imaging and renal angiography—category III by Bosniak: report of 5 cases]. Nippon Hinyokika Gakkai Zasshi 88:826–829 50. Mayo-Smith WW, Boland GW, Noto RB, Lee MJ (2001) State-of-the-art adrenal imaging. Radiographics 21:995–1012 51. Mazumdar A, Siegel MJ, Narra V, Luchtman-Jones L (2002) Whole-body fast inversion recovery MR imaging of small cell neoplasms in pediatric patients: a pilot study. AJR Am J Roentgenol 179:1261–1266 52. Michaely HJ, Herrmann KA, Nael K, Oesingmann N, Reiser MF, Schönberg SO (2006) Functional renal imaging: nonvascular renal disease. Abdom Imaging epub:Jan 30 53. Michaely HJ, Sourbron S, Dietrich O, Attenberger U, Reiser MF, Schönberg SO (2006) Functional renal MR imaging: an overview. Abdom Imaging 54. Nelson CP, Sanda MG (2002) Contemporary diagnosis and management of renal angiomyolipoma. J Urol 168:1315–1325 55. Nishino M, Hayakawa K, Minami M, Yamamoto A, Ueda H, Takasu K (2003) Primary retroperitoneal neoplasms: CT and MR imaging findings with anatomic and pathologic diagnostic clues. Radiographics 23:45–57 56. Outwater EK, Blasbalg R, Siegelman ES, Vala M (1998) Detection of lipid in abdominal tissues with opposed-phase gradient-echo images at 1.5 T: techniques and diagnostic importance. Radiographics 18:1465–1480

57. Pedersen M, Dissing TH, Morkenborg J, Stodkilde-Jorgensen H, Hansen LH, Pedersen LB, Grenier N, Frokiaer J (2005) Validation of quantitative BOLD MRI measurements in kidney: application to unilateral ureteral obstruction. In: Kidney Int. pp 2305–2312 58. Pedrosa I, Naidich JJ, Rofsky NM, Bosniak MA (2001) Renal pseudotumors due to fat necrosis in acute pancreatitis. J Comput Assist Tomogr 25:236–238 59. Pereira JM, Sirlin CB, Pinto PS, Casola G (2005) CT and MR imaging of extrahepatic fatty masses of the abdomen and pelvis: techniques, diagnosis, differential diagnosis, and pitfalls. Radiographics 25:69–85 60. Prando A, Prando D, Prando P (2006) Renal cell carcinoma: unusual imaging manifestations. Radiographics 26:233–244 61. Prasad PV (2006) Evaluation of intra-renal oxygenation by BOLD MRI. Nephron Clin Pract 103:c58–65 62. Prasad SR, Wang H, Rosas H, Menias CO, Narra VR, Middleton WD, Heiken JP (2005) Fat-containing lesions of the liver: radiologic-pathologic correlation. Radiographics 25:321–331 63. Pretorius ES, Siegelman ES, Ramchandani P, Cangiano T, Banner MP (1999) Renal neoplasms amenable to partial nephrectomy: MR imaging. Radiology 212:28–34 64. Pretorius ES, Wickstrom ML, Siegelman ES (2000) MR imaging of renal neoplasms. Magn Reson Imaging Clin N Am 8:813–836 65. Pusl T, Weiss M, Hartmann B, Wendler T, Parhofer K, Michaely H (2006) Malacoplakia in a renal transplant recipient. Eur J Intern Med 17:133–135 66. Rafal RB, Kosovsky PA, Markisz JA (1991) Xanthogranulomatous pyelonephritis in an infant. Urology 37:553–556 67. Rha SE, Byun JY, Jung SE, Oh SN, Choi YJ, Lee A, Lee JM (2004) The renal sinus: pathologic spectrum and multimodality imaging approach. Radiographics 24 Suppl 1: S117–131 68. Rockall AG, Babar SA, Sohaib SA, Isidori AM, Diaz-Cano S, Monson JP, Grossman AB, Reznek RH (2004) CT and MR imaging of the adrenal glands in ACTH-independent cushing syndrome. Radiographics 24:435–452 69. Rofsky NM, Bosniak MA (1997) MR imaging in the evaluation of small (< or = 3.0 cm) renal masses. Magn Reson Imaging Clin N Am 5:67–81 70. Rofsky NM, Bosniak MA, Weinreb JC, Coppa GF (1989) Giant renal cell carcinoma: CT and MR characteristics. J Comput Assist Tomogr 13:1078–1080 71. Rofsky NM, Weinreb JC, Bosniak MA, Libes RB, Birnbaum BA (1991) Renal lesion characterization with gadoliniumenhanced MR imaging: efficacy and safety in patients with renal insufficiency. Radiology 180:85–89 72. Rusinek H, Kaur M, Lee VS (2004) Renal magnetic resonance imaging. Curr Opin Nephrol Hypertens 13:667–673 73. Sadowski EA, Fain SB, Alford SK, Korosec FR, Fine J, Muehrer R, Djamali A, Hofmann RM, Becker BN, Grist TM (2005) Assessment of Acute Renal Transplant Rejection with Blood Oxygen Level-Dependent MR Imaging: Initial Experience. Radiology 236:911–919

961

962

6  Abdomen and Retroperitoneum 74. Semelka RC, Corrigan K, Ascher SM, Brown JJ, Colindres RE (1994) Renal corticomedullary differentiation: observation in patients with differing serum creatinine levels. Radiology 190:149–152 75. Sheth S, Ali S, Fishman E (2006) Imaging of renal lymphoma: patterns of disease with pathologic correlation. Radiographics 26:1151–1168 76. Szolar DH, Preidler K, Ebner F, Kammerhuber F, Horn S, Ratschek M, Ranner G, Petritsch P, Horina JH (1997) Functional magnetic resonance imaging of human renal allografts during the post-transplant period: preliminary observations. Magn Reson Imaging 15:727–735 77. Szolar DH, Korobkin M, Reittner P, Berghold A, Bauern­ hofer T, Trummer H, Schoellnast H, Preidler KW, Samonigg H (2005) Adrenocortical carcinomas and adrenal pheochromocytomas: mass and enhancement loss evaluation at delayed contrast-enhanced CT. Radiology 234:479–485

78. Tanaka K, Kumano Y, Kanomata N, Takeda M, Hara I, Fujisawa M, Kawabata G, Kamidono S (2004) Oncocytic adrenocortical carcinoma. Urology 64:376–377 79. Triantopoulou C, Rizos S, Bourli A, Koulentianos E, Dervenis C (2002) Localized unilateral perirenal fibrosis: CT and MRI appearances. Eur Radiol 12:2743–2746 80. Velchik MG (1985) Radionuclide imaging of the urinary tract. Urol Clin North Am 12:603–631 81. Warren KS, McFarlane J (2005) The Bosniak classification of renal cystic masses. BJU Int 95:939–942 82. Zhang J, Israel GM, Krinsky GA, Lee VS (2004) Masses and pseudomasses of the kidney: imaging spectrum on MR. J Comput Assist Tomogr 28:588–595

Chapter 7

Pelvis

7

7.2.1.4  Pathological Findings  .. . . . . . . . . . . . . . . .   1003

7.1 

Female Pelvis  .. . . . . . . . . . . . . . . . . . . . . . .   964 E. Sala and H. Hricak

7.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . .   964

7.1.2 

Magnetic Resonance Imaging Technique    964

7.2.1.6  Indications (Symptom-Specific   Imaging Modalities)  .. . . . . . . . . . . . . . . . .   1014

7.1.2.1  Patient Positioning and Preparation  . . .   964

7.2.1.7  Diagnostic Procedures  .. . . . . . . . . . . . . . .   1015

7.1.2.2  Selection of Coils  . . . . . . . . . . . . . . . . . . . .   964

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1016

7.1.2.3  Examination Sequences   and Imaging Planes  . . . . . . . . . . . . . . . . . .   964 7.1.3 

Normal Anatomy  . . . . . . . . . . . . . . . . . . . .   966

7.1.4 

Uterus and Vagina  . . . . . . . . . . . . . . . . . . .   967

7.1.4.1  Congenital Anomalies   . . . . . . . . . . . . . . .   967 7.1.4.2  Benign Uterine Conditions  . . . . . . . . . . .   970 7.1.4.3  Malignant Conditions   of the Uterine Corpus  . . . . . . . . . . . . . . . .   973 7.1.4.4  Malignant Conditions of the Cervix  . . .   980 7.1.4.5  Malignant Conditions   of Vagina and Vulva  .. . . . . . . . . . . . . . . . .   985 7.1.5 

7.2.2 

7.1.5.2  Benign Tumors   .. . . . . . . . . . . . . . . . . . . . .   989 7.1.5.3  Malignant Ovarian Tumors   .. . . . . . . . . .   993 7.1.5.4  Ovarian Metastasis   . . . . . . . . . . . . . . . . . .   996 Miscellaneous Conditions  . . . . . . . . . . . .   996

7.1.6.1  Chronic Pelvic Pain  . . . . . . . . . . . . . . . . . .   996 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   997 7.2 

Male Pelvis  .. . . . . . . . . . . . . . . . . . . . . . . . .   999

7.2.1 

Urinary Bladder  . . . . . . . . . . . . . . . . . . . . .   999 U.G. Müller-Lisse and U.L. Müller-Lisse

7.2.1.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . .   999 7.2.1.2  Examination Techniques  .. . . . . . . . . . . . .   999 7.2.1.3  Normal Anatomy  . . . . . . . . . . . . . . . . . . . .   1002

Male Pelvis: Prostate  .. . . . . . . . . . . . . . . . .   1018 U.G. Müller-Lisse, M.K. Scherr, and U.L. Müller-Lisse

7.2.2.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . .   1018 7.2.2.2  Examination Techniques  .. . . . . . . . . . . . .   1018 7.2.2.3

Normal Anatomy  . . . . . . . . . . . . . . . . . . . .   1028

7.2.2.4  Pathological Findings  .. . . . . . . . . . . . . . . .   1029 7.2.2.5  Indications and Value of MRI  . . . . . . . . .   1034 7.2.2.6  Indications (Symptom-Specific   Imaging Modalities)  .. . . . . . . . . . . . . . . . .   1034 7.2.2.7  Diagnostic Procedures  .. . . . . . . . . . . . . . .   1037

Adnexa  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   986

7.1.5.1  Role of MRI in Characterization   of Adnexal Lesions  .. . . . . . . . . . . . . . . . . .   986

7.1.6 

7.2.1.5  Indications and Value of MRI  . . . . . . . . .   1013

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1037 7.2.3 

Male Pelvis: Scrotum  . . . . . . . . . . . . . . . . .   1039 U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse

7.2.3.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . .   1039 7.2.3.2  Examination Techniques  .. . . . . . . . . . . . .   1039 7.2.3.3  Normal Anatomy  . . . . . . . . . . . . . . . . . . . .   1041 7.2.3.4  Pathological Findings  .. . . . . . . . . . . . . . . .   1043 7.2.3.5  Indications and Value of MRI  . . . . . . . . .   1052 7.2.3.6  Diagnostic Procedures  .. . . . . . . . . . . . . . .   1053 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1054 7.2.4 

Male Pelvis: Penis  . . . . . . . . . . . . . . . . . . . .   1055 U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse

7.2.4.1  Introduction   . . . . . . . . . . . . . . . . . . . . . . . .   1055 7.2.4.2  Examination Techniques  .. . . . . . . . . . . . .   1055

964

7  Pelvis 7.2.4.3  Normal Anatomy  . . . . . . . . . . . . . . . . . . . .   1057

7.3.2 

MRI Anatomy  . . . . . . . . . . . . . . . . . . . . . . .   1069

7.2.4.5  Pathological Findings  .. . . . . . . . . . . . . . . .   1059

7.3.3 

Functional Anatomy   of the Pelvic Floor  .. . . . . . . . . . . . . . . . . . .   1069

7.3.4 

MRI Techniques   for Pelvic Floor Dysfunction  . . . . . . . . . .   1070

7.2.4.5  Indications and Value of MRI  . . . . . . . . .   1065 7.2.4.6  Indications (Symptom-Specific   Imaging Modalities)  .. . . . . . . . . . . . . . . . .   1066 7.2.4.7  Diagnostic Procedures  .. . . . . . . . . . . . . . .   1067 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1067 7.3 

7.3.1 

Pelvic Floor Assessment  by Magnetic Resonance Imaging  .. . . . .   1069 A. Maubon, C. Servin-Zardini, M. Pouquet, Y. Aubard, and J.P. Rouanet Introduction  .. . . . . . . . . . . . . . . . . . . . . . . .   1069

7.1  Female Pelvis E. Sala and H. Hricak 7.1.1  Introduction The role of magnetic resonance imaging (MRI) of the female pelvis has evolved during the last two to three decades. There is now a substantial body of evidence that MRI is useful in evaluating Müllerian duct anomalies and both benign and malignant conditions of the female pelvis. MRI has been shown to be superior to CT in the workup of uterine and cervical carcinoma and is a very useful problem-solving tool in the characterization of adnexal lesions. Although MRI is still relatively expensive, it has been shown to minimize costs in some clinical settings by limiting or eliminating the need for further expensive and/or more invasive diagnostic or surgical procedures. Advantages of MRI include superb spatial and tissue contrast resolution, no use of ionizing radiation, multiplanar capability, and fast techniques. 7.1.2  Magnetic Resonance Imaging Technique 7.1.2.1  Patient Positioning and Preparation Patient preparation and positioning are very important to obtain optimal results. Patients are usually instructed to fast for 4–6 hours before the MRI examination in order to limit artifacts due to small bowel peristalsis. An antiperistaltic agent (hyoscine butyl bromide or glucagon) may be administered before imaging as an alternative. Ideally, the patient is asked to empty the bladder prior to going on the MR scanner. A full bladder may degrade T2-weighted images due to ghosting and motion artifacts.

7.3.4.1  Preparation  .. . . . . . . . . . . . . . . . . . . . . . . . .   1070 7.3.4.2  Sequences  . . . . . . . . . . . . . . . . . . . . . . . . . . .   1071 7.3.4.3  Normal Results   .. . . . . . . . . . . . . . . . . . . . .   1072 7.3.4.4  Prolapses   . . . . . . . . . . . . . . . . . . . . . . . . . . .   1073 7.3.5 

Conclusion   . . . . . . . . . . . . . . . . . . . . . . . . .   1075 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1076

7.1.2.2  Selection of Coils Patients are imaged in the supine position using a pelvic or cardiac surface array coil. Although a body coil has been shown to provide phased similar staging accuracy, use of a phase-array coil increases resolution and decreases imaging time (Hawighorst et al. 1998; Yu et al. 1998). Endoluminal coils (endorectal and endovaginal) can provide high-resolution images of small tumors of the cervix or those with adjacent parametrial invasion. However, their field of view is limited in assessing large tumors and those with extrauterine extension to adjacent organs, the pelvic sidewall or distant pelvic lymph nodes. 7.1.2.3  Examination Sequences and Imaging Planes Basic imaging protocols for gynecologic MR imaging include axial T1-weighted and T2-weighted images, and sagittal T2-weighted images. Coronal T2-weighted images are very useful for the evaluation of the Müllerian duct anomalies and adnexal masses. Fat-saturation T1-weighted images are necessary to diagnose hemorrhage, especially when high signal intensity is seen within the adnexa on T1-weighted images and the diagnosis of endometriosis is suspected clinically or at ultrasound. T1-weighted axial images with a large field of view to evaluate the entire pelvis and upper abdomen for lymphadenopathy as well as bone marrow changes are essential in staging gynecologic malignancies. T1-weighted gradient echo sequences are faster than spin-echo sequences but produce images of lower quality (Table 7.1.1). 7.1.2.3.1  Endometrial Carcinoma In addition to the basic imaging protocol described, highresolution axial oblique T2-weighted fast spin-echo (FSE)

7.1  Female Pelvis Table 7.1.1  Main imaging planes and sequences in gynecology Condition

Plane

Sequence

Congenital anomalies

Sagittal Axial Axial Coronal oblique (parallel to long axis of the uterus)

T2-W FSE T1-W SE T2-W FSE T2-W FSE (with or without fat-sat)

Endometrial carcinoma

Sagittal Axial Axial oblique (parallel to short axis of the corpus) Axial Axial Sagittal and axial

T2-W FSE T2-W FSE T2-W FSE

Sagittal Axial Axial oblique (parallel to short axis of the cervix) Axial Axial

T2-W FSE T2-W FSE T2-W FSE

Sagittal Axial Coronal Axial Axial

T2-W FSE T2-W FSE T2-W FSE T1-W SE (pelvis) fat-sat T1-W SE (if high SI lesion on T1-W) T1-W SE (abdomen) T1-W SE T1-W SE before and after i.v. CM at 70 s and 5/10 min

Cervical carcinoma

Adnexal lesions

Axial Axial

T1-W SE (pelvis) T1-W SE (abdomen) T1-W GRE dynamic before and after i.v. CM at 0, 1, 2, 4 min

T1-W SE (pelvis) T1-W SE (abdomen)

T1-weighting can be conventional spin echo, but in the interest of time, most centers use gradient echo sequences, which are faster but yield images of lower quality FSE fast spin echo, SE spin echo, i.v. CM intravenous contrast medium administration, fat-sat fat saturation, T1-W T1 weighted T2-W T2 weighted

images taken parallel to the short axis of the uterine corpus are favored for the evaluation of primary tumor and myometrial invasion (Shibutani et al. 1999; Takahashi et al. 1998). Dynamic multiphase contrast-enhanced 3D T1-weighted sequences through the uterus in the sagittal and axial (parallel to the short axis of the uterine corpus) planes are routinely used to improve staging accuracy in endometrial cancer (Table 7.1.1). The early enhancement phase (0 and 1 minute post-injection) allows identification of the subendometrial zone, which enhances earlier than the bulk of the myometrium and corresponds to the inner junctional zone (JZ): This is especially important in detecting early myometrial invasion, as the JZ often becomes indistinct in postmenopausal women (Yamashita et al. 1993). However, the subendometrial zone is seen in less than 50% of cases. The equilibrium phase (2 min post-injection) allows better evaluation of deep myome-

trial invasion while the delayed phase (4–5 min) enables better evaluation of cervical stroma invasion. The tumor/ myometrium interface should be assessed in at least two planes. 7.1.2.3.2  Cervical Carcinoma Cervical tumors are best seen on T2-weighted images. The sagittal plane allows evaluation of tumor extension into the body of the uterus and vagina. The axial oblique T2-weighted FSE (parallel to the short axis of the cervix) (Table 7.1.1) is important in assessing parametrial invasion (Shiraiwa et al. 1999). True axial T2-weighted FSE with fat saturation can be helpful in the evaluation of parametrial invasion, especially in younger women, in whom there is a prominent peri-cervical/vaginal plexus

965

966

7  Pelvis

(Sala et al. 2007). It is also valuable in the detection of lymph nodes as well as soft tissue and bone edema. The use of contrast medium is optional in staging of cervical carcinoma, as it does not improve staging accuracy compared with T2-weighted images (Scheidler et al. 1998). However, the early phase of enhancement does significantly increase lesion conspicuity compared to T2-weighted images. Dynamic contrast-enhanced MRI (DCE MRI) is a helpful non-invasive technique for assessing tumor response to treatment (Zahra et al. 2007). Early increase in tumor perfusion has been shown to correlate with response to chemo-radiotherapy in carcinoma of the cervix (Mayr et al. 2000). Recently, ultrasmall particles of iron oxide (USPIO) have been shown to improve the detection of lymph node metastases independent of node size in patients with endometrial and cervical cancer (Rockall et al. 2005). 7.1.2.3.3  Characterization of Adnexal Masses Coronal T2-weighted images may be useful in establishing the origin of the mass. Fat-saturated T1-weighted images should be added when a high-signal-intensity lesion is seen within the adnexa on T1-weighted images. Contrastenhanced T1-weighted images are useful for the evaluation of cystic lesions, as they may help differentiate solid components or papillary projections from clots and debris (Table 7.1.1). Contrast studies are essential and fat-saturation techniques should be applied to both T1-weighted and contrast-enhanced T1-weighted sequences in ovarian cancer staging. Subtraction views are particularly useful for detection of areas of enhancement within cystic lesions, which are hyperintense on pre-gadolinium T1-weighted images. Dynamic imaging is not necessary. 7.1.3  Normal Anatomy Pelvic anatomy is exquisitely demonstrated by MRI (Fig. 7.1.1). On T1-weighted sequences, the normal pelvic musculature and viscera demonstrate homogeneous low to

7 Fig. 7.1.1  Zonal anatomy of the uterus. Coronal (a) and sagittal (b) T2-weighted FSE images of the uterus in a premenopausal woman. The central, high-signal-intensity stripe represents the endometrium; the band of low signal intensity adjacent to the endometrial stripe represents the inner myometrium or junctional zone (arrows). The outer layer of the myometrium (M) is of intermediate signal intensity (O ovary, C cervix, * bladder). Sagittal T2-weighted FSE image of the uterus in a postmenopausal woman (c). Note the loss of definition of the zonal anatomy in the postmenopausal woman compared to the premenopausal woman (C cervix, * bladder) and the presence of nabothian cysts (arrowhead)

7.1  Female Pelvis

intermediate signal intensity. However, it is the contrast resolution of T2-weighted imaging that is the basis for the superb tissue characterization of MRI. On T2-weighted sequences the uterus, cervix, and vagina exhibit distinct layers of different signal intensity—the so-called zonal architecture. The endometrium yields high signal intensity on T2-weighted images. In comparison, the signal intensity of the outer myometrium is intermediate, higher than that of striated muscle. Interposed between these two layers is a narrow band of lower signal intensity, the JZ, which corresponds to the innermost myometrium. Its signal properties reflect its lower water content compared with the remainder of the myometrium, which may be a function of its lower extracellular matrix/unit volume (Brown et al. 1991). The width of the endometrium (both leaflets) varies with the menstrual cycle and, on the sagittal plane of section, measures up to 3 mm in the proliferative phase and up to 7 mm in the secretory phase. In postmenopausal women not receiving exogenous hormones, uterine zonal anatomy is often indistinct and the endometrium measures less than 3 mm. On T2-weighted images, the normal cervix demonstrates a central area of high signal intensity (endocervical glands and mucus) surrounded by low-signal-intensity stroma (elastic fibrous tissue). Around the periphery of the cervix, smooth muscle predominates, resulting in a rim of intermediate signal intensity similar to that of myometrium. Occasionally, intermediate signal intensity cervical mucosal folds (plicae palmatae) can be seen interposed between the low-signal-intensity cervical stroma and the high-signal-intensity endocervical canal. Following the administration of intravenous paramagnetic gadolinium chelates, the zonal anatomy of the uterus is demonstrated on T1-weighted images. The endometrium and outer myometrium enhance to a greater extent than does the JZ. Similarly, the inner cervical mucosa and outer smooth muscle enhance more than the fibrocervical stroma. The parametrial tissues, vaginal walls, and submucosa also enhance after intravenous administration of contrast medium. The normal MRI appearance of the ovaries varies depending on the pulse sequence used. On T1-weighted images, the ovaries display homogeneous low to intermediate signal intensity, whereas on T2-weighted images the follicles become brighter than the surrounding stroma. The normal Fallopian tubes are not routinely seen because of their small size and tortuous course. 7.1.4  Uterus and Vagina 7.1.4.1  Congenital Anomalies Müllerian duct anomalies result from non-development or varying degrees of non-fusion or non-resorption of the Müllerian ducts. The clinical classification of Müllerian duct anomalies follows the guidelines proposed by

Buttram and Gibbons/American Fertility Society. Surgical management in patients with female anatomic anomalies is aimed at preventing endometriosis and preserving fertility. The role of imaging is to provide a detailed map of the pelvic anatomy, including the presence and extent of any anomalies (Fig. 7.1.2). MRI is the most accurate imaging modality allowing both precise classification and demonstration of associated complications (Carrington et al. 1990). 7.1.4.1.1  Müllerian Agenesis or Hypoplasia Uterine agenesis or hypoplasia result from non-development or rudimentary development of the Müllerian ducts. A subtype of uterine agenesis is Mayer-RokitanskyKuster-Hauser (MRKH) syndrome. In this syndrome, the presence of vaginal agenesis or hypoplasia with intact ovaries and Fallopian tubes is accompanied by variable anomalies of the uterus, urinary tract, and skeletal system. Uterine remnants may be present. The absence or anomalies of the uterus and upper vagina, with varying degrees of development of the lower vagina, can be detected reliably on a combination of sagittal and axial MR images (Fig. 7.1.3). Normal ovaries are usually present. Uterine hypoplasia is diagnosed when the uterus is small, the endometrium is atrophic, and on T2-weighted images the myometrium is of lower than normal signal intensity. 7.1.4.1.2  Unicornuate Uterus The unicornuate uterus results from non-development or rudimentary development of one Müllerian duct, with the other Müllerian duct fully developed. The role of MRI is to demonstrate the presence of a rudimentary horn and whether it contains endometrium. MRI is also very useful for determining the communication of the rudimentary horn with the main uterine cavity as patients with a rudimentary horn communicating with the uterine cavity might benefit from surgical removal of this rudimentary horn. On T2-weighted MR images, the unicornuate uterus appears as a curved, elongated uterus with tapering of the fundal segment off midline (the “banana-like” configuration). Normal uterine zonal anatomy is maintained. The rudimentary horn, when present, usually demonstrates lower signal intensity on T2-weighted images. 7.1.4.1.3  Uterus Didelphys The uterus didelphys results from non-fusion of the two Müllerian ducts. Two separate normal-sized uterine horns and cervices are demonstrated on T2-weighted MR images. A longitudinal vaginal septum is present in 75% of cases, occasionally complicated by transverse septa

967

968

7  Pelvis

Fig. 7.1.2a–e  Schematic drawing of Müllerian duct anomalies. a Unicornuate uterus. The uterus is “banana shaped.” b Uterus didelphys. Two separate uterine horns, two cervices, and a vaginal septum are present. c Uterus bicornuate (unicollis). There are two uterine cornu and two endometrial cavities separated by myometrium. The outer fundal contour is concave. d Septate uterus (unicollis). There is a complete septum between the two endometrial cavities. e Arcuate uterus. There is “heart-shaped” endometrium, characteristic of this anomaly. The fundal contour is only slightly concave

7.1  Female Pelvis

Fig. 7.1.3a,b  Mayer-Rokitansky-Kuster-Hauser syndrome. Sagittal (a) and axial (b) T2-weighted FSE images show absence of the uterus and upper two thirds of the vagina. Note the posterior displacement of the posterior bladder wall (b) as there is no vagina present between the bladder (*) and the rectum (R)

causing obstruction. The two uterine horns are usually widely separated, with preservation of the endometrial and myometrial widths. An oblique plane of imaging parallel to the long axis of the uterus is useful for delineation of the uterus, whereas the axial oblique plane is useful to delineate the vaginal septum. When hemorrhage is present in the obstructed segment it is best seen on T1-weighted images as high signal intensity. 7.1.4.1.4  Bicornuate Uterus Partial fusion of the Müllerian ducts results in the bicornuate uterus (Fig. 7.1.4). There is incomplete fusion of the cephalad extent of uterovaginal horns with resorption of the uterovaginal septum. MRI shows uterine horns separated by an intervening cleft in the external fundal myometrium of longer than 1 cm. Normal zonal anatomy is seen in each horn and there is a dividing septum composed of myometrium. 7.1.4.1.5  Septate Uterus A septate uterus is seen when there is incomplete resorption of the final fibrous septum between the two uterine horns. The septum may be partial, or it may be complete and extend to the external cervical os. The differentiation between the septate and the bicornuate uterus is clinically important. The septate uterus is associated with a higher rate of reproductive complications, with only 3% of pregnancies delivered at term. Lacking the neces-

sary blood supply, the collagenous septum in the septate uterus cannot support a pregnancy as well as the myometrial septum in the bicornuate uterus. The abortion rate in patients with a septate uterus is twice that of patients with a bicornuate uterus. When clinically warranted, the dividing septum can be removed hysteroscopically. T2-weighted MR images taken parallel to the long axis of the uterus demonstrate a convex, flat or concave (less than 1 cm) external uterine contour and the presence of the fibrous septa. 7.1.4.1.6  Congenital Absence of the Müllerian Ducts (Vaginal Aplasia, MRKH Syndrome) The pathophysiology of vaginal absence may be either a result of failure of the vaginal plate to form, or failure of cavitation. Absence of the uterus and Fallopian tube indicates total failure of Müllerian duct development and is known as MRKH syndrome. The presence of vaginal agenesis or hypoplasia with intact ovaries and Fallopian tubes is accompanied by variable anomalies of the uterus, urinary tract, and skeletal system. See above section on Müllerian duct anomalies for further discussion of MRKH syndrome. 7.1.4.1.7  Disorders of Vertical Fusion The transverse vaginal septum prevents loss of menstrual blood and result in hematocolpos. Most patients present

969

970

7  Pelvis

Fig. 7.1.4a,b  Bicornuate uterus. Coronal T2-weighted FSE images (a,b) demonstrate the right and left corpus uterus (arrows) and a single cervix (C)

as teenagers with cyclical abdominal pain, and a hematocolpos might be palpable within the pelvis. In these patients, a careful pelvic examination and ultrasound (US) are helpful for diagnosis. On MRI, T2-weighted images show dilatation of the vagina with intraluminal fluid of intermediate or high signal intensity and the occasional presence of fluid/debris levels. The lower third of the vagina is replaced by low signal intensity fibrous tissue with loss of normal zonal anatomy. T1-weighted images with fat suppression confirm blood products in hematometrocolpos and associated endometriosis if present. 7.1.4.1.8  Disorder of Lateral Fusion These patients often present with the incidental finding of a vaginal septum that is usually asymptomatic. It may first be diagnosed during pregnancy, and excision will be necessary to ensure a vaginal delivery. This malformation may be missed if careful examination is not performed. 7.1.4.1.9  Cervical Incompetence Cervical incompetence is responsible for approximately 15% of second- and third-trimester abortions. Primary incompetence may be congenital, associated with diethylstilbestrol exposure, or caused by reduced collagen within the cervix. Secondary incompetence usually results from multiple gestations, gynecologic/obstetric trauma, or increased prostaglandin production. US is currently the investigation of choice for diagnosing cervical incompetence during pregnancy.

MRI offers the potential to diagnose cervical incompetence in the non-pregnant as well as the pregnant patient. Four MRI findings have been described as suggestive of cervical incompetence (Hricak et al. 1986). These include shortening of the endocervical canal (less than 3 cm), widening of the internal cervical os (greater than 4 mm), asymmetric widening of the endocervical canal and thinning or absence of the low-signal-intensity cervical stroma. When one or more of these findings are present, cervical incompetence should be suspected. 7.1.4.2  Benign Uterine Conditions 7.1.4.2.1  Adenomyosis Adenomyosis is the presence of endometrial tissue within the myometrium and secondary smooth muscle hypertrophy–hyperplasia. It can be diffuse or focal. The most frequent symptoms are dysmennorhea and dysfunctional uterine bleeding. It is found in 15–27% of hysterectomy specimens, with an increased incidence in multiparous women. Transvaginal US is the initial imaging modality, whereas MRI should be reserved for indeterminate cases or those undergoing uterus sparing surgery (Ascher et al. 2003; Tamai et al. 2005). Pitfalls in the diagnosis of uterine adenomyosis include leiomyoma, endometrial carcinoma, and myometrial contractions (Reinhold et al. 1999). On T2-weighted MRI, adenomyosis appears as areas of low myometrial signal intensity, which presents as focal or diffuse thickening of the JZ (Fig. 7.1.5). When diffuse, a widened low-signal-intensity JZ > 12 mm diagnoses disease with high accuracy, while a JZ < 8 mm excludes

7.1  Female Pelvis

Fig. 7.1.5a,b  Adenomyosis. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted FSE images through the pelvis demonstrate extensive widening of the entire anterior junctional zone (JZ), which contains multiple foci of high signal

intensity that represent endometrial rests. There is a large adenomyoma anteriorly and a small adenomyoma posteriorly (A). E = endometrial canal

disease with high accuracy. For indeterminate cases, (JZ 8–12 mm) ancillary criteria are used. These include the presence of high-signal-intensity linear striations (fingerlike projections) extending out from the endometrium into the myometrium on T2-weighted images and highsignal-intensity foci on T1-weighted images. These foci are believed to represent endometrial rests and/or small punctuate hemorrhages (Ascher et al. 2003; Tamai et al. 1998).

On T1-weighted images, leiomyomas most commonly present as well-circumscribed, rounded lesions with intermediate signal intensity, often indistinguishable from adjacent myometrium. Optimum contrast is achieved on T2-weighted images, where the tumor is of lower signal intensity relative to the myometrium or endometrium (Fig. 7.1.6). The presence of calcification usually results in areas of signal void on both T1-weighted and T2-weighted images; however, on MRI, signal void can also be produced by fast flowing blood and therefore is not specific for calcification as it is on CT. Intravenous gadolinium chelates are not necessary to make the diagnosis of leiomyomas, but may be helpful in determining the myometrial origin in cases of submucosal leiomyomas. It also provides useful information about vascularity of lesions, a factor that may influence the type of treatment undertaken (Ascher et al. 2003).

7.1.4.2.2  Leiomyoma Leiomyomas are the most common uterine tumors. These benign tumors are found in up to 40% of women in their reproductive years. They are usually multiple and may be subserosal, intramural or submucosal in location. Symptoms may be caused by the location of the leiomyoma (e.g., menorrhagia with submucosal leiomyomas) and/or their mass effect (e.g., urinary frequency caused by large subserosal leiomyomas). Diagnostic ultrasound is often the initial radiological evaluation in these patients, while MRI is usually reserved for patients with inconclusive US results or patients undergoing myomectomy, uterine embolization, or MR-guided focused ultrasound (Murase et al. 1999).

Degenerating Leiomyomas A variety of degenerative processes can alter the characteristic appearance of a leiomyoma, making differential diagnosis more difficult. Types of degeneration include hyaline, hemorrhagic, necrotic, myxoid, fatty, calcific, and sarcomatous transformation. On MRI, degenerating leiomyoma appears as a heterogeneous, round, well-defined mass. Fat saturation T1-weighted images may be

971

972

7  Pelvis

Fig. 7.1.6a,b  Degenerating leiomyoma. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted FSE MR images through the pelvis demonstrate a large submucosal leiomyoma (L) with hyaline degeneration. Note the posterior displacement of the endometrium (E) and cervix by the large leiomyoma

helpful in cases of hemorrhagic degeneration. Following administration of intravenous contrast medium, most leiomyomas enhance less than adjacent myometrium, whereas degenerated areas may not enhance, indicating the presence of hemorrhage or necrosis (Ascher et al. 2003; Tamai et al. 2005). Role of MRI in Treatment Selection and Follow-Up Hysterectomy has been the traditional primary treatment for debilitating leiomyomas. While hysterectomy is curative, alternative uterine-sparing procedures may be appropriate for some patients. Specifically, myomectomy has been successfully performed for many years, and transcatheter uterine arterial embolization (UAE) is being used as an alternative. Recently another alternate minimally invasive treatment has been MR-guided highintensity focused ultrasound. MRI allows precise determination of the size, location, and number of leiomyomas. It is also very useful in differentiating a pedunculated subserosal leiomyoma from an adnexal mass (Murase et al. 1999). MRI is the most accurate noninvasive diagnostic imaging test available so far for differentiation of a leiomyoma from adenomyosis (Ascher et al. 2003). This distinction impacts on clinical management, as an accepted surgical treatment of a leiomyoma is myomectomy, whereas standard treatment of

debilitating adenomyosis is hysterectomy. MRI may help in the selection of invasive treatment (i.e., myomectomy versus hysterectomy versus UAE). Hemorrhagic or necrotic leiomyomas do not respond to UAE and are best treated by myomectomy or hysterectomy. Presence of ovarian-uterine artery collaterals on contrast-enhanced MRI may be associated with lower UAE success rates and amenorrhea (Spies et al. 2002). The effects of hormonal therapy on leiomyomas can be monitored by MRI. Gonadotropin-releasing hormone analogues usually provide temporary relief, and leiomyomas regrow once the treatments stops. MRI is valuable in follow-up of patients who have undergone myomectomy as recurrence occurs in up to 15% of patients. MRI may also be useful to monitor the success of UAE and assess its durability and potential complications such as intracavitary-sloughed leiomyomas, endometritis, and pyometra (Spies et al. 2002). Following UAE, leiomyomas are of homogenous high signal intensity on T1-weighted images; the central area may have lower signal intensity. On T2-weighted images, leiomyomas demonstrate homogeneous low signal intensity, but the central area might show higher signal intensity. On contrast-enhanced MRI, leiomyomas do not enhance; however, they may demonstrate heterogeneous signal intensity and enhancement if not completely infarcted (Spies et al. 2002).

7.1  Female Pelvis

7.1.4.2.3  Endometrial Polyps Benign endometrial polyps are common in the endometrial cavity at all ages, with their greatest prevalence after age 50. They occur most commonly in patients using tamoxifen. US hysterosalpingography is highly accurate in detecting endometrial polyps even in the setting of endometrial hyperplasia, but differentiating a broad-based polyp from a submucosal fibroid may be problematic. Outpatient hysteroscopy is the technique of choice for their diagnosis. MRI features are often non-specific. Endometrial polyps usually appear as focal thickening of the endometrium. Visualization of a separate endometrial stripe is diagnostic (Grasel et al. 2000). Endometrial polyps are isointense to endometrium on both T1- and T2-weighted images. Areas of cystic changes show high signal intensity, whereas the central fibrous core demonstrates low signal intensity on T2-weighted images. On contrast-enhanced images, endometrial polyps appear as solid masses with heterogeneous enhancement and cystic changes but without myometrial invasion. Small polyps are best appreciated against hypointense endometrium on early arterial phase of enhancement (Grasel et al. 2000). 7.1.4.2.4  Endometrial Hyperplasia Endometrial hyperplasia may be subdivided into cystic hyperplasia (simple), adenomatous hyperplasia (complex), and atypical hyperplasia. The main presentations are abnormal uterine bleeding, infertility and post-menopausal bleeding. The important practical considerations in the natural history of endometrial hyperplasia are coexisting endometrial carcinomas, coexisting ovarian carcinomas, and the risk of progression to endometrial carcinoma. The main objectives of investigating a woman found to have endometrial hyperplasia are to exclude invasive endometrial cancer or ovarian cancer (Reinhold and Khalili 2002). US evaluation might reveal endometrial thickening, which may be due to endometrial hyperplasia, polyps, or carcinoma. In postmenopausal women, the threshold value for serious endometrial abnormalities on transvaginal US is 5 mm. There is no reliable threshold value for pre-menopausal women, but limited data in the literature suggest that endometrial thickness greater than 8 mm during the proliferative phase or greater than 16 mm during the secretary phase is abnormal. Endometrial hyperplasia usually appears as diffuse thickening of the endometrial stripe on T2-weighted images. The signal intensity of the endometrial stripe is isointense or slightly hypointense relative to normal endometrium. Cystic changes appear as small hyperintense foci. However, these MRI characteristics are nonspecific and are identical to stage IA endometrial carcinoma (Reinhold and Khalili 2002).

7.1.4.3  Malignant Conditions of the Uterine Corpus Adenocarcinomas arise from the uterine epithelium and constitute 90% of endometrial cancers. The remaining histological types of endometrial carcinoma include adenocarcinoma with squamous differentiation, adenosquamous carcinoma, clear cell carcinoma, and papillary serous carcinoma. Uterine sarcomas are rare tumors of mesenchymal origin accounting for 2–5% of all uterine malignant tumors. The most common histological variants are endometrial stromal sarcoma, mixed Müllerian tumors and leiomyosarcoma. Primary uterine lymphoma is very rare occurring in only 1% of patients with lymphoma. Metastasis to the uterus from non-gynecologic neoplasms are rare, with breast and colon being the two most common primary sites. 7.1.4.3.1  Endometrial Carcinoma Endometrial carcinoma is the fourth most common female cancer and the most common malignancy of the female reproductive tract (Jemal et al. 2007). In 2007, 39,080 new cases and 7,400 deaths are expected in the United States (Jemal et al. 2007). Five-year survival rates vary between 96% for stage I disease and 26% for stage IV disease (Jemal et al. 2007). Presenting as post-menopausal bleeding, the disease occurs most frequently in white women, with peak incidence between ages 55 and 65. Risk factors include unopposed estrogen intake, nulliparity, obesity, and diabetes and Stein-Leventhal syndrome. The prognosis of endometrial carcinoma depends on a number of factors, including stage, depth of myometrial invasion, lymphovascular invasion, nodal status, and histological grade. Determination of the depth of myometrial invasion is important, since it correlates with the incidence of lymph node metastases. Preoperative evaluation of prognostic factors helps in sub-specialist treatment planning (Frei et al. 2000). Endometrial cancer primarily presents at stage I (80% of cases), and the standard treatment is total abdominal hysterectomy and bilateral salpingo-oophorectomy. Diagnosis Endometrial carcinomas are typically diagnosed at endometrial biopsy or dilatation and curettage with imaging being reserved to evaluate extent of disease. Dynamic contrast-enhanced MRI offers a “one-stop” examination, with the highest efficacy for pretreatment evaluation in patients with endometrial cancer (Sala et al. 2007; Frei et al. 2000; Hricak et al. 1991; Manfredi et al. 2004). On unenhanced T1-weighted images, endometrial carcinoma is isointense with the normal endometrium. Although endometrial cancer may demonstrate high signal intensity

973

974

7  Pelvis Table 7.1.2  TNM and FIGO classification of endometrial carcinoma TNM

FIGO

Endometrium

Tis

0

Carcinoma in situ

T1

I

Limited to the uterus

T1a

Ia

Limited to endometrium

T1b

Ib

Invasion less than or equal to half of the myometrium

T1c

Ic

Invasion greater than half of the myometrium

T2

II

Invasion of cervix

T2a

IIA

Invasion of endocervical glands

T2b

IIB

Invasion of cervical stroma

T3 and/or N1

III

Local and/or regional spread

T3a

IIIA

Involvement of serosa and/or adnexa and/or positive peritoneal cytology

T3b

IIIB

Vaginal involvement

T3c

IIIC

Metastatic to pelvic and/or para-aortic nodes

T4

IV

Tumor extends outside pelvis or invades bladder or rectal mucosa

T4a

IVA

Invasion of bladder and/or bowel mucosa

M1

IVB

Distant metastasis

on T2-weighted sequences, it is more typically heterogeneous, and may even be of low signal intensity. Routine use of dynamic intravenous contrast enhancement is necessary for state-of-the-art MR evaluation of endometrial carcinoma (Frei et al. 2000; Hricak et al. 1991; Manfredi et al. 2004). Following intravenous contrast medium administration, there is early enhancement of endometrial cancer relative to the normal endometrium, allowing identification of small tumors, even those contained by the endometrium. In the later phases of enhancement, i.e., equilibrium phase, tumor appears hypointense relative to the myometrium. Staging Imaging criteria for staging of endometrial cancer are based on the TNM/FIGO classification (Table 7.1.2). MRI is significantly superior to ultrasound (US) and CT in the evaluation of both tumor extension into the cervix and myometrial invasion (Kim et al. 1995; Kinkel et al. 1999). The overall staging accuracy of MRI has been reported to be between 85% and 93% (Hricak et al. 1987, 1991; Kim et al. 1995; Manfredi et al. 2004; Rockall et al. 2007). Stage I endometrial cancers include tumors confined to the corpus. Stage IA tumors (limited to endometrium)

appear as a normal or widened (focal or diffuse) endometrium. An intact JZ and a band of early subendometrial enhancement exclude deep myometrial invasion (Fig. 7.1.7) (Yamashita et al. 1993). Regardless of sequence, the tumor-myometrium interface is smooth and sharp. In stage IB disease (Fig. 7.1.8), tumor extends less than 50% into the myometrium with associated disruption or irregularity of the JZ and subendometrial enhancement. If these landmarks are not present, stage IB tumor is suggested by an irregular tumor-myometrium interface. Presence of low-signal-intensity tumor (later phases of enhancement) within the outer myometrium or beyond indicates deep myometrial invasion—stage IC disease (Fig. 7.1.9). Erroneous MRI assessment of the depth of myometrial invasion may occur there is a large polypoid endometrial carcinoma, which distends the uterus so that the thin rim of myometrium is stretched over it rather than deeply infiltrated (Yamashita et al. 1993). Other causes include the presence of leiomyomas, congenital anomalies, and indistinct zonal anatomy. Stage II includes tumor extension beyond the uterine corpus into the cervix. In stage IIA, invasion of the endocervix appears as widening of the internal os and endocervical canal with preservation of the normal low-

7.1  Female Pelvis

Fig. 7.1.7a–d  Stage 1a endometrial carcinoma. Sagittal T2weighted FSE (a), sagittal gadolinium-enhanced T1-weighted fat-suppressed (b), un-enhanced (c) and gadolinium-enhanced

axial T1-weighted fat-suppressed (d) MR images show an endometrial carcinoma (T) confined to the endometrium (O ovary). Note the preservation of the junctional zone (arrow in a)

975

976

7  Pelvis

Fig. 7.1.8a,b  Stage 1b endometrial carcinoma. Sagittal T2weighted FSE (a) and sagittal gadolinium-enhanced T1-weighted fat-suppressed (b) MR images show an endometrial carcinoma (T) with superficial myometrial invasion. Note the disruption

of the junctional zone. The tumor involves less than 50% of the myometrium, which is better appreciated on the gadoliniumenhanced T1-weighted image (arrow in b)

signal-intensity fibrocervical stroma on T2-weighted images (Fig. 7.1.10). Disruption of the fibrocervical stroma by high-signal-intensity tumor on T2-weighted images together with disruption of normal enhancement of the cervical mucosa by low-signal-intensity tumor on late dynamic contrast-enhanced MRI indicate cervical stroma invasion—stage IIB disease. In stage III disease, tumor extends outside the uterus but not outside the true pelvis. Parametrial involvement stage IIIA appears as disruption of the serosa with direct extension into the surrounding parametrial fat. In stage IIIB disease, tumor extends into the upper vagina, and there is segmental loss of the low-signal-intensity vaginal wall. In stage IIIC disease lymphadenopathy is present. Stage IV disease is tumor that extends beyond the true pelvis or invades the bladder or rectum. Loss of low signal intensity of the bladder or rectal wall indicates stage IVA disease. In stage IVB disease, distant metastasis, malignant ascites, or peritoneal deposits are present. Peritoneal deposits are better demonstrated on the delayed contrast-enhanced MR images (Low et al. 2005).

Role of MRI in Treatment Selection and Follow-Up MRI can assist in the evaluation of the tumor location and extent, and it therefore can contribute to planning of the treatment approach and subspecialty referral. Preoperative identification of patients who could benefit from hormone therapy or from appropriate surgical expertise, if extensive surgery is required, is an all-important challenge. The major diagnostic factors necessary for the preoperative evaluation of endometrial cancer are: 1 Differentiation between stages IA and IB; this is becoming critical with increased use of hormonal treatment for stage IA disease. 2 Determination of the risk of lymph node metastasis in order to determine surgical management. Differentiation of stage IB from stage IC has prognostic as well as morbidity implications. Stage IB patients would undergo lymph node sampling, while stage IC patients would undergo radical lymph node resection. 3 Diagnosing gross cervical invasion, which requires preoperative radiation therapy or a different surgical plan, i.e., radical hysterectomy instead of total abdominal hysterectomy.

7 Fig. 7.1.10a,b  Stage 2a endometrial carcinoma. Sagittal (a) and axial (b) gadolinium-enhanced T1-weighted fat-suppressed MR images show an endometrial carcinoma (T), with deep myometrial invasion and tumor extension into the cervical ca-

nal (arrow in a). Note the preservation of the low signal intensity cervical stroma (C in a). Incidentally, presence of a uterine leiomyoma (L in a) is noted

7.1  Female Pelvis

Fig. 7.1.9a–c  Stage 1c endometrial carcinoma. Sagittal T2weighted FSE (a) and sagittal (b) and axial (c) gadoliniumenhanced T1-weighted fat-suppressed MR images show an endometrial carcinoma (T). It is difficult to assess the depth of myometrial invasion on sagittal images due to the presence of multiple leiomyomas (L in a–c) and the loss of definition of the JZ. However, on the gadolinium-enhanced axial T1-weighted image (c), it can be appreciated that the tumor involves more than 50% of the myometrium. Note excellent demarcation of the cervix (C in a–c) on contrast-enhanced compared to un-enhanced images

977

978

7  Pelvis

Fig. 7.1.11a–c  Tumor recurrence after hysterectomy for endometrial carcinoma. Sagittal T2-weighted FSE image (a) shows a stage 1c endometrial carcinoma (T in a), which was treated with radical hysterectomy. Sagittal (b) and axial (c) T2-weighted images show an intermediate signal intensity mass at the vaginal vault (T in b, c) and a deposit in the rectal wall, both in keeping with tumor recurrence (T in c)

In summary, the main role of imaging in the evaluation of endometrial cancer is the assessment of morphologic prognostic factors. The depth of myometrial invasion is probably the single most important morphologic prognostic factor as it correlates with tumor grade, tumor extension into the cervix and the prevalence of lymph node metastases. MRI is the modality of choice for detection of tumor recurrence within the pelvis. Approximately 17% of patients with endometrial carcinoma develop local or distant recurrence. The prevalence of recurrence depends on the tumor grade, size, and stage at presentation in addition to the histological type. Recurrence is most common in the first five years after diagnosis, with 70% of patients with endometrial carcinoma presenting with recurrent disease in the first 3 years. The vagina is the sole site of recurrence in 30–50% of patients, the remainder developing pelvic or para-aortic lymph node

involvement or systemic spread manifesting as hepatic, pulmonary or osseous metastasis or peritoneal carcinomatosis. On MRI, recurrent disease at the vaginal vault appears as a mass or nodule on T1-weighted imaging and has high signal intensity on T2-weighted imaging (Fig. 7.1.11). 7.1.4.3.2  Uterine Sarcomas (Leiomyosarcomas, Endometrial Sarcomas, Malignant Mixed Müllerian Tumors) Sarcomas of the uterus are often highly malignant. They are rare, with an incidence of approximately 2 per 100,000 women over the age of 20, and account for 3–5% of all uterine cancers. The tumors are frequently very large at the time of the examination and it is difficult to determine the primary origin of the mass. MRI can provide

7.1  Female Pelvis

Fig. 7.1.12a–c  Stage 3c Malignant Mixed Mullerian Tumour. Sagittal T2-weighted FSE (a), sagittal gadolinium-enhanced T1-weighted fat-suppressed (b) and axial (parallel to short axis of the corpus) T2-weighted FSE (c) MR images show a large heterogeneous tumour (T), with deep myometrial invasion and tumor extension into the cervical stroma (a,b). Note the presence of enlarged bilateral obturator lymph nodes (N in c)

an accurate preoperative assessment of uterine size and degree of involvement. The MR imaging features are non-specific and may be indistinguishable from those of endometrial carcinoma (Sahdev et al. 2001; Shapeero and Hricak 2002). However, uterine sarcomas tend to be large and heterogeneous with areas of hemorrhage and cystic necrosis. Deep myometrial invasion and peritoneal seeding are usually seen at presentation (Fig. 7.1.12). Leiomyosarcomas account for only 1.3% of uterine malignancies. Most leiomyosarcomas arise de novo from the myometrium although malignant transformation of leiomyomas can occur. Although it has been suggested that an irregular margin of uterine leiomyoma may indi-

cate malignant transformation (Pattanti et al. 1995), MRI a leiomyoma undergoing benign degeneration and a leiomyosarcoma cannot be differentiated reliably on MRI. 7.1.4.3.3  Gestational Trophoblastic Disease Gestational trophoblastic neoplasms include the tumor spectrum of hydatidiform mole, invasive mole (choriocarcinoma destruens), and choriocarcinoma. They arise from fetal tissue within the maternal host and are composed of both syncytiotrophoblastic and cytotrophoblastic cells. The role of imaging in gestational trophoblastic

979

980

7  Pelvis

disease (GTD) has been primarily to document metastatic disease at initial diagnosis or to evaluate persistent disease. There are no specific imaging findings yet that allow differentiation of complete mole from invasive mole or choriocarcinoma. On T2-weighted images, GTD is seen as a heterogeneous, predominantly high-signal-intensity mass that obliterates normal uterine zonal anatomy (Green et al. 1996; Hricak et al. 1986). On T1-weighted images, it may be iso- or hyperintense to adjacent myometrium. The tumors are hypervascular, and enlarged vessels in the broad ligament and the uterus are depicted as signal voids on both T1-weighted and T2-weighted images. The tumors avidly enhance following injection of IV contrast medium. 7.1.4.4  Malignant Conditions of the Cervix The most common histological type of cervical carcinoma is squamous cell carcinoma (90%), followed by adenocarcinoma (5–10%). Other rare histological types include small cell carcinoma, adenosquamous carcinoma, and lymphoma. MRI imaging features of these rare tumors are the same as those of squamous cell carcinoma. However, small cell carcinoma usually demonstrates highly aggressive features such as parametrial involvement, pelvic lymphadenopathy, and distant metastasis. Adenoma malignum is a very rare tumor and has specific MR imaging appearances. 7.1.4.4.1  Cervical Carcinoma Cervical carcinoma is the third most common gynecological malignancy (Jemal et al. 2007). In 2007, 11,150 new cases and 3,670 deaths are expected in the United States (Jemal et al. 2007). Five-year survival rates vary between 92% for stage I disease and 17% for stage IV disease (Jemal et al. 2007). During the last 50 years, there has been a steep decline in deaths from cervical cancer. This improvement in mortality has been attributed to the development of the Papanicolaou smear, and only minor improvement has been achieved in the survival of invasive cervical cancer. Established risk factors for cervical cancer include early sexual activity, especially with multiple partners, cigarette smoking, immunosuppression, and infection with human papilloma viruses 16 and 18. Abnormal uterine bleeding, especially after intercourse, and vaginal discharge may be symptoms leading to the diagnosis. Diagnosis MRI is the best single imaging investigation and can accurately determine tumor location (exophytic or endocervical), tumor size, depth of stromal invasion and extension into the lower uterine segment (Nicolet et al. 2000;

Okamoto et al. 2003). On T1-weighted images, tumors are usually isointense with the normal cervix, and may not be visible. On T2-weighted images, cervical cancer appears as a relatively hyperintense mass and is easily distinguishable from low signal intensity cervical stroma. Staging The most important issue in the staging of cervical cancer is to distinguish early disease (stages I and IIA) that can be treated with surgery from advanced disease that must be treated with radiation alone or combined with chemotherapy. Because of a relatively high likelihood of parametrial invasion and/or lymph node metastases, MRI is recommended in evaluating cervical carcinoma patients with clinical stage IB disease or greater when the primary lesion is larger than 2 cm (Nicolet et al. 2000; Okamoto et al. 2003; Scheidler and Heuck 2002). In single-institutional studies, MRI has been shown to be better than either CT or physical examination in demonstrating parametrial invasion and as good as CT in detecting nodal metastases (Hricak et al. 1996). The staging accuracy of MRI ranges from 75–96%. The reported sensitivity of MRI in the evalu­ ation of parametrial invasion is 69%, and the specificity is 93% (Okamoto et al. 2003; Scheidler and Heuck 2002). A recent prospective multi-center study conducted jointly by the American College of Radiology Imaging Network (ACRIN) and the Gynecologic Oncology Group (GOG) compared MR imaging, CT and FIGO clinical staging in the pretreatment assessment of early invasive cervical cancer (Hricak et al. 2005). The study showed that MR was equivalent to CT for overall preoperative staging. However, MRI performed significantly better than CT for preoperative tumor visualization and determination of parametrial invasion. Reader agreement was higher for MRI readers than for CT readers (Hricak et al. 2005). Recommendations for diagnostic evaluation of tumor staging derive from the TNM/FIGO classification system (Table 7.1.3). Stage I tumors are confined to the uterus. Stage IA is defined as a microinvasive tumor that cannot be demonstrated at MRI. Stage IB carcinoma appears as a high-signal-intensity mass in contrast to the low-signal-intensity fibrocervical stroma on T2-weighted images (Fig. 7.1.13). Stage II. In stage IIA tumors, segmental disruption of the upper two-thirds of the vaginal wall without parametrial invasion is demonstrated on T2-weighted images. The lack of preservation of low signal intensity cervical stroma is highly indicative of parametrial invasion—stage IIB disease (Fig. 7.1.14). Stage III. In stage IIIA, vaginal involvement reaches the lower third of the vaginal canal without extending to the pelvic sidewall. When the tumor extends to the pelvic sidewall (pelvic musculature or iliac vessels) or causes hydronephrosis, it is defined as stage IIIB. Stage VI. Once tumor invades the adjacent organs such as the bladder and rectal mucosa, or distant metas-

7.1  Female Pelvis Table 7.1.3  TNM and FIGO classification of cervical carcinoma TNM

FIGO

Cervix

Tis

0

Carcinoma in situ

T1

I

Carcinoma confined to the cervix

T1a

Ia

Invasive carcinoma identified only microscopically

T1a1

Ia1

Stromal invasion no greater than 3 mm in depth and no wider than 7 mm

T1a2

Ia2

Stromal invasion greater than 3 mm but less than 5 mm in depth and no wider than 7 mm

T1b1

Ib1

Clinical lesions no greater than 4 cm in size

T1b2

Ib2

Clinical lesions greater than 4 cm in size

T2

II

Extension beyond cervix and involvement of the vagina (but not the lower vagina)

T2a

IIa

No parametrial invasion

T2b

IIb

Parametrial invasion

T3

III

Invasion of the lower third of the vagina and/or extending to the pelvic sidewall

T3a

IIIa

Invasion of the lower third of the vagina

T3b

IIIb

Extension to the pelvic wall and/or hydronephrosis

T4

Iva

Invasion of the mucosa of bladder/rectum and/or extending beyond the true pelvis

M1

IVb

Spread to distant organs

Fig. 7.1.13a,b  Stage 1b cervical cancer. Sagittal (a) and axial (b) T2-weighted FSE images show an endocervical mass (T). There is preservation of the low-signal-intensity ring, which indicates an intact fibrocervical stroma—stage 1b disease

981

982

7  Pelvis

Fig. 7.1.14a,b  Stage 2b cervical cancer. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted images show a large cervical cancer (T in a, b), which invades the fibrocervical stroma and extends to both parametria (b)

tasis occurs, the stage is defined as IV (Fig. 7.1.15). Disruption of low-signal-intensity bladder, stroma, or rectal wall by high-signal-intensity tumor on T2-weighted images indicates bladder/rectal involvement. In stage IVB, there is distant metastatic disease (Fig. 7.1.16). Although pelvic node metastases do not change the FIGO stage, para-aortic or inguinal node metastases are classified as stage IVB. Role of MRI in Treatment Selection and Follow-Up Staging of cervical cancer is still based on clinical FIGO criteria, which, compared to surgical staging, can be erroneous in up to about 30% of patients with stage IB disease and 64% of patients with stage IIIB disease (Hricak et al. 2005). The greatest difficulties in the clinical evaluation of patients with cervical cancer are the assessment of parametrial and pelvic sidewall invasion, accurate estimation of tumor size (especially if the tumor is primarily endocervical in location), and the evaluation of lymph node metastases. Accurate pretreatment evaluation of these prognostic factors is crucial in determining appropriate therapy in patients with cervical cancer. MRI follow-up examination in patients treated with radiotherapy demonstrates a small size cervix of low signal intensity on both T1- and T2-weighted sequences (Fig. 7.1.17). However, high signal intensity can be seen on T2-weighted images within the first 6–12 months after radiotherapy due to increased capillary permeability, edema, and inflammation. After 12 months, uniformly low-signal-intensity fibrosis is appreciated. If present, retention of secretions within the endometrial canal appears bright on T2-weighted images. Associated post-

radiation changes in pelvis, bladder, and bowel may be present. Cervical stenosis and fistula formation can be accurately detected at MRI. MRI is the modality of choice for detection of tumor recurrence within the pelvis. Approximately 30% of patients with invasive cervical carcinoma die because of recurrent or persistent disease. The likelihood of tumor recurrence depends on the tumor size, stage at presentation as well as tumor grade and histological type. Recurrence is most common in the first few years after diagnosis, with 60% of patients developing recurrent disease within two years. Manifestations of recurrent disease in cervical carcinoma can be divided into typical and atypical. Typical manifestations involve the pelvis and lymph nodes. However, with increasing use of intensive pelvic irradiation in the treatment of this disease, less typical patterns of recurrence are becoming more frequent. These include peritoneal carcinomatosis and solid organ metastasis to the liver, adrenal glands, lung, or bone (Fulcher et al. 1999). Pelvic recurrence may involve other pelvic organs with invasion of the bladder, urethra, ureters, or rectum, causing secondary complications such as fistula formation or hydronephrosis (Jeong et al. 2000). Tumor extension into the bladder or rectal wall is suggested by abnormally high signal intensity on T2-weighted imaging. Post-gadolinium imaging is helpful in assessing bladder and rectal invasion Sugimura and Okizuka 2002). In patients who have had radiotherapy the critical issue is distinguishing recurrent disease from post-radiation changes. On MRI studies recurrent disease appears as a mass or nodule on T1-weighted imaging that has high signal on T2-weighted imaging (Fig. 7.1.18) (Jeong et al. 2003).

7.1  Female Pelvis

Fig. 7.1.15a–c  Stage 4a cervical cancer. Sagittal (a) and parasagittal (b) T2-weighted FSE demonstrate a large cervical cancer (T in a), which extends to the uterine corpus superiorly and to the lower vagina inferiorly. The tumor is invading the posterior bladder wall (T in b) and the lower right ureter (U in b) causing severe hydronephrosis. Axial body T1-weighted images (c) confirm a markedly dilated right renal pelvis (H in c)

However, T2-weighted images have low specificity for recurrent disease, as benign conditions such as inflammation and edema show increased T2-weighted signal as well. Dynamic contrast-enhanced MRI has been shown to be helpful, improving specificity and accuracy (Kinkel et al. 1997), but early radiation change continues to pose a problem as it may show enhancement. Serial imaging, image-guided biopsy, or PET may be required to further clarify the situation.

7.1.4.4.2  Adenoma Malignum Adenoma malignum, which is also known as mucinous minimal deviation adenocarcinoma, is a subtype of mucinous adenocarcinoma of the cervix. It is a rare tumor, representing only 1.3% of invasive adenocarcinomas. The most common clinical symptom is watery vaginal discharge. It has an unfavorable prognosis, as it disseminates into the peritoneal cavity even in the early stage

983

984

7  Pelvis

Fig. 7.1.16a,b  Stage 4b cervical cancer. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted images show a large cervical cancer (T), which invades the posterior wall of the

bladder (a), lower vagina (a, b), and urethra (T in b). Note the area of high signal intensity within the left pubic ramus (M in b) and the adjacent muscle in keeping with bone metastases

Fig. 7.1.17a,b  Appearances of the uterus after radiotherapy and chemotherapy for cervical carcinoma. Sagittal T2-weighted FSE images a prior to and b after completion of radiotherapy and chemotherapy in a patient with locally advanced cervical carcinoma demonstrate an excellent response to treatment. Prior

to start of treatment (a), there is a large cervical tumor (T in a), with invasion of cervical stroma bilaterally and extension to the endometrial canal. After treatment completion, there is little residual tumor present (T in b). Note the classical appearance of the scar from caesarian section anteriorly (arrow in b)

7.1  Female Pelvis

Fig. 7.1.18a,b  Tumor recurrence after surgery for cervical carcinoma. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted images show an intermediate signal intensity mass at the right vaginal vault (T in a,b) in keeping with tumor recurrence

of the disease and its response to radiation or chemotherapy is poor. Adenoma malignum is usually difficult to detect on regular punch biopsy, Papanicolaou smear, or colposcopy due to its location in the deep cervical stroma. MRI is the imaging test of choice and helps to guide deep cervical stromal biopsy. On T2-weighted images, adenoma malignum appears as a multicystic lesion of high signal intensity that extends from the endocervical glands to the deep cervical stroma with solid components, which are of intermediate signal intensity (Okamoto et al. 2003). 7.1.4.4.3  Cervical Lymphoma Cervical lymphoma is a rare cause of cervical malignancy, but the cervix is the most common site of uterine lymphomas. It accounts for 2% of extranodal lymphomas in females and is almost invariably of non-Hodgkin’s type. MRI shows diffuse enlargement of the cervix with preservation of the cervical stroma. On T2-weighted images, the mass is of high signal intensity compared to lowsignal-intensity cervical stroma. There is heterogeneous enhancement following intravenous contrast medium administration (Okamoto et al. 2003). 7.1.4.5  Malignant Conditions of Vagina and Vulva 7.1.4.5.1  Vaginal Carcinoma Primary vaginal malignancies are uncommon, accounting for only 2% of gynecological malignancies. They

usually affect postmenopausal women age 60–70 years old. Ninety percent of vaginal cancers are squamous cell carcinomas and 5–10% adenocarcinomas. Clear cell carcinoma of the vagina is a subtype of adenocarcinoma associated with maternal exposure to diethylstilbestrol during pregnancy. Vaginal carcinoma most commonly occurs in the upper portion of the posterior wall of the upper vagina. It usually spreads by direct extension to adjacent organs, however lymphatic and hematogenous extension can occur. TNM/FIGO classification of vaginal carcinoma is given in Table 7.1.4. MR is the imaging method of choice due to excellent soft tissue contrast (Chang 2002). Vaginal carcinoma appears as a nodular, exophytic mass of high-to-intermediate signal intensity on T2-weighted images and intermediate signal intensity on T1-weighted images. Disruption of the low-signal intensity vaginal wall on T2-weighted images and an irregular tumor-fat interface indicate invasion of paravaginal tissues (Fig. 7.1.19). High-signal intensity tumor extension to pelvic muscles on T2-weighted images indicates pelvic sidewall invasion. Contrast-enhanced MRI is useful in identification of vesicovaginal fistulas. 7.1.4.5.2  Vaginal Metastases Metastases to the vagina are more frequent than the primary vaginal tumors and originate most commonly from carcinoma of the endometrium and cervix, followed by melanoma and carcinoma of the colon and kidney. The MRI appearances are identical to those of primary vaginal carcinoma.

985

986

7  Pelvis Table 7.1.4  TNM and FIGO classification of vaginal carcinoma TNM

FIGO

Vagina

Tis

0

Carcinoma in situ

T1

I

Tumor confined to the vagina

T2

II

Tumor extends to the paravaginal tissue but not to the pelvic sidewall

T3

III

Tumor invasion of the pelvic sidewall

T4

IV

Invasion of the mucosa of bladder/rectum and/or extending beyond the true pelvis

Fig. 7.1.19a,b  Vaginal carcinoma. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted images show an intermediate signal intensity mass involving the posterior wall of the entire vagina (T in a,b). Note that the anterior vaginal wall and urethra (U in b) are preserved

7.1.4.5.3  Vulvar Cancer Vulvar cancer is a rare malignancy accounting for 5% of female genital tract cancers. It is most commonly seen in postmenopausal women age 65–70 years old. The most common histologic type is squamous cell carcinoma (85%). Other types include melanoma, Bartholini gland cancer, Paget’s disease, adenocarcinoma, sarcoma, and basal cell carcinoma. Carcinoma of the vulva most commonly involves the labia majora or minora (70%), clitoris, and perineum. Diagnosis of vulvar carcinoma and assessment of involvement of superficial inguinal lymph nodes are done clinically. MRI is very useful in treatment planning for evaluating the local extent of the disease and deep inguinal and pelvic node involvement. Vulvar cancer appears as a mass of intermediate signal intensity on T1-weighted

images and of high signal intensity on T2-weighted images (Chang 2002) (Fig. 7.1.20). The MRI appearances of metastases to the vulva are identical to those of primary vulvar carcinoma. 7.1.5  Adnexa 7.1.5.1  Role of MRI in Characterization of Adnexal Lesions The strength of MRI is its ability to characterize ovarian lesions, especially in the case of masses that are indeterminate on US. The signal intensity characteristics of ovarian masses can lead to a systematic approach to diagnosis. The signal intensity of a specific tumor depends on the presence, type, and extent of cystic and solid compo-

7.1  Female Pelvis

Fig. 7.1.20a,b  Vulvar carcinoma. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted images show an intermediate signal intensity large lobulated mass in the vulva, which

extends to the lower vagina (T in a,b). Note the presence of an enlarged left inguinal node, which was metastatic at histopathology (N in b)

Table 7.1.5  Criteria for differentiating benign from malignant ovarian lesions Criteria

Benign

Malignant

Cystic Septa Wall/septa thickness

Simple <3 mm <3 mm

Papillary projection >3 mm >3 mm

Solid

Homogeneous

Heterogeneous with necrosis or hemorrhage

Size

<4 cm

>4 cm

Absent Absent No enlarged LN

Present Present Enlarged LN

Primary criteria

a

Ancillary criteria Ascites Peritoneal/omental deposits Lymph nodes (LN)

This is an US criteria that has not been shown to be an independent predictor of malignancy on MRI   (Hricak et al. 2000; Tempany et al. 2000)

a

nents within a lesion. Lesions that have a homogeneous low signal intensity on T1-weighted images and high signal intensity on T2-weighted images are simple fluidfilled structures and are considered benign. In general, benign epithelial ovarian neoplasms are predominantly cystic. Fat, hemorrhage, and mucin-containing lesions have high signal intensity on T1-weighted images. Fatsaturated T1-weighted images help distinguish between hemorrhage and fat within a lesion (e.g., endometriosis versus mature cystic teratoma). An adnexal mass of low

or intermediate signal intensity on T1-weighted images and low signal intensity on T2-weighted images contains fibrosis and smooth muscle components. Such lesions include pedunculated leiomyoma, fibroma, fibrothecoma, cystadenofibroma, and Brenner tumors. The absence of a normal ipsilateral ovary or the presence of small follicles surrounding the mass helps identify the ovarian origin of fibromas (Jeong et a. 2000). It is important to recognize that as there are no MRI signal intensity characteristics that are specific for ma-

987

988

7  Pelvis

Fig. 7.1.21a,b  Vegetation within a cystic lesion. Axial T1weighted (a) image demonstrates a solid nodule/vegetation (V in a) within a left adnexal cystic lesion. Gross histopathology

specimen (b) confirms the presence of the vegetation within the lesion. Diagnosis of clear cell carcinoma of the left ovary was made on histopathology

Fig. 7.1.22a,b  Necrosis within a solid lesion. Coronal T2weighted FSE (a) and gadolinium-enhanced T1-weighted fatsuppressed images demonstrate a solid right ovarian mass (T in

a,b) with an area of tumor necrosis (arrow a,b) best appreciated on enhanced images. Histology rendered the diagnosis of serous papillary adenocarcinoma of the right ovary

lignant epithelial tumors, such tumors must be distinguished based on morphologic criteria (Table 7.1.5). The MRI features most predictive of malignancy are an enhancing solid component or vegetations within a cystic lesion (Fig. 7.1.21), presence of necrosis within a solid lesion (Fig. 7.1.22) as well as presence of ascites and peritoneal deposits (Fig. 7.1.23) (Hricak et al. 1996; Tempany et al. 2000). Utilizing unenhanced T1-, T2- and contrast-

enhanced T1-weighted sequences, the presence of at least one of the primary criteria coupled with a single criterion from the ancillary group correctly characterizes 95% of malignant lesions. Both transvaginal US and contrast-enhanced MRI have high sensitivity (97 and 100%, respectively) in the identification of solid components within an adnexal mass. MRI, however, shows higher accuracy (93%) (Hricak et al. 2000).

7.1  Female Pelvis

Fig. 7.1.23a,b  Peritoneal deposits and ascites. Axial T2-weighted images through the upper abdomen (a, b) demonstrate a large amount of ascites (As) as well as the presence of peritoneal deposits and porta hepatic (Ip) and falciform ligament (If) tumor implants (I in a,b) in a patient with ovarian carcinoma

Fig. 7.1.24a,b  Peritoneal inclusion cyst. Coronal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted FSE images show a large cyst (Cy in a,b) conforming to the shape of the pelvis (* bladder). Note the presence of an entrapped right ovary (O) within the cyst

7.1.5.2  Benign Tumors 7.1.5.2.1  Functional Cysts Most ovarian cysts have low signal intensity on T1-weighted images and very high signal intensity on T2-weighted images, as they contain simple fluid. Cyst walls are thin and featureless on T1-weighted images and clearly delineated on T2-weighted images. These are most commonly physiologic cysts, such as follicular cysts that occur because of failure of ovulation. Corpus luteum cysts formed after ovulation are usually larger than follicular cysts and often have thick, ir-

regular walls. Hemorrhagic corpus luteum cysts have relatively high signal intensity on T1-weighted images and intermediate to low signal intensity on T2-weighted images. They do not demonstrate the significant T2 shortening that is seen with many endometriomas. 7.1.5.2.2  Peritoneal Inclusion Cysts Peritoneal inclusion cysts are usually seen in premenopausal women with a history of prior pelvic or abdominal surgery. Failure of peritoneal resorption of ovarian exudates and peritoneal adhesions following surgery results

989

990

7  Pelvis

in fluid collection around the ovary. This usually conforms to the shape of the pelvis (Fig. 7.1.24). The ovary is frequently entrapped within the fluid collection, which demonstrates thick enhancing septation. Identification of the normal ovary within the fluid collection allows confident diagnosis of peritoneal inclusion cyst. 7.1.5.2.3  Polycystic Ovarian Disease (Stein-Leventhal Syndrome) Polycystic ovarian disease is characterized by bilaterally enlarged polycystic ovaries, secondary amenorrhea or oligomenorrhoea, and infertility. About 50% of patients are hirsute, and many are obese. Many cases of female infertility secondary to failure of ovulation are due to polycystic ovarian disease. The classic appearance on MRI is enlarged ovaries with intermediate signal intensity central stroma on T2-weighted images and more than 10 peripherally placed follicles less than 9 mm in diameter. The uterus may demonstrate thickened endometrium due to hyperplasia or even early endometrial carcinoma. 7.1.5.2.4  Endometrioma and Endometriosis Endometriosis is the presence of endometrial epithelium and stroma outside of endometrium and myometrium. It is a syndrome that can present with endometriomas, adhesions, or endometrial implants. Staging of endometriosis requires a laparoscopic procedure to look for any of these elements. Common locations of endometrial tissue include ovary, uterine ligaments, Fallopian tube, recto-vaginal septum, pouch of Douglas, bladder wall, and umbilicus. Imaging is most commonly used for the diagnosis and follow-up of endometriomas. Laparoscopy, however, remains the gold standard, as it can provide a complete evaluation for endometrial implants within the abdomen as well as the pelvis. Transvaginal US is the first imaging modality, with MRI reserved for masses atypical on US. The most specific (91–98%) MRI findings in endometriomas are multiple cystic masses of high signal intensity on T1-weighted images and low signal intensity on T2-weighted images. A range of low signal intensities (shading) within the lesion or a low-signal-intensity ovarian cyst wall (hemosiderin deposition) can also be seen on T2-weighted images. Occasionally, endometriomas may demonstrate high signal intensity on both T1and T2-weighted images. They retain high signal intensity on fat-suppressed T1-weighted images (Fig. 7.1.25). Endometriomas larger than 1 cm are routinely seen, but imaging small implants remains problematic. Implants commonly appear as solid masses or spiculated bands of intermediate-to-low signal intensity on both T1- and T2-weighted images due to fibrosis surrounding the

glandular island (Fig. 7.1.25). Small hemorrhagic endometrial implants become more obvious on fat-saturated T1-weighted images (Woodward et al. 2001). 7.1.5.2.5  Mature Cystic Teratoma Mature cystic teratomas or dermoid cyst is the only benign germ cell tumors. They are most common in young women with a median age at presentation of 30 years. MRI can be used to diagnose dermoid cysts with confidence as the signal intensity of the fat or sebum within the cyst parallels that of fat on all pulse sequences (i.e., high signal intensity on T1-weighted images and intermediate signal intensity on T2-weighted images). Typical findings of mature cystic teratomas include fat–fluid and/ or fluid–fluid levels; layering debris; low-signal-intensity calcifications (e.g., teeth) and soft-tissue protuberances (Rokitansky nodules or dermoid plugs) attached to the cyst wall (Fig. 7.1.26). Both endometriomas and mature cystic teratomas show high signal intensity on T1-weighted images and therefore must be distinguished from each other. The fat in mature cystic teratomas results in chemical shift artifact at the fat fluid interface, which appears as bright or dark bands along the frequency-encoding direction. Use of frequency selective fat-suppression allows differentiation of hemorrhagic lesions such as endometriomas or hemorrhagic cysts from mature cystic teratomas. Immature teratomas are malignant ovarian tumors that also demonstrate fat on MRI. They are usually very large at presentation and have significant solid components that contain coarse calcification and multiple small foci of fat (Outwater et al. 2001). 7.1.5.2.6  Fibrotic Tumors Fibromas, thecomas, and fibrothecomas are all fibrotic tumors of gonadal stromal origin and account for approximately 5% of all ovarian tumors. They are usually asymptomatic and typically detected in middle-aged women during routine gynecologic examination. Fibrotic tumors appear solid on imaging, thus mimicking malignant ovarian tumors. They are associated with ascites in 15% of the cases and pleural effusion (Meigs syndrome) in 1%. Fibromas show homogeneous low signal intensity on T1-weighted images. On T2-weighted images, they appear as well-defined masses of low signal intensity, which contain scattered areas of high signal intensity, representing cystic degeneration or edema. The low signal intensity in all sequences is a reflection of the predominantly collagen content of these tumors and is diagnostic of fibromas (Fig. 7.1.27). Thecomas without prominent fibrotic components have MR imaging appearances similar to those of malig-

7.1  Female Pelvis

Fig. 7.1.25a,b  Endometriosis. Sagittal T2-weighted FSE (a), axial T1-weighted (b), and axial T1-weighted fat-suppressed (c) images demonstrate the presence of bilateral T1-weighted high-signal intensity cystic ovarian lesions, which retain their high signal intensity on the fat-suppressed T1-weighted image (E in b,c). A range of low signal intensities (shading) is seen within the left ovarian lesion on the T2-weighted image (E in a). Within the superior and posterior bladder wall (* bladder), two solid implants of intermediate signal intensity on both T1and T2-weighted images (EI in a, c) are seen. Diagnosis of endometriosis was confirmed at surgery

nant tumors. The prominent lipid content of thecomas demonstrates intermediate to high signal intensity on both T1- and T2-weighted images. Cystadenofibromas appear as cystic masses with a solid fibrotic component that shows marked enhancement after administration of gadolinium. They are less likely to be borderline or malignant compared with other serous or mucinous tumors. 7.1.5.2.7  Benign Epithelial Tumors All epithelial ovarian neoplasms can be classified as benign, borderline (low malignant potential) or malignant based on their pathologic features and clinical behavior. Benign types of serous and mucinous tumors are com-

mon while benign forms of endometrioid and clear cell tumors are exceptionally rare. The role of MRI in differentiation of benign from malignant ovarian cystic masses is described in Sect. 7.5.1. Serous cystadenoma. Serous cystadenoma is the most common benign ovarian epithelial tumor. It appears as a thin-walled ovarian unilocular cyst and is usually indistinguishable from a follicular cyst on MRI. The signal intensity of cyst contents is variable but is usually low to intermediate on T1-weighted images and high on T2-weighted images. The cyst wall may contain small nodules, which are of high signal intensity peripherally and have thin low-signal-intensity cores on T2-weighted images due to fibrosis and/or calcification. If these solid projections are prominent, then a borderline tumor should be suspected.

991

992

7  Pelvis

Fig. 7.1.26a,b  Mature cystic teratoma. Axial T1-weighted (a) and axial T1-weighted fat-suppressed (b) images demonstrate bilateral hyperintense cystic and solid ovarian masses (T in a,b), which drop in signal intensity after fat suppression (b), allowing confident diagnosis of bilateral mature cystic teratomas

Fig. 7.1.27a,b  Ovarian fibrothecoma. Sagittal T2-weighted FSE (a) and axial (b) fat-suppressed T2-weighted FSE images show a large well-defined pelvic mass (T in a, b) of low signal intensity. Histopathology confirmed the diagnosis of ovarian fibroma (* bladder)

Mucinous cystadenoma. Mucinous cystadenoma contributes to 15–25% of all ovarian tumors and is the second most common epithelial tumor. The most common appearance of a mucinous cystadenoma is a very large multilocular cyst with no solid components. In mucinous

cystadenomas, the signal intensities of the loculi often vary depending on the amount of proteinaceous or mucinous fluid and whether hemorrhage is present. Endometrioid cystadenomas are often malignant and difficult to distinguish from endometriosis.

7.1  Female Pelvis

Brenner tumors are usually benign and account for only 1–2% of all ovarian tumors. They arise from Wolffian metaplasia of the surface epithelium. Thirty percent of Brenner tumors are associated with cystic teratomas or cystadenomas. Brenner tumors may be solid, cystic, or mixed. When solid, they demonstrate low signal intensity on both T1- and T2-weighted images. 7.1.5.3  Malignant Ovarian Tumors Approximately 90% of ovarian cancers are of epithelial origin. Malignant epithelial tumors are subtyped as serous (50%), mucinous (20%), endometrioid (20%), clear cell (10%), or undifferentiated (1%). Non-epithelial cancers include malignant granulosa cell tumor, dysgerminoma, immature teratoma, endodermal sinus tumor, and metastases to the ovary. Ovarian carcinoma accounts for 3% of all cancers in females in the United States (Jemal et al. 2007). It is the second most common malignancy of the female reproductive tract, but the most frequent cause of death from gynecological malignancy (Jemal et al. 2007). In 2007, 22,430 new cases and 15,280 deaths are expected in the United States (Jemal et al. 2007). Ovarian cancer is a genetically heterogeneous disease with a poor prognosis (overall 5-year survival of < 45%). Five-year survival rates vary between 93% for stage I disease and 30% for stage IV disease (Jemal et al. 2007). In general, ovarian cancer is a disease of the postmenopausal woman and, occasionally, prepubescent girls. The cause of ovarian cancer is unknown, although a number of risk factors have been identified. Chronic anovulation, oral contraception, multiparity, and history of breast feeding seem to be protective, whereas genetic factors appear to play an important role in the development of progression of ovarian cancer. Ten percent of ovarian cancers are due to hereditary syndromes such as BRCA1 and BRCA2 mutations (risk of breast cancer) and Lynch syndrome II (risk of colon cancer). 7.1.5.3.1  Characterization of Malignant Lesions Combined transabdominal and transvaginal US is the best imaging modality for the detection and characterization of ovarian carcinoma. These studies provide superb morphologic detail of the adnexa, allowing detection of masses before they are clinically apparent. MRI is better reserved for problem solving when US findings are non-diagnostic or equivocal. It is generally accepted that MRI cannot reliably differentiate borderline from malignant tumors, although borderline tumors have few papillary projections whereas malignant tumors are predominantly solid with some necrosis (Jeong et al. 2000). In addition, MRI cannot confidently differentiate be-

tween specific surface epithelial, germ cell, stromal cell, or metastatic tumors. However, it is possible to suggest the histologic subtype of the epithelial cancer based on the imaging findings. Serous cystadenocarcinomas are the most common type of malignant epithelial tumors. They are frequently bilateral and usually appear as mixed solid and cystic masses with irregularly shaped solid components. The solid components show avid enhancement and areas of necrosis (Fig. 7.1.28). These appearances in combination with a disproportionately large amount of ascites compared to the tumor size, presence of enlarged lymph nodes and peritoneal and/or serosal implants are suggestive of diagnosis. Mucinous cystadenocarcinomas tend to be larger, more often unilateral, occur in an older age group, and usually have better prognosis than do their serous counterparts. They are usually multiloculated and may be of higher signal intensity on T1-weighted images due to high protein concentration within the mucoid material. Presence of ascites and peritoneal implants is rare. Mucinous cystadenocarcinomas may be associated with Pseudomyxoma peritonei although 90% of P. peritonei originate from the appendix. MRI findings include the presence of mucinous loculated collections of low signal intensity on T1weighted images and high signal intensity on T2-weighted images, which cause scalloping of the liver and splenic surfaces and displacement of the bowel loops due to pressure effects (Fig. 7.1.29). Clear cell carcinoma accounts for only 5% of ovarian cancers and it is almost invariably malignant. It is associated with endometriosis in 25% of cases. The diagnosis should be considered when a nodule is seen within a predominantly cystic endometrioma. Endometrioid carcinomas are usually bilateral and associated with endometrial hyperplasia or carcinoma in 20–30% of cases. They are mainly solid with areas of necrosis and avid enhancement. 7.1.5.3.2  Staging TNM FIGO classification of ovarian carcinoma is presented in Table 7.1.6. Cross-sectional imaging is better accepted and more commonly used in the evaluation and staging of ovarian carcinoma than for other gynecologic malignancies (Tempany et al. 2000; Woodward et al. 2004). CT is the most commonly performed study for the preoperative staging of a suspected ovarian carcinoma (Coakley 2002). The role of MRI in patients with known ovarian carcinoma is still evolving. Currently, the appropriate role of MR is characterization of ovarian masses rather than staging of histologically proven ovarian cancer. When MR is used for staging of ovarian carcinoma, imaging of the whole abdomen and pelvis should be performed using both T1-weighted and T2-weighted

993

994

7  Pelvis

Fig. 7.1.28a,b  Bilateral ovarian serous papillary adenocarcinoma. Axial (a) and coronal (b) gadolinium-enhanced T1weighted fat-suppressed images demonstrate bilateral cystic adnexal masses with enhancing solid components (T in a,b). Note

the avid enhancement of the thickened nodular peritoneum (arrows in b), best appreciated on the delayed post-contrast images (5–10 min post-injection). Findings are in keeping with bilateral ovarian carcinoma with peritoneal implants

Fig. 7.1.29a,b  Pseudomyxoma peritonei. Axial (a) and coronal (b) T2-weighted FSE images demonstrate multiple mucinous loculated collections of high signal intensity on T2-weighted

images (T in a,b). Note the scalloping of the liver and splenic surfaces and displacement of the bowel loops due to pressure effects. (Images are courtesy of Dr. Susan Ascher)

sequences in at least two planes. The coronal plane is useful in the evaluation of the liver surface and diaphragm. Contrast-enhanced fat-suppressed T1-weighted imaging is essential for optimal staging of ovarian cancer as it improves tumor delineation and increases the conspicuity of peritoneal or serosal deposits (Fig. 7.1.28). Intraperitoneal dissemination is the most common route of spread of ovarian carcinoma. Peritoneal im-

plants appear as nodular or plaque-like enhancing soft tissue masses of varying size. MRI is very sensitive (95%) for detection of peritoneal metastases, which show delayed enhancement on contrast-enhanced MRI (Ricke et al. 2003). Ascites is a non-specific finding but, in a patient with ovarian cancer, usually indicates peritoneal metastases (Hricak et al. 2000). Ascitic fluid may outline small implants, facilitating detection. Peritoneal im-

7.1  Female Pelvis Table 7.1.6  TNM and FIGO staging of primary ovarian cancer TNM

FIGO

Ovary

T1

I

Tumor confined to ovaries

T1a

Ia

Tumor confined to one ovary, capsule intact

T1b

Ib

Tumor confined to both ovaries, capsule intact

T1c

Ic

Tumor present on the surface of one or both ovaries; capsule ruptured; malignant ascites (positive peritoneal washings)

T2

II

Tumor involving one or both ovaries with pelvic extension

T2a

IIa

Extension and/metastasis to the uterus or tubes

T2b

IIb

Extension to other pelvic tissue

T2c

IIc

Tumor either stage IIa or IIb but present on surface of one or both ovaries; or capsule ruptured; or malignant ascites (positive peritoneal washings)

T3 and/or N1

III

Tumor involving one or both ovaries with peritoneal implants outside the pelvis or positive retroperitoneal or inguinal nodes

T3a

IIIa

Tumor limited to the true pelvis with negative lymph nodes but with histologically confirmed microscopic peritoneal metastases

T3b

IIIb

Peritoneal metastases, none exceeding 2 cm in diameter; negative lymph nodes

T3c and/or N1

IIIc

Abdominal implants greater than 2 cm in diameter or positive retroperitoneal or inguinal nodes

M1

IV

Distant metastases

plants may occur anywhere in the peritoneal cavity, but the most common sites include the pouch of Douglas, paracolic gutters, surface of the small and large bowel, greater omentum, surface of the liver (perihepatic implants), and subphrenic space (Figs. 7.1.23, 7.1.28). MRI is useful in differentiating between subcapsular liver implants and parenchymal liver metastasis, which alters staging and therapy. These implants are best seen on the delayed (5 min) contrast-enhanced fat-suppressed T1-weighted images. Surface implants are usually well defined, biconvex, and peripheral, and they indent the liver. True intraparenchymal metastases are often ill-defined, circular, and partially or completely surrounded by liver tissue. 7.1.5.3.3  Role of MRI in Treatment Selection and Follow-Up The imaging findings that are critical for the management of ovarian cancer may be divided into those related to characterization of the primary tumor, identification of metastatic disease, identification of disease that may be an indication for neoadjuvant chemotherapy, detection of cancer recurrence, and evaluation of the feasibility for

secondary cytoreduction. In each of these groups, imaging information can be utilized to guide better treatment selection. Lesion characterization by MRI assists in treatment planning and subspecialty referral as a malignant lesion will require surgical staging and referral to a gynecologic oncologic surgeon rather than a general gynecologist as in the case of benign lesions. Modern management of ovarian cancer is related to stage and extent of disease and includes primary surgery followed by adjuvant chemotherapy, neoadjuvant chemotherapy followed by interval debulking surgery or primary chemotherapy alone (stage IV disease). Optimal debulking refers to the reductions of all tumor sites to a maximal diameter of less than 1 cm. There is no agreement on the established surgical criteria that indicate inoperable disease. In considering all these issues, the role of imaging is not to describe disease as resectable or non-resectable, but rather to describe the findings that are important for surgical management in alerting the surgeon to the presence of extensive disease that indicates complex surgery or neoadjuvant chemotherapy. Surgically important features of the disease are: • Tumor deposits greater than 1 cm in size in the gastrosplenic ligament, lesser sac, fissure for the ligamentum

995

996

7  Pelvis

teres, porta hepatic, subphrenic space, small bowel mesentery, or retroperitoneal nodes above the renal hila • Invasion of pelvic side wall, rectum, sigmoid colon or urinary bladder • Differentiation of liver surface implants (peritoneal spread—stage III) from true intraparenchymal liver metastases (hematogenous spread—stage IV) MRI is also very useful in the follow-up of treatment response. Gadolinium-enhanced MRI is comparable (sensitivity 90% and specificity 88%) to laparotomy (sensitivity 88% and specificity 100%) but superior to serum carbohydrate antigen-125 (CA-125) (sensitivity 65% and specificity 88%) in the detection of residual or recurrent peritoneal and serosal implants in women who have been treated for ovarian carcinoma (Low et al. 2005; Rockall et al. 2003). MRI plays a crucial role in the detection of recurrent disease. It is important to realize that second look surgery is no longer routine and imaging diagnosis of recurrence may obviate a second look laparotomy since secondary cytoreduction is only justified if resection is possible with no residual tumor. Imaging findings that indicate nonresectable recurrent tumor are pelvic side wall invasion, which should be suspected when the primary tumor lies within 3 mm of the pelvic side wall or when the iliac vessels are surrounded or distorted by tumor. Bone invasion from the adjacent pelvic sidewall recurrence also constitutes non-resectable disease. 7.1.5.4  Ovarian Metastasis The appearance of ovarian metastasis is indistinguishable from that of a primary ovarian neoplasm. Metastases to the ovaries tend to be bilateral, and the oval shape of the ovary is preserved. They are usually solid and appear as high signal intensity heterogeneous masses on T2-weighted images. They may contain cystic areas with strongly enhancing walls after administration of gadolinium. Most common primary malignant tumors that metastasize to the ovary originate from stomach, colon, breast, lung, contralateral ovary, endometrium, melanoma, or pancreas. Krukenberg tumors are metastatic ovarian tumors that usually originate from the gastrointestinal tract, usually the stomach. 7.1.6  Miscellaneous Conditions 7.1.6.1  Chronic Pelvic Pain Chronic pelvic pain (CPP) has been described as noncyclic pelvic pain of greater than 6 months duration that is not relieved by strong analgesics. Radiological evalua-

tion of women with CPP often includes US and MRI. US (transvaginal and transabdominal) is considered the primary imaging modality in the evaluation of CPP, while MRI is reserved as a problem-solving tool (Kuligowska et al. 2005). Common causes of CPP are endometriosis, adenomyosis, pelvic inflammatory disease (PID) and pelvic congestion syndrome. Here, we focus on imaging of PID and pelvic congestion syndrome, as MRI appearances of adenomyosis and endometriosis are described in previous sections (7.3.2 and 7.4.2). 7.1.6.1.1  Pelvic Inflammatory Disease PID is defined as the acute clinical syndrome associated with ascending spread of microorganisms from the vagina or cervix to the endometrium, Fallopian tubes and/or contiguous structures. Chlamydia trachomatis and Neisseria gonorrhoeae are probably the most prevalent sexually transmitted bacteria in the western world. Early diagnosis of PID is of paramount importance in the management of women with acute upper genital tract infection, as delay will result in late treatment, with an attendant increased risk of complications (e.g., infertility). Classically, women with PID present with subacute lower abdominal pain that is dull in nature and usually bilateral. Other clinical features include fever, purulent vaginal discharge, bilateral adnexal tenderness with cervical excitation, and an elevated erythrocyte sedimentation rate. On MRI, an adnexal abscess is seen as a soft-tissue mass with central areas of low attenuation. Thick, irregular walls are commonly present and it may be difficult to differentiate an ovarian abscess from a necrotic tumor or endometrioma based solely on the MRI finding. MRI features of adnexal abscess are similarly nonspecific. Pyosalpinx and hydrosalpinx may present as a flask-shaped cystic adnexal mass that may contain faintly echogenic material due to debris. A large hydrosalpinx may be indistinguishable from an ovarian cyst. MRI may demonstrate a serpentine cluster of cyst. 7.1.6.1.2  Pelvic Congestion Syndrome Pelvic varices are dilated veins in the broad ligament and ovarian plexus. When symptomatic, the condition is called pelvic congestion syndrome. Symptoms have been described as dull, aching pain often occurring during walking, sexual intercourse, or other activities that create increased intra-abdominal pressure. Varices are most often found in multiparous women of reproductive age. The criteria for diagnosis includes venous structures greater than 4 mm in diameter, slower than 3-cm/s velocity flow and connecting arcuate vein within the myometrium. Threedimensional T1 gradient-echo sequences performed af-

7.1  Female Pelvis

ter the intravenous administration of gadolinium are the most effective MR imaging sequence for demonstrating pelvic varices. Blood flow in pelvic varices demonstrates high signal intensity (Kuligowska et al. 2005). References 1.

2.

3.

4. 5. 6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

Ascher SM, Jha RC, Reinhold C (2003) Benign myometrial conditions: leiomyomas and adenomyosis. Top Magn Reson Imaging 14:281–304 Brown HK, Stoll BS, Nicosia SV, Fiorica JV, Hambley PS, Clarke LP, Silbiger ML (1991) Uterine junctional zone: correlation between histologic findings and MR imaging. Radiology 179:409–413 Carrington BM, Hricak H, Nuruddin RN et al (1990) Mullerian duct anomalies: MR imaging evaluation. Radiology 176:715–720 Chang SD (2002) Imaging of the vagina and vulva. Radiol Clin North Am 40:637–658 Coakley FV (2002) Staging ovarian cancer: role of imaging. Radiol Clin North Am 40:609–636 Frei KA, Kinkel K, Bonel HM et al (2000) Prediction of deep myometrial invasion in patients with endometrial cancer: clinical utility of contrast-enhanced MR imaging-a meta-analysis and Bayesian analysis. Radiology 216:444–449 Fulcher AS, O’Sullivan SG, Segreti EM et al (1999) Recurrent cervical carcinoma: typical and atypical manifestations. Radiographics 19(Spec. no.):S103–116; quiz S264–255 Grasel RP, Outwater EK, Siegelman ES et al (2000) Endometrial polyps: MR imaging features and distinction from endometrial carcinoma. Radiology 214:47–52 Green CL, Angtuaco TL, Shah HR et al (1996) Gestational trophoblastic disease: a spectrum of radiologic diagnosis. Radiographics 16:1371–1384 Hawighorst H, Schönberg SO, Knapstein PG et al (1998) Staging of invasive cervical carcinoma and of pelvic lymph nodes by high resolution MRI with a phased-array coil in comparison with pathological findings. J Comput Assist Tomogr 22:75–81 Hricak H, Demas BE, Braga CA et al (1986) Gestational trophoblastic neoplasm of the uterus: MR assessment. Radiology 161:11–16 Hricak H, Stern JL, Fisher MR et al (1987) Endometrial carcinoma staging by MR imaging. Radiology 162:297–305 Hricak H, Chang YC, Cann CE et al (1990) Cervical incompetence: preliminary evaluation with MR imaging. Radiology 174:821–826 Hricak H, Rubinstein LV, Gherman GM et al (1991) MR imaging evaluation of endometrial carcinoma: results of an NCI cooperative study. Radiology 179:829–832 Hricak H, Powell CB, Yu KK et al (1996) Invasive cervical carcinoma: role of MR imaging in pretreatment work-up– cost minimization and diagnostic efficacy analysis. Radiology 198:403–409

16. Hricak H, Chen M, Coakley FV et al (2000) Complex adnexal masses: detection and characterization with MR imaging–multivariate analysis. Radiology 214:39–46 17. Hricak H, Gatsonis C, Chi DS et al (2005) Role of imaging in pretreatment evaluation of early invasive cervical cancer: results of the intergroup study ACRIN 6651/GOG 183. J Clin Oncol 23:9329–9336 18. Jemal A, Murray T, Ward E et al (2007) Cancer statistics, 2005. CA Cancer J Clin 57:43–46 19. Jeong YY, Outwater EK, Kang HK (2000) Imaging evaluation of ovarian masses. Radiographics 20:1445–1470 20. Jeong YY, Kang HK, Chung TW et al (2003) Uterine cervical carcinoma after therapy: CT and MR imaging findings. Radiographics 23:969–981; discussion 981 21. Kim SH, Kim HD, Song YS et al (1995) Detection of deep myometrial invasion in endometrial carcinoma: comparison of transvaginal ultrasound, CT, and MRI. J Comput Assist Tomogr 19:766–772 22. Kinkel K, Ariche M, Tardivon AA et al (1997) Differentiation between recurrent tumor and benign conditions after treatment of gynecologic pelvic carcinoma: value of dynamic contrast-enhanced subtraction MR imaging. Radiology 204:55–63 23. Kinkel K, Kaji Y, Yu KK et al (1999) Radiologic staging in patients with endometrial cancer: a meta-analysis. Radiology 212:711–718 24. Kuligowska E, Deeds L III, Lu K III (2005) Pelvic pain: overlooked and underdiagnosed gynecologic conditions. Radiographics 25:3–20 25. Low RN, Duggan B, Barone RM et al (2005) Treated ovarian cancer: MR imaging, laparotomy reassessment, and serum CA-125 values compared with clinical outcome at 1 year. Radiology 235:918–926 26. Manfredi R, Mirk P, Maresca G et al (2004) Local-regional staging of endometrial carcinoma: role of MR imaging in surgical planning. Radiology 231:372–378 27. Mayr NA, Hawighorst H, Yuh WT et al (1999) MR microcirculation assessment in cervical cancer: correlations with histomorphological tumor markers and clinical outcome. J Magn Reson Imaging 10:267–276 28. Murase E, Siegelman ES, Outwater EK et al (1999) Uterine leiomyomas: histopathologic features, MR imaging findings, differential diagnosis, and treatment. Radiographics 19:1179–1197 29. Nicolet V, Carignan L, Bourdon F et al (2000) MR imaging of cervical carcinoma: a practical staging approach. Radiographics 20:1539–1549 30. Okamoto Y, Tanaka YO, Nishida M et al (2003) MR imaging of the uterine cervix: imaging-pathologic correlation. Radiographics 23:425–445; quiz 534–535 31. Outwater EK, Siegelman ES, Hunt JL (2001) Ovarian teratomas: tumor types and imaging characteristics. Radiographics 21:475–490 32. Pattani SJ, Kier R, Deal R et al (1995) MRI of uterine leiomyosarcoma. Magn Reson Imaging 13:331–333

997

998

7  Pelvis 33. Reinhold C, Tafazoli F, Mehio A et al (1999) Uterine adenomyosis: endovaginal US and MR imaging features with histopathologic correlation. Radiographics 19(Spec no.): S147–S160 34. Reinhold C, Khalili I (2002) Postmenopausal bleeding: value of imaging. Radiol Clin North Am 40:527–562 35. Ricke J, Sehouli J, Hach C et al (2003) Prospective evaluation of contrast-enhanced MRI in the depiction of peritoneal spread in primary or recurrent ovarian cancer. Eur Radiol 13:943–949 36. Rockall AG, Meroni R, Sohaib SA, et al. Evaluation of endometrial carcinoma on magnetic resonance imaging. Int J Gynecol Cancer 2007; 17:188–196 37. Rockall AG, Sohaib SA, Harisinghani MG et al (2005) Diagnostic performance of nanoparticle-enhanced magnetic resonance imaging in the diagnosis of lymph node metastases in patients with endometrial and cervical cancer. J Clin Oncol 23:2813–2821 38. Sahdev A, Sohaib SA, Jacobs I et al (2001) MR imaging of uterine sarcomas. AJR Am J Roentgenol 177:1307–1311 39. Sala E, Wakely S, Senior E, Lomas DJ. Magnetic resoance imaging of malignant neoplasms of uterine corpus and cervix. Am J Roentgenol, 2007; 188:1577–1587 40. Scheidler J, Heuck AF (2002) Imaging of cancer of the cervix. Radiol Clin North Am 40:577–590, vii 41. Scheidler J, Heuck AF, Steinborn M et al (1998) Parametrial invasion in cervical carcinoma: evaluation of detection at MR imaging with fat suppression. Radiology 206:125–129 42. Shapeero LG, Hricak H (1989) Mixed mullerian sarcoma of the uterus: MR imaging findings. AJR Am J Roentgenol 153:317–319 43. Shibutani O, Joja I, Shiraiwa M et al (1999) Endometrial carcinoma: efficacy of thin-section oblique axial MR images for evaluating cervical invasion. Abdom Imaging 24:520–526 44. Shiraiwa M, Joja I, Asakawa T et al (1999) Cervical carcinoma: efficacy of thin-section oblique axial T2-weighted images for evaluating parametrial invasion. Abdom Imaging 24:514–519

45. Spies JB, Roth AR, Jha RC et al (2002) Leiomyomata treated with uterine artery embolization: factors associated with successful symptom and imaging outcome. Radiology 222:45–52 46. Sugimura K, Okizuka H (2002) Postsurgical pelvis: treatment follow-up. Radiol Clin North Am 40:659–680, viii 47. Takahashi S, Murakami T, Narumi Y et al (1998) Preoperative staging of endometrial carcinoma: diagnostic effect of T2-weighted fast spin-echo MR imaging. Radiology 206:539–547 48. Tamai K, Togashi K, Ito T et al (2005) MR imaging findings of adenomyosis: correlation with histopathologic features and diagnostic pitfalls. Radiographics 25:21–40 49. Tempany CM, Zou KH, Silverman SG et al (2000) Staging of advanced ovarian cancer: comparison of imaging modalities–report from the Radiological Diagnostic Oncology Group. Radiology 215:761–767 50. Woodward PJ, Sohaey R, Mezzetti TP Jr (2001) Endometriosis: radiologic-pathologic correlation. Radiographics 21:193–216; questionnaire 288–294 51. Woodward PJ, Hosseinzadeh K, Saenger JS (2004) From the archives of the AFIP: radiologic staging of ovarian carcinoma with pathologic correlation. Radiographics 24:225–246 52. Yamashita Y, Harada M, Sawada T et al (1993) Normal uterus and FIGO stage I endometrial carcinoma: dynamic gadolinium-enhanced MR imaging. Radiology 186:495–501 53. Yu KK, Hricak H, Subak LL et al (1998) Preoperative staging of cervical carcinoma: phased array coil fast spin-echo versus body coil spin-echo T2-weighted MR imaging. AJR Am J Roentgenol 171:707–711 54. Zahra MA, Hollingsworth, K, Sala E, Brenton J, Lomas DJ, Tan LT. Dynamic contrast enhanced magnetic resonance imaging as a preditor of tumor response to radiotherapy. Lancet Oncology, 2007;8:63–74

7.2  Male Pelvis

7.2  Male Pelvis 7.2.1  Urinary Bladder U.G. Müller-Lisse and U.L. Müller-Lisse 7.2.1.1  Introduction The diagnosis of diseases of the urinary bladder frequently involves imaging examinations. The least invasive and best available cross-sectional imaging modalities usually include ultrasonography (US) and computed tomography (CT). Depending on the clinical work environment, US may be performed by the urologist or by the radiologist. Since many diseases of the urinary bladder affect the urothelial layer, cystoscopy is oftentimes carried out by the urologist to detect, localize, and distinguish disorders of the urothelium. As an endoscopic examination, cystoscopy allows the urologist to gather samples from the bladder wall, either for cytology or for histo­path­ology. Morphologic imaging examinations of the urinary bladder by other means of cross-sectional imaging, i.e., computed tomography (CT) or magnetic resonance imaging (MRI), are usually performed when the disorder is complex and cannot be sufficiently diagnosed by US, cystoscopy, or conventional X-ray or fluoroscopy examinations. Examples include complex congenital anomaly, complex trauma, extensive inflammatory disease, invasive tumor, tumor staging, tumor recurrence, and complex postoperative changes. Bladder function can be studied by means of MRI. However, standard methods include clinical examination, ultrasound, uroflow, and urodynamic examination. 7.2.1.2  Examination Techniques 7.2.1.2.1  Patient Positioning and Preparation MRI examinations of the urinary bladder are usually performed with the patient in the supine position. For improved patient comfort, the knees and lower legs can be slightly elevated by means of a leg rest, and the head and neck supported by a neck rest. The advantage of applying head and leg rests is that the patient’s spine and pelvis are in a more comfortable and relaxed position, such that lower back pain is less likely to occur than when the patient lies flat on the scanner table. As a result, shifting of the pelvis in an effort to reduce discomfort is less likely, and image quality is likely to be better. External surface coils should be fixed in their positions by means of fixation straps that are usually supplied by the vendors of MRI systems. Fixing external surface coils in the region of the pelvis provides additional support for the patient and may help to decrease motion artifact.

Patient comfort may be increased by the application of blankets that both separate the external surface coil slightly from the body (which decreases sweating) and cover the patient’s pelvis and legs from sight and draft. Since organs of the male pelvis are approximately halfway between head and toes, the examiner has the choice of putting the patient into the MRI scanner head or feet first. However, from the patient’s point of view, it may be more comfortable to be examined feet first. As part of patient safety measures for the MRI examination, ear protection against MRI noises and an alarm bell should be offered to every patient. Preparation of the patient for an MRI examination of the urinary bladder includes bladder distension. Unless there is disease or disorder that has severely decreased bladder capacity (such as may be the case in status post radiotherapy of the urinary bladder), the urinary bladder of an adult man should contain at least 150 ml, and preferably 200–300 ml, of fluid for the MRI examination. Thus, voiding shortly before the scan should be avoided. To increase fluid volume within the urinary bladder, the patient may be asked to drink 0.5–1 l of clear fluid (such as water or tea) within 30 to 60 min prior to the examination. Alternatively, fluid may be administered intravenously or instilled directly into the bladder, the latter requiring adequate measures to provide sterility of bladder contents. In cases of extensive bladder cancer or other disease of the urinary bladder that is suspected of affecting the rectum or sigmoid colon, it may be useful to distend the rectum and sigmoid colon by means of water (for negative contrast on T1-weighted images) or water doted with small amounts of positive contrast media for MRI (for positive contrast on T1-weighted images). Application of butylscopolamide (Buscopan®) or glucagon has been recommended by some authors to decrease bowel motion during the examination; however, contraindications have to be observed. 7.2.1.2.2  Selection of Coils MRI examinations of the urinary bladder should be performed in whole-body high-field MR scanners. Recent, scientifically evaluated studies were mostly performed at a magnetic field strength of 1.5 T. While historically, the body coil was used as both transmitter and receiver, multi-channel phased array surface coils (PASC) with at least four independent elements have since been applied to the advantage of signal-to-noise ratio and lesion conspicuity. Scheidler et al. (1997) have demonstrated in pelvic MRI examinations that application of a fourchannel PASC improves signal-to-noise ratio by a factor of 2 to 2.5 when compared with the body coil at 1.5 T. For examinations of the urinary bladder, it is not necessary in most instances to use an endorectal surface coil (ERC).

999

1000

7  Pelvis

7.2.1.2.3 Examination Sequences MRI examinations of the urinary bladder have most often been performed at a magnetic field strength of 1.5 T. Early imaging protocols were based on T1-weighted and T2-weighted spin-echo (SE) sequences. Fields of view typically ranged from 28 to 36 cm, and matrix sizes from 128 × 256 to 256 × 256. The section thickness typically amounted to 0.8 cm (0.5–1.2 cm), and the intersection gap to 0.2 cm (0–1 cm) (Amendola et al. 1986; Barentsz et al. 1988; Bryan et al. 1987; Buy et al. 1988; Fisher et al. 1985; Husband et al. 1989; Koelbel et al. 1988: Küper et al. 1986; Rholl et al. 1987). The introduction of turbo or fast spin-echo (TSE or FSE) techniques has made T2-weighted MR imaging considerably faster, particularly at higher magnetic field strengths (Barentsz et al. 1996a; Kim et al. 1994). As a result, imaging time could be reduced, or the matrix or number of slices increased and slice section thickness reduced. Although the use of half-Fourier–acquired singleshot TSE or FSE sequences (HASTE or SSFSE) has been advocated in combination with intravesical administration of small particles of iron oxide (SPIOs) by Beyersdorff and co-workers (2000), there are no published data to compare the respective diagnostic accuracies of T2-weighted HASTE/SSFSE and TSE/FSE MR sequences in diseases of the urinary bladder. Spin-echo sequences with T1-weighting have been replaced by a three-dimensional magnetization-prepared rapid gradient-echo (MPRAGE) sequence by the group of Barentsz et al. (1994, 1996a, b). With a TR of 10 ms, a TE of 4 ms, an inversion time TI of 500 ms, a flip angle of 10°, an effective section thickness of 0.12 cm, two acquisitions, a 192 × 256 matrix, and a 30-cm field of view, the MPRAGE sequence allows for multiplanar image reconstruction and improved visualization of both tumors and perivesical lymph nodes (Barentsz et al. 1996b). Functional MRI of the urinary bladder and pelvic floor has been applied in pelvic floor insufficiency (Lienemann et al. 1997; Pannu 2003), patients with status post prostatectomy (Müller-Lisse et al. 2002), and in children with various congenital anomalies of bladder and pelvic floor morphology and function (Schneider et al. 2001). Functional MRI of the urinary bladder and pelvic floor has been based on rapid, T2-weighted MR sequences, such as HASTE and trueFISP. With the urinary bladder filled with urine or water, and the rectum and vagina filled with ultrasound gel, median sagittal MR images are obtained at intervals of 1 s. Changing strain on the urinary bladder and pelvic floor by alternating between contraction, pressing, and relaxation demonstrates motion and position of pelvic organs in different functional situations as well as evacuation of intraluminal contents of the pelvic organs (Lienemann et al. 1997; Müller-Lisse et al. 2002; Pannu 2003; Schneider et al. 2001; Fig. 7.2.1). MRI examinations of micturition can be performed with rapid, T1-weighted MR images with or without contrast

Fig. 7.2.1a,b  Functional MRI of the pelvic floor and urinary bladder demonstrates the positions of the bladder outlet, vaginal vault, and rectal ampulla (arrows) relative to the pubococcygeal line (white line); rapid, sagittal T2-weighted TrueFISP MR images obtained (a) at rest and (b) under abdominal straining

enhancement of urine (Nolte-Ernsting et al. 1998), or with rapid T2-weighted sequences, such as HASTE or True FISP (Müller-Lisse et al. 2002). 7.2.1.2.4  Imaging Planes In MRI examinations of the urinary bladder, T1-weighted images should be obtained in at least one plane (preferably, axial) before and two planes (axial and sagittal or coronal) after intravenous (i.v.) contrast administration. T2-weighted images should be obtained in at least two planes (axial and sagittal or coronal). Dynamic, contrastenhanced T1-weighted images should be obtained in the plane that allows the best visualization of the lesion under scrutiny (Barentsz et al. 1996b; Kim et al. 1994; Neuerburg et al. 1989; Tachibana et al. 1991; Tanimoto et al. 1992). 7.2.1.2.5  Thickness of Slices It is reasonable to cover the urinary bladder with slice thickness as narrow as reasonably possible (5–6 mm) and a narrow interslice gap (1–2 mm) while keeping TR as low as reasonably possible to improve the chances of MR imaging without motion artifact. Recommended FOV and imaging matrix ranges from FOV with an edge length of 32–40 cm with a matrix of maximum 192 × 256 image points per direction to FOV with an edge length of 30 cm and a matrix of 320 × 512 image points per direction (Tables 7.2.1, 7.2.2). 7.2.1.2.6  Preferred Coverage In MR examinations of the urinary bladder, preferred coverage extends from the aortic bifurcation to the pelvic floor, to cover both the primary region of interest, i.e., the urinary bladder and its adjacent organs and tissues, and

7.2  Male Pelvis Table 7.2.1  MRI of the urinary bladder: examination parameters for spin-echo (SE) sequences   (modified from Müller-Lisse et al. 1998) Author

Magnetic field strength

T1-weighted SE sequence

T2-weighted SE sequence

Fisher et al. 1985

Not reported

TR 0.5 s, TE 28 ms

TR 1–2 s, TE 28–56 ms

Amendola et al. 1986

0.35 T, body coil

TR 0.5 s, TE 28 ms

TR 1–2 s, TE 28–56 ms

Küper et al. 1986

1.5 T, body coil

TR 0.4–0.8 s, TE 30 ms

TR 0.8–3.2 s, TE 90–240 ms

Bryan et al. 1987

0.3 T and 1 T, body coil

TR 0.5 s, TE 30–35 ms

TR 1–2 s, TE 60–90 ms

Rholl et al. 1987

0.35–0.5 T, body coil

TR 0.5 s, TE 30–35 ms

TR 1.5–2.1 s, TE 90 ms

Barentsz et al. 1988

0.5 T, body coil double surface coil

TR 0.25–0.5 s, TE 30 ms

TR 2 s, TE 30/60–150 ms

Buy et al. 1988

0.5 T, body coil

TR 0.4 s, TE 28 ms

TR 1.6 s, TE 40/80/120 ms

Koelbel et al. 1988

1.5 T, body coil

TR 0.8 s, TE 30 ms

TR 2 s, TE 30–90 ms

Husband et al. 1989

1.5 T, body coil

TR 0.5 s, TE 17 ms

TR 2.1 s, TE 30–70 ms

Table 7.2.2  MRI of the urinary bladder: examination parameters for T2-weighted turbo or fast spin echo (FSE or TSE) sequences (modified from Müller-Lisse et al. 1998) Author

Kim et al. 1994

Barentsz et al. 1996b

Magnetic field strength

1.5 T, body phased array coil

1.5 T, Helmholtz surface coil

TR (s)

4–5

3

TE (ms)

102

90

Section thickness (cm)

0.5

0.8

Intersection gap (cm)

0.1

0.2

Slice orientation

Not specified

In plane of best tumor visibility

No. of acquisitions

4

3

Matrix

192 × 256

320 × 512

Field of view (cm)

32–40

30

the regional lymph nodes and skeleton, which are most frequently involved when cancer of the urinary bladder becomes metastatic. 7.2.1.2.7  Use of Contrast Medium For MRI of the urinary bladder, numerous investigators have included contrast-enhanced T1-weighted sequences in their imaging protocols. The dose of contrast medium applied has invariably been the standard dose of gadolinium chelates per kilogram of body weight (Barentsz et al. 1994, 1996b; Doringer et al. 1991; Kim et al. 1994;

Neuerburg et al. 1989; Sohn et al. 1990; Sparenberg et al. 1991; Tachibana et al. 1991; Tanimoto et al. 1992). Later reports have also included dynamic contrast-enhanced MRI (DCE MRI) studies with short TR/short TE sequences repetitively acquired after intravenous bolus injection of gadolinium chelates (Barentsz et al. 1996b; Kim et al. 1994; Neuerburg et al. 1989; Tachibana et al. 1991; Tanimoto et al. 1992; Table 7.2.3). The different approaches to DCE MRI show that this aspect of MRI of the bladder is still in its development. One of the most severe problems of DCE MRI lies in the tradeoff between the number of slices per scan and the repetition rate. The problem of missing the tumor has been addressed

1001

1002

7  Pelvis Table 7.2.3  MRI of the urinary bladder: examination parameters for dynamic, contrast-enhanced, MR Imaging   (modified from Müller-Lisse et al. 1998) Author

Neuerburg et al. 1989

Tachibana et al. 1991

Tanimoto et al. 1992

Kim et al. 1994

Barentsz et al. 1996b

Field strength (T)

1.5

1.5

1.5

1.5

1.5

Sequence

FLASH

FSE

SE

GRE

TurboFLASH

TR (ms)

50

100

100

40–130

7

TE (ms)

10

14

12

4.5–12 ms

3

Flip angle (°)

60

70–80

10

TI (ms)

15

No. of acquisitions

1–2

1

1

1–2

1–2

No. of slices

1

1

1–2

Not reported

1

Acquisition time (s)

13–22

14

6.7–14

20–40

1.25–2.5

Slice thickness (cm)

0.8

1

1

Not reported

0.8

Interslice gap

Not reported

Not reported

Orientation

Various

Various

Various

Not reported

Variousa

Matrix

256 × 256

Not reported

128 × 256

128 × 256

128 × 256

FOV (cm)

Not reported

Not reported

28–32

32–40

35

a

a

a

Perpendicular to tumor base FOV field of view, FLASH fast low-angle shot, FSE fast spin echo, GRE spoiled gradient echo a

by Barentsz et al. (1996b), who reported failure of tumor inclusion in the single slice depicted by their DCE MRI sequence in 2 of 49 cases. In fact, successful DCE MRI with a single- or dual-slice technique requires recognition of tumor location prior to slice selection. It is, therefore, likely to fail in cases of flat, superficial, or superficially invasive tumors, for example, in stages Tis or T1. In pre-biopsy MRI examinations, DCE MRI is likely to miss its target when focal edema or inflammatory reaction leads to local bladder wall thickening. In multifocal TCC of the urinary bladder, it may be difficult to include all foci recognized in preceding multislice sequences in one or two slices for DCE MRI. Also, in recurrent TCC of the urinary bladder, it may be difficult to select the site of recurrence among focal alterations in the bladder wall brought about by earlier therapeutic measures. 7.2.1.3  Normal Anatomy Based on MRI sequences as listed above and in Tables 7.2.1, 7.2.2, and 7.2.3, the normal bladder wall shows intermediate signal intensity on T1-weighted MR images (Fig. 7.2.2a). Urine has low signal intensity, and peri-

vesical fat tissue has high signal intensity. The different layers of the bladder wall cannot be differentiated on T1weighted SE images (Fig. 7.2.2a, c), and the bladder floor and trigone do not contrast sufficiently with the prostate (Fisher et al. 1985; Küper et al. 1986). On T2-weighted MR images, urine exhibits very bright signal that contrasts strongly with the low-signal-intensity bladder wall. The different layers of the wall of the urinary bladder are not distinguished by T2-weighted MR images, independent of the underlying type of sequence (i.e., SE, TSE/FSE, HASTE/SSFSE, or trueFISP sequences, Fig. 7.2.2b). Perivesical fat tissue is moderately bright, and the prostate has slightly higher signal intensity than does the bladder wall. The seminal vesicles are readily distinguished by their bright internal signal and the low signal intensity of their walls (Fisher et al. 1985; Küper et al. 1986) (Fig. 7.2.2b). T2-weighted MR images are prone to chemical shift artifact that may produce low-signal-intensity lines on one side of the bladder and high-signal-intensity lines on the other, due to differences in precession frequencies of water and fat. These lines may be misinterpreted as bladder wall pathology. In case of doubt, the direction of the readout gradient should be swapped and the sequence repeated (Heiken and Lee 1988).

7.2  Male Pelvis

7.2.1.4  Pathological Findings 7.2.1.4.1  Groups of Pathologic Conditions With few exceptions, MRI examinations of the bladder have been limited to tumor patients. Non-neoplastic bladder pathologies, such as uncomplicated inflammatory bladder disease, oftentimes do not require any imaging. However, MRI does have a role in complex congenital disorders of the urinary tract and pelvic floor. Congenital In complex congenital malformations of the pelvic floor and continence organs in children and young adults, such as anorectal malformations, bladder exstrophy, and cloacal exstrophy (Fig. 7.2.3), MRI plays a major role in assessing urinary and fecal incontinence or constipation. Both morphology and function of the pelvic floor and pelvic organs can be demonstrated, particularly when MRI includes a dynamic investigation, such as MRI defecography. The advantages of MRI include direct visualization of pelvic floor muscles and continence organs and their anatomical relationships during different functional actions, and avoidance of radiation. Disadvantages of MRI include associated costs and long investigation time (Boemers et al. 2006). Trauma Organs typically injured by blunt and penetrating trauma to the urinary tract include the kidneys, bladder, and male urethra, while the ureter is hardly ever involved. The assessment of genitourinary tract trauma by means of ultrasound, CT and MRI, including contrast-enhanced magnetic resonance angiography, has recently gained tremendous clinical significance and determines decision making with regard to conservative and surgical management (Obenauer et al. 2006). Although it takes major disruptive forces to fracture the pelvic ring, pelvic injury is an important issue in trauma patients. In the initial detection of fractures, radiography remains important. However, more sophisticated techniques, such as CT (including 3D reconstruction) and MRI, are useful for operative planning. MRI helps to detect occult pelvic fractures, especially in elderly patients. Other means of imaging in pelvic trauma include cystography, which can be used to detect bladder ruptures, and angiography, which detects bleeding sites and guides therapeutic embolization (Geusens et al. 2000). Inflammatory A variety of rare, non-neoplastic disorders of the urinary bladder can cause either focal masses or diffuse mural thickening and mimic malignancy. In turn, findings at MRI or CT of focal masses or diffuse mural thickening of the urinary bladder are suggestive of, but not specific for malignant bladder tumor.

Some of the non-neoplastic entities are poorly understood, such as inflammatory pseudotumor, which produces ulcerated, bleeding polypoid masses in the urinary bladder. These masses may be large and have an extravesical component (Wong-You-Cheong et al. 2006) and may occasionally be found in children, as a differential diagnosis to rhabdomyosarcoma (Schneider et al. 2001). Inflammatory pseudotumor of the urinary bladder may show at MRI with intermediate signal on T1-weighted images, prominent, high signal intensity on T2-weighted images, and marked enhancement after intravenous administration of contrast media (Fujiwara et al. 1999) (Fig. 7.2.4). Imaging features of malacoplakia are non-specific; however, characteristic Michaelis-Gutmann bodies are found at pathologic evaluation. The various types of cystitis of the urinary bladder (cystitis cystica, cystitis glandularis, and eosinophilic cystitis) require pathologic diagnosis for differentiation. Bladder infection with tuberculosis and schistosomiasis produces non-specific bladder wall thickening and ulceration in the acute phase and should be suspected in patients who are immunocompromised or have lived in countries where these infections are common. Cystitis of the urinary bladder may also result from chemotherapy or radiation therapy. While the cause should be clinically evident, imaging may be used to determine severity and to assess complications. Extrinsic inflammatory disease, such as Crohn’s disease or colonic diverticulitis, may be associated with fistulas to the bladder and subsequent focal bladder wall abnormality. The extravesical findings allow the diagnosis to be made easily (Won-You-Cheong et al. 2006). Neoplastic Bladder tumors may be benign or malignant, and derive from epithelial or mesenchymal cells. Primary tumors of the bladder have to be distinguished from secondary tumors that either infiltrate the bladder from the outside or represent metastases of other, distant tumors. Among the latter, tumors infiltrating the bladder include malignancies of the female reproductive organs, the prostate, or the colon and rectum. Metastases often derive from carcinoma of the stomach, melanoma, breast cancer, or lung cancer. Primary epithelial tumors include papilloma, transitional cell carcinoma (urothelial carcinoma, TCC), squamous cell carcinoma, adenocarcinoma, and undifferentiated carcinoma. Mesenchymal primaries include benign fibroma, myxoma, leiomyoma, hemangioma, neurofibroma, neurinoma, pheochromocytoma, and other rare benign tumors as well as malignant leiomyosarcoma, fibrosarcoma, osteochondrosarcoma, rhabdomyosarcoma, and reticuloendothelial tumors. The majority of bladder tumors are of epithelial (urothelial) origin. Normal urothelium in the bladder consists of three to seven layers of transitional cells, mainly of elongated shape with excentric nuclei. Most cells are in

1003

1004

7  Pelvis

Fig. 7.2.2  a MRI of normal urinary bladder on T1-weighted images without intravenous contrast administration (GRE or FLASH MR images) demonstrates wall of urinary bladder with intermediate signal intensity (arrows in A and B) that resembles skeletal muscle and urine with low signal intensity (asterisks in A and B). Bladder wall around the ostia of ureters and at the

7.2  Male Pelvis

trigone of the urinary bladder may show increased width with smooth contours (double-lined arrows in B). Note that width of the normal bladder wall decreases with increasing volume of urine within the urinary bladder (A and B). b MRI of normal urinary bladder on T2-weighted images (A TSE, B and C TSE with fat signal suppression, D and E HASTE, F trueFISP) demonstrates wall of urinary bladder with low signal intensity (arrows) that resembles skeletal muscle and urine with high signal intensity (asterisks). Bladder wall around the ostia of ureters and at the trigone of the urinary bladder may show increased width with smooth contours (double-lined arrows in A). Vaginal wall demonstrates slightly higher signal intensity than bladder wall on T2-weighted TSE images (curved arrow in A). Bladder wall shows with slightly higher signal intensity than fat tissue on T2-weighted TSE images with fat signal suppression (B and C). TSE MR images are susceptible to artifact from bowel motion (C), while HASTE MR images are prone to artifact from inflow of urine into the bladder (E). TrueFISP MR images show decreased soft tissue contrast when compared with other T2weighted MR images (F). c MRI of normal urinary bladder on T1-weighted images after intravenous contrast administration (GRE or FLASH MR images with fat signal suppression) demonstrates wall of urinary bladder with intermediate to high signal intensity (arrows in A and B) that resembles skeletal muscle and urine with low signal intensity (asterisks in A and B). Of note, while T1-weighted MR images obtained immediately after intravenous administration of contrast media clearly delineate the bladder wall from urine (A and B), excretion of contrastenhanced urine severely disturbs interpretation of T1-weighted MR images obtained at later time points (note mixing of contrast-enhanced urine from ureters with remnant urine in the bladder [curved arrow in C])

7 Fig. 7.2.3a–d  Complex congenital anomaly of the pelvic floor and urinary bladder. T2-weighted axial HASTE (A and B) and coronal (C) and sagittal (D) TSE MR images demonstrate cloaca with common outlet of urinary bladder, vagina, and rectum (straight arrows), rudimentary levator ani muscles (double-lined arrows), and isolated right uterine horn (arrowhead). Normal kidneys are identified bilaterally (curved arrows). Note complex anomaly of lower lumbar spine and bony pelvis

1005

1006

7  Pelvis

Fig. 7.2.4a,b  Inflammatory pseudotumor of the urinary bladder (arrows) produces ulcerated, bleeding, polypoid masses, which may show at MRI with intermediate signal on T1-weighted images, intermediate to high signal intensity on T2-weighted im-

ages, and marked enhancement after intravenous administration of contrast media. a axial, T2-weighted TSE MR image, b axial, T1-weighted gradient echo (FLASH) MR image obtained 50 s after intravenous administration of contrast media

Fig. 7.2.5a–d  Cervical cancer invading the urinary bladder demonstrates as a mass that extends from the uterine cervix and continues to the posterior and left walls of the urinary bladder (arrows). Axial, T1-weighted SE (a) and T2-weighted TSE (b)

MR images. Axial (c) and sagittal (d) T1-weighted SE MR images with fat suppression obtained after intravenous administration of contrast media

7.2  Male Pelvis

contact with the basal membrane. Structural and cellular alterations of the urothelium are described as hyperplasia, metaplasia, dysplasia, and carcinoma. Epithelial hyperplasia consists of an increased number of transitional cell layers that are regularly arrayed. Squamous cell metaplasia describes a situation where urothelium is replaced by non-keratinized squamous cell epithelium. Metaplasia in the presence of keratinization and cellular atypia that may extend into deeper layers of the bladder wall is described as leukoplakia. Leukoplakia is considered to represent a pre-cancerous state, with a 20% chance of developing into bladder cancer (Benson, Jr., et al. 1983). Adenocarcinoma of the bladder represents less than 2% of all urinary bladder cancers and may derive from remnants of the urachus, primary adenomatous cells anywhere in the urinary bladder, or secondary metastatic clusters of adenocarcinoma cells (Jocham 1994). Squamous cell epithelial bladder cancer is a rare type of urothelium-derived carcinoma often associated with Schistosoma haematobium and squamous cell metaplasia (Jocham 1994). Non-Hodgkin’s lymphoma rarely presents as primary extranodal disease of the urinary bladder wall. Bladder lymphoma may appear as a large, lobulated submucosal mass. On the basis of MRI signal characteristics or contrast enhancement patterns, however, lymphoma of the bladder wall cannot be differentiated from TCC (Yeoman et al. 1991). Tumorous invasion of the urinary bladder wall by other pelvic malignancies (Figs. 7.2.5, 7.2.6) lends itself to diagnostic MRI investigation due to the high soft tissue contrast and free choice of imaging plane that MRI offers. Various studies show the value of MRI in staging extensive gynecologic cancers (Houvenaeghel et al. 1993; Iwamoto et al. 1994; Kaji et al. 1994; Semelka et al. 1993). However, while MRI staging is correct in over 80% of cases, some authors prefer MRI to CT but find endoscopic ultrasound a reliable alternative (Houvenaeghel et al. 1993; Iwamoto et al. 1994). Benign Neurofibromatosis (von Recklinghausen disease) is a congenital, hereditary dysplasia that affects all three germ cell layers and may involve any organ system in the body. Involvement of the urogenital system is rare. However, in these cases, it often affects the bladder. Neurofibromas usually show signal intensities similar to that of skeletal muscle on T1-weighted images but present with markedly increased signal intensity on T2-weighted images and with strong enhancement following intravenous administration of gadolinium-containing contrast media. Bladder wall involvement may be nodular or diffuse, and is accompanied by pelvic sidewall and adjacent soft tissue alterations (Shonnard et al. 1992). Leiomyomas are seldom found in the retrovesical pouch. A leiomyoma may indent the posterior bladder wall as an extravesical tumorous lesion. Leiomyomas

Fig. 7.2.6a–c  Non-Hodgkin’s lymphoma of the pelvis, involving the uterine cervix, vagina, parametria, anterior rectal wall, and urinary bladder (arrows), appears with intermediate signal intensity on T2-weighted MR images, and marked enhancement after intravenous administration of contrast media. a sagittal, T2-weighted TSE MR image, b sagittal, T1-weighted gradient echo (FLASH) MR image, obtained after intravenous administration of contrast media, and c axial, T2-weighted TSE MR image with fat saturation

are characterized by their low signal-intensity both on T1- and T2-weighted MR images, although they may be inhomogeneous on T2-weighted images when hyaline, myxomatous, or fatty degeneration has occurred (Thurnher et al. 1992). Among various other tumor-like lesions of the urinary bladder, differential diagnosis is histopathological rather than macroscopic. While some of the tumor-like lesions are benign, others represent precursors of malignant disease. Tumorous lesions considered benign include von Brunn’s cell nests, cystitis cystica, and cystitis follicularis. As a differential diagnosis of urothelial metaplasia, von Brunn’s cell nests represent inclusions of normal-looking urothelium into deeper layers of the bladder wall.

1007

1008

7  Pelvis

A similar lesion is found in cystitis cystica, whose urothelial inclusions in the bladder wall surround a center of amorphic, eosinophilic matter. In cystitis follicularis, lymphatic follicles develop in the submucosal layer, as a reaction to chronic inflammatory disorder (Jocham 1994). The various types of cystitis of the urinary bladder (cystitis cystica, cystitis glandularis, and eosinophilic cystitis) require pathologic diagnosis for differentiation (Wong-You-Cheong et al. 2006) and cannot be distinguished with certainty by means of cross-sectional imaging. Endometriosis of the urinary bladder manifests as submucosal masses with characteristic MRI features that consist of hemorrhagic foci and reactive fibrosis. Nephrogenic adenoma has no typical features, and pathologic evaluation is required for diagnosis (WongYou-Cheong et al. 2006). Chronic granulomatous pseudotumor and pseudosarcomatous myofibroblastic (fibromyxoid) tumor represent rare entities that are differential diagnoses to rhabdomyosarcoma of the urinary bladder in children (Schneider et al. 2001). Tumorous lesions considered precursors of bladder cancer include cystitis glandularis, inverted papilloma,

Malignant TCC of the bladder may present with T1 and T2 relaxation times so similar to those of uninvolved bladder wall that differentiation of TCC and healthy surrounding tissue is very difficult on the basis of signal intensity both in T1and T2-weighted SE images (Koelbel et al. 1988). However, TCC may show signal intensity higher than that of unaffected bladder wall on T2-weighted images (Kim et al. 1994). If the echo time TE does not exceed 100 ms in SE sequences, the signal intensity of TCC on T2-weighted images is usually higher than that of normal bladder wall. In the presence of fibrosis, however, the signal intensity

Fig. 7.2.7a–d  Transitional cell carcinoma (urothelial carcinoma), stage T3b, of the urinary bladder (arrows) produces polypoid masses, which may show at MRI with intermediate to high signal intensity on T2-weighted images, and marked enhancement

after intravenous administration of contrast media. a axial, and b coronal T2-weighted TSE MR images. c and d axial T1-weighted gradient echo (FLASH) MR images, obtained 20 and 50 s after intravenous administration of contrast media, respectively

and urothelial dysplasia. In combination with metaplastic changes, urothelial inclusions into the bladder wall are signs of cystitis glandularis, which is considered to represent a precursor to TCC. Inverted papilloma prefers invasion into the fibromuscular stroma and also represents a precancerous lesion. Urothelial dysplasia includes various degrees of nuclear alterations in the absence of an increased number of mitoses or cell layers. It is often difficult to distinguish between urothelial dysplasia and transitional cell (urothelial) carcinoma (TCC) in situ (Jocham 1994).

7.2  Male Pelvis

of TCC may be as low as that of normal bladder wall (Barentsz et al. 1990). Among the types of TCC of the bladder, papillary exophytic tumors are the most common, accounting for about 70% of cases; solid, infiltrating or nodular types account for about 10% of cases and mixed types for about 20% of cases (Jocham 1994; Figs. 7.2.7, 7.2.8). Cell morphology and differentiation within the tumor generally is described as good (G1), intermediate (G2), or poor (G3). With the exception of Tis tumors that are defined by their poor differentiation, tumor size and depth of invasion correlate with both the degree of cell differentiation and the risk of progression (Jocham 1994). As a tumor stage of its own kind, the in situ TCC (Tis) has been defined as an anaplastic (grade III) urothelial tumor without exophytic, papillary protrusion into the bladder lumen or infiltration beyond the basal membrane (“flat tumor”). Loss of the usual array of epithelial cell layers distinguishes carcinoma in situ from urothelial dysplasia. While primary Tis is relatively rare (accounting for 5% of all superficial bladder tumors), its importance lies both in its extreme aggressiveness (almost 80% risk of progression; 38–83% risk of developing into a muscle-infiltrating tumor) (Althausen et al. 1976; Jakse et al. 1989)

and in the high coincidence rate of secondary carcinoma in situ with exophytic, papillary or solid urothelial carcinoma (20–75%) (Jocham, 1994). Also, Tis is frequently found in recurrent or primary multifocal bladder cancer. Bladder tumors require thorough staging that goes beyond the confines of the urinary tract lumina. While CT has been the first cross-sectional imaging method among the non-invasive staging tools, MRI has established a role in the evaluation of the more advanced stages of invasive bladder cancer (stages T2 and higher; Figs. 7.2.7, 7.2.8) (Patel and Hricak 1995). The TNM classification of bladder cancers includes only carcinomas, while papillomas and other tumors are excluded. Local tumor staging (T staging) is guided by clinical examination, various imaging methods, endoscopy, and biopsy. Lymph node staging (N staging) includes clinical examination and imaging methods. The earliest staging system was developed in the 1940s (Jewett and Strong 1946; Marshall 1952) and integrates local tumor extent, lymph node infestation, and metastatic extension under one common staging symbol. The TNM system of the Union Internationale Contre le Cancer (UICC) stages local tumor extent (T stage), lymph node involvement (N stage), and metastasis (M stage) indi-

Fig. 7.2.8a–d  Transitional cell carcinoma (urothelial carcinoma), stage T4a, of the urinary bladder with invasion of the left seminal vesicle and the prostate (arrows) produces a polypoid mass with intermediate signal intensity on both T1- and T2-weighted MR images and shows marked enhancement after intravenous

administration of contrast media. a axial T1-weighted SE MR image, b axial, T1-weighted gradient echo (FLASH) MR images with fat saturation, obtained after intravenous administration of contrast media, c axial T2-weighted TSE MR image, and d coronal T2-weighted TSE MR image with fat saturation

1009

1010

7  Pelvis Table 7.2.4  Comparison of the Jewett-Strong-Marshall system and the TNM system of bladder cancer staging, according to the 6th edition of the UICC in 2002 (modified and amended, from Jewett 1973, Jewett and Strong 1946; Marshall 1952; Müller-Lisse et al. 1998, 2004b; Wittekind et al. 2002) Jewett–Strong– Marshall stages

TNM stages

Staging criteria

No stage

Tx T0

Tumor cannot be evaluated No evidence of tumor

0

Ta

Non-invasive, superficial tumor, confined to mucosal epithelium

0

Tis

Carcinoma in situ, non-invasive, “flat tumor”

A

T1

Invasive tumor, confined to subepithelial connective tissue layer (lamina propria)

B1

T2a

Invasive tumor, confined to inner half of muscle layer

B2

T2b

Invasive tumor, confined to outer half of muscle layer

C

T3a T3b

Microscopic extravesical tumor growth, confined to perivesical fat (no macroscopic imaging criteria) Macroscopic extravesical tumor growth, confined to perivesical fat (macroscopic, extravesical mass)

D1

T4a

Extravesical tumor growth, invading other pelvic organs (prostate, seminal vesicles, uterus, or vagina)

D1

T4b

Extravesical tumor growth, invading pelvic or abdominal wall

D1

N1

Solitary regional lymph node metastasis, less than 2 cm in largest extent

D1

N2

Solitary regional lymph node metastasis, with lymph nodes more than 2 cm, but less than 5 cm in largest extent, or multiple lymph node metastases with lymph nodes of no more than 5 cm in largest extent

D1

N3

Lymph node metastasis, with lymph nodes more than 5 cm in largest extent

D2

M1

Juxtaregional (common iliac, inguinal, or aortic) lymph node metastases

D2

M1

Distant organ metastases

vidually. Table 7.2.4 lists both staging systems and their staging criteria. Much of the accuracy of MRI in the staging of TCC of the urinary bladder will depend on the individual experience of the radiologist with bladder cancer as a pathological and a radiological entity. However, careful selection among the ever-increasing spectrum of MRI sequences and staging criteria as well as regard to the biopsy status of the respective urinary bladder are likely to determine the clinical value of MRI. The classical staging criteria for unenhanced MR images refer to Spin Echo (SE) sequences with a matrix size of 128 × 256 or 256 × 256 and a field of view of 32 × 32 cm to 36 × 36 cm, acquired with the body coil at field strengths ranging from 0.35–1.5 T (Fisher et al. 1985; Küper et al. 1986). While criteria for the absence of tumor (T0) or the presence of low-grade superficial tumors (Ta) are not included, Fisher et al. (1985) list in-situ (Tis) TCC as “too small for current resolution.”

According to Fisher et al. (1985), invasive tumor is best recognized as a high-signal-intensity defect in the bladder wall on T2-weighted MR images that either leaves an undisrupted layer of low signal intensity, representing the outer bladder wall and marking stage T1 or T2a bladder cancer, or involves most of its width, marking T2b bladder cancer, according to the TNM classification of 2002 (Müller-Lisse et al. 2004a; Wittekind et al. 2002). Complete disruption of the low signal intensity line of the bladder wall on T2-weighted MR images or abnormal tissue external to the bladder wall in perivesical fat are hallmarks of stage T3b (Fig. 7.2.7) or higher of TCC of the urinary bladder. Extension of abnormal tissue to other pelvic organs (e.g. seminal vesicles, prostate, vagina, and rectum) is suggestive of T4a disease (Fig. 7.2.8). Stage T4b bladder cancer is marked by the extension of abnormal tissue to distant organs or to the pelvic sidewalls (Fisher et al. 1985; Müller-Lisse et al. 2004b). While the criteria

7.2  Male Pelvis

marking stages T3b and above are based on findings on both T1- and T2-weighted images, recognition of tumor confined to the bladder greatly relies on the presence of a high-signal-intensity defect on a T2-weighted image. Possible reasons for staging error in the lower stages of invasive TCC of the urinary bladder include generation of a high-signal-intensity lesion by entities other than tumor (e.g. inflammation and edema alone) in the presence of tumor or after tumor biopsy or by motion artifact or by chemical shift artifact, and erroneous assumption of another depth of infiltration due to the high or low filling state of the bladder and depending wall thickness. It is not possible to reliably differentiate between TCC and surrounding inflammatory reaction or edema of the bladder wall with any of the unenhanced sequences that have been applied in MRI of the bladder (Fisher et al. 1985; Heiken and Lee 1988; Hricak et al. 1983; Küper et al. 1986; Rholl et al. 1987). MRI-assisted staging in cases of tumor absence (TNM 2002 stage T0), either before or after tumor resection at cystoscopy, has received relatively little attention. One early study (Sohn et al. 1990) reports seven patients without evidence of bladder cancer who underwent contrastenhanced MRI of the bladder. While MRI staging of T0 carcinoma was correct in six of seven cases, the authors do not list the criteria they used to rule out cancer on MR images. The other studies (Barentsz et al. 1994, 1996b) include only post-biopsy patients. Their staging criteria include the presence of a biopsy defect in the bladder wall and the absence of tumor signs. Without contrast enhancement, the absence of tumor was correctly diagnosed in only 4 of 12 patients, while the other 8 were being over-staged (stages Ta–T2, according to TNM 2002 staging criteria). However, with DCE MRI of the urinary bladder, 11 of 12 cases were correctly staged as being free from tumor, the criteria being absence of bladder wall enhancement in the suspicious area within 10 s of arterial peak enhancement (Barentsz et al. 1996b). Results indicate that staging TCC of the urinary bladder as T0 or absent is virtually impossible without contrast enhancement, and in patients who have undergone cystoscopy and biopsy prior to MRI. Bladder cancers whose most deeply growing parts do not infiltrate the muscle layer of the bladder wall are called superficial. The superficial bladder cancers include TNM 2002 stages Ta, Tis, and T1. Together, superficial TCC comprise about 80% of all bladder cancers (Jewett 1973; Jocham 1994). In TNM 2002 stages Ta, Tis, and T1, TCC of the urinary bladder in principle can be cured with local therapy, and the urinary bladder can be preserved. Therapy regimens include electro coagulation, laser therapy, thermal or photodynamic tumor destruction, topical chemotherapy or immunotherapy, or radiation therapy, either alone or in combination with the most frequently applied diagnostic and therapeutic modality, transurethral electroresection of the bladder (TUR-B).

However, 5-year survival rate, the rate of metastasization, and rate of progression differ among the stages. In stage Ta tumors, about 95% of patients survive over a follow-up period of 5 years. Metastases occur in 0.7%, and tumor progression is found in 4.4% of these patients (Jocham 1994). Five-year survival is equal among patients with and without Ta-tumor recurrence (Jocham 1994). For an imaging method applied in the staging of bladder cancer, it is therefore important to distinguish between superficial and muscle-infiltrating urothelial carcinoma, and to be as accurate as possible in the differentiation of the three stages of superficial tumor. MRI staging of TNM 2002 stage Ta and Tis tumors is difficult because the tumors may be small and flat and may not differ much from metaplastic or dysplastic urothelium. Thus, particularly flat tumors may escape detection by MRI, which is currently based on focal bladder wall thickening or thinning and altered contrast enhancement. The only MRI series to include more than 10 Ta/ Tis tumors under-staged 7 of 12 tumors as T0 and overstaged 2 of 12, while only 3 of 12 were correctly identified (Sohn et al. 1990). However, the criteria applied to stage the tumors have not been listed in this study (Sohn et al. 1990). Fisher et al. (Fisher et al. 1985) have suggested that TCC with TNM stage Ta/Tis was too small for the spatial resolution of MRI at the time of their study. However, there is yet no study to show that even current MRI methods are capable of correctly staging Ta/Tis bladder cancer. TNM stage T1 TCC is most difficult for surgical decision-making. There is a vast prognostic difference between highly differentiated (histologic grades G1 and G2; 5-year survival rate approximately 81%, metastasization rate approximately 14%, progression rate approximately 19% after transurethral electroresection) and poorly differentiated or anaplastic (histologic grades G3 and G4; 5-year survival rate approximately 64%, metastasization rate approximately 22%, progression rate approximately 31% after transurethral electroresection) stage T1 tumors (Jocham 1994). While TUR-B is likely to be the adequate therapy mode to treat the better-differentiated T1 bladder cancers, cystectomy will in many instances be recommended for poorly differentiated T1 tumors. The importance of differentiating between superficial Ta/Tis tumor, superficially invasive T1 tumor, and muscle-infiltrating tumor that is unlikely to be cured by electroresection alone has not yet been addressed by many MRI studies. Only one series includes more than 10 patients with stage T1 bladder cancers. In this series, unenhanced MRI alone failed to contribute significantly to the correct staging of Ta–T1 carcinoma of the bladder. With contrast enhancement, about 70% of Ta–T1 tumors were correctly staged by MRI (Sparenberg et al. 1991). The treatment of choice in muscle-infiltrating TCC that does not invade extravesical tissues is radical cystectomy. Cystectomy shows a high local efficacy, with a local

1011

1012

7  Pelvis

tumor recurrence rate of 10–20% (Jocham 1994). In principle, invasive tumors of stages pT2a and pT2b can also be cured by transurethral electro-resection. Up to 45% of solitary TNM 2002 stage pT2a tumors can be transurethrally resected with tumor-free margins, while another 30% can be fully resected in up to three follow-up operations. However, in TNM 2002 stage pT2b tumors, more than 60% cannot be completely resected by TUR-B (Barens et al. 1977; Henry et al. 1988; Herr 1987; O’Flynn et al. 1975; Solsona et al. 1992). Partial resection of the bladder in open surgery is restricted to tumors in bladder diverticula, tumors at the vertex of the bladder (carcinoma of the urachus), circumscribed lesions in the vicinity of a ureteral orifice, and patients with superficial tumors who cannot undergo transurethral resection, e.g., due to severe coxarthritis (Jocham 1994). Five-year survival in TNM 2002 pT2a tumors averages 58% after TUR-B, and 59% after radical cystectomy. In TNM 2002 pT2b tumors, the respective rates are 30 and 30% (Raghavan et al. 1990). The different therapy options and prognoses for muscle-infiltrating bladder cancers make it desirable to locate exactly the cancerous lesion and to distinguish TNM 2002 tumor stages pT2a and pT2b both from superficial and from extravesical TCC. Also, exact differentiation of TNM 2002 stages pT2a and pT2b can be crucial for determining prognosis and, therefore, for decisions on additional follow-up examinations and adjuvant therapy, such as chemotherapy, or radiation therapy. In the largest series published, 8 of 9 patients (89%) with TNM 2002 stage T2a bladder cancer were correctly staged with MRI when dynamic contrast-enhanced images obtained prior to biopsy were evaluated (Tanimoto et al. 1992). Various authors have published series of eight or more patients with TNM 2002 stage T2b TCC. Independent of both the MRI technique applied (unenhanced, contrast-enhanced, or dynamic, contrast-enhanced imaging) and the biopsy status of the patient’s bladder, the staging accuracy varied between 64% and 100%. Two of the studies (Kim et al. 1994; Tanimoto et al. 1992) included comparison with CT staging results. In these studies, the accuracy of CT in the staging of TNM 2002 stage T2b TCC was 43–50%. Similar accuracies were reached with unenhanced or delayed postcontrast MRI sequences (Kim et al. 1994; Tanimoto et al. 1992). Urological centers with an aggressive surgical approach to TCC of the urinary bladder even in superficially infiltrating (T1) tumors may benefit from cumulative evaluation of imaging results for TNM 2002 stages T1 to T2b, because TCC in those stages in theory can be cured by cystectomy alone. With the exception of one group of three patients whose bladder cancers were all incorrectly staged by MRI (Bryan et al. 1987), results in groups of more than 10 patients show an average 90% cumulative

staging accuracy of MRI for TNM 2002 stages T1 to T2b. Rholl et al (1987) examined 10 of 14 patients with TCC of TNM 2002 stages T1 to T2b with both MRI and CT. In their study, CT correctly staged 9 of 10 of these tumors. In view of the morbidity associated with locally progressive, infiltrating TCC, which includes recurrent hematuria, decreased bladder capacity, intestinovesical fistula, and formation of a cloaca, palliative cystectomy is often considered. Since higher tumor stages bear a higher risk of local and distant metastasis, adjuvant chemotherapy with schemes including methotrexate and cisplatin frequently follows radical cystectomy in stage T3b and T4a tumors, particularly when lymph node metastasis is found at surgery. Sometimes, inductive or neoadjuvant chemotherapy is used for tumor reduction (down-staging) or to sterilize occult micrometastasis prior to radical cystectomy or TUR-B (Jocham 1994). The combination of cystectomy with radiation therapy has also been discussed (Whitmore et al. 1977). MRI detects perivesical fat tissue infiltration with TCC in 74–100% of patients with TNM 2002 stage pT3b bladder cancer when larger patient groups are examined. Although, oftentimes, TNM 2002 stage T3b TCC is correctly diagnosed with unenhanced MRI, contrastenhancement and dynamic imaging tend to improve accuracy (Barentsz et al. 1996b; Tanimoto et al. 1992). The accuracy of CT in the staging of TNM 2002 stage T3b TCC in studies also reporting MRI results varies between 0% (0 of 2 patients, Kim et al. 1994) and 100% (5 of 5 patients, Rholl et al. 1987; 10 of 10 patients, Husband et al. 1989). Infiltration into neighboring organs (TNM 2002 stage T4a) and into the pelvic or abdominal wall (TNM 2002 stage T4b) is the hallmark of TCC tumor stages that most often cannot be treated with a curative intention. The 5-year survival rate of stage T4 tumors lies between 0 and 29% with cystectomy, and is greater than 10% with definitive radiation therapy (60 Gy) (Raghavan et al. 1990). On the one hand, survival is linked with the aggressiveness of treatment in each individual case. Tumors extending into the prostate in men can be fully excised with radical cysto-prostato-vesiculectomy as long as they are organ-confined. Partial or complete exenteration of the small pelvis is another surgical means to treat locally extensive disease that does not infiltrate the pelvic sidewalls. In stage T4b tumors, however, surgery oftentimes is not radical, since tumor tissue can remain after sharp tumor detachment from the pelvic or abdominal wall. On the other hand, the likelihood of occult lymphatic or hematogenous metastasis increases with increasing tumor stage. Five-year survival data are thus a reflection of both local and distant tumor control. Correct MRI staging of TNM 2002 stage T4a tumors requires contrast-enhanced imaging. Barentsz et al. (1996b) showed that the addition of DCE MRI to the protocol improves accuracy in the staging of T4 tumors

7.2  Male Pelvis

by more than 25%. In series including ten or more patients with TNM 2002 stage T4a TCC, the accuracy of DCE MRI is approximately 90% (Barentsz et al. 1996b; Koelbel et al. 1988). The accuracy of CT in studies reporting both MRI and CT staging results in their patients is similar to the accuracy of MRI of approximately 60–80% (Husband et al. 1989; Kim et al. 1994; Tachibana et al. 1991). Although the numbers of patients examined in individual studies are rather small, there remains little doubt that, independent of both the MRI technique used and the biopsy status of the respective patients, T staging by means of MRI is correct in almost all patients with TNM 2002 stage T4b TCC (approximately mean 90%, range of 70–100%, in nine different studies: Barentsz et al. 1988, 1990, 1996b: Husband et al. 1989; Kim et al. 1994; Küper et al. 1986; Sohn et al. 1990; Tachibana et al. 1991; Tavares et al. 1990). Results of CT staging of T4b tumors in studies also including MRI examinations are very similar (Husband et al. 1989; Kim et al. 1994; Tachibana et al. 1991). The prevalence of lymph node metastasis of TCC correlates with its respective T stage. While TNM 2002 stage pT1 TCC of the urinary bladder presents with positive lymph nodes in about 5% of cases, the likelihood of nodal metastasis increases to about 30% in TNM 2002 stage pT2 and is higher than 60% in TNM 2002 stage pT3b bladder cancer (Jocham 1994). TCC cancer cells prefer certain routes of lymph node spread to others. Metastases are most frequently found in lymph nodes of the obturator muscle group (75%) and along the external iliac arteries (65%). Common iliac artery nodes are involved in 19%, and internal iliac artery nodes in 15% of patients. Lymph nodes along the hypogastric artery and the perivesical vessels are infested in 15% (Jocham 1994). In a study of 57 patients, Barentsz et al. (1996b) detected 12/14 cases of pelvic lymph node metastasis with combined unenhanced MRI and DCE MRI. Criteria were rapid contrast-uptake (less than 10 s after peak arterial enhancement at a sequence repetition rate of one in 1.25–2.50 s) and a shortest-axial-size-to-long-axis ratio of more than 0.8 in nodes of more than 8 mm in diameter. Other pathologic signs were a nodal diameter of more than 10 mm or an asymmetric cluster of small nodes. They report an accuracy of 93%, sensitivity of 86%, specificity of 95%, positive predictive value of 86%, and negative predictive value of 95%. Without contrastenhancement, Barentsz et al. (1996b) recognized lymph node metastasis in 10 of 14 cases. Comparison in 58 patients, (172 pelvic lymph nodes, 122 benign, and 50 malignant) between precontrast MR images (based on nodal size and shape criteria, round node, >8 mm; oval, >10 mm axial diameter) and postcontrast MR images obtained 24 hours after the intravenous administration of ultrasmall particles of iron oxide

(USPIOs, ferumoxtran-10) demonstrated that sensitivity and negative predictive value were significantly higher at postcontrast imaging than at precontrast imaging, increasing from 76 to 96% (p < 0.001) and from 91 to 98% (p < 0.01), respectively (Deserno et al. 2004). Hematogenous metastasis is found in more than 60% of patients with TCC of TNM 2002 stage T3b and higher. About 40% of distant metastases are exclusively skeletal. Other locations of frequent metastasis include the liver, the lung, the central nervous system, and endocrine organs such as the thyroid gland and the adrenal glands (Jocham 1994). Due to the spatial restriction of MRI examinations, most of these locations cannot be covered during an MRI examination dedicated to the urinary bladder. However, with whole-body MRI, it appears likely that both T staging and whole-body N and M staging of TCC can be performed within the same examination. 7.2.1.4.2  Description of Single Pathologic Conditions MRI signal intensities, epidemiologic data, and differential diagnoses for tumorous lesions of the urinary bladder are listed in Table 7.2.5. 7.2.1.5  Indications and Value of MRI Visualization of lesions of the urinary bladder is frequently necessary to assess the kind and distribution pattern of disease. Whenever disease is considered to primarily affect the urothelial layer of the bladder wall, cystoscopy is the diagnostic modality of choice. Particularly because cystoscopy allows the urologist to take biopsy samples, it represents a means of definitive diagnosis. However, cystoscopy only provides a look at the urothelial surface of the urinary bladder. Thus, information on disease considered to affect deeper tissue layers must be obtained by other means. While trans-abdominal ultrasonography has the advantages of being readily available in most clinical settings and being non-invasive and well tolerated by most patients, it is usually insufficient to determine the depth of tissue invasion in inflammatory or neoplastic bladder disease. Thus, the staging of neoplastic bladder disease usually involves cross-sectional imaging by means of CT or MRI. In the majority of instances, the information sought from CT or MRI is the presence and extent of extravesical tumor. Currently, there is no evidence to suggest that one of the modalities is superior to the other in this regard. MR angiography may be used to assess the respective courses and branching patterns of major pelvic blood vessels, which helps in the planning of pelvic intravascular catheter-guided embolization therapy. In complex congenital disorders and status post pelvic trauma or pelvic surgery, it is often helpful to ob-

1013

1014

7  Pelvis Table 7.2.5  Differential diagnosis of tumorous urinary bladder lesions at MRI Lesion type

Urothelial carcinoma

Tumor invasion (other malignancy)

Inflammatory pseudotumor

Incidence

2004 populationa (male/female, per 100.000) Munich: 29.2/11.9 Germany: 32.0/9.3 Europe: 23.4/7.1 World: 15.1/4.6

In stages T4a/T4b of various malignancies, including colorectal cancer, prostate cancer, uterine cervix cancer, uterine corpus cancer, ovarian cancer, lymphoma, rhabdomyosarcoma, and others

Rare

Age

Male, 70.7±11.1 yearsa Female, 74.1±12.7 yearsa (mean±SD, Munich 2004)

Various

Adults and children

Sex Localization

Male > female Urinary bladder

Male and female Urinary bladder

Male and female Urinary bladder

Prognosis   (survival after therapy)

All stages (1995-2001)b 5 years: 82% 10 years: 77% 15 years: 73%

Depends on underlying tumor entity and radicality of therapy

Non-neoplastic

MRI signals   (signal Intensity [SI],   compared with bladder wall)

TlwMRI Intermediate SI May extend beyond bladder

TlwMRI Intermediate SI Invades bladder

TlwMRI Intermediate SI May extend beyond bladder

T2wMRI Intermediate-to-high SI polypoid or flat or lobulated mass May exulcerate

T2wMRI Various polypoid or flat or lobulated mass May exulcerate

T2wMRI Intermediate-to-high SI polypoid or flat or lobulated mass May exulcerate

DCE-MRI Rapid contrast enhancement Marked SI post-contrast

DCE-MRI Rapid contrast enhancement

DCE-MRI Rapid contrast enhancement

Other pelvic tumor invading urinary bladder, other bladder tumor, inflammatory pseudotumor

Urothelial carcinoma, other bladder tumor, inflammatory pseudotumor

Urothelial carcinoma, other bladder tumor, other pelvic tumor invading urinary bladder

Differential diagnosis

T1w T1 weighted, T2w T2-weighted a http://www.tumorregister-muenchen.de/facts/base/GGTSPU-hydra3.fw.med.uni-muenchen.de-4835–2842314-DAT/base_C61_ _G.pdf b American Cancer Society (2006) Cancer facts and figures 2006. American Cancer Society, 1599 Clifton Road, NE, Atlanta, GA 30329-4251, USA, pp 17–19. www.cancer.org

tain cross-sectional images. When the anomaly to be assessed primarily involves soft tissue structures, MRI can be considered the imaging modality of choice. However, when bony structures are under scrutiny, it is often best to first obtain CT images.

7.2.1.6  Indications (Symptom-Specific Imaging Modalities) The two cardinal symptoms that bring patients to the attention of an urologist are pelvic pain and hematuria. Neither of these symptoms is specific for a particular

7.2  Male Pelvis Table 7.2.6  Diagnostic imaging procedures in diseases of the urinary bladder Ultrasonography (US) Abdominal and/or pelvic plain film radiography (CR) Cystography or cystourethrography (CUG) Intravenous urography (IVU) Cystoscopy or endoscopy Computed tomography (CT) Magnetic resonance imaging (MRI)

on the list of differential diagnoses and the order of most likely disease entities. Cross-sectional imaging by means of CT or MRI is usually a second-line imaging test, the first-line imaging test frequently being ultrasonography. However, since the late 1990s, unenhanced CT has taken over from ultrasound and plain film radiography in adult patients presenting with signs and symptoms suggestive of urinary calculus disease (Müller-Lisse et al. 2004a). 7.2.1.7  Diagnostic Procedures

underlying disease. However, since hematuria indicates that blood has transgressed the urothelial lining of the urinary tract, cystoscopy is likely to be the first imaging test to be applied. Pelvic pain, on the other hand, requires thorough history taking and physical examination to determine the most likely differential diagnoses. The imaging modalities of choice (if any) are therefore dependent

The diagnostic procedures involved in the investigation of lesions of the urinary bladder are manifold and include urinalysis, blood tests, microbiology, cytology, and molecular biology, endoscopy (cystoscopy) with or without biopsy, functional tests such as uroflow and urodynamic examinations, and various imaging tests. The latter are included in Table 7.2.6 and in a schematic diagram (Fig. 7.2.9).

Fig. 7.2.9  Imaging of the urinary bladder. Schematic diagram of diagnostic procedures (asterisk selection of subsequent imaging test will depend on clinical history and presentation, patient sex and age, and most likely order of differential diagnoses)

1015

1016

7  Pelvis

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Althausen AF, Prout GR Jr, Daly JJ (1976) Noninvasive papillary carcinoma of the bladder associated with carcinoma in situ. J Urol 116:575–580 Amendola MA, Glazer GM, Grossman HB, Aisen AM, Francis IR (1986) Staging of bladder carcinoma: MRI-CTsurgical correlation. Am J Roentgenol 146:1179–1183 Barentsz JO, Lemmens JAM, Ruijs SHJ, Boskamp EB, Hendrikx AJM, Karthaus HFM, Kaanders JHAM, Rosenbusch G (1988) Carcinoma of the urinary bladder: MR imaging with a double surface coil. Am J Roentgenol 151:107–112 Barentsz JO, Debruyne FMJ, Ruijs SHJ (1990) MR imaging of carcinoma of the urinary bladder. Kluwer, Dordrecht, The Netherlands Barentsz JO, Jager G, Mugler JP III, Oosterdorf G, Peters H, van Erning LTJO, Ruijs SHJ (1994) Staging urinary bladder cancer: value of T1-weighted three-dimensional magnetization prepared-rapid gradient-echo and two-dimensional spin-echo sequences. Am J Roentgenol 164:109–115 Barentsz JO, Jager GJ, Witjes JA, Ruijs SHJ (1996a) Primary staging of urinary bladder carcinoma: the role of MRI and a comparison with CT. Eur Radiol 6:129–133 Barentsz JO, Jager GJ, van Vierzen PBJ, Witjes JA, Strijk SP, Peters H, Karssemeijer N, Ruijs SHJ (1996b) Staging urinary bladder cancer after transurethral biopsy: value of dynamic contrast-enhanced MR imaging. Radiology 201:185–193 Barnes RW, Dick AL, Hadley HL, Johnston OL (1977) Survival following transurethral resection of bladder carcinoma. Cancer Res 37:2895–2897 Benson RC Jr, Tomera KM, Kelalis PP (1983) Transitional cell carcinoma of the bladder in children and adolescents. J Urol 130:54–55 Beyersdorff D, Taupitz M, Giessing M, Turk I, Schnorr D, Loening S, Hamm B (2000) [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 172:504–508 Boemers TM, Ludwikowski B, Forstner R, Schimke C, Ardelean MA (2006) Dynamic magnetic resonance imaging of the pelvic floor in children and adolescents with vesical and anorectal malformations. J Pediatr Surg. 41:1267–1271 Bryan PJ, Butler HE, LiPuma JP, Resnick MI, Kursh ED (1987) CT and MR imaging in staging bladder neoplasms. J Comput Assist Tomogr 11:96–101 Buy JN, Moss AA, Guinet C, Ghossain MA, Malbec L, Arrive L, Vadrot D (1988) MR staging of bladder carcinoma: correlation with pathologic findings. Radiology 169:695–700 Debruyne FMJ (1991) Induktive Chemotherapie des Blasenkarzinoms. Eine kritische Beurteilung. Urologe (A) 30:81–84

15. Deserno WM, Harisinghani MG, Taupitz M, Jager GJ, Witjes JA, Mulders PF, Hulsbergen van de Kaa CA, Kaufmann D, Barentsz JO (2004) Urinary bladder cancer: preoperative nodal staging with ferumoxtran-10-enhanced MR imaging. Radiology 233:449–456 16. Doringer E, Joos H, Forstner R, Schmoller HJ (1991) MRT des Harnblasenkarzinoms: Tumorstaging und Gadoliniumkontrastverhalten. Fortschr Roentgenstr 154:357–363 17. Fisher MR, Hricak H, Tanagho EA (1985) Urinary bladder MR imaging. Part II. Neoplasm. Radiology 157:471–477 18. Fujiwara T, Sugimura K, Imaoka I, Igawa M (1999) Inflammatory pseudotumor of the bladder: MR findings. J Comput Assist Tomogr 23:558–561 19. Geusens E, Brys P, Maleux G, Janzing H (2000) Imaging in pelvic trauma. JBR-BTR 83:173–180 20. Heiken JP, Lee JKT (1988) MR Imaging of the pelvis. Radiology 166:11–16 21. Henry K, Miller J, Mori M (1988) Comparison of transurethral resection to radical therapies for stage B bladder tumors. J Urol 140:964–967 22. Herr HW (1987) Conservative management of muscleinfiltrating bladder cancer: Prospective experience. J Urol 138:1162–1163 23. Houvenaeghel G, Delpero JR, Rosello R, Resbeut M, Viens P, Jacquemier J, Noirclerc M, Guerinel G (1993) Results of a prospective study with comparison of clinical, endosonographic, computed tomography, magnetic resonance imaging and pathologic staging of advanced gynecologic carcinoma and recurrence. Surg Gynecol Obstet 177:231–236 24. Hricak H, Williams RD, Spring DB, Moon KL Jr, Hedgcock MW, Watson RA, Crooks LE (1983) Anatomy and pathology of the male pelvis by magnetic resonance imaging. Am J Roentgenol 141:1101–1110 25. Husband JES, Olliff JFC, Williams MP, Heron CW, Cherryman GR (1989) Bladder cancer: staging with CT and MR imaging. Radiology 173:435–440 26. Iwamoto K, Kigawa J, Minagawa Y, Miura H, Terakawa N (1994) Transvaginal ultrasonographic diagnosis of bladder-wall invasion in patients with cervical cancer. Obstet Gynecol (United States) 83:217–219 27. Jager GJ, Barentsz JO, Oosterhof G, Ruijs JHJ (1996) 3D MR imaging in nodal staging of bladder and prostate cancer. AJR Am J Radiol 167:1503–1507 28. Jakse G, Putz A, Feichtinger J (1989) Cystectomy: the treatment of choice in patients with carcinoma in situ of the urinary bladder? Eur J Surg Oncol 15:211–216 29. Jewett HJ (1973) Cancer of the bladder. Cancer 32:1072–1074 30. Jewett HJ, Strong GH (1946) Infiltrating carcinoma of the bladder. Relation of depth of penetration of the bladder wall to incidence of local extension and metastases. J Urol 55:366–372 31. Jocham D (1994) Maligne Tumoren der Harnblase. In: Jocham D, Miller K (eds) Praxis der Urologie, Band II, Teil 2. Thieme, Stuttgart, pp 49–115

7.2  Male Pelvis 32. Kaji Y, Sugimura K, Kitao M, Ishida T (1994) Histopathology of uterine cervical carcinoma: diagnostic comparison of endorectal surface coil and standard body coil MRI. J Comput Assist Tomogr 18:785–792 33. Kawakami S, Togashi K, Kojima N, Morikawa K, Mori T, Konishi J (1995) MR appearance of malignant lymphoma of the uterus. J Comput Assist Tomogr 19:238–242 34. Kim B, Semelka RC, Ascher SM, Chalpin DB, Carroll PR, Hricak H (1994) Bladder tumor staging: comparison of contrast-enhanced CT, T1- and T2-weighted MR imaging, dynamic gadolinium-enhanced imaging, and late gadolinium-enhanced imaging. Radiology 193:239–245 35. Koelbel G, Schmiedl U, Griebel J, Hess CF, Kueper K (1988) MR imaging of urinary bladder neoplasms. J Comput Assist Tomogr 12:98–103 36. Küper K, Kölbel G, Schmiedl U (1986) Kernspintomographische Untersuchungen von Harnblasenkarzinomen bei 1,5 T. Fortschr. Röntgenstr. 144:674–680 37. Lienemann A, Anthuber C, Baron A, Kohz P, Reiser M (1997) Dynamic MR colpocystorectography assessing pelvic-floor descent. Eur Radiol 7: 1309–1317 38. Marshall VF (1952) The relation of the preoperative estimate to the pathologic demonstration of the extent of vesical neoplasms. J Urol 68:714–723 39. Müller-Lisse GU, Heuck AF, Barentsz JO (1998) Bladder. In: Heuck A, Reiser M (eds) Abdominal and Pelvic MRI. Medical Radiology – Diagnostic Imaging and Radiation Oncology (Baert AL, Heuck FHW, Youker JE, series eds), Springer, Berlin, Heidelberg, New York, Barcelona, Budapest, Hong Kong, London, Milan, Paris, Santa Clara, Singapore, Tokyo, pp 209–228 40. Müller-Lisse UL, Müller-Lisse UG, Lienemann A, Schneede P, M.F. Reiser MF (2002) Feasibility of True-FISP MR voiding cystourethrography. Proceedings of the International Society for Magnetic Resonance in Medicine 10th scientific meeting and exhibition, 2002, p 1901 41. Müller-Lisse UG, Müller-Lisse UL (2004a) MDCT of the kidney. In: Multislice CT. 2nd revised edition. Reiser MF, Takahashi M, Modic M, Becker CR (eds). Medical Radiology – Diagnostic Imaging (Baert AL, Sartor K, series eds). Springer-Verlag, Berlin, Heidelberg, New York, Hong Kong, London, Milan, Paris, Tokyo, pp 211–232 42. Müller-Lisse UG, Müller-Lisse UL (2004b) Harnblase und Prostata. In: Lemke AJ, Felix R (eds) Kontrastverstärkte MRT-Bildgebung. Kompendium der modernen MRT-Diagnostik. Schering, Berlin, pp 444–465 43. Neuerburg JM, Bohndorf K, Sohn M, Teufl F, Guenther RW, Daus HJ (1989) Urinary bladder neoplasms: evaluation with contrast-enhanced MR imaging. Radiology 172:739–743 44. Nicolas V, Harder T, Steudel A, Krahe T, Schindler G, van Ahlen H, Jaeger N (1988) Die Wertigkeit bildgebender Verfahren bei der Diagnostik und dem Staging von Harnblasentumoren. Fortschr Röntgenstr 148:234–239

45. Nolte-Ernsting CC, Glowinski A, Katterbach FJ, Adam G, Rasche V, Günther RW (1998) MR-Miktionszysturethrographie mit radialer k-Raum-Abtastung. RöFo – Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 168: 385–389 46. O’Flynn JD, Smith JD, Hanson JS (1975) Transurethral resection for the assessment and treatment of vesical neoplasm. A review of 800 consecutive cases. Europ Urol 1:38 47. Obenauer S, Plothe KD, Ringert RH, Heuser M (2006) Imaging of genitourinary trauma. Scand J Urol Nephrol 40:416–422 48. Pannu HK (2003) Dynamic MR imaging of female organ prolapse. Radiol Clin North Am 41:409–423 49. Patel MD, Hricak H (1995) Current role of magnetic resonance imaging in urology. Current Opinion in Urology 5:67–74 50. Raghavan D, Shipley WU, Garnick MB, Russell PJ, Richie JP (1990) Biology and management of bladder cancer. N Engl J Med. 322:1129–1138 51. Rholl KS, Lee JKT, Heiken JP, Ling D, Glazer HS (1987) Primary bladder carcinoma: evaluation with MR imaging. Radiology 163:117–121 52. Scheidler J, Heuck AF, Bruening R, Kohz P, Kimmig R, Stehling MK, Reiser MF (1997) Magnetic resonance imaging of the female pelvis. New circularly polarized body array coil versus standard body coil. Invest Radiol. 32:1–6 53. Scher HI, Yagoda A, Herr HW, Sternberg CN, Reutter V, Geller N, Hollander PS, Vaughan ED (1988) Neoadjuvant M-VAC (methotrexate, vinblastine, adriamycin and cisplatin) effect on the primary bladder lesion. J Urol 139:470–474 54. Schneider G, Ahlhelm F, Altmeyer K, Aliani S, Remberger K, Schoenhofen H, Kramann B, Uder M (2001) Rare pseudotumors of the urinary bladder in childhood. Eur Radiol 11:1024–1029 55. Semelka RC, Lawrence PH, Shoenut JP, Heywood M, Kroeker MA, Lotocki R (1993) Primary ovarian cancer: prospective comparison of contrast-enhanced CT and preand postcontrast, fat-suppressed MR imaging, with histologic correlation. J Magn Reson Imaging 3:99–106 56. Shonnard KM, Jelinek JS, Benedikt RA, Kransdorf MJ (1992) CT and MR of neurofibromatosis of the bladder. J Comput Assist Tomogr 16:433–438 57. Sohn M, Neuerburg J, Teufl F, Bohndorf K (1990) Gadolinium-enhanced magnetic resonance imaging in the staging of urinary bladder neoplasms. Urol Int 45:142–147 58. Solsona E, Iborra I, Ricos JV, Monros JL, Dumont R (1992) Feasibility of transurethral resection for muscle-infiltrating carcinoma of the bladder: Prospective study. J Urol 147:1513–1515 59. Sparenberg A, Hamm B, Hammerer P, Samberger V, Wolf KJ (1991) Diagnostik von Harnblasenkarzinomen in der Kernspintomographie: Verbesserung mit Gd-DTPA? Fortschr Röntgenstr 155:117–122 60. Splinter TAW (1990) Neoadjuvante Chemotherapie des invasiven Blasenkarzinoms. Akt Urol 21:173–174

1017

1018

7  Pelvis 61. Tachibana M, Baba S, Deguchi N, Jitsukawa S, Hata M, Tazaki H, Tanimoto A, Yuasa Y, Hiramatsu K (1991) Efficacy of gadolinium-diethylenetriamninepentaacetatic acid-enhanced magnetic resonance imaging for differentiation between superficial and muscle-invasive tumor of the bladder: a comparative study with computerized tomography and transurethral ultrasonography. J Urol 145:1169–1173 62. Tanimoto A, Yuasa Y, Imai Y, Izutsu M, Hiramatsu K, Tachibana M, Tazaki H (1992) Bladder tumor staging: comparison of conventional and gadolinium-enhanced dynamic MR imaging and CT. Radiology 185:741–747 63. Tavares NJ, Demas BE, Hricak H (1990) MR imaging of bladder neoplasms: correlation with pathologic staging. Urol Radiol 12:27–33 64. Thurnher S, Marincek B, Hauri D (1992) Retrovesical leiomyoma: CT and contrast-enhanced MR imaging findings. Urol Radiol 13:190–193 65. Whitmore WF, Batata MA, Ghoneim MA, Grabstald H, Unal A (1977) Radical cystectomy with or without prior irradiation in the treatment of bladder cancer. J Urol 118:184–187 66. Wittekind C, Mezer HJ, Bootz F (eds) (2002) UICC–International Union Against Cancer: TNM-Klassifikation maligner Tumoren (6th edn.). Springer, Berlin Heidelberg New York 67. Wong-You-Cheong JJ, Woodward PJ, Manning MA, Davis CJ (2006) From the archives of the AFIP: Inflammatory and nonneoplastic bladder masses: radiologic-pathologic correlation. Radiographics 26:1847–1868 68. Yeoman LJ, Mason MD, Olliff JFC (1991) Non-Hodgkin’s lymphoma of the bladder—CT and MRI appearances. Clinical Radiol 44:389–392

7.2.2  Male Pelvis: Prostate U.G. Müller-Lisse, M.K. Scherr, and U.L. Müller-Lisse 7.2.2.1  Introduction The diagnosis of diseases of the prostate frequently involves imaging examinations. The least invasive and best available imaging modality usually is ultrasonography (US). Depending on the clinical work environment, US may be performed by the urologist or by the radiologist. Transrectal ultrasonography (TRUS) is the imaging modality most frequently applied to examine the prostate. TRUS is usually applied by a urologist and primarily serves the purposes of determining prostate volume and guiding prostate biopsy. The ability of TRUS to detect and localize prostate cancer is limited (Müller-Lisse and Hofstetter 2003). Recent improvements of TRUS include tissue harmonic imaging, color Doppler imaging, and contrast-enhanced color Doppler imaging

(Frauscher et al. 2003; Halpern et al. 2005). However, it has long been recognized that MRI is the imaging modality that best depicts prostatic zonal anatomy and best detects and localizes prostate cancer. MRI allows for T staging of prostate cancer according to the TNM classification of the UICC (Heuck et al. 2003). Recent improvements of MRI of the prostate that increase specificity in the detection of prostate cancer include MR spectroscopy (MRS) (Kurhanewicz et al. 1995, 1996a; Müller-Lisse et al. 2001a, b; Müller-Lisse and Scherr 2003; Scheidler et al. 1999) and dynamic, contrast-enhanced MRI (DCE MR) (Kiessling et al. 2003; Schlemmer et al. 2004). Improved N staging of prostate cancer according to the TNM classification of the UICC is achieved by means of MRI with intravenously (i.v.) applied ultra-small particles of iron oxide (USPIOs) that accumulate in lymph nodes unless the nodes are affected by metastasis (Harisinghani et al. 2003; Heesakkers et al. 2006; Hovels et al. 2004). MR examinations of the prostate can also be used to follow up on prostate cancer after various treatments, both to assess post-therapeutic changes and to detect or rule out residual or recurrent prostate cancer (Coakley et al. 2003; Kurhanewicz et al. 1996b; Müller-Lisse 2001a, b). Initial experience extends to MRI-guided biopsy of the prostate (Anastasiadis et al. 2006; Beyersdorff et al. 2005a). 7.2.2.2  Examination Techniques 7.2.2.2.1  Patient Positioning and, if Applicable, Patient Preparation General positioning for MRI examinations of the pelvis is described in the chapter on MRI of the urinary bladder. MRI examinations of the prostate are usually performed with the patient in supine position. In particular, preparation of the patient for an MRI examination of the prostate includes preparation for application of an endorectal surface coil ([ERC] see below, selection of coils). Bowel preparation may be requested to include a cleansing enema of the sigmoid colon and rectum; however, in clinical practice, it is most often sufficient that the patient empties his bowels and urinary bladder prior to the MR examination. Application of butylscopolamide (Buscopan®) or glucagon has been recommended to decrease bowel motion during the examination when an ERC is applied; however, contraindications have to be observed. Also, it is usually sufficient to inflate the balloon of the endorectal coil with 80 to 100 ml of air once it has been entered into the empty rectum (Fig. 7.2.10). In all instances, digital rectal examination should precede placement of an ERC, to detect or rule out severe narrowing or other lesions of the anus and rectum potentially harmful to the patient when an ERC is applied.

7.2  Male Pelvis

Fig. 7.2.10a–g  MRI of the prostate and seminal vesicles demonstrates optimal position of endorectal coil inside rectal ampulla (arrows in a) and examples of typical MRI sequences, including axial. T1-weighted SE MR images of the prostate (b), axial, T1-weighted SE or GRE or FLASH MR images of the pelvis below the aortic bifurcation (c), axial STIR of the pelvis in case of suspected bony lesions (double-lined arrows in d), and axial (e), sagittal (f), and coronal (g) T2-weighted TSE MR images of the prostate

7.2.2.2.2  Selection of Coils MRI examinations of the prostate should be performed in whole-body high-field MR scanners. The majority of scientifically evaluated studies were performed at a magnetic field strength of 1.5 T. However, examinations at 3 T have been the subject of some recent studies (Beyersdorff et al. 2005b; Futterer et al. 2006). While historically, the body coil was used as both transmitter and receiver, multi-channel phased-array surface coils (PASCs) with at least four independent elements have since been applied to the advantage of signal-to-noise ratio and image homogeneity. Using the body-coil or a PASC system is usually sufficient if the examination aims at determining

prostate volume (e.g., in patients with benign prostatic hyperplasia [BPH] prior to and after therapy). However, in a recent meta-analysis, the application of an ERC was shown to significantly improve prostate cancer staging by means of MRI of the prostate at 1.5 T (Engelbrecht et al. 2002). Application of the ERC in combination with a phased array body coil or pelvic surface coil has been shown to improve prostate imaging significantly (Heuck et al. 2003; Hricak et al. 1994). Currently, ERCs are commercially available for whole-body MRI systems operating at a magnetic field strength of 1.5 T. ERCs for use at 3 T are not yet commercially available on the market. When an ERC is being used in MRI examinations of the prostate, motion artifact can be reduced by filling

1019

1020

7  Pelvis

the balloon of the ERC with 80–100 ml of air after correct placement in the rectum. The balloon of the ERC should be completely placed within the rectum, such that its concave part covers the dorsal aspect of the prostate from base to apex. Currently, there is only one kind of ERC available in the market. This ERC has a shaft that is marked with a straight line. The ERC is optimally placed when the straight line shows straight up (12 o’ clock position in lithotomy position) and the ERC has been pulled back toward the anal sphincter after filling of the balloon with air (Fig. 7.2.10). 7.2.2.2.3  Examination Sequences The majority of MRI examinations of the prostate and seminal vesicles have been performed at a magnetic field strength of 1.5 T. Thus, while examinations at 3 T have been the subject of some recent studies, examination sequences suggested here refer to a magnetic field strength of 1.5 T (Tables 7.2.7, 7.2.8; Fig. 7.2.10). For tumor detection, tumor localization within the prostate, and locoregional tumor staging, MRI protocols including T1-weighted and T2-weighted sequences are applied. These are usually preceded by localizer or scout images in three different planes of imaging that are performed with either rapid, T1-weighted sequences such as gradient-echo (GRE) or turbo-FLASH sequences, or rapid, T2-weighted

sequences such as HASTE or SSFSE. T1-weighted images are usually obtained from the aortic bifurcation to the pelvic floor. While it appears best for high spatial resolution and signal-to-noise ratio to apply T1-weighted SE (alternatively, turbo SE or fast SE) sequences for the prostate and seminal vesicles, it is a viable alternative to apply GRE sequences to cover the upper part of the pelvis, from the aortic bifurcation down to the seminal vesicles. The latter may prove to be advantageous in patients with strong bowel movement or respiratory movement in the lower abdomen, since GRE sequences may allow for breath holding or respiratory gating. T1-weighted MR images of the pelvis are obtained to demonstrate pathologic enlargement (> 0.8 cm, Heuck et al. 1997) or pathologic clustering (three or more nodes of > 0.3 cm immediately adjacent to one another) of lymph nodes, pathologic signal change within the pelvic bone marrow, which may indicate metastasis, and hemorrhage within the prostate and seminal vesicles. T1-weighted MR imaging is followed by T2-weighted turbo SE or fast SE sequences in two or more planes of imaging. It is advantageous for the evaluation of local tumor extent in prostate cancer to obtain additional T2-weighted images in the coronal plane, which cover the small pelvis with narrow slice thickness, narrow interslice gap, and high imaging matrix and focus on the prostate and seminal vesicles. Sagittal T2-weighted images provide additional information on the bladder floor, blad-

Table 7.2.7  Examination protocol for magnetic resonance imaging of the prostate and seminal vesicles for imaging with a combined endorectal/body phased-array coil system at 1.5 T Sequence type

T1-weighted a SE sequence

T1-weighted SE sequence

T2-weighted TSE sequence

T2-weighted TSE sequence

T2-weighted TSE sequence

Anatomical region

Pelvis

Prostate

Prostate

Pelvis

Prostate

Imaging plane

Axial

Axial

Axial

Coronal

Sagittal

TR (ms)

500–700

500–700

3,500–5,000

3,500–5,000

3,500–5,000

TE (ms)

15–17

15–17

90–110

90–110

90–110

Imaging matrix

192–256 × 256

256 × 256

224 × 256–256 × 512

256–512 × 512

256–512 × 512

FOV (mm)

225 × 300

160 × 160–200 × 200

120 × 120–200 × 200

250 × 250–300 × 300

250 × 250–300 × 300

Slice thickness (mm)

7–8

3–4

3–4

3–4

3–4

Interslice gap (mm)

1–2

0–1

0–1

0–1

0–1

 Alternatively, the pelvis may be examined with a gradient echo- (GRE) or fast low-angle shot (FLASH) sequence. For FLASH, suggested parameters would be TR 140–150 ms, TE 4,1–4,5 ms, flip angle 75–90°, imaging matrix 163–192 × 256, FOV 225 × 300 mm, slice thickness 7–8 mm, interslice gap 1–2 mm

a

7.2  Male Pelvis Table 7.2.8  Examples of examination protocols suggested by various authors for dynamic, contrast-enhanced magnetic resonance imaging (DCE MRI) of the prostate at 1.5 T Author

Kiessling et a al. 2003

Schlemmer b et al. 2004

Prochnow et c al. 2005

Kozlowski et d al. 2006

Villers et e al. 2006

Field strength (T)

1.5

1.5

1.5

1.5

1.5

Sequence

FLASH 2D

Inversion recovery turboFLASH

Inversion-prepared dual-contrast turboFLASH

FSPGR

GRE

TR (s)

125 ms

1300 ms

3.8 / 28.3 ms

18.5 ms

120 ms

TE (ms)

3.11 ms

4.2 ms

2.1 ms / 7 ms

3 ms

4.6 ms

Flip angle˚

Not reported

13

25

50

TI (ms)

654

280

No. of acquisitions

Not reported

Not reported

Not reported

Not reported

2

No. of slices

10

10

2×1

Not reported

14

Time resolution (s)

13

13

1.65

22

15

Section thickness (mm)

Not reported

4

5

5

4

Intersection gap (mm)

Not reported

0.8

NA

None

Not reported

Slice orientation

Not reported

Not reported

Not reported

Not reported

Axial

Matrix

128 × 128

128 × 128

128 × 90

128 × 256

128 × 109

Field of view (cm)

20 × 20 cm

36 × 36 cm

6/8 × 22.8 cm

24 × 24 cm

16 × 16 cm

FLASH fast low-angle shot, FSE fast spin echo, GRE spoiled gradient echo, FSPGR multislice fast spoiled gradient-recalled sequence a  Kiessling et al. 2003: 25 repetitions; contrast injection of 0.1 mmol/kg body weight of gadopentetate dimeglumine (Magnevist, Schering, Berlin) after third repetition through a cubital vein within 30 s, using a commercially available speed infusion pump b  Schlemmer et al. 2004: 22 repetitions; contrast injection of 0.1 mmol/kg body weight of gadopentetate dimeglumine (Magnevist, Schering) with beginning of third repetition through a cubital vein by a short constant rate infusion within 30 s, using a variable speed infusion pump (Spectris MR Injector, Medrad, Maastricht, The Netherlands) c  Prochnow et al. 2005: experimental inversion-prepared dual-contrast turbo fast low-angle shot (FLASH) sequence for the interleaved acquisition of an inversion recovery prepared short TE image and an image with a long TE; non-slice-selective inversion preparation, adjusted to suppress signal contributions from non-contrast-enhanced arterial blood; PAT with GeneRalized Autocalibrating Partially Parallel Acquisition (GRAPPA) algorithm, acceleration factor of 2 and 24 reference lines; 513 dynamic scans for each echo time, total measurement time of 846 s; bolus of gadopentetate dimeglumine (Magnevist, Schering) at a dose of 0.1 mmol Gd-DTPA/kg body weight was injected intravenously, followed by infusion of 20 ml saline, at a flow rate of 6 ml/s, using a power injector (Medrad, Volkach, Germany). The infusion of the CM started simultaneously with the acquisition of the 10th TFI1 image d  Kozlowski et al. 2006: Three T1-weighted baseline images prior to the injection and 42 T1-weighted images following bolus injection of Gadodiamide (Omniscan, Nycomed-Amersham, Norway; dose: 0.1 mmol/kg injected within 10 s, followed by a 20-mL flush of normal saline) e  Villers et al. 2006: six identical sequences, started immediately after intravenous bolus administration at 2 ml/s of 0.1 mmol/kg body weight of gadoteric acid, followed by 20 ml normal saline flush

1021

1022

7  Pelvis

der neck, and seminal vesicles when they are performed with narrow slice thickness. Due to shorter acquisition times and higher signal-to-noise ratios at an increased number of acquisitions, and because they allow more accurate evaluation of the prostate, T2-weighted turbo SE

(or fast SE FSE) sequences are preferable to SE sequences in MRI of the prostate (Engelbrecht et al. 2002; Heuck et al. 2003). T2-weighted MR images of the prostate serve to demonstrate zonal anatomy of the prostate and seminal vesicles (Fig. 7.2.11), detect and localize tumors, cysts,

7.2  Male Pelvis

9 Fig. 7.2.11  a Zonal anatomy of the prostate as demonstrated on axial T2-weighted TSE MR images obtained at the base (A), middle (B), and apex (C) of the prostate, respectively. MR images demonstrate respective positions of bladder floor (hatched double-lined arrows in A), pre-prostatic mixed connective and adipose tissue (dot), periprostatic venous plexus (striped arrowheads) dorsal periprostatic fat layer, Denonvilliers’ fascia, and perirectal space (arrowheads), rectum (ring), neurovascular bundles of the prostate (arrows), prostatic capsule (double-lined arrows in B), anterior fibromuscular band (open block arrow), peripheral zone of the prostate (asterisks), prostatic pseudocapsule (curved arrows), central compartments of the prostate gland, including the ejaculatory ducts (open curved arrows), seminal collicle (verumontanum, chevron arrow), prostatic urethra with the periurethral zone (open arrowheads), the central zone (open chevron arrow), and the bilateral transition zones (block arrows). Note nodule of benign prostatic hyperplasia (BPH) within peripheral zone on the left side (open ring in B). b Zonal anatomy of the prostate as demonstrated on T2-weighted TSE MR images obtained in the paramedian sagittal (A), and coronal (B) plane of imaging. MR images demonstrate respective positions of bladder floor (hatched, double-lined arrows in A), pre-prostatic mixed connective and adipose tissue (dot), periprostatic venous plexus (striped arrowheads) dorsal periprostatic fat layer, Denonvilliers’ fascia, and perirectal space (arrowheads), rectum with endorectal coil (ring), seminal vesicles (arrows), prostatic capsule (double-lined arrows in B), anterior fibromuscular band (open block arrow), peripheral zone of the prostate (asterisks), prostatic urethra with the periurethral zone (open arrowheads), the central zone (open chevron arrow), and the bilateral transition zones (block arrows). Note nodule of benign prostatic hyperplasia (BPH) within peripheral zone on the left side (open ring in B). c MRI of the normal seminal vesicles demonstrates bilateral, symmetric seminal vesicles with slightly lobulated contour (asterisks), clear delineation from surrounding fat and fibroconnective tissue (dots), and convoluted duct system (arrows in B). Seminal vesicles are dorsal to urinary bladder (ring) and show with homogenous, intermediate signal on T1-weighted SE MR image (A). Coronal T2-weighted TSE MR image (B) delineates low signal of fibromuscular walls of seminal vesicles from bright signal of ducts (arrows). Note bilateral ampullae of vas deferens centrally in (B) (open arrowheads)

8 Fig. 7.2.12a,b  MRI of the prostate with bilateral prostate cancer confined to the prostatic capsule demonstrates circumscribed tumor in the peripheral zone bilaterally (arrows), which does not breach the prostatic capsule (double-lined arrows). Signal intensity of prostate cancer is low on axial (a) and coronal (b) T2-weighted TSE MR images

and other pathologic entities within the prostate (Figs. 7.2.11, 7.2.12), and show extra-prostatic extension of tumor (Fig. 7.2.13). Hemorrhage within the prostate (Fig. 7.2.14) and seminal vesicles and loco-regional lymphadenopathy are identified with a combination of T2- and T1-weighted MR images of the prostate and small pelvis. Phase encoding in axial T1- and T2-weighted sequences that cover the prostate and seminal vesicles and their immediate surroundings should be directed from left to right rather than from anterior to posterior, to minimize phase-encoding artifact. If patient history and previous examinations or previously obtained T1- and T2-weighted MR images of the pelvis imply that bone metastasis may be present, the MR protocol should be extended to include a short-tau inversion recovery (STIR) sequence with an inversion time that allows for recognition of bone marrow edema (inversion time, TI approximately 150 ms at 1.5 T). In this instance, STIR images should ideally be obtained in the same imaging plane and with the same slice thickness and interslice gap as the T1-weighted MR images of the pelvis and prostate (Table 7.2.7; Fig. 7.2.10). The i.v. application of gadolinium-based contrast media may improve the detection of seminal vesicle infiltration by prostate cancer (Huch-Böni et al. 1995). However, few reports have been published to date, and thus the use of contrast media to detect seminal vesicle infiltration in MR examinations of the prostate cannot be generally recommended. Magnetic resonance spectroscopy (MRS) of the prostate, which is currently being performed by means of three-dimensional hydrogen-proton MRS (3D MRS, also

1023

1024

7  Pelvis

Fig. 7.2.13a,b  MRI of extraprostatic extension of prostate cancer demonstrates left-sided prostate cancer (arrows) with extension beyond the prostatic capsule and invasion of the peripros-

tatic fat tissue and periprostatic venous plexus (arrowheads), and into the levator ani muscle (double-lined arrows) on axial (a), and coronal (b) T2-weighted TSE MR images

Fig. 7.2.14a,b  MRI of the prostate after biopsy demonstrates severe hemorrhage (arrows), which may mask or mimic prostate cancer. Hemorrhage shows with heterogenous, bright signal on T1-weighted SE MR image (a), and with heterogenous, intermediate-to-low signal on T2-weighted TSE MR image (b)

referred to as 3D MR spectroscopic imaging, 3D MRSI; or 3D chemical-shift imaging , 3DCSI) allows for the simultaneous acquisition of multiple proton spectra of the prostate, with volume element (voxel) sizes of approximately 0.25 cm3 (Costello et al. 1999; Kurhanewicz et al. 1995, 1996a; Müller-Lisse et al. 2001a,b; Müller-Lisse and Scherr 2003; Scheidler et al. 1999). MRS of the prostate in vivo typically distinguishes between peaks of different intraprostatic metabolites, namely citrate, choline, and creatine (Fig. 7.2.15a). Concentration ratios of these intraprostatic metabolites and their respective distribution throughout the prostate allow for an analysis of tissue composition and biochemical activity (Costello et al. 1999; Kurhanewicz et al. 1995, 1996a) (Fig. 7.2.15b). Typically, 3D MRS of the prostate is based on point-resolved

spectroscopy (PRESS) or SE sequences (Heerschap et al. 1997; Kurhanewicz et al. 1995). Currently, preferred sequence parameters include TR 1,000 (650–1,500) ms, and TE 120–130 ms (Kurhanewicz et al. 1996a; Müller-Lisse et al. 2001a,b; Müller-Lisse and Scherr 2003). Besides bandselective suppression of signal contributions from water and lipids, 3D MRS of the prostate involves suppression of signal contributions from extraprostatic tissue (also referred to as outer-volume suppression) by means of saturation bars that are interactively placed around the prostate (Kurhanewicz et al. 1996a Müller-Lisse and Scherr 2003) (Fig. 7.2.15b). New MR imaging methods to detect, localize, and characterize tumors within the prostate include dynamic, contrast-enhanced MR imaging (DCE MRI) (Fig. 7.2.16)

7.2  Male Pelvis

with computer-assisted calculation of contrast exchange rates between different tissue compartments (Futterer et al. 2005; Schlemmer et al. 2004). Correlation of contrast exchange rates with microvessel density within the prostate has been demonstrated for rapid GRE and turboFLASH sequences (Kiessling et al. 2003; Schlemmer et al. 2004). However, current evidence suggests that DCE MRI of the prostate improves sensitivity and specificity of detecting extracapsular extension (ECE) of prostate cancer by means of MRI only in less experienced radiologists (Futterer et al. 2005). Different sequences that have been applied for DCE MRI of the prostate are listed in Table 7.2.8. Another new approach to characterizing lesions within the prostate is diffusion-weighted MR imaging (Gibbs et al. 2006; Kozlowski et al. 2006; Sato et al. 2005). Reduced apparent diffusion coefficient (ADC) (Gibbs et al. 2006; Kozlowski et al. 2006; Sato et al. 2005) and increased fractional anisotropy values have been noted in prostate cancer at 3 T (Gibbs et al. 2006). However, at present, not enough evidence is available to suggest that clinical improvement would be gained by routine application of diffusion-weighted MR imaging of the prostate and seminal vesicles. 7.2.2.2.4  Imaging Planes In MRI examinations of the prostate and seminal vesicles, T1-weighted images should be obtained in the axial plane. T2-weighted images should be obtained in at least two planes (axial and coronal) (Engelbrecht et al. 2002; Heuck et al. 2003), while adding sagittal images may improve analysis of the seminal vesicles and prostatic apex. When STIR images are obtained because T1-weighted images suggest the presence of an osseous lesion in the bony pelvis or lower lumbar spine, the image orientation of choice would be axial, for best comparison with the T1-weighted images obtained previously. MR spectroscopy of the prostate has been available as a three-dimensional examination since the mid 1990s (Heerschap et al. 1997; Kurhanewicz et al. 1995, 1996a). Thus, there is no selection of a primary imaging plane. However, for several years now, it has been possible to angulate 3D-MR spectroscopy of the prostate in two spatial dimensions to better match the position of the prostate in the individual patient. Thus, since MR spectroscopy of the prostate is usually planned on the basis of T2-weighted images of the prostate in two or three planes, it is important to ascertain optimal coverage of the prostate by MR spectroscopy interactively, by iterating through different angulations and rotations of the three-dimensional field of view of MR spectroscopy (Fig. 7.2.15). Dynamic, contrast-enhanced MR imaging (DCE MRI) is usually performed in only one plane of imaging. The axial plane is preferred, because axial images provide the

best overall separation of the peripheral zone of the prostate from the central and transition zones, and prostate cancer, which often demonstrates with rapid, increased uptake of contrast medium, most frequently occurs in the peripheral zone while benign prostatic hyperplasia (BPH), which may also demonstrate with rapid, increased uptake of contrast media, most frequently occurs in the transition zone of the prostate (Futterer et al. 2005; Kiessling et al. 2003; Schlemmer et al. 2004) (Fig. 7.2.16). 7.2.2.2.5  Thickness of Slices It is reasonable to cover the prostate and seminal vesicles with a narrow slice thickness (3–4 mm) and a narrow interslice gap (0–1 mm), and the remainder of the pelvis, from the aortic bifurcation to the seminal vesicles, with increased slice thickness (7–8 mm) and increased interslice gap (2–3 mm, if necessary) to keep TR as low as reasonably possible. Most important for the demonstration of prostatic anatomy and pathology are axial, T2-weighted images with high spatial resolution, i.e., images obtained with a narrow slice thickness (3–4 mm) and a narrow interslice gap (0–1 mm) within a small field of view (FOV) that only covers the prostate and its immediate anatomical surroundings. Recommended FOV and imaging matrix ranges from FOV with an edge length of 12 cm with a matrix of maximum 256 image points per direction to FOV with an edge length of 20 cm with a matrix of 512 image points per direction (Table 7.2.7). 7.2.2.2.6  Preferred Coverage In MR examinations of the prostate and seminal vesicles, preferred coverage extends from the aortic bifurcation to the pelvic floor, to cover both the primary region of interest, i.e., the prostate and its adjacent organs and tissues, and the regional lymph nodes and skeleton, which are most frequently involved when prostate cancer is metastatic. However, for sequences with narrow slice thickness (i.e., 3 mm) that focus on the prostate and seminal vesicles, coverage is limited to the prostate and seminal vesicles in axial images. Coronal and sagittal images are obtained with fields of view that cover the entire pelvis in a craniocaudal direction, but are limited in extent to the prostate and seminal vesicles and their immediate surroundings in the anterior-posterior (coronal) or lateral (sagittal) dimension, respectively (Table 7.2.7). 7.2.2.2.7  Use of Contrast Medium New MR imaging methods to detect, localize, and characterize tumors within the prostate include DCE MRI (Fig. 7.2.16) with computer-assisted calculation of contrast ex-

1025

1026

7  Pelvis

7.2  Male Pelvis

98 Fig. 7.2.15  a MRS of the prostate in vivo typically distinguishes between peaks of different intraprostatic metabolites, namely citrate (Cit, at 2.6 ppm, arrow), choline (Cho, at 3.2 ppm, double-lined arrow), and creatine (Cr, at 3.0 ppm, arrowhead) (A). Typically, 3D-MRS of the prostate (B, with array of multiple spectra superimposed on T2-weighted MR image of prostate) is based on point-resolved spectroscopy (PRESS) or spin-echo (SE) sequences. b Combined MRI and 3D-MRS of the prostate in vivo. T2-weighted MRI with superimposed grid and saturation bars for outer volume suppression (A). Magnification view with MR spectra superimposed onto axial T2-weighted MR image of the prostate (B). Individual spectra within the peripheral

zone of the prostate demonstrate high choline and low citrate, suggestive of prostate cancer (C), and low choline and high citrate, suggestive of healthy prostate tissue (D), respectively (modified from Müller-Lisse and Scherr 2003). c MRS of the prostate in vivo at 3 T typically distinguishes between peaks of different intraprostatic metabolites, namely citrate (Cit, at 2.6 ppm, arrows), choline (Cho, at 3.2 ppm, double-lined arrow), and creatine (Cr, at 3.0 ppm, arrowhead) (A). Note separation of citrate peak into three peaks, including a tall, central peak and two smaller, lateral peaks. Squares in MR images show position of spectrum within prostate

Fig. 7.2.16a,b  Dynamic, contrast-enhanced MRI (DCE MRI) of the prostate demonstrates rapid, increased uptake of intravenously administered contrast media in area of biopsy-proven prostate cancer (arrows), and slower, less pronounced uptake of contrast media in tissue without evidence of prostate cancer at

biopsy (double-lined arrow). T1-weighted turboFLASH MR image with superimposed map demonstrating velocity and level of contrast uptake (a), and respective signal-intensity-over-time curves of biopsy-proven prostate cancer (arrow in b) and tissue without evidence of prostate cancer (double-lined arrow in b)

1027

1028

7  Pelvis

change rates between different tissue compartments (Futterer et al. 2005; Kiessling et al. 2003; Schlemmer et al. 2004). Injection protocols for DCE MRI usually involve rapid intravenous injection of contrast medium as a bolus, to be followed by a bolus of normal saline solution, which both increases the amount of contrast media in the circulating blood and pushes the initial contrast medium bolus to a more central position, such that arterial inflow in the prostate increases. Different approaches are listed at the bottom of Table 7.2.8. 7.2.2.3

Normal Anatomy

In its shape and position within the body, the human prostate resembles a pear that has been turned upside down. The broad base of the prostate is located cranially and borders on the floor of the urinary bladder. Thus, the floor of the urinary bladder may be elevated by BPH. The narrow apex of the prostate points downward toward the pelvic floor and ends at the urogenital diaphragm. At the urogenital diaphragm, the prostatic urethra leaves the prostate and passes through the external sphincter in its membranous part. Ventrally, the prostate is separated from the symphysis pubis by mixed connective and adipose tissue that contains some venous blood vessels. However, the periproctitic venous plexus, which is rich in venous blood vessels, extends along both lateral aspects of the prostate. Dorsally, the prostate and its narrow periprostatic fat layer are separated from the rectum and perirectal space by a dense fascial layer, Denonvilliers’ fascia. Bilaterally, the neurovascular bundles of the prostate are located in the dorso-lateral angles, lateral to Denonvilliers’ fascia and in between the two layers of the lateral pelvic fascia (the prostatic fascia, which extends along the prostatic capsule, and the levator fascia, which extends laterally along the rectum). Each neurovascular bundle includes arterial and venous blood supply to the prostate and neural supply to both the prostate and the penis. At the prostate, each neurovascular bundle enters through the prostatic capsule at the larger upper pole, which is located at the prostatic base, caudal to the seminal vesicles, and at the smaller lower pole, which is located at the prostatic apex. In the area of the neurovascular poles, the prostatic capsule is less dense than elsewhere. This makes the poles particularly vulnerable for extracapsular extension of prostate cancer (ECE). Specific anatomical findings on unenhanced, T1weighted MR images of the normal prostate include a homogenous, intermediate signal intensity that does not allow the different zones of internal prostatic anatomy to be distinguished (Fig. 7.2.10b). The normal seminal vesicles demonstrate homogenous, medium signal intensity on unenhanced, T1weighted MR images and do not contrast with prostate tissue (Fig 7.2.11c).

On contrast-enhanced, T1-weighted images, the normal prostate shows homogenously high signal throughout the peripheral zone, while the central zone, the periurethral zone, and the transition zones usually demonstrate inhomogeneous contrast uptake (Fig. 7.2.11d). After intravenous administration of gadolinium chelates, the normal seminal vesicles show high signal intensity along their muscular walls and no signal increase within the glandular ducts on T1-weighted MR images. The anatomical model of the zonal anatomy of the normal prostate, which was first described by McNeal (McNeal 1968, 1988) can be seen best on T2-weighted MR images (Fig. 7.2.11). On T2-weighted MR images, the prostatic capsule with its very low signal intensity stands out against both the periprostatic venous plexus and the periprostatic and perirectal adipose and fibroconnective tissues with their mixed high-to-medium signal intensities. The width of the prostatic capsule is approximately 0.1 cm (Fig. 7.2.11). Adjacent to the prostatic capsule at its laterodorsal aspect, the neurovascular bundles are recognized bilaterally as groups of small, low-signal-intensity dots or short lines on T2-weighted MR images. Distinction from the prostatic capsule may be limited at the entry sites of the neurovascular bundles into the prostate, at the upper and lower poles bilaterally. It is here that the dense prostatic capsule has narrow gaps (Fig. 7.2.11). Ventrally to anterolaterally, the prostatic capsule is replaced by the anterior fibromuscular band, which represents a downward extension of the anterior wall of the urinary bladder and shows low to very low signal intensity on T2-weighted MR images (Fig. 7.2.11). Within the confines of the prostatic capsule, the peripheral zone of the prostate makes up approximately 70% of the volume of the prostatic gland. Most of the peripheral zone tissue is located dorsally and laterally within the prostate. On T2-weighted MR images, normal prostate tissue of the peripheral zone shows homogenously bright signal (Fig. 7.2.11). The peripheral zone of the prostate is separated from the central compartments of the prostate gland by a narrow strip of condensed tissue, which is oftentimes referred to as the prostatic pseudocapsule. The prostatic pseudocapsule demonstrates low signal intensity on T2weighted MR images (Fig. 7.2.11). The central compartments of the prostate gland include the ejaculatory ducts, the seminal collicle (verumontanum), the prostatic urethra within the periurethral zone, the central zone, and the transition zone (Fig. 7.2.11). Within the base and the middle of the prostate, the ejaculatory ducts are located bilaterally at the paramedian, dorsal aspect of the prostatic pseudocapsule. The ejaculatory ducts extend to the verumontanum, where their orifices open into the prostatic urethra. On axial, T2weighted MR images, the normal ejaculatory ducts may appear as dots of bright signal intensity (Fig. 7.2.11).

7.2  Male Pelvis

Normally, the prostatic urethra is a median, central structure of high signal intensity on T2-weighted MR images, which extends from the base to the apex of the prostate. The prostatic urethra is surrounded by the periurethral zone, which is composed of fibromuscular tissue and appears as a narrow ring of low signal intensity on axial T2-weighted MR images (Fig. 7.2.11). Lateral and ventral to the periurethral zone, the central zone of the prostate extends from the base to the apex as a tapering structure. The central zone comprises approximately 5 to 10% of prostate gland volume. It is composed of connective tissue and some glandular components and presents with inhomogeneous, medium-to-high signal intensity on T2-weighted MR images (Fig. 7.2.11). Lateral to the central zone, within the confines of the prostatic pseudocapsule, the transition zone is seen bilaterally. Normally, the transition zone makes up approximately 20–25% of prostate gland volume and are similarly large on both sides. The transition zone is composed of varying amounts of stromal, fibromuscular tissue, which demonstrates with low signal intensity on T2-weighted MR images, and glandular tissue with high signal intensity on T2-weighted MR images (Fig. 7.2.11). The seminal vesicles are located cranial to the base of the prostate, at its posterior aspect, dorsal to the posterior wall of the urinary bladder, and bilaterally, lateral to the ampulla of the deferent duct. The excretory ducts of the seminal vesicles and the deferent ducts continue into the ejaculatory ducts of the prostate. The seminal vesicles are bilateral, symmetric, glandular structures that are composed of wide, convoluted glandular ducts, which demonstrate bright signal intensity on T2-weighted MR images, and narrow, fibromuscular walls that show low signal intensity on T2-weighted MR images. Similarly, the ampullae of the deferent ducts are bilateral, symmetric, wide glandular duct structures with narrow, fibromuscular walls whose signal characteristics closely resemble those of the seminal vesicles (Fig. 7.2.11). 7.2.2.4  Pathological Findings 7.2.2.4.1  Groups of Pathologic Conditions (in Order) Congenital Abnormalities Congenital abnormalities of the prostate and seminal vesicles include congenital cysts, which may be associated with other abnormalities of the urogenital system, or with infertility. Congenital cysts are located at the dorsal aspect of the prostate, in median or paramedian positions. Utricle cysts and cysts of the ejaculatory ducts are located inside the prostate, while Müllerian duct cysts are seen on the outside, at the base of the prostate. Differential diagnosis of cystic lesions of the prostate and seminal vesicles is primarily based on the respective

anatomic location of the cyst. In general, simple cystic lesions of the prostate and seminal vesicles demonstrate with homogeneously low signal intensity on T1-weighted MR images and homogeneously high signal intensity on T2-weighted MR images. However, hemorrhage within a cyst may cause signal inhomogeneity and alter signal intensity, which may become high on T1-weighted MR images and may be of any level on T2-weighted MR images. Prostate Trauma While pelvic trauma may alter the position of the prostate, particularly when the membranous urethra is disrupted, MRI is hardly ever applied to investigate trauma to the pelvis or prostate. Prostatic Hemorrhage Hemorrhage inside the prostate may occur spontaneously in patients with BPH, or as a sequel of prostatitis or hematospermia (Nickel 2002). However, most frequently in patients undergoing MRI of the prostate for detection, localization, or staging of prostate cancer, intraprostatic hemorrhage is due to previous biopsy of the prostate. Depending on the respective extent, size, shape, and signal intensities, intraprostatic hemorrhage may mask or mimic other prostatic lesions, particularly prostate cancer (Fig. 7.2.14). It has previously been recommended that MRI of the prostate should be postponed until at least 3–6 weeks after prostate biopsy (Heuck et al. 2003; Kaji et al. 1998). However, more recent research has demonstrated that the interval between prostate biopsy and MR examinations of the prostate should be 6–8 weeks (Qayyum et al. 2004). Inflammatory and Infectious Disease of the Prostate and Seminal Vesicles Prostatitis usually does not represent an indication for MR imaging of the prostate and seminal vesicles. Patients with acute prostatitis (category I of the NIH classification of 1995) usually present with micturition problems and general symptoms of severe infectious disease. Severe pelvic pain may make it impossible to perform an MRI examination, which involves application of an endorectal MRI coil. Chronic prostatitis may be infectious (category II of the NIH classification of 1995) or noninfectious, with different clinical symptoms and laboratory findings (categories III, IIIa, IIIb, and IV of the NIH classification of 1995) and may be found in association with other prostatic disease, which warrants MRI examination. Increase in thickness of intraprostatic septa, inhomogeneity of intraprostatic tissue texture, elongated seminal vesicles, and increase in volume of the periprostatic venous plexus are signs associated with chronic prostatitis. However, those signs are not uniformly found in chronic prostatitis (Nickel 2002). MRI of the prostate in chronic prostatitis may demonstrate thickening of intraprostatic septa and tissue inhomogeneity in the peripheral zone of

1029

1030

7  Pelvis Fig. 7.2.17a,b  MRI of the prostate in chronic prostatitis demonstrates thickening of intraprostatic septa and tissue inhomogeneity in the peripheral zone of the prostate as structures of low signal intensity and diffuse signal decrease, respectively, on axial T2-weighted MR image (square in a). MR spectroscopy usually shows normal ratios of citrate and choline in areas of chronic prostatitis (b)

the prostate as structures of low signal intensity and diffuse signal decrease, respectively, on T2-weighted images (Fig. 7.2.17). Occasionally, chronic prostatitis may be associated with diffuse signal loss throughout the peripheral zone and volume decrease of one side of the prostate. On T2weighted MR images, it may be difficult to distinguish between chronic prostatitis and prostate cancer in such cases (Heuck et al. 2003). In focal, granulomatous prostatitis, the inflammatory reaction within the peripheral zone of the prostate mimics an intraprostatic mass, which may be indistinguishable from prostate cancer on T2weighted MR images (Heuck et al. 2003). Prostate abscesses demonstrate high signal intensity in the liquefied, necrotic center and increased signal intensity in the edematous, inflammatory wall on T2-weighted MR images, while on contrast-enhanced T1-weighted MR images they demonstrate low signal intensity in the liquefied, necrotic center and very high signal intensity in the edematous, inflammatory wall. However, transrectal ultrasonography is usually sufficient to detect and localize a prostate abscess. Chronic inflammation with or without infection of the seminal vesicles frequently results in tissue shrinkage and loss of the high signal intensity within the glandular lumen on T2-weighted MR images. However, it is not possible by means of unenhanced MRI to distinguish between post-inflammatory changes and the consequences of previous hormone deprivation therapy of the prostate, previous radiation therapy, longstanding diabetes mellitus, amyloidosis, and alcoholism. Also, infiltration of the seminal vesicles by prostate cancer, via the ejaculatory ducts or by immediate spread from the base of the prostate to the seminal vesicles, may have an appearance similar to that of post-inflammatory tissue alterations. On contrast-enhanced, T1-weighted MR images of the prostate and seminal vesicles, however, spread of prostate cancer to the seminal vesicles is frequently associated with a contrast-enhancing mass (Futterer et al. 2005; Huch-Böni et al. 1995).

Benign Tumorous Lesions of the Prostate Benign prostatic hyperplasia (BPH) is frequently associated with micturition problems, due to lower urinary tract outflow obstruction, including decrease of urine flow, urge symptoms, and increased frequency of micturition with low voiding volumes and large post-voiding volumes of remnant urine within the urinary bladder. BPH usually affects men of more than 50 years of age, and is prevalent in approximately 60% of men over the age of 60. BPH is associated with a pathologic increase in the volume of the transition zone of the prostate, which may grow symmetrically or asymmetrically. BPH may be glandular and cystic, with a predominant increase of gland tissue, or fibromuscular, with a predominant increase of stromal components of the transition zone. However, frequently, both forms of BPH coexist within the same prostate and cause inhomogeneity of the central gland structures, which makes identification of prostate cancer within the transition zones very difficult for MRI (Maßmann et al. 2003; Schiebler et al. 1989). On T2-weighted MR images, predominantly glandular and cystic BPH presents with inhomogeneous high SI (Fig. 7.2.18). On contrast-enhanced, T1-weighted MR images, cystic and ductal, glandular structures stand out because of their low signal intensity. Predominantly fibromuscular BPH has inhomogeneous, low signal intensity on T2-weighted MR images (Fig. 7.2.18) and may show varying degrees of contrast enhancement on post-contrast T1-weighted MR images. Malignant Tumorous Lesions of the Prostate and Seminal Vesicles Prostate cancer, which is usually composed of adenocarcinoma (Maßmann et al. 2003), is by far the most common malignant tumor entity in men throughout the European Union and North America. The number of men newly diagnosed with prostate cancer exceeds 200,000 per year in the United States and 30,000 per year in Germany. Tumor-related mortality in men affected with prostate cancer approximates 30,000 men per year in the United

7.2  Male Pelvis Fig. 7.2.18a,b  MRI of benign prostatic hyperplasia (BPH) demonstrates enlarged transition zones of the prostate bilaterally with heterogenous, low-to-high signal on axial (a) and coronal (b) T2-weighted TSE MR images (block arrows). Note delineation of central zone (open chevron arrows) prostatic pseudocapsule around enlarged transition zones (curved arrows)

States and exceeds 10,000 in Germany (American Cancer Society 2006; Müller-Lisse and Hofstetter 2003). Prostate cancer prognosis is determined by the tumor stage (Table 7.2.9) and the grade of biological aggressiveness at the time of diagnosis. Improvements in the early detection of PCA are primarily due to reliable laboratory tests for prostate-specific antigen (PSA), which has been applied to detect and follow up on prostate cancer since the early 1990s (Müller-Lisse and Hofstetter 2003). The majority of prostate cancer is now detected at early tumor stages that are clinically asymptomatic and pathologically restricted to the prostate gland and its immediate surroundings (Maßmann et al. 2003, Müller-Lisse and Hofstetter 2003). Locally confined prostate cancer is associated with long-term survival after diagnosis and treatment. While 5-year-survival rate was approximately 67% in 1974–1976, it has increased to almost 100% in 1995– 2001. Respective 10- and 15-year survival rates of 93 and 77% are partly attributable to earlier diagnosis, and partly to improvements in treatment (American Cancer Society). In Germany, PSA testing is not covered by programs for early detection of malignant tumors. Rather, early detection of PCA is restricted to digital rectal examination (DRE), whose sensitivity and specificity for prostate cancer are limited. Only a minor proportion of German men who are entitled to DRE for PCA prevention through their health insurance plan actually make use of this offer (Müller-Lisse and Hofstetter 2003). Punch biopsy of the prostate is indicated in patients with findings suggestive of prostate cancer at DRE and patients with suspicious PSA test results, i.e., PSA serum concentration exceeding 4 ng/ml or PSA increase exceeding 0.75 ng/ml/year. Punch biopsy of the prostate is usually performed transrectally, under guidance by trans­ rectal ultrasonography (TRUS). Systematic, randomized punch biopsy of the prostate is directed at the different sextants of the prostate, namely, the right and left base, middle, and apex (Fig. 7.2.19). Additional biopsies are usually directed at lesions suspicious at DRE (Miller and Weißbach 1999; Müller-Lisse and Hofstetter 2003).

Although MRI, particularly when combined with MR spectroscopy of the prostate, currently represents the most accurate clinical imaging modality for localizing prostate cancer and providing loco-regional staging information, there is yet no clearly established clinical indication for MR examination of the prostate and seminal vesicles. For example, the guidelines for the detection, localization, and staging of prostate cancer issued by Deutsche Gesellschaft für Urologie (German Society for Urology) do not recommend routine use of MR examinations in patients with prostate cancer diagnosed by biopsy of the prostate (Miller and Weißbach 1999). One possible indication for MR examination of the prostate, however, is the need to determine whether radical prostatectomy or radiation therapy is more suitable for patients at medium to high risk for extension of prostate cancer beyond the prostatic capsule. The staging accuracy of MRI currently ranges from 82 to 88%, with a sensitivity of 80–95% and a specificity of 82–93% (Engelbrecht et al. 2002; Heuck et al. 2003). In considering the results of scientific evaluations of MR examinations of the prostate, the examination technique and the quality of images used (Engelbrecht et al. 2002; Heuck et al. 2003) as well as the experience of the radiologists involved (Graser et al. 2007; Mullerad et al. 2004) need to be taken into account. In addition, one must keep in mind that extracapsular extension (ECE) of prostate cancer by more than 1 mm is much easier to detect than ECE of less than 1 mm, although the prognostic significance is currently considered similar. Comparison with TRUS demonstrates the diagnostic superiority of MRI for the localization and staging of prostate cancer. Differential therapy including brachytherapy alone, in prostate cancer of very limited local extent, or brachytherapy in combination with external beam irradiation, in locally more extensive prostate cancer, represents a similar indication for the application of MR examinations of the prostate and seminal vesicles. Biochemical therapy failure, which may be defined as increasing PSA levels at three subsequent tests after completion of therapy, essen-

1031

1032

7  Pelvis Table 7.2.9  TNM classification of prostate cancer, according to the 6th edition of the UICC in 2002, and corresponding cardinal MRI findings (modified and amended, from Heuck et al. 2003 and Wittekind et al. 2002) T stage

Primary tumor

MRI correlation

Tx

Primary tumor cannot be evaluated

Artifact due to motion, therapy, or hemorrhage

T0

No evidence of primary tumor

No evidence of low-signal intensity lesions on T2-weighted MR images

T1

Clinically non-detectable tumor, which is neither palpable nor visible at imaging “Incidental carcinoma“ in 5% or less of resected tissue at transurethral resection of prostate (TUR-P) performed for benign prostatic hyperplasia “Incidental carcinoma“ in more than 5% of resected tissue at transurethral resection of prostate (TUR-P) Tumor diagnosed by punch biopsy or needle biopsy of the prostate, performed because of elevated serum levels of prostate-specific antigen (PSA)

Tumor not detectable; T2-weighted MR images may show small foci of low signal intensity without differentiation between inflammatory lesions, prostatic interepithelial neoplasia, and focal fibrosis

T2-weighted MR images demonstrate lesions of low signal intensity that are confined to the prostate. Lesions may reach into the prostatic capsule without penetrating it

T2c

Tumor confined to prostate Tumor infiltrates less than half of one lobe of the prostate Tumor infiltrates more than half of one lobe of the prostate Tumor infiltrates both lobes of the prostate

T3

Tumor extends beyond the prostatic capsule

T3a

Extracapsular extension (ECE), unilateral or bilateral

T3b

Tumor infiltrates seminal vesicles

T2-weighted MR images demonstrate lesions of low signal intensity that penetrate the prostatic capsule or infiltrate into the seminal vesicles Extracapsular tissue with low signal intensity on T2-weighted MR images, most often found at the neurovascular bundles Lesions of low signal intensity on T2-weighted MR images present as nodules or wall thickening in one or both seminal vesicles, with or without immediate contact to prostate

T4

Tumor infiltrates neighboring structures   (bladder neck, external sphincter muscle,   rectum, levator ani muscle, pelvic side wall)

N stage

Regional lymph nodes

Nx

Regional lymph nodes cannot be evaluated

N0

No evidence of regional lymphadenopathy

N1

Evidence of regional lymphadenopathy

M stage

Distant metastasis

Mx

distant metastasis cannot be evaluated

M0

No evidence of distant metastasis

M1 M1a M1b M1c

Distant metastasis Extraregional lymphadenopathy Bone metastasis Other manifestations

T1a

T1b T1c

T2 T2a T2b

T1c tumors may look like T2-tumors at MRI!

Extracapsular tissue with low signal intensity on T2-weighted MR images extends from the prostate into neighboring structures

7.2  Male Pelvis

tially correlates only with the proportion of tumor tissue in punch biopsies of the prostate, and with the decision for or against MRI in the staging process prior to radiation therapy (Coakley et al. 2003). Still another indication for MR examinations of the prostate lies in the detection and localization of areas within the prostate that are suspicious for harboring PCA, in patients with a suspicion of PCA, and previously

negative results of punch biopsy of the prostate. When MRI and MR spectroscopic imaging findings within the prostate agree, the positive predictive value of combined MRI/MRSI for localizing PCA to a sextant of the prostate ranges between 80% and more than 90% (Amsellem-Ouazana et al. 2005; Scheidler et al. 1999; Yuen et al. 2004), while the negative predictive value exceeds 80% (Scheidler et al. 1999) (Fig. 7.2.20).

Fig. 7.2.19  Division of prostate into sextants (right and left base, middle gland, and apex), as demonstrated on axial, T2-weighted TSE MR images. The base resembles the shape of a cloverleaf and extends from the bladder floor to the level cranial to the axial MR image section with the largest transverse diameter of the prostate. The middle gland resembles the shape of an ellipse

and extends from the level of the largest transverse diameter of the prostate to the caudal (inferior) level of the verumontanum. The apex resembles the shape of a trapezoid and extends from the next caudal level to the urogenital diaphragm. Note prostate cancer with extracapsular extension (arrows). (Modified from Müller-Lisse et al. 2005)

Fig. 7.2.20  Combined MRI and 3D MR spectroscopy at 1.5 T (3D CSI SE sequence, TR/TE 1,050/130 ms). Axial, T2-weighted TSE MR image of middle of prostate (A) with unremarkable right peripheral zone (R) and mass of low signal intensity suspicious for prostate cancer in left peripheral zone (L). MR spectroscopy findings confirm MRI impression. Spectrum from right peripheral zone (B) demonstrates high level of citrate (Integral I, 0.720), low levels of choline and creatine (I, 0.059),

and a (choline + creatine)/citrate ratio of 0.082, which appears typical for healthy peripheral zone tissue. Spectrum from left peripheral zone (C) demonstrates low level of citrate (Integral I, 0.319), high level of choline and low level of creatine (I, 1.140 and I, 0.103, respectively), and a (choline + creatine)/citrate ratio of 3,897, which appears typical for prostate cancer in the peripheral zone. (Modified from Müller-Lisse and Scherr 2003)

1033

1034

7  Pelvis

7.2.2.4.2  Description of Single Pathologic Conditions, Each with Illustrated Example The differential diagnosis of prostate cancer (Figs. 7.2.12, 7.2.13, 7.2.19), benign prostatic hyperplasia (Figs. 7.2.11 and 7.2.18), chronic prostatitis (Fig. 7.2.17), and prostatic hemorrhage (Fig. 7.2.14) is summarized in Table 7.2.10. 7.2.2.5  Indications and Value of MRI The particular value of MRI in the examination of the prostate and seminal vesicles lies in its unique ability to provide cross-sectional images with both high spatial resolution and superb contrast between different tissue structures. On T2-weighted MR images, the zonal anatomy of the prostate as described by McNeal (1968, 1988) is fully appreciated, with high contrast between the different zones of the prostate gland and separating capsular or pseudocapsular structures. The respective ducts of the seminal vesicles and deferent ducts are demonstrated with high contrast when compared to their fibromuscular walls. Among the other cross-sectional imaging modalities, neither ultrasound nor CT distinguishes between different zones of the prostate and different tissue components of the seminal vesicles with the precision of MRI (Figs. 7.2.11, 7.2.19). Consequently, among the cross-sectional imaging modalities, MRI is the modality of choice for the non-invasive localization of tumorous lesions inside the prostate gland. However, in clinical practice, because of the limited access to the prostate for fine needle aspiration or punch biopsy within MRI scanners, TRUS is preferred to guide prostate biopsy. It has long been recognized that the sensitivity of TRUS for tumors within the prostate gland is low (Müller-Lisse and Hofstetter 2003). However, TRUS-guided biopsy of the prostate follows a systematic pattern of obtaining random biopsies throughout the peripheral zone of the gland (Miller and Weißbach 1999; Müller-Lisse and Hofstetter 2003), which does not require that the actual tumor is seen. In most instances, finding any evidence of prostate cancer by means of TRUS-guided prostate biopsy is the only result required by the urologist to suggest a course of therapeutic action to the patient. Still, the proportion of false-negative results of TRUS-guided biopsy has been reported to reach 17–34% (Terris 2002). Thus, an emerging indication for MRI of the prostate lies in the detection and localization within the prostate of tumorous lesions suspicious for prostate cancer (Amsellem-Ouazana et al. 2005; Yuen et al. 2004). MR spectroscopy of the prostate (Kurhanewicz et al. 1996a; Müller-Lisse and Scherr 2003; Scheidler et al. 1999) and dynamic, contrast-enhanced MRI (DCE MRI) of the prostate (Futterer et al. 2005; Kiessling et al. 2003; Schlemmer et al. 2004) are auxiliary methods that increase the specificity of MRI findings, the former by means of biochemical markers, choline and ci-

trate, and the latter by detection of increased micro-vascular density within prostate cancer. It has been demonstrated for combined MRI and MR spectroscopy of the prostate that they are more sensitive and specific than TRUS-guided biopsy in the detection and localization of prostate cancer by sextant (Wefer et al. 2000). The indication first associated with MRI in patients with prostate cancer was the detection of ECE of tumor, either through penetration of the prostatic capsule, or through extension into the seminal vesicles. Recent reviews of literature on MRI of the prostate demonstrate that with application of an endorectal coil and T2-weighted fast or turbo SE sequences in at least two planes of imaging, ECE is detected with a sensitivity of 80–95% and a specificity of 82–93% (Bartolozzi et al. 1996; Engelbrecht et al. 2002; Heuck et al. 2003; Huch-Böni et al. 1995; Nicolas et al. 1994). Although the sensitivity and specificity of MRI for ECE in prostate cancer exceed that of other cross-sectional imaging modalities, MRI is infrequently requested in clinical practice. Rather, urologists and radiation oncologists rely on nomograms that evaluate the information provided by clinical examination (including DRE), PSA serum level, and punch biopsy cores (tumor grade, proportion of biopsy cores affected with prostate cancer) to determine the likelihood of ECE in prostate cancer (Müller-Lisse and Hofstetter 2003; Wang et al. 2004, 2006). While differential therapy suggestions for prostate cancer are often based on such nomograms, it has been shown that the integration of findings at endorectal-coil MRI of the prostate with the information provided by the nomograms leads to significant improvement in the accuracy of prediction (Wang et al. 2004, 2006). The addition of MRSI to MRI further increased accuracy in the prediction of organ-confined prostate cancer, though not by a statistically significant amount (n = 383 patients with combined MRI and MR spectroscopy of the prostate) (Wang et al. 2004) 7.2.2.6  Indications (Symptom-Specific Imaging Modalities) 7.2.2.6.1  Congenital Abnormalities Congenital and cystic lesions of the prostate and seminal vesicles are primarily examined by means of ultrasound. However, complex or unclear situations may require MRI, which offers clearer differentiation between the various zones of the prostate and between the prostate and adjacent pelvic structures, as well as the option of including the entire urogenital tract in one examination. 7.2.2.6.2  Trauma of the Prostate Imaging in pelvic trauma is usually adapted to the expected and observed degree of injury. Ultrasound examination, plain film radiography, and computed to-

7.2  Male Pelvis Table 7.2.10  Differential Diagnosis of Prostate Lesions at MRI Lesion type

Prostate cancer

Benign prostatic hvperplasia

Chronic prostatitis

Hemorrhage

Incidence

2004 populationa   (per 100,000)  Munich: 134.0  Germany: 134.8  Europe: 107.1  World: 72.2

Prevalencec   Age-dependent  60–69 years of age: 50%  >85 years of age: 90%

>80% prevalence in prostate biopsy samples in men  >45 years of age,  with PSA >4 ng/ml or suspicious nodules at digital rectal exame

Variable, frequent   after prostate biopsy

Age

69.3 ± 9.1 yearsa (mean ± SD, Munich 2004)

Usually >50 years

Sex

Male

Male

Male

Male

Localization

Prostate

Prostate

Prostate

Prostate

Prognosis   (survival after therapy)

Locally confined (1995–2001)b  5 years: 100%  10 years: 93%  15 years: 77%

Rarely life-threateningc

Increased risk of prostate cancerc within next 5 years

MRI signals

T2-wMRI Low signal intensity mass Mostly peripheral zone

T2-wMRI Inhomogeneous mostly transition zone

T2-wMRI Low signal intensity Streak- or wedge-like Mostly peripheral zone

T2-wMRI Inhomogeneous May mask or mimic mass

DCEMRI Rapid, increased uptake of contrast media Mostly peripheral Zone

Tl-wMRI Homogenous, unless associated with hemorrhage, Intermediate

Tl-wMRI Homogenous, unless associated with hemorrhage, Intermediate

Tl-wMRI Inhomogeneous May show high signal

MR spectroscopy Decreased citrate Increased choline Mostly peripheral zone

MR spectroscopy Choline may be increased when compared with healthy peripheral zone tissued

MR spectroscopy MR spectrum may be normal; however, signal may be decreased due to scarring

MR spectroscopy MR spectrum may be normal; however, signal maybe decreasedf,g

Granulomatous prostatitis, Chronic prostatitis, Status post-therapy, Hemorrhage, Other pelvic mass invading prostate

Prostate cancer within or invading transitional zone, Hemorrhage, Other pelvic mass invading prostate

Prostate cancer, Status post-therapy, Hemorrhage, Other pelvic mass invading prostate

Prostate cancer, Chronic prostatitis, Other pelvic mass invading prostate

Differential diagnosis

 http://www.tumorregister-muenchen.de/facts/base/GGTSPU-hydra3.fw.med.uni-muenchen.de-4835–2842314-DAT/base_C61_ _G.pdf b  American Cancer Society (2006) Cancer facts and figures 2006. American Cancer Society, 1599 Clifton Road, NE, Atlanta, GA 30329–4251, USA, pp 17–19. www.cancer.org c  Kirby RS (1999) The prostate, men and men’s health. Prostate Cancer Prostatic Dis 2(S4):S2–S4 d  Kurhanewicz J, Vigneron DB, Hricak H, Narayan P, Carroll P, Nelson SJ (1996a) Three-dimensional H-1 MR spectroscopic imaging of the in situ human prostate with high (0.24–0.7-cm3) spatial resolution. Radiology 198: 795–805 e MacLennan GT, Eisenberg R, Fleshman RL, Taylor JM, Fu P, Resnick MI, Gupta S (2006) The influence of chronic inflammation in prostatic carcinogenesis: a 5-year follow-up study. J Urol 176:1012–1016 f  Kaji Y, Kurhanewicz J, Hricak H, Sokolov DL, Huang LR, Nelson SJ, Vigneron DB (1998) Localizing prostate cancer in the presence of postbiopsy changes on MR images: role of proton MR spectroscopic imaging. Radiology 206:785–790 g  Qayyum A, Coakley FV, Lu Y, Olpin JD, Wu L, Yeh BM, Carroll PR, Kurhanewicz J (2004) Organ-confined prostate cancer: effect of prior transrectal biopsy on endorectal MRI and MR spectroscopic imaging. AJR Am J Roentgenol 183:1079–1083 a

1035

1036

7  Pelvis

mography represent first-line imaging tests. They may be followed by functional tests such as retrograde cystourethrography to rule out disruption of the membranous urethra. MRI is hardly ever applied to investigate trauma to the pelvis or prostate but may be used to detect and investigate posttraumatic alterations of pelvic structures. 7.2.2.6.3  Prostatic Hemorrhage Hemorrhage inside the prostate or seminal vesicles can be detected and localized by means of MRI, due to its high sensitivity for blood remnants. Among the cross-sectional imaging modalities, MRI is best suited for finding hemorrhage of the prostate and seminal vesicles. It must be emphasized, though, that intraprostatic hemorrhage may mask or mimic other prostatic lesions, particularly prostate cancer.

7.2.2.6.4  Inflammatory and Infectious Disease of the Prostate and Seminal Vesicles Prostatitis usually does not represent an indication for MRI of the prostate and seminal vesicles. However, since chronic prostatitis is usually symptom-free and may elevate PSA to levels suggestive of prostate cancer, MRI may be used to aide in the differential diagnosis. Abscess of the prostate is usually detected and localized by means of transrectal ultrasonography. MRI may help to detect fluid collections and inflammatory, reactive tissue associated with abscess when other means of examination fail or produce dubious results. 7.2.2.6.5  Benign Tumorous Lesions of the Prostate BPH is usually a clinical diagnosis. Imaging may be used to determine the size and shape of the prostate, its relation to the bladder neck, and secondary bladder wall

Fig. 7.2.21  Schematic diagram for the detection of prostate cancer. PH Patient History,   IC Informed Consent,   DRE Digital Rectal Examination, PSA Prostate-Specific Antigen, TRUS Transrectal Ultra­sono­ graphy, TRUS-Bx TRUS-Guided Punch Biopsy of the Prostate,   MRI Magnetic Resonance Spectroscopy, DCE-MRI Dynamic, Contrast-Enhanced Magnetic Resonance Imaging (Modified and amended from: Deutsche Gesellschaft für Urologie 2002)

7.2  Male Pelvis

hypertrophy and remnant urine volume after voiding. Transabdominal or transrectal ultrasound examinations are usually sufficient for imaging in BPH. However, determination of prostate volume is more precise when based on MRI (Rahmouni et al. 1992). 7.2.2.6.6  Malignant Tumorous Lesions of the Prostate and Seminal Vesicles DRE and PSA testing represent measures in the early detection of PCA. Punch biopsy of the prostate is indicated in patients with findings suspicious of PCA at DRE and patients with suspicious PSA test results, i.e., PSA serum concentration exceeding 4 ng/ml or PSA increase exceeding 0.75 ng/ml/year. Punch biopsy of the prostate is usually performed transrectally, under guidance by trans­ rectal ultrasonography (TRUS). Although there is as yet no clearly established clinical indication for MR examination of the prostate and seminal vesicles, MRI, particularly when combined with MR spectroscopic imaging of the prostate, currently represents the most accurate clinical imaging modality to localize PCA and provide loco-regional staging information on PCA. One indication for MR examinations of the prostate and seminal vesicles under current debate lies in directing targeted biopsy in patients with a suspicion of PCA but with negative previous biopsy results.

3.

4.

5.

6.

7.

8. 9.

10.

7.2.2.7  Diagnostic Procedures • • • • • • • •

Patient history Digital rectal examination (DRE) Transrectal ultrasonography (TRUS) Prostate-specific antigen (PSA) TRUS-guided punch biopsy Magnetic resonance imaging (MRI) Magnetic resonance spectroscopy (MRS) Dynamic, contrast-enhanced magnetic resonance imaging (DCE MRI)

A schematic diagram for the detection of prostate cancer is presented in Fig. 7.2.21.

11.

12.

13. 14.

References 1.

2.

American Cancer Society (2006) Cancer facts and figures 2006. American Cancer Society, 1599 Clifton Road, NE, Atlanta, GA 30329–4251, USA, pp 17–19. www.cancer.org Amsellem-Ouazana D, Younes P, Conquy S, Peyromaure M, Flam T, Debre B, Zerbib M (2005) Negative prostatic biopsies in patients with a high risk of prostate cancer. Is the combination of endorectal MRI and magnetic resonance spectroscopy imaging (MRSI) a useful tool? A preliminary study. Eur Urol 47:582–586

15.

16.

Anastasiadis AG, Lichy MP, Nagele U, Kuczyk MA, Merseburger AS, Hennenlotter J, Corvin S, Sievert KD, Claussen CD, Stenzl A, Schlemmer HP (2006) MRI-guided biopsy of the prostate increases diagnostic performance in men with elevated or increasing PSA levels after previous negative TRUS biopsies. Eur Urol 50:738–749. Epub 2006 Mar 24 Bartolozzi C, Menchi I, Lencioni R et al. (1996) Local staging of prostate carcinoma with endorectal coil MRI: correlation with whole-mount radical prostatectomy specimen. Eur Radiol 6:339–345 Beyersdorff D, Winkel A, Hamm B, Lenk S, Loening SA, Taupitz M (2005a) MR imaging-guided prostate biopsy with a closed MR unit at 1.5 T: initial results. Radiology 234:576–581 Beyersdorff D, Taymoorian K, Knosel T, Schnorr D, Felix R, Hamm B, Bruhn H (2005b) MRI of prostate cancer at 1.5 and 3.0 T: comparison of image quality in tumor detection and staging. AJR Am J Roentgenol 185:1214–1220 Coakley FV, Qayyum A, Kurhanewicz J (2003) Magnetic resonance imaging and spectroscopic imaging of prostate cancer. J Urol 170:69–76 Costello LC, Franklin RB, Narayan P (1999) Citrate in the diagnosis of prostate cancer. Prostate 38:237–245 Engelbrecht MR, Jager GJ, Laheij RJ, Verbeek AL, van Lier HJ, Barentsz JO (2002) Local staging of prostate cancer using magnetic resonance imaging: a meta-analysis. Eur Radiol 12:2294–2302 Frauscher F, Klauser A, Berger AP, Halpern EJ, Feuchtner G, Koppelstaetter F, Pallwein L, Pinggera GM, Weirich H, Horninger W, Bartsch G, zur Nedden D (2003) Sonographie des Prostatakarzinoms. Derzeitiger Stand und Zu­ kunfts­perspektiven. Radiologe 43:455–463 Futterer JJ, Engelbrecht MR, Huisman HJ, Jager GJ, Hulsbergen-van De Kaa CA, Witjes JA, Barentsz JO (2005) Staging prostate cancer with dynamic contrast-enhanced endorectal MR imaging prior to radical prostatectomy: experienced versus less experienced readers. Radiology 237:541–549 Futterer JJ, Heijmink SW, Scheenen TW, Jager GJ, Hulsbergen-Van de Kaa CA, Witjes JA, Barentsz JO (2006) Prostate cancer: local staging at 3-T endorectal MR imaging—early experience. Radiology 238:184–191 Gibbs P, Pickles MD, Turnbull LW (2006) Diffusion imaging of the prostate at 3.0 tesla. Invest Radiol 41:185–88 Graser A, Heuck A, Sommer B, Massmann J, Scheidler J, Reiser M, Müller-Lisse U (2007) Per-sextant localization and staging of prostate cancer: correlation of imaging findings with whole-mount step section histopathology. AJR Am J Roentgenol 188:84–90 Halpern EJ, Ramey JR, Strup SE, Frauscher F, McCue P, Gomella LG (2005) Detection of prostate carcinoma with contrast-enhanced sonography using intermittent harmonic imaging. Cancer 104:2373–2383 Harisinghani MG, Barentsz J, Hahn PF, Deserno WM, Tabatabaei S, van de Kaa CH, de la Rosette J, Weissleder R (2003) Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 348:2491–2499. Erratum in: N Engl J Med 349:1010

1037

1038

7  Pelvis 17. Heerschap A, Jager GJ, van der Graaf M, Barentsz JO, Ruijs SH (1997) Proton MR spectroscopy of the normal human prostate with an endorectal coil and a double spin-echo pulse sequence. Magn Reson Med 37:204–213 18. Heesakkers RA, Futterer JJ, Hovels AM, van den Bosch HC, Scheenen TW, Hoogeveen YL, Barentsz JO (2006) Prostate cancer evaluated with ferumoxtran-10-enhanced T2*-weighted MR Imaging at 1.5 and 3.0 T: early experience. Radiology 239:481–487 19. Heuck A, Scheidler J, Kimmig R, Muller-Lisse U, Steinborn M, Helmberger T, Reiser M (1997) Lymph node staging in cervix carcinomas: the results of high-resolution magnetic resonance tomography (MRT) with a phased-array body coil. RoFo 166:210–214 20. Heuck A, Scheidler J, Sommer B, Graser A, Müller-Lisse UG, Maßmann J (2003) MR-Tomographie des Prostatakarzinoms. Radiologe 43:464–473 21. Hovels AM, Heesakkers RA, Adang EM, Jager GJ, Barentsz JO (2004) Cost-analysis of staging methods for lymph nodes in patients with prostate cancer: MRI with a lymph node-specific contrast agent compared to pelvic lymph node dissection or CT. Eur Radiol 14:1707–1712 22. Hricak H, White S, Vigneron D, Kurhanewicz J, Kosco A, Levin D, Weiss J, Narayan P, Carroll PR (1994) Carcinoma of the prostate gland: MR imaging with pelvic phased-array coils versus integrated endorectal—pelvic phased-array coils. Radiology 193:703–709 23. Huch-Böni RA, Boner JA, Lutolf UM, et al. (1995) Contrast-enhanced endorectal coil MR in local staging of prostate carcinoma. J Comput Assist Tomogr 19:232–237 24. Kaji Y, Kurhanewicz J, Hricak H, Sokolov DL, Huang LR, Nelson SJ, Vigneron DB (1998) Localizing prostate cancer in the presence of postbiopsy changes on MR images: role of proton MR spectroscopic imaging. Radiology 206:785–790 25. Kiessling F, Lichy M, Grobholz R, Farhan N, Heilmann M, Michel MS, Trojan L, Werner A, Rabe J, Delorme S, Kauczor HU, Schlemmer HP (2003) Detektion von Prostatakarzinomen mit T1-gewichteter Kontrastmittel-unterstützter dynamischer MRT. Wertigkeit des Zweikompartimentenmodells. Radiologe 43:474–480 26. Kozlowski P, Chang SD, Jones EC, Berean KW, Chen H, Goldenberg SL (2006) Combined diffusion-weighted and dynamic contrast-enhanced MRI for prostate cancer diagnosis—correlation with biopsy and histopathology. J Magn Reson Imaging 24:108–113 27. Kurhanewicz J, Vigneron DB, Nelson SJ, Hricak H, MacDonald JM, Konety B, Narayan P (1995) Citrate as an in vivo marker to discriminate prostate cancer from benign prostatic hyperplasia and normal prostate peripheral zone: detection via localized proton spectroscopy. Urology 45:459–66 28. Kurhanewicz J, Vigneron DB, Hricak H, Narayan P, Carroll P, Nelson SJ (1996a) Three-dimensional H-1 MR spectroscopic imaging of the in situ human prostate with high (0.24–0.7-cm3) spatial resolution. Radiology 198:795–805

29. Kurhanewicz J, Vigneron DB, Hricak H, Parivar F, Nelson SJ, Shinohara K, Carroll PR (1996b) Prostate cancer: metabolic response to cryosurgery as detected with 3D H-1 MR spectroscopic imaging. Radiology 200:489–496 30. Maßmann J, Funk A, Altwein J, Praetorius M (2003) Prostatakarzinom (PC)—eine organspezifische Neoplasie aus Sicht der Pathologie. Radiologe 43:423–431 31. McNeal J (1988) Normal histology of the prostate. Am J Surg Pathol 12:619–633 32. McNeal JE (1968) Regional morphology and pathology of the prostate. Am J Clin Pathol 49:347–357 33. Miller K, Weißbach L (eds) (1999) Leitlinien zur Diagnostik von Prostatakarzinomen der Deutschen Gesellschaft für Urologie. Urologe [A] 38:388–401 34. Müller-Lisse UG, Swanson MG, Vigneron DB, Hricak H, Bessette A, Males RG, Wood PJ, Noworolski S, Nelson SJ, Barken I, Carroll PR, Kurhanewicz J (2001a) 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. 46:49–57 35. Müller-Lisse UG, Vigneron DB, Hricak H, Swanson MG, Carroll PR, Bessette A, Scheidler J, Srivastava A, Males RG, Cha I, Kurhanewicz J (2001b) Localized prostate cancer: effect of hormone deprivation therapy measured by using combined three-dimensional 1H MR spectroscopy and MR imaging: clinicopathologic case-controlled study. Radiology 221:380–390 36. Müller-Lisse UG, Scherr M (2003) 1H-MR-Spektroskopie der Prostata: Ein Überblick. Radiologe 43:481–488 37. Müller-Lisse UL, Hofstetter A (2003) Urologische Diagnostik des Prostatakarzinoms. Radiologe 43:432–440 38. Müller-Lisse U, Müller-Lisse U, Scheidler J, Klein G, Reiser M (2005) Reproducibility of image interpretation in MRI of the prostate: application of the sextant framework by two different radiologists. Eur Radiol 15:1826–1833 39. Mullerad M, Hricak H, Wang L, Chen HN, Kattan MW, Scardino PT (2004) Prostate cancer: detection of extracapsular extension by genitourinary and general body radiologists at MR imaging. Radiology (232):140–146. Epub 2004 May 27 40. Nickel JC (2002) Prostatitis and related conditions. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ (eds) Campbell’s Urology, 8th ed. Saunders, Philadelphia, London, New York, St. Louis, Sydney, Toronto, pp 603–630 41. Nicolas V, Beese M, Keulers A, Bressel M, Kastendieck H, Huland H (1994) MR-Tomographie des Prostatakarzinoms – Vergleich konventionelle und endorektale MRT. RoFo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 161:319–326 42. Prochnow D, Beyersdorff D, Warmuth C, Taupitz M, Gemeinhardt O, Ludemann L (2005) Implementation of a rapid inversion-prepared dual-contrast gradient echo sequence for quantitative dynamic contrast-enhanced magnetic resonance imaging of the human prostate. Magn Reson Imaging 23:983–990

7.2  Male Pelvis 43. Qayyum A, Coakley FV, Lu Y, Olpin JD, Wu L, Yeh BM, Carroll PR, Kurhanewicz J (2004) Organ-confined prostate cancer: effect of prior transrectal biopsy on endorectal MRI and MR spectroscopic imaging. AJR Am J Roentgenol 183:1079–1083 44. Rahmouni A, Yang A, Tempany CM, Frenkel T, Epstein J, Walsh P, Leichner PK, Ricci C, Zerhouni E (1992) Accuracy of in-vivo assessment of prostatic volume by MRI and transrectal ultrasonography. J Comput Assist Tomogr 16:935–940 45. Sato C, Naganawa S, Nakamura T, Kumada H, Miura S, Takizawa O, Ishigaki T (2005) Differentiation of noncancerous tissue and cancer lesions by apparent diffusion coefficient values in transition and peripheral zones of the prostate. J Magn Reson Imaging 21:258–262 46. Scheidler J, Hricak H, Vigneron DB, Yu KK, Sokolov DL, Huang RL, Zaloudek CJ, Nelson SJ, Carroll PR, Kurhanewicz J (1999) Prostate Cancer: localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study. Radiology 213:473–480 47. Schiebler ML, Tomaszewski JE, Bezzi M, Pollack HM, Kressel HY, Cohen EK, Altman HG, Gefter WB, Wein AJ, Axel L (1989) Prostatic carcinoma and benign prostatic hyperplasia: correlation of high-resolution MR and histopathologic findings. Radiology 172:131–137 48. Schlemmer HP, Merkle J, Grobholz R, Jaeger T, Michel MS, Werner A, Rabe J, van Kaick G (2004) Can pre-operative contrast-enhanced dynamic MR imaging for prostate cancer predict microvessel density in prostatectomy specimens? Eur Radiol 14:309–17 49. Terris MK (2002) Ultrasonography and biopsy of the prostate. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ (eds) Campbell’s urology, 8th edn. Saunders, Philadelphia, pp 3038–3054 50. Villers A, Puech P, Mouton D, Leroy X, Ballereau C, Lemaitre L (2006) Dynamic contrast enhanced, pelvic phased array magnetic resonance imaging of localized prostate cancer for predicting tumor volume: correlation with radical prostatectomy findings. J Urol 176:2432–2437 51. Wang L, Mullerad M, Chen HN, Eberhardt SC, Kattan MW, Scardino PT, Hricak H (2004) Prostate cancer: incremental value of endorectal MR imaging findings for prediction of extracapsular extension. Radiology 232:133–139 52. Wang L, Hricak H, Kattan MW, Chen HN, Scardino PT, Kuroiwa K (2006) Prediction of organ-confined prostate cancer: incremental value of MR imaging and MR spectroscopic imaging to staging nomograms. Radiology 238:597–603 53. Wefer AE, Hricak H, Vigneron DB, Coakley FV, Lu Y, Wefer J, Müller-Lisse U, Carroll PR, Kurhanewicz J (2000) Sextant localization of prostate cancer: comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectroscopy with step section histology. J Urol 164:400–404

54. Wittekind C, Mezer HJ, Bootz F (eds) (2002) UICC–International Union Against Cancer: TNM-Klassifikation maligner Tumoren (6. edn.). Springer, Berlin Heidelberg New York 55. Yuen JS, Thng CH, Tan PH, Khin LW, Phee SJ, Xiao D, Lau WK, Ng WS, Cheng CW (2004) Endorectal magnetic resonance imaging and spectroscopy for the detection of tumor foci in men with prior negative transrectal ultrasound prostate biopsy. J Urol 171:1482–1486

7.2.3  Male Pelvis: Scrotum U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse 7.2.3.1  Introduction The diagnosis of diseases of the scrotum frequently involves imaging examinations. The least invasive and best available imaging modality usually is ultrasonography (US). Depending on the clinical work environment, US may be performed by the urologist or by the radiologist. Cross-sectional imaging of the scrotum by means of MRI is usually reserved for complex or unclear clinical situations, including complex congenital anomaly, unclear cryptorchidism, or invasive cancer, particularly when deemed metastatic. It must be emphasized, though, that local cancer staging by means of imaging is frequently left to US examinations. 7.2.3.2  Examination Techniques 7.2.3.2.1  Patient Positioning and, if Applicable, Patient Preparation General positioning for MRI of the male pelvis is described in Sect. 7.2.1. MRI examinations of the scrotum are usually performed with the patient in the supine position. Preparation of the patient for an MRI examination of the scrotum may be performed in a similar way as preparation for MRI of the penis. However, when a surface coil is used for the examination, elevating the scrotum and supporting it with a foam rubber block or similar device, which is placed between the patient’s inner thighs, may improve imaging results. 7.2.3.2.2  Selection of Coils MRI examinations of the scrotum should be performed in whole-body high-field MR scanners. Recent scientifically evaluated studies were mostly performed at a magnetic field strength of 1.5 T (Andipa et al. 2004; Skiadas

1039

1040

7  Pelvis

Fig. 7.2.22  a Sagittal T2-weighted TSE MR images of the scrotum obtained with a six-element body phased array coil system (A) and with a one-element 20-cm loop coil (B). The testicle (asterisk) demonstrates homogeneously high signal intensity. The rete testis and testicular mediastinum are faintly outlined in A (curved arrows). Except for the testicular hilum in the region of its mediastinum, the testicle is completely surrounded by the low-signal-intensity tunica albuginea testis (double-lined arrows). The epididymis (arrows) is connected to the testis at the hilum through the efferent ductules (open arrowhead in A). b Axial T2-weighted TSE MR images of the scrotum obtained

with a six-element body phased array coil system (C and D). The testicle (asterisk) demonstrates homogeneously high signal intensity. The rete testis and testicular mediastinum are faintly outlined (curved arrows). Except for the testicular hilum in the region of its mediastinum, the testicle is completely surrounded by the low-signal-intensity tunica albuginea testis (double-lined arrows). Along the testicle, the tunica albuginea is inseparable from the tunica vaginalis testis, whose parietal layer shows with low signal intensity (arrowheads). The epididymis (arrows) is connected to the testis at the hilum through the efferent ductules (open arrowhead in C)

et al. 2006; Terai et al. 2006). While historically, small surface coils with diameters of 15–20 cm (6–8 in.) or less were used as receivers, multi-channel phased-array surface coils (PASC) with at least four independent elements are an alternative that can be applied to the advantage of signal-to-noise ratio and image homogeneity (Fig. 7.2.22). In particular, when abdominal or pelvic lymph nodes are included in the examination, it is advantageous to use PASC systems.

7.2.3.2.3  Examination Sequences In many instances, MRI examinations of the scrotum can be performed with T1-weighted and T2-weighted spinecho (SE) or turbo SE (TSE), or fast SE (FSE) sequences. Short-tau inversion recovery (STIR) MR images may be helpful in the detection of fluid collections within the scrotum, e.g., hydroceles, spermatoceles, or pus collections (Table 7.2.11).

7.2  Male Pelvis Table 7.2.11  Standard examination protocol for magnetic resonance imaging of the scrotum with a 15–20 cm (6–8 in.) surface coil at 1.5 T Sequence type

T1-weighted SE

Anatomical region

a

T1-weighted a SE + CM

T2-weighted b TSE

T2-weighted b TSE

T2-weighted b TSE

Scrotum

Scrotum

Scrotum

Scrotum

Scrotum

Imaging plane

Axial

Axial (may add coronal or sagittal)

Axial

Coronal

Sagittal

TR (ms)

500–700

500–700

3,500–5,000

3,500–5,000

3,500–5,000

TE (ms)

15–17

15–17

90–110

90–110

90–110

Imaging matrix

192–256 × 256

192–256 × 256

192–256 × 256

192–256 × 256

192–256 × 256

FOV (mm)

160 × 160–200 × 200

160 × 160–200 × 200

120 × 120–200 × 200

160 × 160–200 × 200

160 × 160–200 × 200

Slice thickness (mm)

3–5

3–5

3–5

3–5

3–5

Interslice gap (mm)

0–1

0–1

0–1

0–1

0–1

CM intravenous contrast media a Alternatively, the scrotum may be examined with a T1-weighted turbo or fast SE (TSE or FSE) sequence b A short-tau inversion recovery (STIR) sequence may alternatively or additionally be applied in at least one plane of imaging for the examination of fluid collections within the scrotum, such as testicular hydrocele

7.2.3.2.4  Imaging Planes In MRI examinations of the scrotum, it will usually be better to obtain MR images in more than one plane of imaging. Protocol suggestions are provided in Table 7.2.11. 7.2.3.2.5  Thickness of Slices Since the majority of structures within the scrotum are of rather small size, MR imaging should be performed with thin slices. At 1.5 T, suggested slice thickness would be 3–5 mm, with an interslice gap of 0–1 mm (Table 7.2.11). However, when a PASC system is applied, and coverage is extended to the pelvis and lower abdomen, slice thickness may be increased. Respective protocol extensions can be adapted from suggestions for MR imaging of the prostate or urinary bladder. 7.2.3.2.6  Preferred Coverage In MRI examinations of the scrotum, preferred coverage extends from the middle of the penile shaft anteriorly to the perineum posteriorly, and from the cranial edge of the symphysis pubis to the most caudal point of the scro-

tum. Within the confines of a 15–20 cm (6–8 in.) surface coil, it will be necessary to include the entire scrotum and the anatomic structures of its immediate surroundings, such as the inguinal part of the spermatic cord and the inguinal lymph nodes. However, when a PASC system is applied, it may be possible to cover both the primary region of interest, i.e., the scrotum and its adjacent organs and tissues, and the pelvis and lower abdomen, possibly to the level of the renal hilus, to include inguinal, parailiac, and para-aortic lymph node stations. 7.2.3.2.7  Use of Contrast Medium Intravenous contrast media that are based on gadolinium (Gd) chelates have been applied in the majority of studies of MRI of the scrotum. 7.2.3.3  Normal Anatomy The scrotum is composed of a skin layer and a subcutaneous layer, the tunica dartos, which contains both fibroelastic tissue and abundant smooth muscle, but very little or no adipose tissue. Since the contents of the scrotum des­ cend from the abdominal peritoneal cavity during their

1041

1042

7  Pelvis

embryological development, they are covered by a duplicate layer of peritoneum, the parietal, and the visceral peritoneal layers, which together are referred to as the tunica vaginalis testis. The fully developed, normal testis has fully descended into the tunica vaginalis testis, and the latter is usually stated in textbooks of human anatomy to have no remnant aperture to the peritoneal cavity, but to exist as a cavity of its own, the cavum scroti. However, this view has recently been challenged by the argument that the adult human testis is really intraperitoneal but may appear extraperitoneal, and that differences in appearance are likely to result from differences in the relative size of the tunica vaginalis testis between infant boys and elderly men (Pham et al. 2005). In adults, the normal testicle is an oval structure that measures 4–5 cm along its craniocaudal axis and 2–3 cm along its anteroposterior axis. Each testicle is composed of 200–300 pyramid-shaped lobules, and each lobule contains 400–600 seminiferous tubules (Andipa et al. 2004). Inside the cavum scroti, the testis is fully

covered by a dense, fibrous capsule, the tunica albuginea testis. Parts of the tunica albuginea testis enter the testicle at the mediastinum. Septa extend from the mediastinum and divide the testicle into pyramid-shaped lobules (Andipa et al. 2004; Fig. 7.2.22). The mediastinum is located at the posterior aspect of the testis and connects the testis to both its blood supply through the rete testis, and to the epididymis through the seminiferous tubules (tubuli seminiferi) and the efferent ductules (ductuli efferentes). The efferent ductules enter the head of the epididymis (caput epididymis) at the posteromedial aspect of the testis, and collect in the duct of the epididymis (ductus epididymis) in its body (corpus epididymis) and tail (cauda epididymis). The duct of the epididymis continues into the vas deferens (ductus deferens), which is part of the spermatic cord. The spermatic cord connects the testicle and the epididymis to the body. It begins at the deep inguinal ring and descends vertically into the scrotum. The spermatic cord is composed of the deferent duct, the testicular artery, the

Fig. 7.2.23  a Coronal, T2weighted MR images with fat suppression (A and B) were obtained with a 20-cm surface coil. Spermatic cords (arrows) include vas deferens, accompanying arterial and venous testicular vessels, sympathetic nerves of the testicular plexus, and the different layers of the tunicae funiculi spermatici et testis. Note superficial inguinal lymph node (asterisk in A). b Axial, T1-weighted MR images (C and D) were obtained with a phased-array surface coil (PASC) system. Spermatic cords (arrows) include vas deferens, accompanying arterial and venous testicular vessels, sympathetic nerves of the testicular plexus, and the different layers of the tunicae funiculi spermatici et testis. Note right-sided inguinal canal (doublelined arrow in C) and left-sided inguinal hernia containing fat tissue (asterisk in C)

7.2  Male Pelvis

cremasteric artery, the deferential artery, the pampiniform plexus, the genitofemoral nerve, and the lymphatic vessels (Dogra et al. 2003; Fig. 7.2.23). Testicular arteries contribute most to the vascular supply of the scrotum. They arise from the abdominal aorta below the renal arteries, cross the ureters and the external iliac arteries and enter the spermatic cord at the inguinal ring. Testicular arteries branch at the upper pole of the testicle, penetrate the tunica albuginea, and form the tunica vasculosa. Scrotal veins correspond to the arteries. The pampiniform plexus surrounds the epididymis and continues through the inguinal ring as the testicular vein. The right testicular vein empties into the inferior vena cava and the left into the left renal vein (Andipa et al. 2004). The scrotal wall usually appears with low signal intensity on T2-weighted MR images (Andipa et al. 2004). On T1-weighted MR images, the normal testis shows homogenous, intermediate signal intensity. On T2-weighted MR images, the normal testis demonstrates homogenous, high signal intensity (Barbaric 1994), which is slightly lower than the signal intensity of water (Andipa et al. 2004). Contrast enhancement of the normal testis after the intravenous administration of contrast media is usually strong and homogenous. The tunica albuginea measures about 1 mm in thickness and has low signal intensity on both T1-weighted and T2-weighted MR images (Andipa et al. 2004; Fig. 7.2.22). The epididymis demonstrates intermediate signal intensity on T1-weighted MR images and is relatively isointense to the testicle. On T2-weighted MR images, the epididymis is of intermediate intensity, lower than that of the testis. The vas deferens may be distinguished on T2weighted MR images as tubular formations, with a dark wall and slightly bright lumen (Andipa et al. 2004). The vascular structures of the spermatic cord may show with low or intermediate signal intensity on T1-weighted or T2-weighted MR images (Andipa et al. 2004). 7.2.3.4  Pathological Findings 7.2.3.4.1  Groups of Pathologic Conditions (in Order) Congenital and Developmental Congenital anomalies affecting the scrotum include missing or supernumerary testes, such as in anorchism, monorchism, polyorchism, undescended testis (cryptorchidism) and ectopic testis. Other, rare anomalies include penoscrotal transposition, cystic dysplasia, and testicular microlithiasis (Barbaric 1994). The absence of one testis (monorchism) or both testes (anorchism) represents a rare differential diagnosis of undescended testis. The incidence of undescended testis (cryptorchidism) may be as high as 33% in premature neonates, while it is about 3% in full-term neonates, and 1% in boys and young adults. Treatment differs between

patients of different ages. At one year of age, orchidopexy is usually performed, while in patients between puberty and 32 years of age, orchiectomy has been advised due to the sevenfold risk of testicular cancer associated with retained cryptorchid testis (Barbaric 1994). Particularly in low-lying undescended testis near the inguinal canal, MRI has been favored as the method of choice to detect the testis. However, higher up in the abdominal cavity, differentiation of an undescended testis from loops of bowel or lymphadenopathy may be impossible. The undescended testis may present with normal signal intensity at MRI (Figs. 7.2.24, 7.2.25). However, when the undescended testis is atrophied, signal intensity may be lower than normal, particularly on T2-weighted MR images (Barbaric 1994). As a differential diagnosis to the undescended testis, the testis may be ectopic. Common sites for testicular ectopy include the femoral triangle and the perineum. Since ectopic testes are usually readily found on physical examination (Barbaric 1994), imaging by means of MRI is not necessary in such instances. Polyorchidism is a rare congenital anomaly in which there are three or more testes. Since polyorchidism has been associated with increased risks of inguinal hernia, testicular torsion, undescended or maldescended testis, and malignant testicular tumor, imaging by means of US or MRI is warranted to find additional testicular tissue (Amodio et al. 2004; Barbaric 1994; Oner et al. 2005). Since polyorchidism has characteristic sonographic features, the diagnosis is often made on the basis of sono­ graphy. MR imaging can be diagnostic, but is more helpful in cases associated with cryptorchidism or neoplasia (Oner et al. 2005). Testes may remain intraperitoneal when the tunica vaginalis testis maintains a large aperture toward the peritoneal cavity or is ectopic. Examples of intraperitoneal testes are found in cases of gastroschisis, where the testis may protrude through the periumbilical defect along with the bowel, testicular torsion, with increased mobility of the testis within the peritoneum, and “bellclapper” testis (Pham et al. 2005). Bell-clapper testis refers to a capacious tunica vaginalis testis, which completely envelops both the testis and the epididymis and promotes intravaginal testicular torsion (Barbaric 1994). Trauma Trauma to the scrotum may be blunt or penetrating. While patients with acute, penetrating trauma to the scrotum do not usually come to the attention of a radiologist for an MRI examination, patients with blunt trauma of the scrotum may be referred to MRI, particularly when the extent of trauma is considered large or complex. Blunt trauma is frequently associated with hemorrhage into one or more of the scrotal compartments. While subcutaneous hemorrhage results in hematoma, blood loss into the scrotal cavity induces a hematocele. Contusion of the testis may lead to intratesticular hemorrhage with subsequent he-

1043

1044

7  Pelvis

Fig. 7.2.24a–d  Undescended testicles (asterisks) and associated structures of the spermatic cord and epididymis (arrows) are demonstrated at the external aspects of the inguinal canals on coronal (a) and axial (b) T2-weighted MR images with fat

Fig. 7.2.25a,b  Undescended left testicle (asterisk) and associated structures of the spermatic cord and epididymis (arrows) are demonstrated at the external aspect of the left inguinal canal on axial (a) and coronal (b) T2-weighted MR images

suppression, and on axial T1-weighted MR images prior to (c) and after contrast administration and fat suppression (d). Note empty scrotal sack (arrowheads in a)

7.2  Male Pelvis

Fig. 7.2.26a–e  Hematoma of the scrotum demonstrates with increased signal intensity along the spermatic cord (arrows in a) and within the epididymis (arrow in b) on unenhanced, axial T1-weighted MR images. Fat-suppressed, contrast-enhanced T1-weighted MR images (c, e axial; d coronal) prove hematoma and associated inflammatory reaction with contrast enhancement (arrowheads in c–e) and rule out fat tissue as a source of bright signal in a and b. Small remnants of blood within the testicle (curved arrows) appear with bright signal in b and slightly decreased contrast uptake in d and e

1045

1046

7  Pelvis

matoma, while rupture of the tunica albuginea testis with or without fracture of the testis may cause a combination of both testicular hematoma and hematocele. On T1-weighted MR images, acute hemorrhage with hematoma or hematocele demonstrates intermediate to high signal intensity (Fig. 7.2.26). On T2-weighted MR images, acute hemorrhage in the scrotum is associated with signal increase. On both T1-weighted and T2weighted MR images, chronic hematocele demonstrates high signal intensity. Rupture of the tunica albuginea testis may present with a breach of the low-signal-intensity layer of the tunica albuginea and is frequently associated with parenchymal hematoma within the testis. Vascularity and Perfusion Testicular Torsion Testicular torsion occurs when the testis rotates around its longitudinal axis, such that its neurovascular bundle and the spermatic cord become twisted and compressed in themselves. In intravaginal torsion, the testis and epididymis turn within an unusually wide tunica vaginalis testis. The underlying disorder is insufficient attachment of the epididymis to the posterior scrotal wall (Barbaric 1994). In extravaginal torsion, there is insufficient or lax attachment of the tunica vaginalis testis to the scrotal wall, which permits torsion of the testis, epididymis, and tunica vaginalis testis around the spermatic cord (Barbaric 1994). Acute testicular torsion with complete interruption of blood supply to the affected testis is usually associated with an increase of testicular volume. However, subsequently, within 48 to 72 h, testicular swelling subsides, and the affected testis may even diminish in size when compared to the unaffected, normal, contralateral testis. On both T1-weighted and T2-weighted MR images, the testis that is affected by testicular torsion initially demonstrates normal, homogenous signal intensity. However, inhomogeneous signal intensity may be observed on T2-weighted MR images in individual cases. On contrast-enhanced, T1-weighted MR images, there is usually a lack of contrast uptake by the affected testis and epididymis when testicular torsion is complete. However, in individual cases of complete testicular torsion, markedly reduced contrast uptake by the affected testis has been observed (Czipull and Asmussen 2006; Terai et al. 2006). Another sign that is characteristic of testicular torsion consists in visible torsion of the spermatic cord, with rotation of the neurovascular bundle around the long axis of the vas deferens (Czipull and Asmussen 2006). Finally, after prolonged ischemia of the testicle, i.e., more than 4–6 h, hemorrhagic transformation of the testicle and epididymis may result in inhomogeneous signal intensity on both T1-weighted and T2-weighted MR images (Czipull and Asmussen 2006). When compared to surgical findings and/or subsequent clinical outcomes, complete lack of testicular contrast enhancement at MRI in pa-

tients presenting with an acute scrotum correctly diagnosed testicular torsion in 93% (13 of 14) of patients in the study by Terai et al. (2006). One of ten patients with nonspecific MRI findings of reduced contrast enhancement had testicular torsion. Three patients with appendiceal torsion and 13 patients with epididymitis were correctly ruled out from having testicular torsion. However, five of six patients with clinical suspicion of intermittent torsion underwent surgical exploration, although MRI detected testicular perfusion. Thus, although MRI proves highly accurate for the diagnosis of testicular torsion, it apparently cannot be used to rule out intermittent torsion. Also, the clinical use of MRI findings negative for testicular torsion appears less than 100% specific (Terai et al. 2006). Incomplete testicular torsion may result in inhomogeneous uptake of contrast media in the testis, due to inhomogeneous perfusion. While the epididymis is usually normal in its appearance and enhancement pattern, the testis may show distinctly non-enhancing areas, which appear band-like (Czipull and Asmussen 2006). Post-ischemic alterations of the testis include involution, fibrosis, and a loss of volume of the affected testicle. These changes occur within weeks to months after testicular torsion. On T2-weighted MR images, fibrosis within the testis shows with decreased signal intensity when compared with normal testicular tissue and may appear band-like in partial fibrosis and homogenously dark in complete testicular fibrosis (Czipull and Asmussen 2006). The differential diagnosis of testicular torsion and its sequels includes iatrogenic injury to the testis and epididymis, such as in inguinal hernia repair, ligature of the vas deferens, or antegrade occlusion of varicoceles, and trauma, when arterial or venous vessels along the spermatic cord are injured or obliterated (Czipull and Asmussen 2006). Subacute or partial testicular torsion may be difficult to distinguish by means of MRI from inflammatory changes associated with epididymitis or orchitis (Terai et al. 2006). However, normal appearance of the epididymis in the presence of decreased testicular volume and low testicular signal intensity imply subacute or chronic, incomplete testicular torsion rather than epididymitis or orchitis (Czipull and Asmussen 2006; Terai et al. 2006). Testicular Infarction Radiological imaging in patients presenting with testicular pain and no prior history of trauma aims at avoiding unnecessary surgical exploration and, in particular, unnecessary orchiectomy (Fitoz et al. 2006; Sentilhes et al. 2002). It has been demonstrated that segmental testicular infarction can be detected by means of MR imaging. MRI signs suggestive of segmental testicular infarction include the presence of a triangular-shaped, avascular, intra-testicular lesion and enhancement of the surrounding borders after intravenous contrast administration (Fitoz et

7.2  Male Pelvis

al. 2006). When surgical exploration is forgone in favor of follow-up with ultrasonography after 3 months, re-vascularization of testicular vessels and reduction of testicular lesion size may be observed (Sentilhes et al. 2002). Testicular Varicoceles Testicular varicocele is a relatively common condition, which may present with scrotal pain and swelling. Testicular varicocele may be defined as an abnormal degree of venous dilatation of the pampiniform plexus, which may be associated with male subfertility. Clinically, testicular varicocele can be diagnosed when a Valsalva maneuver (expiration against a closed glottis) results in abnormal distension of the pampiniform plexus. Further diagnosis is currently based on ultrasonography with a high-frequency (7 MHz or more) transducer. According to Beddy et al. (2005), different imaging modalities and techniques and respective diagnostic criteria are available. B-mode ultrasonography may demonstrate tortuous, anechoic, tubular structures adjacent to the testis. At least one of two or more prominent veins (more than 2–3 mm in diameter) in the pampiniform plexus should expand when the Valsalva maneuver is performed in an upright position. Color Doppler ultrasonography demonstrates reflux in the spermatic vein, which increases with Valsalva maneuver. Venography shows incompetence and dilation of the internal spermatic vein, with reflux into the abdominal, inguinal, scrotal or pelvic portions of the spermatic vein, and, possibly, venous collaterals. Contrastenhanced MRI (with gadolinium chelates) in the venous phase demonstrates dilated veins and prominence of the pampiniform plexus. Scintigraphy with technetium-99mlabeled red blood cells may show intrascrotal accumulation of the labeled red cells on supine and/or erect static images. Reflux in the spermatic vein may be shown on dynamic images (Beddy et al. 2005).

associated with multiple bilateral, elastic, firm nodules of the epididymis. B-mode ultrasonography showed irregularly shaped hypoechoic masses of the epididymis. MR imaging demonstrated enlarged, heterogeneous, nodular epididymis bilaterally, with slightly increased signal intensity on T2-weighted MR images, but without any signs of testicular involvement. Iida et al. (2003) describe the application of MR imaging in the diagnosis of a urethral fistula with a large subcutaneous abscess affecting the perineum and scrotum in a patient who complained of left scrotal swelling and fever. The urethral fistula as the underlying condition was proved by endoscopy and open surgical debridement. Senzaki et al. (2001) report on a rare case of testicular tuberculosis that did not affect the epididymis and presented clinically with painless swelling of the right scrotal contents, and elastic hard and smooth induration in the right testis was palpable. On imaging, both ultra­ sono­graphy and MRI revealed a well-defined nodule in the right testis. Orchiectomy was performed for suspicion of testicular malignancy. Pathology revealed tuberculotic granuloma with necrotic caseation. However, the epididymis was histologically normal.

Intratesticular Arteriovenous Malformation Arteriovenous malformations (AVM) of the male genitalia are rare and may affect the penis, scrotum, spermatic cord, and epididymis. Association of intratesticular AVM with infertility has been reported (Skiadas et al. 2006). Intratesticular AVM and its peripheral feeding and draining vessels may delineate clearly on T2-weighted MR images and show homogenous, early and intense enhancement on dynamic contrast-enhanced T1-weighted MR images.

Neoplastic Conditions Much neoplastic pathology may cause scrotal masses. While ultrasonography is frequently the first and only cross-sectional imaging modality applied to detect, localize, and characterize a scrotal mass, MRI has been applied to study various neoplastic conditions of the scrotum. Benign testicular neoplasms include teratoma, fibroma, rhabdomyoma, and adenoma. Malignant testicular tumors generally fall into one of three categories. Germ cell tumors account for up to 95% of malignant testicular neoplasms and are either seminomatous (testicular seminoma) or non-seminomatous. The latter include teratocarcinoma, embryonal tumors, chorionic carcinoma (chorionic epithelial carcinoma), and mixed tumors that include seminomatous parts. Gonadal stroma tumors account for about 4% of malignant testicular neoplasms and are frequently less malignant. They include Leydig cell tumors, Sertoli cell tumors, and granulosa cell tumors. Secondary testicular neoplasms are rare and account for about 1% of malignant testicular neoplasms. Among the secondary malignant testicular neoplasms, malignant lymphoma is relatively frequent (Haag et al. 2007).

Inflammatory Disease MR imaging has rarely been applied to the diagnosis of scrotal inflammatory disease. However, several case reports imply that MRI may be useful in some instances of inflammatory disease affecting the scrotum. Although sarcoidosis is a multisystem disorder, it rarely involves the genitourinary tract. Kodama et al. (2004) present a patient with bilateral epididymal sarcoidosis without radiographic evidence of intrathoracic lesion. Painless, non-tender, bilateral scrotal swelling was

Benign Neoplasms MR imaging of the testicles may help in the evaluation of male patients with congenital adrenal hyperplasia (CAH) caused by 21-hydroxylase deficiency, both to find or rule out testicular adrenal rest tumors (TART), and to follow up on TART tumors either under glucocorticoid therapy or after testis-sparing surgery (Claahsen et al. 2007). Findings at physical examination of the scrotum and testicles may be normal in patients with CAH even in the presence of testicular lesions that can be demonstrated

1047

1048

7  Pelvis

Fig. 7.2.27a–e  Testicular adrenal rest tumors (TART) in a male patient with congenital adrenal hyperplasia (CAH) caused by 21-hydroxylase deficiency. TART tumors (arrows) usually demarcate well and present with hypointensity on T2-weighted MR images (a), iso- or hyperintensity on T1-weighted MR images (b) and with remarkable contrast enhancement (c) when compared with normal testicular tissue. They are usually found adjacent to the mediastinum testis (double-lined arrows)

by ultrasonography and MR imaging (Fitoz et al. 2006). TART tumors are typically located around the mediastinum testis and usually delineate well (Fitoz et al. 2006; Stikkelbroeck et al. 2003). However, delineation of lesion margins appears better at MRI than at ultrasonography (Stikkelbroeck et al. 2003). TART tumors show with homogenous hypo-echogenicity on ultrasound (Fitoz et al. 2006). In the study by Stikkelbroeck et al. (2003), 17 of 20 TART tumors that were smaller than 2 cm in diameter were hypoechoic, while all 11 lesions larger than 2 cm showed hyperechoic reflections. TART tumors present with iso- or hyperintensity on T1-weighted and hypointensity on T2-weighted MR images when compared to normal testicular parenchyma (Fitoz et al. 2006; Stikkelbroeck et al. 2003; Fig. 7.2.27). TART tumors demonstrate increased vascularization on power Doppler ultrasonography (Fitoz et al. 2006) and usually present with remarkable contrast enhancement at MRI (Fitoz et al. 2006; Stikkelbroeck et al. 2003; Fig. 7.2.27). Contrast

resolution of TART tumors is better at MRI than with ultrasonography (Fitoz et al. 2006). Testicular epidermoid cyst has also been analyzed by means of MRI of the scrotum. T2-weighted MR images typically demonstrate a testicular mass of high signal intensity, with or without foci of low signal intensity, which is surrounded by a rim of low signal intensity (Cho et al. 2002; Yoshida 2004). Typically, the intratesticular mass presents with slightly decreased signal intensity and a peripheral low-signal-intensity rim on T1-weighted MR images (Yoshida 2004). Testicular epidermoid cyst usually does not show enhancement on contrast-enhanced T1-weighted MR images (Cho et al. 2002; Yoshida 2004). Testicular epidermoid cysts may present with a bull’s-eye appearance, which has been considered to depend on the presence of calcification (Yoshida 2004). Clinically, testicular epidermoid cyst may present as a painless, nontender, elastic hard mass in the testicle (Yoshida 2004). B-mode ultrasonography findings include a markedly

7.2  Male Pelvis

Fig. 7.2.28a–c  Left-sided testicular hydrocele (asterisk) with watery, serous contents presents with low SI on T1-weighted MR images (a sagittal, fat-suppressed, contrast-enhanced T1weighted MR image) and high SI on T2-weighted MR images (b sagittal, c axial). The contents do not enhance with contrast media (a). Note left testicle (arrow)

heterogeneous intratesticular mass, with or without alternating hypo- and hyperechoic layers, which is surrounded by a hypoechoic or echogenic rim. Color Doppler ultrasonography reveals absence of flow within the mass (Cho et al. 2002). Testicular hydrocele is an unusually large collection of serous fluid (transudate, with low protein content) within the tunica vaginalis testis. Hydrocele may also be found along the processus vaginalis peritonei, the extension of the peritoneum toward the cavum scroti. In these instances, hydroceles are also being referred to as funiculoceles (Haag et al. 2007). Due to their watery, serous contents, testicular hydroceles and funiculoceles present

with low SI on T1-weighted and high SI on T2-weighted MR images. The contents do not enhance with contrast media (Fig. 7.2.28). Malignant Neoplasms Among primary malignant testicular neoplasms, distinction can be made between seminomatous and non-seminomatous testicular malignancies. Currently, treatment differs between the two tumor types. Usually, surgical pathology after orchiectomy determines both whether a primary malignant testicular neoplasm is seminomatous or non-seminomatous and which tumor stage applies to the respective malignant neoplasm (Table 7.2.12). How-

1049

1050

7  Pelvis Table 7.2.12  Staging of primary malignant testicular neoplasms (modified from Wittekind et al. 2002) TNM class

Description

T-classification (primary tumor)a pTx

Primary tumor cannot be evaluated

pT0

No evidence of primary tumor

pTis

Intratubular tumor (carcinoma in situ)

pT1

Tumor limited to testis and epididymis, without infiltration of blood or lymph vessels; tumor may infiltrate tunica albuginea, but not tunica vaginalis testis

pT2

Tumor limited to testis and epididymis, with infiltration of blood or lymph vessels or tumor infiltration of tunica albuginea and tunica vaginalis testis

pT3

Tumor infiltrates spermatic cord, with or without infiltration of blood or lymph vessels

pT4

Tumor infiltrates scrotum with or without infiltration of blood or lymph vessels

N classification (regional lymph nodes) Nx

Regional lymph nodes cannot be evaluated

N0

No evidence of regional lymphadenopathy

N1

Metastasis in single or multiple lymph nodes of up to 2 cm in maximum diameter

N2

Metastasis in single or multiple lymph nodes, each with maximum diameter between 2 and 5 cm

N3

Metastasis in single or multiple lymph nodes with diameters of more than 5 cm

M classification (distant metastasis) Mx

Distant metastasis cannot be evaluated

M0

No evidence of distant metastasis

M1 M1a M1b

Distant metastasis Distant metastasis in non-regional lymph nodes or pulmonary metastasis Other distant metastasis

S classification (serum tumor markers) Sx

Serum tumor markers not available or not measured

S0

Serum tumor markers within normal ranges LDHb

B-HCG (mIU/ml)b

AFP (ng/ml)b

S1

<1.5 × N

and

<5,000

<1,000

S2

1.5–10 × N

or

5,000–50,000

1,000–10,000

S3

>10 × N

or

>50,000

>10,000

N upper limit of normal range a T classification is based on surgical pathology after radical orchiectomy b  When compared with previous versions of the TNM classification system of primary malignant testicular tumors, the TNM 2002 now includes serum tumor markers LDH, B-HCG, and AFP. Serum tumor markers should be measured immediately after surgery and then serially in the follow-up

7.2  Male Pelvis

ever, attempts have been made to distinguish between seminomatous and non-seminomatous testicular neoplasms preoperatively. Johnson et al. (1990) describe the MR imaging features of six seminomatous and nine non-seminomatous testicular tumors, the latter including teratoma, teratocarcinoma, embryonal cell carcinoma, and choriocarcinoma. Non-seminomatous tumors showed marked heterogeneity of signals on both proton-density and T2-weighted MR images. Within the non-seminomatous tumors, Johnson et al. (1990) describe the typical background signal as being nearly equal to normal testicular tissue, while some regions were less intense and others more intense than normal testicular tissue. Johnson et al. (1990) characterized non-seminomatous tumors as having a dark, peripheral band that correlates with a fibrous tumor capsule on histologic examination. In contradistinction, seminomatous testicular tumors typically lack a capsule on MR images and are isointense with normal testis on proton-density images and consistently hypointense and relatively homogeneous on T2-weighted MR images (Johnson et al. 1990; Menzner et al. 1997; Fig. 7.2.29). However, hemorrhage within a seminoma may mimic a

non-seminomatous testicular neoplasm (Johnson et al. 1990). Menzner et al. (1997) found that differentiation of scrotal diseases, including seminoma, teratoma and inflammation, improved when visual MR image analysis was supported by using a statistical score that was based upon the distribution of the variate extensions of elements inside the pathological area and their maximal and minimal signal intensities, the contrast pattern of the pathological area, and the visibility of healthy tissue in the pathological testicle. Secondary tumors of the testicles are rare. Metastasis of malignant melanoma accounts for about 15% of secondary testicular neoplasms. Very rarely, testicular metastasis may be the first manifestation of malignant melanoma. Clinically, testicular metastasis of malignant melanoma may present with rapidly developing, painless swelling of the testicle. Ultrasonography demonstrates an inhomogeneous testicular tumor (Bothig et al. 2006). Non-Hodgkin lymphoma may affect the testicle (Fig. 7.2.30). Ultrasonography and MRI findings, along with normal serum markers (alpha-fetoprotein [AFP] and beta-human chorionic gonadotropin [B-HCG]) help to differentiate non-Hodgkin’s lymphoma and other sec-

Fig. 7.2.29a–d  Grossly enlarged, right-sided testicle in a baby boy (arrows) demonstrates homogenously high signal intensity on both T2-weighted, fat-suppressed coronal (a,d) and axial (c) MR images and coronal T1-weighted MR images after contrast administration (b). MRI findings do not characterize the tumor sufficiently to allow a diagnosis to be made. Note normal left testicle (arrowheads in a–d)

1051

1052

7  Pelvis

ondary malignant testicular neoplasms from germinal testicular tumors (Haag et al. 2007; Vara et al. 2006). However, it must be emphasized that, based on MRI findings alone, differentiation between primary and secondary solid testicular neoplasms, or between benign and malignant solid testicular neoplasms is currently not reliably possible. 7.2.3.4.2  Description of Single Pathologic Conditions Signal intensities and enhancement patterns of different benign and malignant testicular tumors are summarized in Table 7.2.13. 7.2.3.5  Indications and Value of MRI In general, the imaging method of choice in scrotal lesions is ultrasonography, with MRI being an adjunct method in complex or complicating cases. In particular, ultrasonography is a sensitive imaging technique for the detection and exact localization of scrotal masses. Ultrasonography appearance may be useful in the differentiation of benign and malignant abnormalities.

Presence or clinical suspicion of a scrotal mass is an indication for scrotal imaging. The goal of imaging in a scrotal mass lesion is to determine if the lesion is intratesticular or extratesticular, and if the mass is cystic or solid. Although it depends on the individual case, cystic intratesticular mass lesions and both cystic and solid extratesticular mass lesions tend to be benign. However, solid intratesticular mass lesions are frequently malignant. Thus, imaging may contribute to making the choice between surgical and conservative treatment of the scrotal mass under investigation. Acute testicular torsion is usually not an indication for MRI, since it is most often diagnosed by patient history, physical examination, and color Doppler ultrasonography. However, in those instances when clinical and ultrasonography findings remain unclear, MRI may be helpful (Terai et al. 2006). In patients with congenital adrenal hyperplasia (CAH) caused by 21-hydroxylase deficiency, ultra­sono­graphy and MR imaging show associated testicular lesions equally well. It has been advocated that ultrasonography be considered the method of choice for detection and follow-up of these lesions, because it is the cheapest and quickest imaging technique. However, when partial orchiectomy is considered for treatment of testicular lesions of CAH, MRI has been recommended because it shows lesion margins optimally (Stikkelbroeck et al. 2003).

Table 7.2.13  Description of single pathologic conditions in MRI of the scrotum Disease entity

Signal intensityb T2-weighted  Proton density-weighted T1-weighted T1-weighted + CM Signal pattern T2-weighted  Proton density-weighted T1-weighted T1-weighted + CM References

Seminomatous

Non-seminomatous

Testicular

Testicular

Malignant

Malignant

Adrenal

Epidermoid

Testicular

Testicular

Rest tumor

Cyst

(Benign)

(Benign)

↓

↑

↔ or ↑ ↑↑

↓ ↔ or ↓

Neoplasm

Neoplasm

↓ ↔

↔ ↔

a

a

Homogeneous Homogeneous

Heterogeneous Heterogeneous

Homogeneous

Various

Homogeneous Homogeneous

Various Various

Johnson et al. 1990

Johnson et al. 1990

Fitoz et al. 2006 Stikkelbroeck et al. 2003

Cho et al. 2002 Yoshida 2004

 Both non-seminomatous malignant testicular neoplasm and testicular epidermoid cyst are typically surrounded by a capsule of low signal intensity, while seminomatous testicular tumors typically lack a capsule on MR images (Cho et al. 2002; Johnson et al. 1990; Yoshida 2004) b  Relative to normal testicular tissue a

7.2  Male Pelvis Fig. 7.2.30a–c  Grossly enlarged testicles bilaterally (arrows) in a small boy diagnosed with nonHodgkin’s lymphoma demonstrate homogenously intermediateto-low signal intensity on both fat-suppressed T2-weighted (a coronal) and fat-suppressed contrast-enhanced T1-weighted (b coronal, c axial) MR images

Derouet et al. (1993) suggest that, compared with ultrasonography, MRI shows diagnostic advantages in the detection and characterization of malignant tumors and in the exclusion of non-tumorous testicular pathology, such as old torsion, periorchitis, or old hematoma.

7.2.3.6  Diagnostic Procedures • • • • • •

Patient history Physical examination Blood work / laboratory tests Ultrasonography of the scrotum MRI of the scrotum (in unclear or complex cases) Computed tomography (CT) of the chest, abdomen, and pelvis (staging) • 18F-FDG positron emission tomography (PET) (staging; compare Jana and Blaufox 2006) The diagnostic algorithm is shown as a schematic diagram in Fig. 7.2.31.

1053

1054

7  Pelvis Fig. 7.2.31  Diagnostic algorithm in diseases affecting the scrotum (schematic diagram)

References 1.

2.

3.

4. 5. 6.

7.

Akbar SA, Sayyed TA, Jafri SZ, Hasteh F, Neill JS (2003) Multimodality imaging of paratesticular neoplasms and their rare mimics. Radiographics 23:1461–1476 Amodio JB, Maybody M, Slowotsky C, Fried K, Foresto C (2004) Polyorchidism: report of 3 cases and review of the literature. J Ultrasound Med 23:951–957 Andipa E, Liberopoulos K, Asvestis C (2004) Magnetic resonance imaging and ultrasound evaluation of penile and testicular masses. World J Urol 22:382–391 Barbaric ZL (1994) Principles of genitourinary radiology, 2nd edn. Thieme, New York Beddy P, Geoghegan T, Browne RF, Torreggiani WC (2005) Testicular varicoceles. Clin Radiol 60:1248–1255 Bothig R, Rogosch KU, Mach P, Mahn B, Burgdorfer H (2006) [Testicular metastases as primary manifestation of malignant melanoma at known melanocytoma]. Aktuelle Urol 37:138–140 Cho JH, Chang JC, Park BH, Lee JG, Son CH (2002) Sonographic and MR imaging findings of testicular epidermoid cysts. AJR Am J Roentgenol 178:743–748

8.

Claahsen-van der Grinten HL, Otten BJ, Takahashi S, Meuleman EJ, Hulsbergen-van de Kaa C, Sweep FC, Hermus AR (2007) Testicular adrenal rest tumors in adult males with congenital adrenal hyperplasia: evaluation of pituitary-gonadal function before and after successful testis-sparing surgery in eight patients. J Clin Endocrinol Metab 92:612–615 9. Czipull C, Asmussen M (2006) Hoden und Nebenhoden. In: Rummeny Reimer Heindel (eds) Ganzkoerper-MR-Tomographie. Thieme, Stuttgart, pp 401–408 10. Derouet H, Braedel HU, Brill G, Hinkeldey K, Steffens J, Ziegler M (1993) [Nuclear magnetic resonance tomography for improving the differential diagnosis of pathologic changes in the scrotal contents.] Urologe A 32:327–333 11. Dogra VS, Gottlieb RH, Mayumi O, Rubens DJ (2003) Sonography of the scrotum. Radiology 227:18–36 12. Fernandez-Perez GC, Tardaguila FM, Velasco M, Rivas C, Dos Santos J, Cambronero J, Trinidad C, San Miguel P (2005) Radiologic findings of segmental testicular infarction. AJR Am J Roentgenol 184:1587–1593

7.2  Male Pelvis 13. Fitoz S, Atasoy C, Adiyaman P, Berberoglu M, Erden I, Ocal G (2006) Testicular adrenal rests in a patient with congenital adrenal hyperplasia: US and MRI features. Comput Med Imaging Graph 30:465–468 14. Haag P, Hanhart N, Mueller M (2007) Hoden, Nebenhoden u. Samenleiter. In: Mueller M (ed) Gynäkologie und Urologie für Studium und Praxis. Medizinische Verlags- und Informationsdienste, Breisach, pp 357–370 15. Hormann M, Balassy C, Philipp MO, Pumberger W (2004) Imaging of the scrotum in children. Eur Radiol 14:974–983 16. Iida S, Iuchi H, Sasaki Y, Chujyo T, Nakata Y, Saga Y, Kaneko S, Yachiku S (2003) [A case of perineal abscess due to urethral fistula in a patient with spinal cord injury]. Hinyokika Kiyo 49:567–569 17. Jana S, Blaufox MD (2006) Nuclear medicine studies of the prostate, testes, and bladder. Semin Nucl Med 36:51–72 18. Johnson JO, Mattrey RF, Phillipson J (1990) Differentiation of seminomatous from nonseminomatous testicular tumors with MR imaging. AJR Am J Roentgenol 154:539–543 19. Kodama K, Hasegawa T, Egawa M, Tomosugi N, Mukai A, Namiki M (2004) Bilateral epididymal sarcoidosis presenting without radiographic evidence of intrathoracic lesion: Review of sarcoidosis involving the male reproductive tract. Int J Urol 11:345–348 20. Menzner A, Kujat C, Konig J, Pahl S, Kramann B (1997) [MRI in testicular diagnosis: differentiation of seminoma, teratoma and inflammation using a statistical score] RoFo 166:514–521 21. Oner AY, Sahin C, Pocan S, Kizilkaya E (2005) Polyorchidism: sonographic and magnetic resonance image findings. Acta Radiol 46:769–771 22. Pham SB, Hong MK, Teague JA, Hutson JM (2005) Is the testis intraperitoneal? Pediatr Surg Int 21:231–239 23. Sentilhes L, Dunet F, Thoumas D, Khalaf A, Grise P, Pfister C (2002) Segmental testicular infarction: diagnosis and strategy. Can J Urol 9:1698–1701 24. Senzaki H, Watanabe H, Ishiguro Y (2001) [A case of very rare tuberculosis of the testis]. Nippon Hinyokika Gakkai Zasshi 92:534–537 25. Skiadas V, Antoniou A, Primetis H, Moulopoulos L, Vlahos L (2006) Intratesticular arteriovenous malformation. Clinical course, ultrasound and MRI findings of an extremely rare lesion on a 7 year follow-up basis. Int Urol Nephrol 38:119–122 26. Stikkelbroeck NM, Suliman HM, Otten BJ, Hermus AR, Blickman JG, Jager GJ (2003) Testicular adrenal rest tumours in postpubertal males with congenital adrenal hyperplasia: sonographic and MR features. Eur Radiol 13:1597–1603 27. Terai A, Yoshimura K, Ichioka K, Ueda N, Utsunomiya N, Kohei N, Arai Y, Watanabe Y (2006) Dynamic contrast-enhanced subtraction magnetic resonance imaging in diagnostics of testicular torsion. Urology 67:1278–1282 28. Vara Castrodeza A, Torres Nieto A, Mendo Gonzalez M, Rodriguez Toves A, Penarrubia Ponce MJ, de la Fuente Bobillo MA (2006) Diagnosis of a primary testicular lymphoma by echography and magnetic resonance imaging. Clin Transl Oncol 8:456–458

29. Wittekind C, Mezer HJ, Bootz F (eds) (2002) UICC—International Union Against Cancer: TNM-Klassifikation maligner Tumoren (6. edn.). Springer, Berlin Heidelberg New York, 295 pp 30. Yoshida T (2004) MRI of testicular epidermoid cyst. Radiat Med 22:354–356

7.2.4  Male Pelvis: Penis U.G. Müller-Lisse, M.K. Scherr, C. Degenhart, and U.L. Müller-Lisse 7.2.4.1  Introduction The diagnosis of diseases of the penis occasionally involves imaging examinations. The most frequently used imaging modality is ultrasonography (US). Depending on the clinical work environment, US may be performed by the urologist or by the radiologist. Since many diseases of the penis affect the penile urethra, cystourethrography and endoscopy are often times carried out by the urologist. Particularly, endoscopy serves to detect, localize, and distinguish disorders affecting the urothelium, and to take biopsy samples or dissect scar tissue when necessary. Cavernosography has been the mainstay of imaging evaluation of the penile corpora; however, with the development of color Doppler US and magnetic resonance imaging (MRI), it has decreased in clinical importance. MRI of the penis is usually performed when the disorder is complex and cannot be sufficiently diagnosed by US or endoscopy, or radiography or fluoroscopy examinations. Examples include complex congenital anomaly, complex trauma, extensive inflammatory disease, invasive tumor, tumor staging, tumor recurrence, and complex postoperative changes, including status post penile prosthesis. However, standard methods include clinical examination, ultrasound, cystourethrography, and endoscopy. 7.2.4.2  Examination Techniques 7.2.4.2.1  Patient Positioning and, if Applicable, Patient Preparation General positioning for MRI examinations of the pelvis is outlined in the chapter on MRI of the urinary bladder. MRI examinations of the penis are usually performed with the patient in supine position. Particular preparation of the patient for an MRI examination of the penis includes placement of the penis in such position that it can be examined by means of an external surface coil. Stretching the penis such that the glans penis points toward the patient’s navel, the corpus spongiosum is ventral to the corpora cavernosa, and the penis does not lie over the scrotum is most likely to result in the best positioning

1055

1056

7  Pelvis Fig. 7.2.32a–c  Positioning for MRI of the penis, demonstrated in a phantom model (a) with respective positions of ring coil or similar design with a diameter of 15–20 cm (6–8 in., arrows), penis (asterisk), patient’s legs (doublelined arrows), and patient’s abdomen (hatched arrow). Position of penis and field-of-view of ring coil are demonstrated on sagittal (b) and axial (c) localizer gradientecho MR images

(Fig. 7.2.32). If need be, the shaft of the penis can be fixed to the skin of the lower abdomen by means of band-aids or hospital adhesive tape. Another option is to stretch out the penis such that the glans penis points toward the patient’s toes, and the corpora cavernosa are ventral to the corpus spongiosum. 7.2.4.2.2  Selection of Coils MRI examinations of the penis should be performed in whole-body high-field MR scanners. If local and regional imaging information is sufficient for the diagnostic purposes of the examination, it may suffice to apply a singlechannel surface coil with a diameter large enough to fully cover the region of interest. Depending on the anatomical situation in the individual patient, this may be a ring coil or similar design with a diameter of 15–20 cm (6–8 in.) (Fig. 7.2.32). The advantage of using such singlechannel coil designs with a relatively small coil is that the field of view of the coil is physically limited to the region of interest, such that application of saturation bands to limit signal from outside the region of interest may be unnecessary. The disadvantage is that additional information, such as information on lymph node status beyond the superficial inguinal nodes or on the perineum, may be limited or not available. If clinically necessary, a

multichannel pelvic phased-array coil may therefore be applied instead of the single-channel surface coil. 7.2.4.2.3  Examination Sequences The standard examination protocol for MRI examinations of the penis with a small surface coil of approximately 15–20 cm (6–8 in.) in diameter includes T1weighted spin-echo (SE), or turbo or fast SE (TSE or FSE) images before and after the intravenous administration of gadolinium-containing contrast media, and T2-weighted TSE or FSE images. Suggested sequence parameters are included in Table 7.2.14. 7.2.4.2.4  Imaging Planes It is important to cover the penis in at least two different imaging planes in order to obtain an overview of both penile anatomy and morphologic lesions. While imaging in the axial plane, perpendicular to the long axis of the penis, allows for comparison between the right and left corpus cavernosum and reveals asymmetry as an indicator of altered morphology, additional imaging in the sagittal plane provides a better impression of the course and continuity of the different penile structures (Table 7.2.14).

7.2  Male Pelvis

7.2.4.2.5  Thickness of Slices Since the penis is a relatively small structure, slice thickness should be reduced to 3–5 mm, with small interslice gaps (Table 7.2.14).

lineate better on post-contrast MR images. Usually, T1weighted MR images after intravenous administration of contrast media are obtained by means of SE or TSE/FSE sequences (Table 7.2.14). 7.2.4.3  Normal Anatomy

7.2.4.2.6  Preferred Coverage When the MRI examination is limited to the penis and its immediate surroundings, it is sufficient to apply a small surface coil whose coverage is restricted to its diameter. As a rule of thumb, the depth of tissue penetration in a (circular) surface coil is limited to its radius from the center of the coil at each point. Thus, for covering the penis, adjacent scrotum and perineum, and pubic bone close to the midline, it is sufficient to apply a ring-shaped surface coil of 15–20 cm (6–8 in.) in diameter. 7.2.4.2.7  Use of Contrast Medium Intravenous contrast media that contain gadolinium are helpful in MRI examinations of the penis for two reasons. One is that the corpora cavernosa and the bulbus and corpus spongiosum fill in with blood containing contrast media, which helps to delineate structures that do not enhance with contrast, such as fibrotic tissue or blood clots after longstanding priapism. The other is that penile lesions may be hypervascularized, such that they take up more contrast than healthy, surrounding tissue and de-

The penis is composed of the distal urethra, the corpora cavernosa, the bulbus and corpus spongiosum, the glans penis, the skin and subcutaneous loose connective tissue, and arterial and venous blood vessels. The corpora cavernosa and the corpus and bulbus spongiosum are cylindrical, sponge-like, cavernous structures without septations, which fill with blood during erection (Jordan and Schlossberg 2002). MR signal intensity within the penile corpora is dependent on the rate of blood flow within the cavernous spaces. Generally, the penile corpora demonstrate with intermediate signal intensity on T1-weighted and high signal intensity on T2-weighted MR images. The corpora cavernosa usually have the same signal intensity on MR images, since they are interconnected via septal fenestrations. However, the bulbus and corpus spongiosum may have different signal intensity, since they have an independent blood supply (Pretorius et al. 2001). The corpora cavernosa extend bilaterally from the inferior ramus of the pubic bone, with the crus corporis cavernosum, to the glans penis and are located on the dorsal side of the penis. The bulbus spongiosum is located on the caudal side of the urogenital diaphragm, between the crura of the corpora cavernosa. The bulbus spongiosum

Table 7.2.14  Standard examination protocol for magnetic resonance imaging of the penis with a 15–20 cm (6–8 in.) surface coil at 1.5 T Sequence type

T1-weighted SE

Anatomical region

a

T1-weighted a SE + CM

T2-weighted TSE

T2-weighted TSE

T2-weighted TSE

Penis

Penis

Penis

Penis

Penis

Imaging plane

Axial + sagittal

Axial

Axial

Coronal

Sagittal

TR (ms)

500–700

500–700

3,500–5,000

3,500–5,000

3,500–5,000

TE (ms)

15–17

15–17

90–110

90–110

90–110

Imaging matrix

192–256 × 256

192–256 × 256

192–256 × 256

192–256 × 256

192–256 × 256

FOV (mm)

160 × 160–200 × 200

160 × 160–200 × 200

120 × 120–200 × 200

160 × 160–200 × 200

160 × 160–200 × 200

Slice thickness (mm)

3–5

3–5

3–5

3–5

3–5

Interslice gap (mm)

0–1

0–1

0–1

0–1

0–1

CM intravenous contrast media a Alternatively, the penis may be examined with a T1-weighted turbo or fast SE (TSE or FSE) sequence

1057

1058

7  Pelvis

Fig. 7.2.33  a MRI of normal anatomy of the penis on T2weighted TSE MR images, including axial slices obtained in the pars pendulans at middle of penile shaft (A), at the junction of the shaft and glans penis (B), and at the glans penis (C), and a paramedian sagittal slice through one corpus cavernosum (D). Respective positions of (A), (B), and (C), are marked in (D). T2-weighted MRI demonstrates the distal urethra (curved arrow), the corpora cavernosa (asterisks), the bulbus and corpus spongiosum (chevron arrow), the glans penis (arrowhead), superficial (striped arrowhead) and deep (open arrowheads) blood vessels, the tunica albuginea (arrows), the fascia penis profunda or Buck’s fascia (double-lined arrow), loose connective tissue

without an interspersed fat layer (open curved arrow), the tunica dartos (hatched double-lined arrow), and the skin (open block arrow). b MRI of normal anatomy of the penis on T2weighted TSE MR images, including an axial slice at the level of the perineum (A) and coronal images obtained along the penile shaft and glans penis at the level of the corpus spongiosum (B), the level of the penile urethra (C), and the level of the corpora cavernosa (D). T2-weighted MRI demonstrates the penile urethra (curved arrow), the corpora cavernosa (asterisks) with their crura (double-lined arrows) and roots (arrows), the bulbus and corpus spongiosum (chevron arrow), and the glans penis (arrowhead)

7.2  Male Pelvis

Fig. 7.2.34a–d  MRI of the penis in epispadia-exstrophy complex on axial (a,b) and coronal (c,d) T2-weighted MR images demonstrates status post lower abdominal and pelvic midline closure

with scar tissue (double-lined arrows in a and c), wide separation of corpora cavernosa (arrows in b and d), and penile urethra exposed at the dorsal side of the penis (curved arrows in b and d)

extends distally into the corpus spongiosum, which is located on the ventral side of the penis and surrounds the penile urethra (Jordan and Schlossberg 2002). Each of the penile corpora is surrounded by a dense layer of fibroelastic connective tissue, the tunica albuginea, which has low signal intensity on all MRI sequences and meas­ ures approximately 1 mm in width (Fig. 7.2.33). Around the penile corpora and in between the corpora cavernosa and the corpus spongiosum is another layer of dense, fibroconnective tissue, the fascia penis profunda or Buck’s fascia. Like the tunica albuginea, Buck’s fascia demonstrates with low signal intensity on both T1-weighted and T2-weighted MR images. Thus, the tunica albuginea and the fascia penis profunda are usually indistinguishable from one another on MR images. Loose connective tissue without an interspersed fat layer separates the fascia penis profunda from the tunica dartos, the immediate subcutaneous connective tissue layer, which also shows with low signal intensity on both T1- and T2-weighted MR images (Pretorius et al. 2001). Around the tunica dartos lies the skin of the shaft of the penis. The glans penis is the anterior extension of the corpus spongiosum. The glans penis consists of cavernous, spongiform tissue,

which is also capable of filling with blood during erection, and is covered with skin without underlying loose connective tissue (Fig. 7.2.33). A duplicate layer of skin and subcutaneous loose connective tissue, the prepuce, extends from the shaft of the penis to cover the glans penis. 7.2.4.5  Pathological Findings 7.2.4.5.1  Groups of Pathologic Conditions (in Order) Benign Diseases of the Penis Congenital Among the most frequent congenital anomalies affecting the penis are hypospadia, a condition in which the urethral orifice does not open at the tip of the glans penis, but on the ventral side of the penis, somewhere along the course of the corpus spongiosum, and epispadia, in which there is no urethral orifice at the tip of the glans penis, and the urethra opens to the dorsal side of the penis, usually in between the corpora cavernosa (Fig. 7.2.34). Epi-

1059

1060

7  Pelvis

spadia is frequently associated with other defects of lower anterior midline closure, such as bladder exstrophy. The various manifestations can be summarized as representing different degrees of severity of the epispadia–exstrophy complex of genitourinary malformations (Gearhart 2002). Trauma In general, trauma to the penis can be blunt or penetrating. Unless the penis is partially or completely separated from the body by traumatic force, trauma to the penis usually results in hematoma. Hematoma may affect the subcutaneous compartment alone or involve one or more of the penile corpora. When blunt trauma occurs to the erect penis, it is frequently most severe at the base of the penis, close to the pubic bone. Most likely reasons include shearing or bending of the proximal shaft of the penis at the bone. Also, the base of the penis is somewhat protected by the suspensory ligament, which extends from the symphysis pubis and ends at the base of the penis. Fracture of the penis is a rare blunt penile trauma, with an annual frequency of 0.33 to 1.36 per 100,000 inhabitants, and less than 2,000 cases described in the literature. Etiology varies, but the trauma mainly occurs during coitus. Fracture of the penis is an emergency that requires immediate surgical treatment. Immediate surgery is the only chance for complete recovery and good functional results. While history and physical examination are essential for the correct diagnosis, imaging is used in unclear and atypical cases. Imaging methods applied include cavernosography, urethrography, ultrasound, and MRI (Khinev 2004). MRI findings in blunt penile trauma include a visible, subcutaneous mass whose signal is intermediate to high on both T1- and T2-weighted images in the acute and subacute phases of hemorrhage and may become inhomogeneous later on. When trauma involves rupture of one or more of the penile corpora, the low-signal-intensity layer of the tunica albuginea is obscured or disrupted. Hematoma within the corpora is usually associated with a decrease of signal intensity on T2-weighted and contrastenhanced, T1-weighted MR images. Post-traumatic alterations in the penile corpora include scarring, which may change the width and shape of the tunica albuginea on the one hand. On the other hand, trauma may result in the occurrence of pathologic septations within the penile corpora that may partially or completely interrupt the continuity of the blood-containing compartment of a penile corpus. When blood flow within the corpus is compromised, thrombosis may result from hemostasis. It has been discussed that partial priapism (Fig. 7.2.35) may be the consequence of posttraumatic septation of a penile corpus, and that both straddle injury and repetitive microtrauma to the base of the penis may be the underlying cause.

Fibrosis of the Penile Corpora On the one hand, fibrosis of penile corpora that were previously capable of erection may occur as a sequel of prolonged priapism or intravenous injections (Fig. 7.2.36). In both instances, compromised blood flow, due to hemostasis or scarring within the penile corpora, is causative for the fibrosis. However, fibrosis may also be a complication of severe penile trauma that involves the penile corpora or the tunica albuginea. Furthermore, fibrosis of the penile corpora may be associated with connective tissue disease and has been diagnosed in patients with systemic lupus erythematosus. On T1-weighted MR images, fibrosis of the penile corpora demonstrates with heterogenous, intermediate-to-low signal intensity, without contrast enhancement, and may be inseparable from the tunica albuginea. On T2-weighted MR images, fibrosis of the penile corpora shows with heterogenous, intermediate-to-low signal intensity and usually delineates clearly from surrounding tissue, but may be inseparable from the tunica albuginea. On the other hand, fibrosis of penile corpora may be the underlying condition in young men who complain of impotence and may never have had an erection. In such cases, the penile corpora may never have developed normally, or may have regressed during their development to a state of string-like, fibrotic bands instead of the normal anatomic structures. Transverse fibrotic membranes that partially or completely occlude a corpus cavernosum are the underlying defect in partial priapism. Surgical resection of the membrane typically reveals dense fibroconnective tissue and blood vessels, but no signs of malignancy. It remains unclear if the presence of transverse membranes within the corpora cavernosa represents a congenital anomaly or an acquired condition, which—according to one theory—may be brought about by recurrent microtrauma (Schneede et al. 1999). Thrombosis of the affected part of the corpus cavernosum and clinically partial priapism may be the consequences of membranous occlusion (Fig. 7.2.35). Penile Prosthesis Penile implants offer a dependable way of restoring erections in many patients. However, although they are not very frequent, complications of penile implant surgery may be clinically significant and include infection of the device, which is quite frequent, and some other, less frequent but important complications, such as distal and proximal perforation of the tunica albuginea, deformity of the penis, erosion of a component, and mechanical malfunction of the implanted device. Diagnosis of complications is based on clinical history and physical examination, but imaging techniques are also needed to explore the prosthesis and plan the surgical approach if necessary. MRI is considered the most valuable method for

7.2  Male Pelvis

Fig. 7.2.35  MRI of the penis in partial priapism demonstrates thin, membrane-like septum transversing the left corpus cavernosum (arrows in a and c) and associated thrombosis and dilation of proximal left corpus cavernosum (asterisks in a–d). While T2-weighted TSE MR images (a, b) and contrast-enhanced, T1-

weighted SE MR images (d) show thrombosis of proximal left corpus cavernosum with decreased signal intensity when compared with the normal, contra-lateral corpus cavernosum, unenhanced T1-weighted SE MR images (c) show increased signal of thrombus (modified from Schneede et al. 1999)

the diagnosis of penile prosthesis complications, since it demonstrates penile anatomy in three orthogonal planes, with superior definition of soft-tissue contrast (Moncada et al. 2004).

asymmetric thickening of the tunica albuginea. On postcontrast, T1-weighted MR images, acute plaques of induratio penis plastica may demonstrate contrast enhancement (Pretorius et al. 2001).

Inflammatory Disease of the Penis Induratio penis plastica or Peyronie’s disease is an inflammatory condition of unknown etiology, which affects the tunica albuginea and may extend to the penile corpora. Normal, elastic connective tissue of the tunica albuginea is replaced by fibrous or hyaline scar tissue, whose appearance may be sheet-like or plaque-like. On T2-weighted MR images, plaques of induratio penis plastica may show with decreased signal intensity when compared with adjacent, normal tunica albuginea. On T1-weighted MR images, plaques of induratio penis plastica may show with

Infection of the Penis and Perineum Infection of the penis and perineum may result from trauma or be a sequel of underlying inflammatory disease, e.g., Crohn’s disease. Whatever the underlying cause, infection usually involves penetration into the subcutaneous tissue or deeper tissue layers of pathologic microorganisms that bring about an inflammatory reaction, and, eventually, the formation of pus-containing, localized abscess or phlegmonous spread. In inflammatory bowel disease, such as Crohn’s disease, fistulous tracts may connect the skin surface or the mucosal sur-

1061

1062

7  Pelvis

face of the intestinal tract with subcutaneous or deeper tissue layers of the penis and perineum. On T2-weighted images, tissue edema and inflammatory reaction usually demonstrate with intermediate-to-high signal intensity, while pus collections most frequently appear homogenously bright. However, when pus is mixed with blood or blood remnants, signal intensity may be inhomogeneous. On unenhanced, T1-weighted MR images, inflammatory reaction and surrounding tissue edema demonstrate with intermediate-to-low signal intensity, while pus collections within the abscess or phlegmon show with low signal intensity. On contrast-enhanced, T1-weighted images, contrast uptake and resulting signal usually is high in the inflammatory reaction and surrounding edema. However, within the pus collections, there is usually no contrast uptake. 7.2.4.5.2  Malignant Diseases of the Penis Penile Cancer Penile cancer is a rare malignant condition. Histopathology reveals squamous cell epithelial carcinoma in more than 95% of cases. Clinically, penile cancer usually affects the glans penis (48% of cases) or the prepuce (21% of cases), and frequently presents as an ulcer. Less frequently involved are the glans penis and the prepuce (9%), the coronal sulcus (6%), and the shaft (2%, Fig. 7.2.37) (Singh et al. 2005). Penile cancer usually spreads via lymphatic vessels, since Buck’s fascia (fascia penis profunda) acts as a barrier to invasion of the penile corpora and to he-

Fig. 7.2.36a–d  MRI of fibrosis of the corpora cavernosa of the penis secondary to prolonged priapism and untreated thrombosis is demonstrated on sagittal (a,b) and axial (c,d) T2-weighted TSE MR images (a,c) and T1-weighted SE MR images after in-

matogenous spread (Singh et al. 2005). Regional lymphadenopathy involves superficial and deep inguinal lymph nodes (Table 7.2.15; Fig. 7.2.38). The lymphatic vessels of the skin of the penis and prepuce drain primarily into the superficial inguinal lymph nodes. The lymphatic vessels of the glans penis drain into the deep inguinal and external iliac lymph nodes, and the lymphatic vessels of the erectile tissue and penile urethra drain into the internal iliac lymph nodes. Since lymphatic vessels communicate across the midline, bilateral lymphadenopathy may occur with unilateral penile cancer. Invariably, however, the lymphatic vessels of the penis are said to first drain into the inguinal nodes before reaching the pelvic nodes (Singh et al. 2005). Cancer of the Penile Urethra Primary cancer of the male urethra is very seldom found. The majority of urethral cancers consist of squamous cell epithelial carcinoma at histopathology. Cancer of the penile urethra may invade periurethral loose connective tissue and eventually penetrates into the penile corpora (Fig. 7.2.39); Table 7.2.16). Metastasis to the Penis Although rare in general, metachronic affection of the penile urethra by transitional cell carcinoma occurs in some patients with a history of transitional cell carcinoma (Solsona et al. 1996). Metastasis to the penis of other tumors is unusual. The majority of penile metastases originate from tumors of the urinary bladder, prostate, and rectum, while metastases from renal and respiratory tract

travenous administration of contrast media (b,d). Fibrosis (arrowheads) occurs in the center of the corpora cavernosa, around the cavernosal arteries, and shows with heterogenous, intermediate-to-low signal intensity

7.2  Male Pelvis

Fig. 7.2.37a–e  MRI of the penis with squamous cell carcinoma at level of the penile shaft demonstrates tumor (arrows) invasive into the corpus spongiosum (chevron arrow). Signal intensity of the tumor is heterogenous and intermediate-to-low on coronal (a–c)

and sagittal (d,e) MR images, including T1-weighted SE (a), contrast-enhanced, T1-weighted SE (b,d), and T2-weighted TSE (c,e). Note respective positions of penile urethra (curved arrows) and corpora cavernosa with marked fibrosis (arrowheads)

Fig. 7.2.38a,b  MRI of lymphadenopathy in carcinoma of the penis on contrast-enhanced, T1-weighted SE MR images with fat signal suppression demonstrates coronal image with super-

ficial inguinal lymphadenopathy (arrows in a) associated with ulcerating carcinoma of the penis (open arrow in a) and deep inguinal lymphadenopathy in axial image (b)

1063

1064

7  Pelvis

primaries have also been reported. In individual cases, the route of metastasis to the penis may be hard to define, unless the lesion is clearly located outside of the penile urethra. The presence of hematogenous metastasis in other locations may imply that penile metastasis was blood-borne, although it does not provide definite proof. Clinically, the most frequent sign of penile metastasis is priapism. Penile metastases are usually associated with advanced, rather aggressive neoplastic disease, and they usually appear rather rapidly after the recognition and treatment of the primary tumor. Survival after detection of penile metastasis is usually very limited (Lynch and Pettaway 2002). On T2-weighted MR images, penile tumors and metastasis often demonstrate as solitary, infiltrating tumors with low signal intensity. T2-weighted MR images allow

for delineation of the tumor margin and of any extension into the penile shaft (Singh et al. 2005). On unenhanced, T1-weighted MR images, tumors or metastases of the penis usually have lower signal intensity than the penile corpora. Primary penile cancers usually manifest as solitary, ill-defined, hypointense, infiltrating tumors (Singh et al. 2005). On contrast-enhanced, T1-weighted MR images, tumors or metastases of the penis frequently present with contrast uptake. It is controversial whether, on gadolinium-enhanced, T1-weighted images, contrast enhancement is stronger in primary penile cancer than within the penile corpora (Singh et al. 2005) or stronger within the penile corpora (Asmussen and Czipull 2006). When obtained with fat signal suppression, contrast-enhanced MR images of the penis help to delineate penile tumor or metastasis from the perineum or pelvic floor.

Fig. 7.2.39a–d  MRI of the penis in cancer of the penile urethra at level of fossa navicularis demonstrates respective positions of penile urethra (curved arrows), periurethral tumor (arrows), and glans penis (arrowheads) on axial MR images (T1-weighted SE

[a], T2-weighted TSE [b], contrast-enhanced, T1-weighted SE [c]), and sagittal, contrast-enhanced, T1-weighted SE MR images (d). This tumor demonstrates lower uptake of intravenously administered contrast media than does the glans penis (c,d)

7.2  Male Pelvis

However, since accompanying inflammatory reaction may be severe, clear-cut definition of tumor margins can pose a problem for MRI. 7.2.4.5  Indications and Value of MRI On the whole, MRI is seldom used in pathologic conditions of the penis. The imaging modality most frequently applied to examine the penis is ultrasonography. The advantages of ultrasonography in imaging examinations of the penis include (1) its rapid availability in most clinical settings, (2) its ability to depict penile anatomy with high spatial resolution when high-frequency transducers (5–13 MHz) are being used, (3) the availability of color-coded Doppler duplex ultrasonography modes

to examine arterial and venous vessels of the penis and the penile corpora in modern ultrasonography scanners, and (4) the ease with which modern ultrasonography units can be handled in the patient examination room. Digital ultrasonography has brought about the ability to store meaningful ultrasound images and digital video sequences in the same ways in which it is possible to store MR images. However, the range of image contrast in different MRI sequences, the ability of MRI to cover larger regions of the human body, and its high reproducibility of imaging planes make MRI the modality of choice whenever penile disease appears complex or complicating. In status post penile trauma, its simplicity, precision, and availability make sonourethrography a valuable tool for the reconstructive urologist. However, MRI is consid-

Table 7.2.15  TNM 2002 classification of carcinoma of the penis (modified from Singh et al. 2005) TNM class

Description

Tumor (T) Tx

Primary tumor cannot be evaluated

T0

No evidence of primary tumor

Ta

Non-invasive, verrucous carcinoma

Tis

Carcinoma in situ

T1

Invasion of subepithelial connective tissue

T2

Invasion of one or more corpora

T3

Invasion of urethra or prostate gland

T4

Invasion of other adjacent structures

Lymph nodes (N) Nx

Regional lymph nodes cannot be assessed

N0

No evidence of regional lymphadenopathy

N1

Metastasis in a single, superficial inguinal lymph node

N2

Metastases in multiple or bilateral superficial inguinal lymph nodes

N3

Unilateral or bilateral metastases in deep inguinal or pelvic lymph nodes

Metastasis (M) Mx

Distant metastasis cannot be assessed

M0

No evidence of distant metastasis

M1

Distant metastasis

1065

1066

7  Pelvis Table 7.2.16  TNM 2002 classification of carcinoma of the urethra (modified from Wittekind et al. 2002) TNM class

Description

Tumor (T) Tx

Primary tumor cannot be assessed

T0

No evidence of primary tumor

Ta

Non-invasive, papillary or verrucous carcinoma

Tis

Carcinoma in situ

T1

Invasion of subepithelial connective tissue

T2

Invasion of one of the following organs: corpus spongiosum, prostate, periurethral muscle tissue

T3

Invasion of one of the following organs: corpus cavernosum, prostate with extension beyond the prostatic capsule, anterior vagina, bladder neck

T4

Invasion of other adjacent structures or neighboring organs

Lymph nodes (N) Nx

Regional lymph nodes cannot be assessed

N0

No evidence of regional lymphadenopathy

N1

Metastasis in a solitary lymph node, no larger than 2 cm in maximum diameter

N2

Metastases in a solitary lymph node, larger than 2 cm in maximum diameter, or metastases in multiple lymph nodes

Metastasis (M) Mx

Distant metastasis cannot be assessed

M0

No evidence of distant metastasis

M1

Distant metastasis

ered particularly valuable for defining the distorted pelvic anatomy that is frequently associated with posttraumatic posterior urethral strictures. By showing the location of the prostate and the length of the prostatomembranous defect, MRI may help determine whether a transperineal or transpubic approach for reconstruction is necessary (Gallentine and Morey 2002). A recently published series of seven patients with squamous cell carcinoma of the penis demonstrated that MRI after the intravenous administration of lymphotropic nanoparticles (ferumoxtran-10), when correlated with histology, has a sensitivity of 100% and a specificity of 97% in predicting the presence of regional lymph node metastases and may accurately triage patients for regional lymphadenectomy (Tabatabaei et al. 2005). In whole-body staging of malignant penile disease, computed tomography (CT) is advantageous because of its wide availability, its ability to cover the entire body within less than 1 min, and its high reproducibility of imaging planes and ranges. Recent work implies that

CT, when combined with positron emission tomography (PET) with 18F-desoxy-glucose (FDG), is highly useful in the detection of metastasis of penile cancer (Scher et al. 2005). 7.2.4.6  Indications (Symptom-Specific Imaging Modalities) MRI of the penis is usually applied as a secondary imaging modality, when ultrasonography, cystourethrography, and/or endoscopy fail to deliver sufficient imaging information to cover all clinical questions. In particular, while ultrasonography, including grey scale and color Doppler duplex techniques, is considered irreplaceable in the diagnostic workup of penile masses, MRI helps in the detection and staging of penile cancer and serves as a problem-solving diagnostic modality (Andipa et al. 2004). In complex congenital anomalies of the urinary tract, which may involve the penis, and in post-traumatic

7.2  Male Pelvis Fig. 7.2.40  Diagnostic algorithm in diseases affecting the penis (schematic diagram)

or post-surgical situations, MRI may be an indispensable means of depicting morphologic alterations.

4.

7.2.4.7  Diagnostic Procedures

5.

The diagnostic algorithm in diseases affecting the penis is shown in a schematic diagram in Fig. 7.2.40. References 1.

2.

3.

Andipa E, Liberopoulos K, Asvestis C (2004) Magnetic resonance imaging and ultrasound evaluation of penile and testicular masses. World J Urol 22:382–391 Asmussen M, Czipull C (2006) Penis. In: Rummeny E, ­Reimer P, Heindel W et al (eds) Ganzkoerper-MR-Tomo­ graphie. Thieme, Stuttgart, pp 408–412 Gallentine ML, Morey AF (2002) Imaging of the male urethra for stricture disease. Urol Clin North Am 29:361–372

6. 7.

8.

9.

Gearhart JP (2002) Exstrophy, epispadias, and other bladder anomalies. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ (eds) Campbell’s urology, 8th edn. Saunders, Philadelphia, pp 2136–2196 Jordan GH, Schlossberg SM (2002) Surgery of the penis and urethra. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ (eds) Campbell’s urology, 8th edn. Saunders, Philadelphia, pp 3886–3954 Khinev A (2004) [Penile fracture]. Khirurgiia (Sofiia) 60:32–41 Lynch DF and Pettaway CA (2002) Tumors of the penis. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ (eds) Campbell’s urology, 8th edn. Saunders, Philadelphia, pp 2945–2973 Moncada I, Jara J, Cabello R, Monzo JI, Hernandez C (2004) Radiological assessment of penile prosthesis: the role of magnetic resonance imaging. World J Urol 22:371–377 Pretorius ES, Siegelman ES, Ramchandani P, Banner MP (2001) MR imaging of the penis. Radiographics 21(Spec no.):S283–S298

1067

1068

7  Pelvis 10. Scher B, Seitz M, Reiser M, Hungerhuber E, Hahn K, Tiling R, Herzog P, Reiser M, Schneede P, Dresel S (2005) 18F-FDG PET/CT for staging of penile cancer. J Nucl Med 46:1460–1465 11. Schneede P, Schmeller N, Müller-Lisse UG, Reiser MF, Hofstetter AG (1999) Partieller Priapismus. Kasuistik und Literaturübersicht über diagnostisches und therapeutisches Vorgehen. Urologe [A] 38:179–183 12. Singh AK, Saokar A, Hahn PF, Harisinghani MG (2005) Imaging of penile neoplasms. Radiographics 25:1629–1638

13. Solsona E, Iborra I, Ricos JV, Monros JL, Dumont R, Almenar S (1996) Extravesical involvement in patients with bladder carcinoma in situ: biological and therapy implications. J Urol 155:895–900 14. Tabatabaei S, Harisinghani M, McDougal WS (2005) Regional lymph node staging using lymphotropic nanoparticle enhanced magnetic resonance imaging with ferumoxtran-10 in patients with penile cancer. J Urol 174:923–927 15. Wittekind C, Mezer HJ, Bootz F (eds) (2002)UICC–International Union Against Cancer: TNM-Klassifikation maligner Tumoren (6. edn.). Springer, Berlin Heidelberg New York

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging A. Maubon, C. Servin-Zardini, M. Pouquet, Y. Aubard, and J.P. Rouanet 7.3.1  Introduction Pelvic floor dysfunction affects millions of women worldwide, with approximately 400,000 interventions each year in the United States, 30% of them being re-interventions. Nonetheless, causative mechanisms and preventive meas­ ures remain elusive (DeLancey 2005). The pelvic floor falls into the domains of three surgical specialties: gynecology, urology, and digestive surgery. Until recently, its assessment relied on clinical examination and on ionizing imaging modalities. These examinations were either annoying to the patient and the radiologist (colpocystodefecogram) or gave information about only one of the pelvic compartments (e.g., cystogram, defecogram). Nonetheless, in most cases pelvic floor dysfunction concerns several or all of the pelvic compartments (Maglinte et al. 1999). MRI allows a multicompartmental approach. The first papers about MRI and pelvic floor dysfunction appeared in 1993 (Goodrich et al. 1993), and the number of publications on the topic has recently increased due to the advent of fast imaging sequences. Obviously, MRI offers several potential advantages: innocuousness, painlessness, the possibility of exploring the whole pelvis in one examination, high contrast between anatomical components of the pelvis and, above all, dynamic capabilities (Lienemann et al. 2000; Bump and Cundiff 1988; Maubon et al. 2000; Stokeret al. 2001). The main potential drawback is the prone position of the patient in the magnet, potentially less favorable to the display of prolapse than is the standing or sitting position (Schoenenberger et al. 1998). Nonetheless, clinical examination by gynecologists and urologists is generally performed in the prone position too, and a paper comparing the supine and upright positions at MR imaging did not show a significant difference (Fielding et al. 1996). 7.3.2  MRI Anatomy Sagittal sequences display the various pelvic compartments (Fig. 7.3.1): anterior urinary (bladder, bladder neck, urethra), middle genital (uterus, cervix, vagina), and posterior digestive (sigmoid colon, rectum, anal canal) (Piloni et al. 1997; Perez et al. 1999). Assessment of the peritoneal compartment (the fourth pelvic compartment) is of high importance and is one of the major contributions of MRI; the lowermost part of the peritoneum is usually easily seen, behind the uterus, as a thin low-sig-

Fig. 7.3.1  Midline sagittal T2-weighted sequence, with rectal ultrasound gel, displays the four pelvic compartments: urinary (1 urethra, 2 bladder neck, 3 bladder), genital (4 vagina, 5 uterine cervix, 6 uterus), peritoneal (7 pouch of Douglas), and digestive (8 rectum, 9 anal canal)

nal-intensity band outlined by the high signal intensity of fat (Fig. 7.3.2) (Lienemann et al. 2000). 7.3.3  Functional Anatomy of the Pelvic Floor The most recent anatomofunctional theories admit an active and a passive support system for pelvic organs (DeLancey 1994; Strohbehn et al. 1996; Klutke and Siegel 1995).The active support system is represented mainly by the levator ani muscles. They constitute the pelvic floor and separate the ischiorectal fossa inferiorly from the infraperitoneal space superiorly (Fielding et al. 2000; Hoyte et al. 2001). Levator ani muscles are composed of two portions: 1 The iliococcygeal portion is the elevating portion, lateral and well seen in frontal and sagittal planes. It attaches to the pubis, the arcus tendineus lateralis and dorsally to the coccyx and anococcygeal junction (Fig. 3.3.3).

1069

1070

7  Pelvis

Fig. 7.3.2  Sagittal T2 view displays the peritoneal wall itself in the pouch of Douglas (arrow), in a patient with prior interadnexal hysteretomy. Ultrasound gel delineates vagina and rectum

Fig. 7.3.3  Frontal T2 view with rectal ultrasound gel through the iliococcygeal portions of the levator ani (arrows) at rest. Note the bird-wing shape on the frontal view

2 The puborectal or pubovisceral portion is the sphincteric portion. It is thicker and delineates the urogenital hiatus. It attaches to the pubis and surrounds the midline organs (bladder, urethra, vagina, uterus, rectum, anal canal). It makes a puborectal sling around the rectum, and its contraction closes the anorectal angle. Its posterolateral fibers are intertwined with the fibers of the external anal sphincter; its anteromedial fibers attach to the vaginal walls and to the urethra (Fig. 3.3.4).

views and is constant, and its use allows the specific anatomy of any given patient to be taken into account (Fig. 7.3.5). Some authors advocate the use of the mid-pubic line that corresponds to the hymeneal ring marker used in clinical examination (Singh et al. 2001; Lienemann et al. 2004). • Mobile landmarks include the bladder neck, urethra, uterine cervix or (in case of hysterectomy) vaginal vault, and the lowest point of the peritoneum. For the posterior compartment, most investigators use the posterior rectoanal junction, corresponding to the posterior portion of the puborectal portion of the levator ani (levator sling). It is sometimes called the levator plate, and it delineates the urogenital hiatus.

The passive support system consists in the bony pelvis (sacrum, ischia, pubic rami), and in connective structures (pelvic fascia, with the arcus tendineus of the pelvic fascia, and the arcus tendineus levator ani that attaches the levator ani). 7.3.3.1  Anatomical Landmarks Analysis of the examinations will rely on the position of pelvic organs relative to anatomical landmarks at rest, during contraction, when straining, and during rectal evacuation. • A useful fixed landmark is the pubococcygeal line (PCL), between the inferior aspect of the pubis and the sacrococcygeal joint. It is easy to draw on sagittal

7.3.4  MRI Techniques for Pelvic Floor Dysfunction 7.3.4.1  Preparation Talking to the patient before the examination is essential in order to make her feel comfortable with the examination, confirm the indication, and have the patient rehearse the various movements that she will have to perform in the magnet. No special preparation is necessary prior to the examination. The patient lies supine in the

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging

Fig. 7.3.4a,b  Axial and frontal T2 views through the puborectal portions of the levator ani (arrows)

magnet, with a protection cloth under her buttocks. The bladder should not be emptied. Rectal opacification is compulsory in order to better visualize the recto-vaginal wall and the rectum. We use plain ultrasound gel, introduced with a wide-tip 60-ml syringe. It appears as high signal intensity on T2-weighted sequences. Some authors mix it with diluted gadolinium contrast media. 7.3.4.2  Sequences

Fig. 7.3.5  Sagittal T2 view at rest, with rectal and vaginal gel. Mobile landmarks evaluated with respect to the PCL in a patient with prior hysterectomy. 1 bladder neck, 2 vaginal vault, 3 lowest part of the peritoneum, 4 rectoanal junction

Two types of sequences can be used, the “classical” T2weighted sequences, which have high spatial resolution, to assess anatomy and, specifically, levator ani trophicity and defects; and the fast “dynamic “ sequences (Gousse et al. 2000) to assess morphological changes between the various positions (contraction, rest, straining, evacuation). A common examination for pelvic floor dysfunction at our center includes: • Three T2-weighted turbo spin-echo (TSE) sequences in the frontal, sagittal, and transverse planes (TE/TR = 90/5,321 (shortest), turbo factor = 18, NSA = 2, FOV = 300 mm, rec FOV = 70%, matrix = 384 × 512, scan percentage = 80%, acquisition time = 1 min 25 s) with respiratory gating, a synergy pelvic coil, and parallel imaging. • At least two sagittal dynamic sequences with at least four positions: contraction, at rest, mild straining,

1071

1072

7  Pelvis Fig. 7.3.6a–d  Example of the four dynamic positions in a patient with normal pelvic floor function: a contraction, b at rest, c straining mildly, and d straining fully. PCL is figured on all slices. Note angulation and position changes in the morphology of the iliococcygeal portion of the levator ani

maximum straining (Fig. 7.3.6) and at rest, mild straining, maximum straining, and evacuation. The evacuation position is of high importance, as evacuation is the only position in which all portions of the levator ani are totally relaxed, thus giving a more exact indication of prolapse extent. • We use a modified T2 TSE sequence allowing three slices in each of the four positions in 12 s, but many different sequences can be used, specifically singleshot spin-echo sequences and fast gradient-echo sequences, as long as they are fast enough to allow dynamic imaging. During those sequences, special attention must be given to the patient holding her breath so images are not blurred by motion artifact. • Clinical examination can be performed into the magnet by applying a plastic speculum valve in the anterior and posterior vaginal cul-de-sac during the dynamic acquisitions. • Total examination time is 20–25 min.

7.3.4.3  Normal Results At rest, the bladder base, the uterine cervix, and the lowest part of the peritoneum are above the PCL, the urethropubic angle is around 45°, and the anorectal angle is around 110°. The urogenital hiatus (distance from the inferior aspect of the pubis to the posterior aspect of the anal canal) is +5 to –1.5 cm. It corresponds to the levator plate. During straining and defecation, the bladder and the urethra are displaced posteriorly; the uterine cervix and the vagina show posterior and inferior displacement, with enlargement of the vaginal angle (Goh et al. 2000). The bladder neck, uterine cervix, and the lowest portion of the peritoneum stay above the PCL. The length of the urogenital hiatus remains less than 7 cm. The rectum is displaced inferiorly. The iliococcygeal portion of the levator ani displays a dome shape at rest on frontal views; during straining,

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging

Fig. 7.3.7a–c  Sagittal dynamic acquisitions on the midline a at rest, b during mild straining, and c during maximum straining. Cervicocystoptosis with cystocele. The bladder neck and blad-

der base pass under the PCL with eversion of the urethra during maximum straining. Note the presence of an associated minor hysteroptosis, and no rectocele

it tends to become more horizontal and then inverts its shape with widening. At the same time, during straining, the puborectal portions move away from each other, widening and enlarging the urogenital hiatus.

from the axis of the anal canal (same definition as in defecography) (Fig. 7.3.9). • Peritoneocele. The lowest portion of the peritoneum passes under the PCL. The peritoneocele can be composed of the peritoneal wall alone, or with peritoneal fat or fluid, or contain the sigmoid colon (sigmatocele), or contain small-bowel loops (enterocele) (Fig. 7.3.10). • Descending perineum. The ano-rectal junction descends more than 5 cm during straining or defecation (Fig. 7.3.11). • Descended perineum. At rest, the anorectal junction descends more than 5 cm below the PCL.

7.3.4.4  Prolapses The technique of MRI is still being evaluated for these indications, although it is now widely accepted as a preoperative imaging modality (Lienemann and Fischer 2003). MRI classifications for pelvic prolapse are derived from clinical classifications, and are not yet validated. Generally accepted definitions of the various pelvic prolapses are listed below (Healy et al. 1997; Kirschner-Hermanns et al. 1993; Vanbeckevoort 1999). • Cervicocystoptosis. There is a caudal displacement of the bladder and bladder cervix under the PCL, either spontaneously or during straining. It is generally accompanied by a an exacerbated posterior displacement of the urethra with opening of the pubourethral angle to more than 90° (equivalent to the clinical “Qtip test”) and opening of the posterior urethrovesical angle to more than 150° (Fig. 7.3.7). • Cystoptosis. The bladder cervix passes under the PCL, while the bladder base remains above the PCL. • Cystocele. The bladder base passes under the PCL. • Hysteroptosis and vaginal vault descent. The uterine cervix, or, in case of hysterectomy, the vaginal vault, goes under the PCL during straining. Hysteroptosis and vaginal vault descent is generally associated with an elongation of the anterior margin of the uterine cervix (Fig. 7.3.8). • Rectocele. During straining, the anterior aspect of the rectal wall makes an anterior bulge, more than 3 cm

MRI classifications of prolapse are derived from clinical classifications: stage 0, no prolapse; stage 1, minor prolapse, above the level of the hymen; stage 2, moderate prolapse, down to the level of the hymen; stage 3, exteriorized prolapse; stage 4, complete eversion. These clinical classifications can easily be adapted to MRI. Recently the HMO classification has been proposed for MRI, based on the size of the urogenital hiatus (H), the size of the muscular relaxation (M) and the degree of organ prolapse (O) in four stages (Comiter et al. 1999). The International Continence Society Committee on Standardization of Terminology (Bump et al. 1996) can be adapted to MRI (POPQ: pelvic organ prolapse quantification), but it appears to be difficult to precisely detect the non-osseous landmarks. Whatever the classification, the report must specify the state of the levator ani muscles, the presence of an incidental lesion (e.g., of the uterus, ovary, digestive tract), the position of pelvic organs at rest, in contraction, and during maximum straining, the degree of prolapse, and the possibility of prolapse masked by a prominent prolapse.

1073

1074

7  Pelvis

Fig. 7.3.8a,b  Dynamic acquisition a at rest and b during maximum straining. Prominent hysteroptosis is seen, with a lengthening of the anterior uterine cervix. No associated cystocele, rectocele, or peritoneocele

Fig. 7.3.9a,b  Dynamic acquisition a at rest and b during maximum straining. Prominent rectocele with small peritoneocele, small cystocele, and hysteroptosis. Note the bulge of the anterior rectal wall, outlined by the ultrasound gel mixed with bowel gas; arrow show the axis of the anal canal

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging

Fig. 7.3.10a–c  Dynamic acquisition a at rest, b mild straining, and c during maximum straining with defecation. Patient had a prior hysterectomy. Prominent peritoneocele between bladder

and rectum, occupied by small-bowel and sigmoid loops (arrows) associated with a small cystocele, and a small rectocele. Note ultrasound gel evacuated by defecation (star)

Fig. 7.3.11a–c  Dynamic acquisition during maximum straining. Major pelvic floor dysfunction with cystocele, vaginal vault prolapse with peritoneocele, rectocele, and descending perineum

At present, MRI for pelvic prolapse appears equivalent to clinical examination for the detection of prolapse in anterior compartments (bladder, uterus) and superior to clinical examination for detection of prolapse in posterior compartments (peritoneal, digestive) (Kelvin et al. 1999, 2000; Rentsch et al. 2006). Its precise role in the preoperative assessment of pelvic floor dysfunction is still to be determined, although most authors agree on its usefulness, and it is now recognized that performing a preoperative assessment by MRI can modify the surgical approach (Heetzer et al. 2006) Furthermore cost–benefit

analyses are warranted (Barbaric et al. 2001). MRI clearly helps in the understanding of complex multicompartmental prolapses and surgical failures or recurrences (Pannu et al. 2000). 7.3.5  Conclusion MRI for pelvic prolapse is still in development. Its most important potential advantage over conventional examinations is that it offers the possibility of scanning all

1075

1076

7  Pelvis

pelvic compartments at once; furthermore, unlike the conventional examinations, it is non-invasive and simple, and allows a pathophysiological approach to assessing pelvic floor dysfunction. The MRI examination must be thorough and include muscle assessment, static and dynamic sequences. Comparison with clinical examination results is essential. References 1.

Barbaric ZL, Marumoto AK, Raz S (2001) Magnetic resonance imaging of the perineum and pelvic floor. Top Magn Reson Imaging 12:83–92 2. Bump RC, Cundiff WC (1998) Urogynecology and pelvic floor dysfunction. Obstetrics and gynecology clinics of North America. Saunders, Philadelphia 3. Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey JOL, Klarskov P, Shull BL, Smith ARB (1996) The standardization of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 175:10–17 4. Comiter CV, Vasvada SP, Barbaric ZL, Gousse AE, Raz S (1999) Grading pelvic prolapse and pelvic floor relaxation using dynamic magnetic resonance imaging. Urology 54:454–457 5. DeLancey JO (1994) The anatomy of the pelvic floor. Curr Opin Obstet Gynecol 6:313–316 6. DeLancey JO (2005) The hidden epidemic of pelvic floor dysfunction: achievable goals for improved prevention and treatment. Am J Obstet Gynecol 192:1488–1495 7. Fielding J, Dumanli H, Schreyer AG, Okuda S, Gering D, Zou K, Kikinis R, Jolesz F (2000) MR-based three dimensional modeling of the normal pelvic floor in women. Quantification of muscle mass. AJR Am J Roentgenol 174:657–660 8. Fielding JR, Versi E, Mulkern RV, Lerner MH, Griffiths DJ, Jolesz FA (1996) MR imaging of the female pelvic floor in the supine and upright position. J Magn Reson Imaging 6:961–963 9. Goh V, Halligan S, Kaplan G, Healy JC, Bartram CI (2000) Dynamic MR imaging of the pelvic floor in asymptomatic subjects. Am J Roentgenol 174:661–666 10. Goodrich MA, Webb MJ, King BF, Bampton AE, Campeau NG, Riederer SJ (1993) Magnetic resonance imaging of pelvic floor relaxation: dynamic analysis and evaluation of patients before and after surgical repair. Obstet Gynecol 82:883–891 11. Gousse AE, Barbaric ZL, Safir MH, Madjar S, Marumoto AK, Raz S (2000) Dynamic half-Fourier acquisition, single shot turbo spin-echo magnetic resonance imaging for evaluating the female pelvis. J Urol 164:1606–1613 12. Healy JC, Halligan S, Reznek RH, Watson S, Phillips RKS, Armstrong P (1997) Patterns of prolapse in women with symptoms of pelvic floor weakness: assessment with MR imaging. Radiology 203:77–81

13. Hetzer FH, Andreisek G, Tsagari C, Sahrbacher U, Weishaupt D (2006) MR defecography in patients with fecal incontinence: imaging findings and their effect on surgical management. Radiology 240:449–457 14. Hoyte L, Schierlitz L, Zou K, Flesh G, Fielding JR (2001) Two- and 3-dimensional MRI comparison of levator ani structure, volume, and integrity in women with stress incontinence and prolapse. Am J Obstet Gynecol 185:11–19 15. Kelvin FM, Hale DS, Maglinte DD, Patten BJ, Benson JT (1999) Female pelvic organ prolapse: contribution of dynamic cystoproctography and comparison with physical examination. Am J Roentgenol 174:31–37 16. Kelvin FM, Maglinte DD, Hale DS, Benson JT (2000) Female pelvic organ prolapse: a comparison of triphasic dynamic MR imaging and triphasic fluoroscopic cystocolpoproctography. Am J Roentgenol 174:81–88 17. Kirschner-Hermanns R, Wein B, Niehaus S, Schaefer W, Jakse G (1993) The contribution of magnetic resonance imaging of the pelvic floor to the understanding of urinary incontinence. Br J Urol 72:715–718 18. Klutke CG, Siegel CL (1995) Functional female pelvic anatomy. Urologic clinics of North America 3:487–498 19. Lienemann A, Fischer T (2003) Functional imaging of the pelvic floor. Eur J Radiol 47:117–122 20. Lienemann A, Anthuber C, Baron A, Reiser M (2000) Diagnosing enteroceles using dynamic magnetic resonance imaging. Dis Colon Rectum 43:205–212; discussion 212–213 21. Lienemann A, Sprenger D, Janssen U, Grosch E, Pellengahr C, Anthuber C (2004) Assessment of pelvic organ descent by use of functional cine-MRI: which reference line should be used? Neurourol Urodyn 23:33–37 22. Maglinte DD, Kelvin FM, Fitzgerald K, Hale DS, Benson JT (1999) Association of compartment defects in pelvic floor dysfunction. AJR Am J Roentgenol 172:439–444 23. Maubon A, Martel-Boncoeur MP, Juhan V, Courtieu C, Meny R, Mares P, Rouanet JP (2000) Static and dynamic magnetic resonance imaging of the pelvic floor. J Radiol 81:1875–1886 24. Pannu HK, Kaufman HS, Cundiff GW, Genadry R, Bluemke DA, Fishman EK (2000) Dynamic MR imaging of pelvic organ prolapse: spectrum of abnormalities. Radiographics 20:1567–1582 25. Perez N, Garcier JM, Pin-Leveugle J, Lhoste-Trouilloud A, Ravel A, McLaughlin P, Viallet JF, Boyer L. Dynamic magnetic resonance imaging of the female pelvis: radioanatomy and pathologic applications (1999) Preliminary results. Surg Radiol Anat 21:133–138 26. Piloni V, Bassotti G, Fioravanti P, Amadio L, Montesi A (1997) Dynamic imaging of the normal pelvic floor. Int J Colorectal Dis 12:246–253 27. Rentsch M, Paetzel C, Lenhart M, Feuerbach S, Jauch KW, Furst A (2001) Dynamic magnetic resonance imaging defecography: a diagnostic alternative in the assessment of pelvic floor disorders in proctology. Dis Colon Rectum 44:999–1007

7.3  Pelvic Floor Assessment by Magnetic Resonance Imaging 28. Schoenenberger AW, Debatin JF, Guldenschuh I, Hany TF, Steiner P, Krestin GP (1998) Dynamic MR defecography with a superconducting, open-configuration MR system. Radiology 206:641–646 29. Singh K, Reid WM, Berger LA (2001) Assessment and grading of pelvic organ prolapse by use of dynamic magnetic resonance imaging. Am J Obstet Gynecol 185:71–77 30. Stoker J, Halligan S, Bartram CI (2001) Pelvic floor imaging. Radiology 218:621–641

31. Strohbehn K, Elllis JH, Strohbehn JA, DeLancey JO (1996) Magnetic resonance imaging of the levator ani with anatomic correlation. Obstet Gynecol 87:277–285 32. Vanbeckevoort D, Van Hoe L, Oyen R, Ponette E, De Ridder D, Deprest J (1999) Pelvic floor descent in females: comparative study of colpocystodefecography and dynamic fast MR imaging. J Magn Reson Imaging 9:373–377

1077

Chapter 8

Musculoskeletal System

8

8.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1081 C. Glaser, S. Weckbach, and M. Reiser

8.5 

Inflammatory Diseases  of Bone and Joints  . . . . . . . . . . . . . . . . . . .   1094

8.2 

Examination Technique  .. . . . . . . . . . . . .   1081

8.5.1 

Bacterial/Viral Infections  . . . . . . . . . . . . .   1094

8.2.1 

Patient Preparation and Positioning   .. .   1081

8.5.1.1  Osteomyelitis  .. . . . . . . . . . . . . . . . . . . . . . .   1094

8.2.2 

Coil Selection   . . . . . . . . . . . . . . . . . . . . . . .   1082

8.5.1.2  Septic Arthritis  . . . . . . . . . . . . . . . . . . . . . .   1097

8.2.3 

Imaging Planes   .. . . . . . . . . . . . . . . . . . . . .   1082

8.5.1.3  Spondylitis, Spondylodiskitis  .. . . . . . . . .   1100

8.2.4 

MRI Sequences   .. . . . . . . . . . . . . . . . . . . . .   1082

8.5.1.4  Soft-Tissue Infections   .. . . . . . . . . . . . . . .   1102

8.2.4.1  T1-Weighted SE Sequences  .. . . . . . . . . . .   1084

8.5.1.5  Rheumatoid Diseases  .. . . . . . . . . . . . . . . .   1102

8.2.4.2  T2-weighted SE Sequences   .. . . . . . . . . . .   1084

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1106

8.2.4.3  TSE Sequences  .. . . . . . . . . . . . . . . . . . . . . .   1084

Suggested Reading  . . . . . . . . . . . . . . . . . . .   1106

8.2.4.4  GRE Sequences   .. . . . . . . . . . . . . . . . . . . . .   1084

8.6 

Avascular Necrosis  .. . . . . . . . . . . . . . . . . .   1106

8.2.4.5  MTC Sequences  . . . . . . . . . . . . . . . . . . . . .   1085

8.6.1 

Avascular Necrosis of the Hip and Transient Bone Marrow Edema Syndrome  . . . . . . . . . . . . . . . . . . . .   1107

8.2.5 

Contrast Medium Application   . . . . . . . .   1085

8.3 

Relaxation Times, Signal Intensities, and Contrast Behavior  .. . . . . . . . . . . . . .   1086 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1086

8.4 

Diseases of the Bone Marrow and Hematopoietic System  . . . . . . . . . . .   1086

8.4.1 

Leukemia   .. . . . . . . . . . . . . . . . . . . . . . . . . .   1087

8.4.2 

Multiple Myeloma   .. . . . . . . . . . . . . . . . . .   1088

8.4.3 

Metastases and Malignant Lymphomas   .. . . . . . . . . . . . . . . . . . . . . . . .   1090

8.4.4 

8.6.1.1  Avascular Necrosis of the Hip   . . . . . . . .   1107 8.6.1.2  Bone Marrow Edema Syndrome  .. . . . . .   1110 8.6.1.3  Subchondral Insufficiency Fracture of the Femoral Head   . . . . . . . . . . . . . . . . .   1110 8.6.2 

Perthes Disease and Coxitis Fugax   . . . .   1110

8.6.2.1  Perthes Disease   .. . . . . . . . . . . . . . . . . . . . .   1110 8.6.2.2  Coxitis Fugax   . . . . . . . . . . . . . . . . . . . . . . .   1113 8.6.3 

Bone Infarction   . . . . . . . . . . . . . . . . . . . . .   1113

Storage Diseases   .. . . . . . . . . . . . . . . . . . . .   1091

8.6.4 

Kienboeck’s Disease   .. . . . . . . . . . . . . . . . .   1113

8.4.5 

Osteomyelofibrosis   . . . . . . . . . . . . . . . . . .   1093

8.6.5 

Necrosis of the Scaphoid Bone   .. . . . . . .   1114

8.4.6 

Aplastic Anemia and Sequelae of Chemotherapy and Radiotherapy   .. .   1093

8.6.6 

8.4.7 

Hemosiderosis, Hemochromatosis, and Sickle Cell Anemia   . . . . . . . . . . . . . .   1093

Osteochondritis Dissecans   and Spontaneous Osteonecrosis   of the Knee  .. . . . . . . . . . . . . . . . . . . . . . . . .   1114

8.4.7.1  Hemosiderosis and Hemochromatosis     1093 8.4.7.2  Sickle Cell Anemia  .. . . . . . . . . . . . . . . . . .   1094 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1094

8.6.6.1  Osteochondritis Dissecans   . . . . . . . . . . .   1114 8.6.6.2  Spontaneous Osteonecrosis   of the Knee  .. . . . . . . . . . . . . . . . . . . . . . . . .   1114 8.6.7 

Osteonecrosis and Osteochondritis in Other Locations  . . . . . . . . . . . . . . . . . . .   1116 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1116

1080

8  Musculoskeletal System 8.7 

Imaging of Internal  Joint Derangement  . . . . . . . . . . . . . . . . . .   1117

8.7.1 

Imaging of Normal Joint Structures   . . .   1117

8.7.2 

Knee   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1117

8.7.2.1  Menisci   .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1117 8.7.2.2  Cruciate and Collateral Ligaments,   Patellar Tendon,   and Quadriceps Tendon   .. . . . . . . . . . . . .   1121 8.7.2.3  Cartilage Imaging   . . . . . . . . . . . . . . . . . . .   1126

8.7.7 

8.7.7.1  Tendon Pathologies   .. . . . . . . . . . . . . . . . .   1152 8.7.7.2  Collateral Ligaments   .. . . . . . . . . . . . . . . .   1153 8.7.7.3  Neuropathies   .. . . . . . . . . . . . . . . . . . . . . . .   1154 8.7.7.4  Osteochondritis Dissecans and Osteonecrosis  .. . . . . . . . . . . . . . . . . . .   1154 8.7.7.5  Synovial Pathologies and Bursitis  .. . . . .   1154 8.7.7.6  Intra-Articular Loose Bodies   . . . . . . . . .   1154 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1155

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1129 8.7.3 

Shoulder   .. . . . . . . . . . . . . . . . . . . . . . . . . . .   1130

8.7.3.1  Lesions to the Rotator Cuff   . . . . . . . . . . .   1133 8.7.3.2  Capsulolabral Tears   .. . . . . . . . . . . . . . . . .   1137 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1140 8.7.4 

Wrist   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1141

8.7.4.1  Intrinsic Ligaments: Scapholunate and Lunotriquetral Ligaments  .. . . . . . . .   1141 8.7.4.2  Triangular Fibrocartilage Complex   .. . .   1142

Elbow   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1151

8.8 

Bone and Soft Tissue Tumors  .. . . . . . . .   1155

8.8.1 

Intraosseous Tumor Extension  . . . . . . . .   1155

8.8.2 

Compact Bone  .. . . . . . . . . . . . . . . . . . . . . .   1158

8.8.3 

Extraosseous Tumor Extension  .. . . . . . .   1158

8.8.4 

Soft Tissue Tumors  .. . . . . . . . . . . . . . . . . .   1161

8.8.5 

Evaluation of Tumor Nature  . . . . . . . . . .   1162

8.8.6 

Tumor Classification  . . . . . . . . . . . . . . . . .   1164

8.8.7 

Therapy Control  . . . . . . . . . . . . . . . . . . . . .   1167

8.7.4.3  Ulnocarpal Impaction   . . . . . . . . . . . . . . .   1143

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1168

8.7.4.4  Ganglia   .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1143

8.9 

Posttraumatic Alterations  .. . . . . . . . . . .   1169

8.7.4.5  Nerve Compression Syndromes: Carpal Tunnel Syndrome and Loge de Guyon  .. . . . . . . . . . . . . . . . . .   1143

8.9.1 

Bone Injuries  . . . . . . . . . . . . . . . . . . . . . . . .   1169

8.7.4.6  Tendinopathy   . . . . . . . . . . . . . . . . . . . . . . .   1145 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1145 8.7.5 

Temporomandibular Joint   .. . . . . . . . . . .   1146

8.7.6 

Ankle and Foot   .. . . . . . . . . . . . . . . . . . . . .   1147

8.7.6.1  Osteochondral Lesion of the Talus: OLT   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1147 8.7.6.2  Osteonecrosis   . . . . . . . . . . . . . . . . . . . . . . .   1147 8.7.6.3  Stress Fractures   .. . . . . . . . . . . . . . . . . . . . .   1147 8.7.6.4  Transient Edema-Like Bone Marrow Abnormalities  .. . . . . . . . .   1148 8.7.6.5  Tendon Pathologies   .. . . . . . . . . . . . . . . . .   1149 8.7.6.6  Capsuloligamentous Pathologies   .. . . . .   1149 8.7.6.7  Sinus-Tarsi Syndrome   .. . . . . . . . . . . . . . .   1149 8.7.6.8  Plantar Fasciitis, Plantar Fibromatosis     1150 8.7.6.9  Morton Neuroma   .. . . . . . . . . . . . . . . . . . .   1150 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1151

8.9.1.1  Stress Fractures  . . . . . . . . . . . . . . . . . . . . . .   1169 8.9.1.2  Occult Fractures   .. . . . . . . . . . . . . . . . . . . .   1171 8.9.1.3  Bone Contusions   . . . . . . . . . . . . . . . . . . . .   1171 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1172 8.10 

Differential Diagnosis  .. . . . . . . . . . . . . . .   1173

8.11 

Diagnostic Value of MRI and Comparison with Other Imaging Modalities  .. . . . . . . . . . . . . . . . .   1173

8.11.1 

Bone Marrow   . . . . . . . . . . . . . . . . . . . . . . .   1174

8.11.2 

Bone Tumors and Bone Metastases   .. . .   1174

8.11.3 

Infections   .. . . . . . . . . . . . . . . . . . . . . . . . . .   1174

8.11.4 

Aseptic Osteonecroses   . . . . . . . . . . . . . . .   1174

8.11.5 

Joints   .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1174

8.11.6 

Bone and Soft Tissue Tumors  . . . . . . . . .   1174

8.11.7 

Traumatology   . . . . . . . . . . . . . . . . . . . . . . .   1175

8.12 

Diagnostic Process  .. . . . . . . . . . . . . . . . . .   1175

8.2  Examination Technique

8.1  Introduction C. Glaser, S. Weckbach, and M. Reiser Together with the examination of the brain and imaging of the spinal canal, MR imaging of the musculo­skeletal system has been established as a clinically valuable indication for MRI for several decades. There is broad agreement concerning indications, clinical value, and examination technique. Usually MRI is applied in cases in which no definite clarification is achieved by conventional diagnostic procedures. However, more and more clinical settings are arising in which MRI is included early in the diagnostic process. The diseases of the musculoskeletal system involve numerous medical specialties such as orthopedics, trauma surgery, rheumatology, oncology, nephrology, hematology, and pediatrics. Cooperation with the corresponding clinical specialist and consideration of clinical findings is of the utmost importance for interpreting the sometimes non-specific findings of MRI. 8.2  Examination Technique 8.2.1  Patient Preparation and Positioning Special preparation of the patient usually is not required. Informed consent must be obtained before the examination and the lack of contraindications (e.g., cardiac pacemakers, neurostimulating implants, metal implants with ferromagnetic components, shell splinters in the body) must be ensured. Metal parts in the clothes or on the body of the patient must be removed prior to the exam. Ferromagnetic joint prostheses and osteosynthetic material in themselves do not represent an absolute contrain-

dication. However, one must be clear about the fact that locally restricted artifacts will appear in the environment of these metal bodies. Artifact intensity depends on the strength of the main magnetic field and pulse sequence design. GRE sequences, particularly T2*-weighted GRE sequences show extensive artifacts. TSE sequences, however, are considerably less sensitive to susceptibility effects and therefore are degraded to a lesser extent by metal artifacts (Fig. 8.2.1). After surgery, metal abrasion that is not recognizable radiologically may have remained in the tissue and may cause (sometimes prominent) signal voids on MRI. If a contrast-enhanced scan is anticipated, an intravenous (i.v.) injection route should be introduced before the patient is positioned on the exam table in order not to delay the post-contrast MR sequences and in order to facilitate identical patient positioning before and after contrast medium injection. In most cases, comfortable supine positioning of the patient in the magnet is feasible. Examining the extremities one must decide whether it is more important to compare and therefore to image both sides simultaneously, or whether maximal spatial resolution shall be obtained from one region alone. The examination should be performed accordingly with the body coil, other coils covering a large field of view (FOV) or with dedicated extremity coils. The patient must be positioned as exactly as possible in order to facilitate optimal depiction of anatomy. For example, the spinal canal should be aligned in the sagittal plane, the pelvis should not be rotated, and the extremities should be placed in a neutral position. If this is not possible, a proper section angulation compensating for such deviations can be planned from the scout view. Dynamic examinations, e.g., flexion and extension of the cervical spine, abduction, and adduction in the shoulder joint as well as flexion and extension in the knee Fig. 8.2.1a,b  Tibial medullary nail with fixation screws after osteotomy. a T1-weighted SE image, b moderately T2-weighted frequency-selective fat-presaturated TSE sequence at 1.5 T. In both sequences, severe artifacts are shown that restrict the evaluability of the surrounding tissue. In b, the magnetic field inhomogeneity leads to loss of fat saturation in significant parts of the image and does not allow reliable evaluation of bone marrow signal

1081

1082

8  Musculoskeletal System

joint can be carried out in an open magnet with different functional positions targeting analysis of movement disturbances. Until now, these procedures have not gained general acceptance in clinical routine applications. 8.2.2  Coil Selection Coils receive the MR signal from the body segment to be examined. It is important to choose their size according to the size of the investigated body segment. The signalto-noise ratio (SNR) profits from small-diameter coils since they ensure a higher fill factor. Surface coils have to be positioned as near as possible to the region of interest. With circularly polarized coils, the SNR can be improved. Phased-array coils represent a series of several coil systems that overlap to some degree (Table 8.2.1). Each of these coils is connected to its own readout channel. These phased-array coils allow imaging of larger body segments with the higher SNR characteristics of surface coils. On the other hand, multichannel dedicated coils, such as wrist, small extremity, or knee coils, allow considerable shortening of the imaging time with only a moderate penalty in SNR using parallel imaging techniques, thus facilitating dynamic contrast-enhanced studies. 8.2.3  Imaging Planes The choice of the imaging planes depends on the respective anatomical region and the clinical question (Table 8.2.2). Usually the field of view is imaged in at least two perpendicular planes. If a truly isotropic 3D acquisition with

Table 8.2.1  Coils for MRI of the musculoskeletal system Anatomical region

Coil(s)

Spine

Phased-array coil, Helmholtz coil (cervical spine)

Pelvis, hip joints

Phased-array coil, body coil

Thigh, calf

Phased-array coil, body coil, surface coil

Knee, ankle

Surface coil, wrap-around coil (flexible surface coil)

Upper and lower arm

Wrap-around coil   (flexible surface coil)

Elbow

Flexible surface coil

Hand, wrist

Flexible surface coil, phased-array coil

adequate (in our opinion: ≤ 0.6 mm3) spatial resolution is chosen, all desired specific planes/angulations can be reconstructed from this data set without relevant loss of information. For the evaluation of the extraosseous extension of a pathological process (infection, tumors) the transverse plane usually is indispensable, while longitudinal (coronal, sagittal) planes parallel to the long axis of the affected bone/region are particularly instructive regarding bone marrow changes (bone tumors, leukemia, lymphoma, myeloma) and give a good impression of the overall extent of disease. In the spine, sagittal sections give a rapid survey of the longitudinal extension and localization of a pathological process as well as about the prevertebral and intraspinal extension of the process. Coronal sections are particularly valuable for the evaluation of a paravertebral abscess and paravertebral tumor extension while transverse sections are able to depict alterations within the spinal canal in better detail. For the examination of the joints, special conditions have to be taken into account depending on the clinical question and body region. Nowadays there is the clear recommendation to routinely use sequences in all three perpendicular imaging planes in any joint in order to have a complete depiction of the (pathological) anatomy. The multiplanar imaging capabilities of MRI are particularly valuable for the visualization of various structures with their full extension within the imaging plane and even—preferably—within one image. Examples are the depiction of the supraspinous muscle belly, tendon, and tendon insertion to the humeral head within the same section using a paracoronal section orientation. The sacroiliac joint is best depicted in sections parallel and perpendicular to the sacral bone (paracoronal, paraxial). The anterior cruciate ligament is visualized in its whole length on sagittal sections in the slightly (approximately 15°) externally rotated leg. If there is any doubt on the status of the ligament, especially in anterior cruciate ligament (ACL) reconstruction, then an additional paracoronal imaging plane may be helpful. 8.2.4  MRI Sequences Principally, pulse sequences depend on the available technique (scanner field strength, gradients, coils) and on the clinical question. Although there are no generally binding recommendations for MRI of the musculoskeletal system, there is growing agreement on the need to apply (moderately: TE ~ 30–50 ms) T2-weighted fat suppressed TSE sequences in at least two imaging planes in combination with a T1-weighted SE sequence. It is essential to choose a combination of pulse sequences that ensure a diagnostically adequate spatial resolution, high SNR, and meaningful image contrast (Table 8.2.3).

8.2  Examination Technique Table 8.2.2  Imaging planes relevant to anatomical site and purpose Region

Orientation

Clinical request

Spine

Sagittal

Longitudinal overview Disks Width of the spinal canal Prevertebral and intraspinal pathologies

Axial (resp. parallel to vertebral body endplates)

Disc herniation and bulging Width of the spinal canal Intervertebral foramen

Coronal

Paravertebral pathologies Intervertebral foramen

Axial

Overview Hip joints and femoral head Adductor muscles

Coronal

Hip joints and femoral head Adductor muscles

Parallel to femoral neck

Exact evaluation of the femoral neck

Parallel and perpendicular to sacrum

Sacroiliac joints

Axial

Comparison to opposite side of body Size of extremity Bone marrow Cortical bone Paraosseous soft tissue Neurovascular bundles Compartments

Sagittal/coronal

Longitudinal extent in bone marrow or soft tissue Affection of pelvis or thorax

Pelvis

Extremities

Joints

Dependent on joint and clinical request

Table 8.2.3  MRI sequences for musculoskeletal exams Sequence

Advantageous for:

T1-weighted SE (± fat saturation)

Morphological details Spatial orientation Paramagnetic substances Contrast enhancement

T2-weighted SE (± fat saturation), STIR

Sensitive detection of: • Edema • Inflammation • Tumor • Fluid

GRE

Joint examination 3D acquisition + multiplanar reconstruction Dynamic studies with contrast media Bone marrow reconversion and infiltration (opposed phase)

1083

1084

8  Musculoskeletal System

8.2.4.1  T1-Weighted SE Sequences T1-weighted SE sequences are robust with little susceptibility to artifacts, providing high SNR and facilitating a detailed morphological analysis of anatomy and bone marrow. Fat, paramagnetic substances (e.g., methemoglobin), and fluids rich in proteins show high signal intensity (SI). Muscle, compact bone, solid calcifications, and most fluids as well as inflammations and tumors are depicted with lower SI. 8.2.4.2  T2-weighted SE Sequences In conventional T2-weighted SE images, fat and muscle are visualized with lower SI than in T1-weighted SE images. Pathologies like inflammations and tumors as well as fluid exhibit high SI contrasting well with fat and musculature. However, these conventional T2-weighted SE sequences are very time-consuming and therefore have been replaced by TSE sequences. 8.2.4.3  TSE Sequences For the sake of acquisition time, shortening rapid (turbo, FAST) SE sequences have widely gained acceptance in MRI of the musculoskeletal system (Dawson et al. 1992), too. An echo train length of 3–16 is considered appropriate; in our routine experience an echo train length of 5–7 yields the best compromise between image blurring and time savings. Especially with short echo spacing and low-field-strengths fat shows very high SI in the TSE (as compared with the conventional SE) images because of the spin–spin coupling effects (Vahlensieck et al. 1993). The timesaving effects of the TSE sequences can be invested into the improvement of both resolution and SNR (more averaging). Since the echoes are generated by a train of 180° pulses in the TSE sequences, they are considerably less sensitive to susceptibility effects. However, the relatively high SI of fatty tissue in the TSE sequences can compromise the delineation of pathological processes within or adjacent to fatty tissue. Therefore, fat saturation techniques are considered indispensible in musculoskeletal imaging. 8.2.4.3.1  Fat Suppression Chemical frequency selective fat presaturation (CHESS) uses a preparatory pulse on the exact precession frequency of fat-bound protons, followed by spoiling gradients prior to the proper excitation pulse in order to eliminate the fat signal from the MR image. This technique can be combined with T1-weighted and T2-weighted SE sequences. It delineates processes with increased T2 re-

laxation time or increased CM enhancement with very high contrast from surrounding tissues. Another method of fat suppression is the inversion recovery sequence with short inversion time (STIR) in which an inversion (180°) pulse is applied before the excitation pulse. The inverted longitudinal magnetization recovers during the inversion time (time between the inversion pulse and the excitation pulse). The duration of the inversion time is chosen to match the time point when the longitudinal magnetization of the fat-associated protons has recovered back to zero. Thus, those protons do not yield any signal at the time of the (subsequent) 90° excitation pulse. With this technique, both a robust fat suppression and an additive T1 and T2 weighting are achieved. Therefore, structures with long T1 and/or T2 relaxation times can be depicted with high SI. FAST STIR (short TR and TI) or TSE-STIR sequences further contribute (Vahlensieck et al. 1993) to the shortening of acquisition times. The STIR sequence with its modifications (as well as (moderately) T2-weighted fat-saturated TSE sequences) is excellent for a quick, efficient, and sensitive depiction of numerous pathological changes, like edema, traumatized tissue, and tumor. To our experience, when there is a homogeneous main magnetic field, TSE FS sequences yield better SNR than so STIR sequences, which can be invested into higher anatomical resolution. Another technique of fat suppression that is increasingly gaining acceptance in musculoskeletal diagnostic MRI is so-called water excitation. In routine clinical practice, it is based on the combination of three very short excitation pulses with an amplitude relationship of 1-2-1 (binomial pulse scheme). Temporal pulse spacing is adapted to the difference in the rotational frequency between water- and fat-associated protons (π; π/2) so that only the signal of the water-bound protons contributes to the image at the end of the pulse train. The technique provides a stable fat suppression (Hardy et al. 1998) and is applicable in both SE and GRE sequences. It enables a considerable reduction in imaging time as compared to the frequency-selective fat presaturation. 8.2.4.4  GRE Sequences With GRE sequences, a considerable shortening of the imaging time can be obtained. The image contrasts depend on the type of pulse sequences (steady-state GRE, spoiled GRE), the flip angle, as well as TR and TE. At longer echo times, GRE sequences show distinctive susceptibility artifacts that, for example, emphasize calcifications as signal voids. The phase shift between fat and water protons depends on the echo time of the sequence and therefore on whether there are “in-phase,” “out-ofphase,” or “opposed-phase” conditions. Whereas the SIs of fat and water are added together under “in-phase”

8.2  Examination Technique

conditions, they are subtracted from each other under “opposed-phase” conditions. Opposed-phase GRE sequences are suitable for the examination of bone marrow. In the hematopoeitic marrow, the SI of the water component (hematopoeitic cells) and the SI of the fat component (fat cells) extinguish each other, so that normal bone marrow exhibits low SI (Lang et al. 1992). In neoplastic infiltration, however, the water component predominates resulting in high SI. In musculoskeletal MRI GRE, pulse sequences are also used to image the hyaline articular cartilage and to perform dynamic contrast-enhanced examinations. They are mostly applied as three-dimensional acquisitions, yielding high-resolution reconstructions and enabling multiplanar reformations. 8.2.4.5  MTC Sequences In magnetization transfer contrast (MTC) sequences the broad band resonance of protons associated to macromolecules is saturated by an MT prepulse and, depending on the tissue imaged, entails a signal loss (related to collagen-rich structures) as compared with the same sequence without MT prepulse. Subtracting data sets obtained with an MT prepulse from corresponding data sets without an MT prepulse will thus produce low SI in all tissues that do not exhibit a high MTC effect. This can be useful for MRI of the musculoskeletal system since muscle, cartilage, and tendons show a strong MTC effect and therefore appear with low SI. Artificial SI increase in tendons running oblique to the main magnetic field (“magic-angle phenomenon”) is suppressed by the MTC contrast and thus can be differentiated from true tendon degeneration (e.g., in the rotator cuff). The MTC subtraction technique has been used for the depiction of superficial cartilage lesions (Vahlensieck et al. 1994).

8.2.5  Contrast Medium Application For MRI of the musculoskeletal system, usually the standard dose (0.1 mmol/kg) of paramagnetic gadolinium chelates is administered. Dynamic examinations using GRE sequences may be advantageous when assessing the degree of vascularization and permeability in bone and soft-tissue tumors, the differentiation of vital and necrotic tumor portions, and inflammatory proliferations of the synovium (Table 8.2.4). However, pathologic findings in the fatty tissues or in the bone marrow, which clearly are depicted on non-enhanced T1-weighted sequences, may be masked and therefore no longer recognizable after contrast medium application. Therefore, comparison with precontrast imaging is indispensable. Contrast medium enhancement is most prominent on fat-saturated T1-weighted sequences. Nonetheless, comparison with precontrast images is required, since many lesions show low signal intensity on non-enhanced fat saturated sequences (cf. overview). The intra-articular injection of paramagnetic contrast medium (direct MR arthrography) is particularly useful for the diagnosis of chondral and osteochondral lesions and contributes to the depiction of capsuloligamentous injury. Distension of the joint capsule and contrast in MR arthrography improve the delineation of the surface of hyaline articular cartilage and other intra-articular structures (menisci, ligaments). Furthermore, pathological communications between the joint space and extra-articular spaces may become recognizable (e.g., rotator cuff tear) by extravasation of the contrast medium. An interesting alternative to direct MR arthrography may be indirect MR arthrography (Semmler et al. 1988). After an intravenous injection of paramagnetic contrast medium, there is accumulation (by diffusion) of this contrast medium into/within the joint space. This process is considerably faster in inflammatory arthropathies

Table 8.2.4  Examinations using GRE sequences Anatomical site

Advantageous for:

Bone and soft tissue tumors

Differentiation of vital from necrotic tumor components Better differentiation of tumor, muscles and edema Response to chemotherapy Differentiation of tumor recurrence, reactive changes and fibrosis

Inflammation

Differentiation of abscess membrane and content Detection of synovial proliferations

Avascular necrosis

Vascularization of necrotic bone tissue Early diagnosis of bone necrosis

Joints

Direct or indirect MR arthrography Visualization of superficial cartilage lesions and internal joints derangement Rotator cuff lesions

1085

1086

8  Musculoskeletal System

or with (active or passive) motion of the corresponding joint. Thus, with a standard dose of i.v. contrast medium diagnostically relevant demarcation of the intra-articular structures becomes feasible in T1-weighted (in particular fat-saturated) sequences.

References 1.

2.

8.3  Relaxation Times, Signal Intensities, and Contrast Behavior To date, quantitative tissue relaxation times are not obtained routinely with the techniques implemented in whole-body scanners. They do not permit a clear differentiation between different pathologies (e.g., tumor versus inflammation) nor do they allow determination of the dignity of bone and soft-tissue tumors. However, there is growing (experimental) evidence that they may well be useful for the diagnosis and follow-up of cartilage degeneration. Bone marrow and fat show a high SI in T1w SE sequences. The SI of the bone marrow depends on the content of hematopoeitic cells and fat cells and is age dependent (Vogler and Murphy 1988). Muscle shows intermediate SI on T1-weighted images. Cancellous bone exhibits a low SI independent of the sequence parameters (Table 8.3.1). T2-weighted SE sequences show a decrease of the SI of fat and musculature as compared to T1-weighted sequences, with fat maintaining high SI in TSE sequences. With few exceptions, pathologic changes of the musculoskeletal system are characterized by codirectional changes of the SI, independent of the underlying pathology. Bone and soft-tissue tumors, just like inflammations, appear hypointense on T1-weighting and have high SI in the T2-weighted sequences. Accordingly, they reveal high contrast to the bone marrow and fat on T1-weighted sequences. This contrast is even more prominent in T2weighted fat-saturated sequences or STIR sequences.

4.

5.

6.

8.4  Diseases of the Bone Marrow and Hematopoietic System The bone marrow is one of the most important and greatest organs of the human body, with approximately 3,000 g of it in the adult. Besides immunologic functions, it has the task of providing the organism with red and white blood corpuscles and platelets. A complex of vessels and the erythropoietic and granulopoietic system are stored between the trabeculae of the spongy bone. Moreover, fat and reticular cells are found there (Table 8.4.1). While the complete bone marrow serves for hematopoiesis in the newborn, in the adult—usually after

Table 8.4.1  Distribution of red and yellow marrow

Table 8.3.1  Physiologic signal intensities T1-weighted

T2-weighted

STIR

Muscle

_

_

_

Fat

+

+

_

Bone marrow

+

+

_

Tumor

_

+

+

Inflammation

_

+

+

Compact bone

_

_

_

+ hyperintense signal, – hypointense signal

3.

Dawson KL, Moore SG, Rowland JM (1992) Age-related marrow changes in the pelvis: MR and anatomic findings. Radiology 183:47–51 Lang P, Fritz R. Vahlensieck M et al (1992) Residuales und rekonvertiertes haematopeotisches Knockenmark im distalen Femure-Spinecho und gegenphasierte Gradientenecho-MRT. RoFo 156:89–95 Semmler W, Gademann G, Bachert-BaumannP et al (1988) Monitoring of human tumor response to therapy by means of 31P-MR-spectroscopy. Radiology 166:533–539 Vahlensieck M, Seelos K, Traeber F et al (1993) Magnetresonanztomogographie mit schneller STIR-Technik: Optimierung und Vergleich mit anderen Sequenzen an einem 0.5 T: Einfluss von Echodistanz und Echozahl auf den Bildkontrast. RoFo 158:260–264 Vahlensieck M, Dombrowski R, Leutner C et al (1994) Magnetization transfer contrast (MTC) and MTC-subtraction enhances cartilage lesions and intrasubstance degeneration in vitro. Skelet Radiol 23:535–539 Vogler JB, Murphy WA (1988) Bone marrow imaging. Radiology 168:679–693

Red marrow: hematopoietic

Yellow marrow:  fatty

In all marrow spaces in the newborn

Replaces red marrow from appendicular to axial sites

In the short and flat bones in the adult

In tubular bones from diaphysis to metaphysis Soon after ossification in epi- and apophyses

40% water, 40% fat, 20% protein

15% water, 80% fat, 5% protein

8.4  Diseases of the Bone Marrow and Hematopoietic System

the 25th year of life—hematopoiesis is largely limited to the axial skeleton with vertebral bodies, ribs, sternum, scapulae, pelvis and calcaneus, to the skull calotte, and the proximal portions of the humerus and femur. Soon after birth a centripetal transformation of the hematopoietic bone marrow into fat marrow begins, which progresses in the long bones from the diaphysis to the metaphysis (at first distal, then proximal). In the epi- and apophyses, fatty marrow is detected already a few months after ossification. In addition, in the axial skeleton, the relative fat content increases with age (Table 8.4.2). The distribution of red and yellow bone marrow as represented in Tables 8.4.1 and 8.4.2 was analyzed extensively in the last few years and particular age-dependent changes were detected. These must be taken into account so that normal results are not interpreted as pathological, e.g., as a bone marrow–infiltrating process. In the pelvis confluent fat marrow islands (Cova et al. 1993) develop at first adjacent to the acetabulum, then in the ischium, and finally in the portions of the sacrum and ileum adjacent to the sacroiliac joints. The vertebral bodies—primarily in the lumbar spine—may have band-shaped fat islands adjacent to the intervertebral disks, which are frequently associated with degenerative changes of the intervertebral discs. Moreover, the fat marrow can also be multifocally distributed (Ricci et al. 1990) in the vertebral bodies. In the proximal metaphyses of the humerus a transformation into fat marrow, most frequently in the lateral portion, is found with advancing age. This also applies to the lower and middle segment of the scapula adjacent to the glenoid. With teenagers, a streaky image of the distal femur metaphysis is often seen prior to the complete conversion of the red marrow (Lang et al. 1992). If the need for elements of the hematopoiesis increases, a reconversion (i.e., transformation of fat marrow into hematopoietic bone marrow) in a centrifugal direction begins. This is also detected when the available bone marrow cannot meet the requirements of the body anymore, e.g., in chronic anemia and infections, cardiac insufficiency, or repression of greater portions of the hematopoietic marrow by tumor infiltration (myeloma, lymphoma, metastases, leukemia). Also after chemotherapy, radiation therapy or in osteomyelosclerosis reconversion of the fat marrow can be seen, as well as

in competitive athletes, heavy smokers, premenopausal women (menstruation), and after a severe loss of blood (Fig. 8.4.1). 8.4.1  Leukemia In leukemia, diffuse osteopenia, radiolucent and radiodense metaphyseal bands, osteolytic lesions, periostitis, and osteosclerosis are observed in 5–100% of patients. In chronic leukemia, osseous manifestations are less common and less pronounced than in acute leukemia. MRI may be used to determine the extent of the bone marrow changes, the response to therapy, the detection of skeletal complications such as septic arthritis, osteonecrosis, and bony destruction and the investigation of

Table 8.4.2  Age dependency of red and yellow marrow Relation of hematopoietic marrow to fat marrow After puberty

1:1

Middle-aged adult

1:2

Seniors

1:4

Fig. 8.4.1a,b  Marrow reconversion in the proximal femur and acetabulum in a 43-year-old patient after chemotherapy. a Coronal STIR-image with confluent “patchy” areas of hyperintense signal, and b coronal T1-weighted SE image with corresponding hypointense signal

1087

1088

8  Musculoskeletal System

specific types such as granulocytic sarcoma (chloroma) (Fig. 8.4.2). In various types of leukemia, a reduction of the SI of bone marrow is noted in T1-weighted images, while T2weighted SE images do not exhibit distinctive SI changes. These changes are similar to other infiltrative bone marrow diseases such as lymphoma or multiple myeloma. While in acute leukemia of children alterations in T1 relaxation times may be useful for detection of recurrence, no benefit was found in adult patients with leukemia.

When MRI is performed before, during, and at the end of chemotherapy, no differentiation of aplasia and leukemic infiltration could be obtained. In acute lymphoblastic leukemia, no aplasiogenic therapy is employed, and widely normal signal intensities are noted within the bone marrow already at the first control examination. The normocellular or hypocellular bone marrow of patients in complete remission is mostly not differentiable from healthy bone marrow in T1-weighted images. In some cases, it shows a slight, mostly spotted SI reduction. Granulocytic sarcoma (chloroma) is most often associated with acute leukemia of the myeloid type and is more common in children than in adults. Granulocytic sarcomas are composed of immature cells of the granulocytic series and typically involve bone, soft tissue, lymph nodes, orbits, and the skin. 8.4.2  Multiple Myeloma Multiple myeloma is a malignant bone marrow neoplasia in which atypical plasma cells proliferate unrestrainedly and typically produce monoclonal immunoglobulins within the bone marrow. It is a common disease, repre-

Fig. 8.4.2a,b  29-year-old patient with acute lymphatic leukemia. a Coronal T1-weighted SE image before chemotherapy with diffuse hypointense signal in both femur shafts. b After 4 weeks of chemotherapy, normal signal intensity of the bone marrow is noted

Fig. 8.4.3a,b  Focal infiltration of bone marrow in a patient with multiple myeloma. a Sagittal STIR and b sagittal T1-weighted image. Please note the lesion in L1 with hyperintense signal in the STIR and hypointense signal in the T1-weighted image. Additionally, a pathologic fracture of L2 has occurred

8.4  Diseases of the Bone Marrow and Hematopoietic System

senting 10–15% of the malignancies of the hematopoietic system. The evaluation of the extent of infiltration and the pattern of infiltration can be carried out by means of MRI after the diagnosis is established by bone marrow biopsy. The spine, particularly the thoracic and lumbar spine, is often involved. When there is a diffuse bone marrow infiltration, conventional radiographs may exhibit a nonspecific osteopenia or are completely normal. According to Baur-Melnyk et al. (2005), five different patterns of infiltration can be distinguished: 1 Normal appearance of the bone marrow in low-grade diffuse bone marrow infiltration 2 Focal involvement (Fig. 8.4.3) 3 Homogenous, diffuse infiltration (Fig. 8.4.4) 4 Mixed focal and diffuse

Fig. 8.4.4a,b  Diffuse infiltration of bone marrow in a patient with multiple myeloma. a Sagittal STIR- and b sagittal T1weighted image. Diffuse islands of hyperintense signal in STIRweighted image and hypointense signal in the T1-weighted image can be seen

5 “Salt-and-pepper pattern” with inhomogeneous bone marrow signals due to multiple fat islands (Fig. 8.4.5) The combination of a T1-w SE sequence and a fat suppression technique, such as STIR sequences is best suited for detecting the infiltration patterns and excluding differential diagnoses. In low-grade diffuse interstitial infiltration of atypical plasma cells (below 20 vol%), no SI alteration of the bone marrow is recognizable. In intermediate to severe infiltration, bone marrow shows a variable, homogenous SI change. SI is reduced in the T1-weighted image, whereas other pulse sequences show less pronounced alterations. Focal foci are best depicted as hyperintense areas in STIR sequences. Focal foci in simultaneously diffuse bone marrow infiltration show the best contrast in opposed-phase GRE-sequences. Contrast-enhanced T1-weighted images are appropriate (Stäbler et al. 1996) to detect diffuse bone marrow infiltration. In cases of intermediate to severe infiltration, they show a homogenous enhancement ranging above 40%. Comparing whole-body MRI versus whole-body CT, MRI is superior to CT in detecting early infiltrations with myeloma cells, without osteolyses due to its ability to directly visualize the soft-tissue components of the bone marrow. CT, on the other hand, is more precise in

Fig. 8.4.5a,b  “Salt-and-pepper pattern” in multiple myeloma. a Sagittal STIR and b sagittal T1-weighted image. The typical salt-and-pepper pattern arises from multiple fat depositions within the bone marrow. Even if this pattern appears highly suggestive of high-grade involvement, it is indicative of fatty conversion following successful therapy

1089

1090

8  Musculoskeletal System

advanced myeloma since it enables assessment of bony destruction and evaluation of fracture risk. Indications for MRI in multiple myeloma • Unclear osteopenia • Detection of paraproteins, without radiological correlate • Myeloma without secretion of paraproteins • Preoperative evaluation of the bone marrow infiltration • Staging under therapy Since parallel imaging and special coil devices allow a whole-body exam within a reasonable amount of time, MRI has been implemented in the clinical staging process of multiple myeloma within the last two years (BaurMelnyk et al. 2005). 8.4.3  Metastases and Malignant Lymphomas Metastases of various primary cancers, e.g., breast or prostate cancer, are associated with an increase in tracer uptake in bone scans and permit a clear diagnosis particularly when multiple foci of increased uptake are present. In some malignancies, however, bone scans and radiographs are negative in the early stage of metastatic involvement. Moreover, bone scans have a high rate of false positives. In these cases and in primarily skeletal malignant lymphomas or skeletal manifestation of nodal malignant lymphoma, MRI can contribute to establish

the diagnosis. With modern MRI systems at 1.5 and 3 T, a whole-body exam of the complete skeletal system can be obtained in less than 45 min. Lymphomas include Hodgkin’s disease and nonHodgkin’s malignant lymphomas, which occur approximately three times more frequently than does Hodgkin’s disease. Skeletal changes are common in all lymphomas and occur in 5–50% of cases, depending on the stage of the disease and the method of investigation. With nonHodgkin’s lymphoma diffuse infiltration of the bone marrow is more frequently observed than a primary lesion is and usually appears after the initial manifestation. The prevalence is estimated 10–20% in adults and 20–30% in children. In Hodgkin’s disease, skeletal involvement is quite common; however, diffuse SI alterations are less often observed (Sugisawa et al. 2006). Metastases and malignant lymphoma of the bone exhibit a SI reduction in T1-weighted images and a SI increase in T2-weighted images (TSE sequences are less than the conventional T2-weighted SE technique). However, STIR and opposed-phase GRE sequences as well as fat-saturated contrast-enhanced images are considerably more sensitive for the detection of metastases and lymphoma manifestations in the bone marrow (Figs. 8.4.6, 8.4.7). An exception is formed by metastases of malignant melanoma since the paramagnetic effect of the melanin may cause high SI in T1-weighted images. MRI is considerably superior to other examination procedures for intraosseous metastases and intraosseous manifestations of malignant lymphomas (Ghanem et al. 2006; Frat et al. 2006). Fig. 8.4.6a–c  38-year-old female patient with hepatic and osseous metastases of breast cancer. a Sagittal STIR image of the cervical and thoracic spine with hyperintense metastatic foci, and b sagittal T1-weighted image pre-contrast with hypointense lesions. c Sagittal T1-weighted image post-contrast exhibits significant contrast uptake of multiple metastases

8.4  Diseases of the Bone Marrow and Hematopoietic System

In a study performed by Schmidt et al. (2007), diagnostic accuracy of screening for bone metastases was evaluated using whole-body magnetic resonance imaging compared with combined FDG PET and CT (FDG PET CT). Thirty patients with different oncological diseases and suspected skeletal metastases underwent FDG PET CT as well as whole-body MRI. The standard of reference was constituted by radiological follow-up within at least 6 months. Whole-body MRI showed a sensitivity of 94%; PET CT exams achieved a sensitivity of 78%. Specificities were 76% for whole-body MRI and 80% for PET CT. Diagnostic accuracy was 91 and 78%, respectively, for whole-body MRI and PET CT. Cut-off size for the detection of malignant bone lesions was 2 mm for MRI and 5 mm for PET CT. MRI revealed ten additional bone metastases due to the larger field of view. 8.4.4  Storage Diseases Musculoskeletal findings are a frequent manifestation of lipidoses. Lipid storage diseases includes Gaucher’s dis-

Fig. 8.4.7a–c  40-year-old male patient with biopsy-proven B-cell non-Hodgkin’s lymphoma of the fibula. a Coronal STIR image with a small hyperintense lesion, b coronal T1-weighted image before i.v. Gd application showing hypointense SI, and c coronal T1-weighted image after Gd application with Gd enhancement

ease, Niemann-Pick-disease, Fabry’s disease, Krabbe’s disease, and Tay-Sachs disease. Gaucher’s disease is an autosomal recessive disorder that results in an abnormal accumulation of glucocerebrosides in the bone marrow, brain, and spleen due to a deficit of glucocerebroside hydrolase or betaglucosidase. The glucocerebrosides are stored in the cells of the reticuloendothelial system (RES). The Gaucher cells replace the normal fat and hematopoietic cells in the bone marrow. Pancytopenia, hepatosplenomegaly, and bone changes are found clinically with osteoporosis, marrow infiltration, subarticular osteonecrosis, fractures, and deformations (“Erlenmeyer-flask deformity”), which is very suggestive of the diagnosis (Fig. 8.4.8). So far, MRI is the best diagnostic modality for assessing skeletal complications in Gaucher’s disease and monitoring response to enzyme replacement therapy (Maas et al. 2002). A pronounced SI reduction of the bone marrow is found in T1-, PD- and T2-weighted images. Spectrometric examinations of surgically harvested bone marrow showed very short T2 times. This signal pattern is almost

1091

1092

8  Musculoskeletal System

unique since most bone marrow diseases show an increase in T1 and T2 time. The SI reduction of the bone marrow at Gaucher’s disease usually spares the epiphyses. An affection of the

epiphyses is only seen in advanced cases. These alterations are found more frequently and distinctively in the proximal segments of the long bones than in the distal skeleton portions.

Fig. 8.4.8a–d  12-year-old patient with Gaucher’s disease. Sagittal a T1-weighted and b STIR images of the cervical and thoracic spine with hypointense bone marrow; c and d are T1-weighted coronal images of the thighs and knees of the same patient suggesting “Erlenmeyer-flask deformity”. Additionally, a bone marrow infarct is found in the right distal femur

8.4  Diseases of the Bone Marrow and Hematopoietic System

8.4.5  Osteomyelofibrosis Osteomyelofibrosis is a rare entity in which the normal elements of the bone marrow are replaced by fibrotic connective-tissue elements. Consequently, the SI of the bone marrow in T1- and T2-weighted images is reduced (Fig. 8.4.9). In some cases, almost the entire bone marrow is replaced. The degree of bone marrow fibrosis is generally thought to correlate with the severity of the disease. Since large skeleton segments can be depicted in one examination by MRI, prognostic indications can be derived from the extent of the SI changes. 8.4.6  Aplastic Anemia and Sequelae of Chemotherapy and Radiotherapy In aplastic anemia as well as after aggressive chemotherapy and radiation therapy, hematopoietic elements are replaced by fat cells. After radiation, these changes are limited exactly to the irradiation field, with a typical sequence of different stages. In the first week after the irradiation is completed, edema develops with low SI in T1-weighted and high SI in T2-weighted and STIR images. After 2–3 weeks, the edema resolves and a transformation into fat marrow begins, which can persist at high doses (>30–40 Gy), while at low doses hematopoietic marrow reappears. In T1-weighted images, the bone marrow shows high SI, which exceeds that of the normal bone marrow and corresponds largely to that of the subcutaneous fatty tissue. These changes are particularly conspicuous in patients with radiation treatment since a direct comparison with the lower SI of normal bone marrow is possible at the field margins. Codirectional changes can appear at the field borders because of the scattered radiation, however. If hematopoietic foci develop in the bone marrow in aplastic anemia under therapy, then confluent islands of low SI in T1-weighted images can be found, which finally will extend throughout the entire medullary cavity. In patients with bone marrow transplantation, a repopulation of the bone marrow is recognizable through band-shaped SI reductions of the vertebral body portions adjacent to the endplates after approximately 2–3 months. A central zone of high SI in T1-weighted images may persist. 8.4.7  Hemosiderosis, Hemochromatosis, and Sickle Cell Anemia 8.4.7.1  Hemosiderosis and Hemochromatosis The skeletal manifestations of hemochromatosis include chondrocalcinosis, osteoporosis, and arthropathy with

Fig. 8.4.9a,b  Patient with osteomyelofibrosis. a STIR and b T1weighted images of the thoracolumbar spine, with strongly reduced signal intensity of bone marrow. Please note the two hyper­ intense lesions in T9 and T10, which show hyperintense signal in both sequences indicating the presence of hem­an­giomas

joint-space narrowing, subchondral cysts, and osteophytes formation. This type of arthropathy is more frequent in genetic “primary” hemochromatosis than it is in secondary hemosiderosis and most often involves the small joints of the hands. The SI of the bone marrow is reduced particularly in T2*- and T2-weighted images since hemosiderin and ferritin deposits cause susceptibility artifacts. The SI changes are dependent on the concentration of the stored hemosiderin. Multifocal or diffuse SI changes can be found leading to the image of “black bone marrow.” MRI may reveal meniscal chondrocalcinosis with findings resembling a meniscal tear.

1093

1094

8  Musculoskeletal System

8.4.7.2  Sickle Cell Anemia

8.5  Inflammatory Diseases of Bone and Joints

Sickle cell disease is characterized by the presence of abnormal hemoglobin (HbS). Expansion of the hematopoietic bone marrow due to chronic anemia as well as secondary hemosiderosis due to chronic hemolysis and repeated blood transfusions contribute to a significant SI reduction of the bone marrow in T1-weighted images. In sickle cell anemia osteomyelitises and bone infarcts are frequently observed, which are often difficult to detect in T1-weighted images since there is no or only low contrast to the hypointense bone marrow (Patel et al. 2006).

Inflammatory diseases of the musculoskeletal system may have various etiologies, including bacterial (specific and non-specific) and viral infections and rheumatoid diseases. According to the primary site of microorganism seeding, bacterial infections of the skeletal system are classified as osteomyelitis, arthritis, and spondylitis. Acute, subacute, and chronic forms exist. Soft-tissue infections also are included in this section. The various rheumatoid diseases are characterized by synovial proliferations. The main categories are rheumatoid arthritis and seronegative spondyloarthropathies.

References 1.

Baur-Melnyk A, Buhmann S, Durr HR, Reiser M (2005) Role of MRI for the diagnosis and prognosis of multiple myeloma. Eur J Radiol 55:56–63 2. Cova MA, Dalla Palma L, Pozzi-Mucelli RS, Ricci C (1993) MRI of spondylodiskitis: contribution of gadolinium-DTPA and fat suppression sequence. Eur Radiol 3:541–547 3. Frat A, Agildere M, Gencoglu A, Cakir B, Akin O, Akcali Z, Aktas A (2006) Value of whole-body turbo short tau inversion recovery magnetic resonance imaging with panoramic table for detecting bone metastases: comparison with 99MTc-methylene diphosphonate scintigraphy. J Comput Assist Tomogr 30:151–156 4. Ghanem N, Altehoefer C, Kelly T, Lohrmann C, Winterer J, Schafer O, Bley TA, Moser E, Langer M (2006) Wholebody MRI in comparison to skeletal scintigraphy in detection of skeletal metastases in patients with solid tumors. In Vivo 20:173–182 5. Lang P, Fritz R. Vahlensieck M et al (1992) Residuales und rekonvertiertes haematopeotisches Knockenmark im distalen Femure-Spinecho und gegenphasierte Gradientenecho-MRT. RoFo 156:89–95 6. Maas M, Poll LW, Terk MR (2002) Imaging and quantifying skeletal involvement in Gaucher disease. Br J Radiol 75(Suppl 1):A13–A24 7. Patel A, Klassen C, Griffiths HJ (2006) The case: bone disease in sickle cell anemia. Orthopedics 29:470, 552–554 8. Ricci C, Cava M, Kang YS et al (1990) Normal age related patterns of cellular and fatty bone marrow distribution in the axial skeleton: MR imaging study. Radiology 177:83–88 9. Schmidt GP, Schönberg SO, Schmid R, Stahl R, Tiling R, Becker CR, Reiser MF, Baur-Melnyk A (2007) Screening for bone metastases: whole-body MRI using a 32-channel system versus dual-modality PET-CT. Eur Radiol 17:939–949 10. Stäbler A, Baur A, Bartl R et al (1996) Contrast enhancement and quantitative signal analysis in MR imaging of multiple myeloma AJR 167:1029–1036 11. Sugisawa N, Suzuki T, Hiroi N, Yamane T, Natori K, Kiguchi H, Kuraishi Y, Higa M (2006)Primary bone malignant lymphoma: radiographic and magnetic resonance images. Intern Med 45:665–666

8.5.1  Bacterial/Viral Infections 8.5.1.1  Osteomyelitis Osteomyelitis is defined as an infection of bone and marrow. The main causes are bacterial, although viruses, fungi, and parasites also can infect bony structures and bone marrow. 8.5.1.1.1  Acute Osteomyelitis A single pathogenic organism usually is responsible for acute hematogenous osteomyelitis. Single or multiple bones can be affected, with infection of multiple sites seemingly more common in children than in adults. There are three main routes of contamination: • Hematogenous infection (infection via bloodstream from sources such as pharyngitis, pyodermia, otitis) • Spread from an adjacent infection (e.g., sinusitis, or cutaneous or dental infection) • Direct implantation of infectious material (penetrating injury) The fourth route is postoperative infection, which may occur by the above-mentioned three principal routes. Osteomyelitis often manifests as a severe clinical illness, with swelling and fever. Local pain and reduced range of motion of an extremity are suggestive of osteomyelitis. The metaphyses of the long bones are the most common site of infection. Since early treatment is required in order to avoid long-term sequelae, diagnosis should be made at an early stage. Conventional radiography may be negative for up to 14 days after onset of symptoms, whereas three-phase bone scintigraphy and leukocyte scintigraphy show early intense tracer uptake. In osteomyelitis, the amount of intramedullary fluid and inflammatory cells is increased. Therefore, MRI (Fig. 8.5.1; Table 8.5.1) shows a signal intensity (SI) decrease in

8.5  Inflammatory Diseases of Bone and Joints

the bone marrow on T1-weighted images. In T2-weighted images, the inflammatory process results in high signal intensity, so that contrast between inflammation and normal bone marrow is reduced or absent. This is also true for contrast-enhanced T1-weighted images. Thus, short-tau inversion recovery (STIR) and proton density (PD) fat-suppressed sequences are highly sensitive for the detection of osteomyelitis; in these sequences the inflammatory process shows high signal contrasted against the low signal of normal bone marrow. Also, contrast-en-

hanced T1-weighted fat-suppressed images are excellent for the detection of osteomyelitis. Inflammatory tissue is depicted with strong contrast enhancement, and abscesses are clearly visualized with marked enhancement of the abscess membrane. Any extraosseous spread of infection is best detected on STIR, PD fat-suppressed, and contrast-enhanced T1wfat-suppressed images; it might be missed in non-enhanced T1-weighted images. In contrast to the SI increase in tumors, the SI increase in paraosseous soft tissues is

Table 8.5.1  Acute osteomyelitis: signal of the inflammatory process relative to that of normal bone marrow and muscle Sequence

Bone marrow edema

Muscle

T1-weighted

Hypointense

Isointense or slightly hyperintense

T2-weighted

Iso- or slightly hyperintense

Hyperintense

STIR, PD fat-suppressed

Hyperintense

Hyperintense

T1-weighted CE

Iso- or slightly hyperintense

Hyperintense

Fat-suppresses T1-weighted CE

Hyperintense

Hyperintense

Fig. 8.5.1a–d  Acute osteomyelitis after resection of the second toe in a 58-year-old patient with peripheral arterial disease. a The STIR image shows extensive bone marrow edema in the second and third metacarpal bones and within the adjacent soft tissues. b Coronal and c axial T1-weighted images demonstrate an ill-defined signal decrease of the bone marrow. d Coronal T1-weighted

contrast-enhanced image and e axial T1-weighted fat-suppressed image after contrast administration show significant contrast enhancement with reduction of the difference between normal and inflamed marrow. Note the absent marrow signal on T1-weighted images that “returns” on T1-weighted contrast-enhanced images. This phenomenon is called “vanishing-bone sign”

1095

1096

8  Musculoskeletal System

indistinct and poorly demarcated from the surrounding muscles and fat. If a soft tissue abscess is present, then demonstration of its fluid content and surrounding abscess membrane allow for clear diagnosis. In order to select the appropriate therapy, it is essential to differentiate between isolated soft tissue infection, osteomyelitis with concomitant soft tissue infection, and osteomyelitis alone. MRI is superior to all other imaging methods in making this distinction. Especially in patients with diabetes (Fig. 8.5.2), this differentiation is crucial because local operative procedures together with antibiotics for a short time are sufficient therapy for soft tissue infections, whereas more extensive surgery and long-term antibiotic therapy are required when bone also is involved.

Osteomyelitis lasting for more than 6 weeks is generally referred to as chronic osteomyelitis. Radiographs

exhibit periosteal reactions, sclerosis, and osteolysis. Findings in active inflammation include sequestra and sinus tracts. T1-weighted MR images show extensive SI decrease of the bone marrow. T1-weighted images, however, do not allow differentiation of active infection from changes caused by previous infection and healing, such as sclerosis or marrow fibrosis. However, on STIR and T2-weighted images areas of high signal intensity are found in active processes (Table 8.5.2). Again, contrast-enhanced T1-weighted fat-suppressed sequences are most useful since a central non-enhancing area with peripheral enhancement is almost always proof of active infection. Sinus tracts can be identified on T2-weighted images as linear SI increases extending from the bone to the skin. On contrast-enhanced T1-weighted fat-suppressed images, prominent contrast enhancement can be noted. Sequestrations (bony fragments with increased sclerosis) show low SI on T1-weighted and T2-weighted images and therefore stand out relative to the surrounding inflamed tissue.

Fig. 8.5.2a,b  62-year-old patient with insulin-dependent diabetes and diabetic arthropathy. a Sagittal T1-weighted image with complete disintegration of the ankle joint and signal reduction in para-articular soft tissue. b Sagittal T1-weighted fat-sup-

pressed image after contrast administration. Prominent contrast enhancement in the para-articular soft tissue as well as within the destroyed joints. Central areas without enhancement are due to abscess formation

8.5.1.1.2  Chronic and Subacute Osteomyelitis

Table 8.5.2  Chronic osteomyelitis: signal intensity of involved bone marrow relative to normal marrow Sequence

Active infection

Healed infection

T1-weighted

Hypointense

Hypointense

T2-weighted/STIR

Hyperintense (central)

Hypointense

T1-weighted CE

Hyperintense (peripheral)

Hypointense

8.5  Inflammatory Diseases of Bone and Joints

Subacute Types of Osteomyelitis Subacute types are Brodie’s abscess (Fig. 8.5.3) as well as Garré ’s chronic sclerosing osteomyelitis. These types develop when the host has moderately high resistance to infection, or when the infecting organisms have a somewhat reduced virulence. Brodie’s abscesses appear as well-circumscribed areas of low signal intensity on T1-weighted images and of high signal intensity on T2-weighted images (Fig. 8.5.4). A rim of low signal intensity is found in all sequences, resulting from bone sclerosis. Surrounding bone marrow edema may be seen on T2-weighted and STIR images. A double line on T2-weighted images may be produced by granulation tissue on the inner wall of the abscess. Delineation of Brodie’s abscess is facilitated by intravenous contrast material, which causes peripheral rim enhancement. 8.5.1.1.3  Posttraumatic and Postoperative Osteitis and Osteomyelitis In the early posttraumatic/postoperative phase, changes caused by fracture healing or osteosynthesis can raise diagnostic problems. In bone scintigraphy, the increased bone turnover leads to a long-lasting, nonspecific increased activity. When patients undergo osteosynthesis, the diagnostic role of CT is limited to a greater degree than that of MRI. In areas of former screw holes and in zones of fracture healing a band-like signal intensity reduction is found. In recent fractures, an inhomogeneous pattern of signal intensities results from the simultaneous presence of hematoma, granulation tissue and fracture callus formation. Again, contrast-enhanced T1-weighted fat-suppressed sequences are very valuable. Central regions of fluid-like

Fig. 8.5.3a–d  Brodie’s abscess in the proximal tibial metaphysis with extension into the epiphyseal plate in a 10-year-old girl. a The coronal T1-weighted image shows a multilobulated hypointense lesion with hyperintense septations in the tibial

signal with peripheral enhancement are indicative of an active inflammatory process, especially when surrounding edema is present (Fig. 8.5.5). 8.5.1.1.4  Chronic Recurrent Multifocal Osteomyelitis Chronic recurrent multifocal osteomyelitis (CRMO) predominantly affects children and adolescents (Fig. 8.5.6). CRMO is characterized by a prolonged course (over several years), with multifocal lesions. CRMO is a chronic, systemic aseptic inflammation and constitutes approximately 2–5% of all osteomyelitis cases. The disease predominantly affects the metaphyses of the long bones adjacent to the growth plates, and the clavicle and spine. The etiology is still unknown. Females are more frequently affected (F:M = 5:1). The skeletal manifestations may be associated with skin lesions. The disease is self-limiting and has a good prognosis. 8.5.1.2  Septic Arthritis Septic arthritis may result from spread from a contiguous focus of osteomyelitis, hematogenous inoculation of the joint capsule and synovial membrane, or direct inoculation of bacteria after traumatic or iatrogenic injury to a joint. On radiographs, soft-tissue alterations such as swelling, joint effusion, and obliteration of the normal fat planes are detected in the early phase. When proteolytic enzymes are released from leukocytes and cartilage is destroyed, narrowing of the joint space, often associated with juxta-articular osteoporosis (which is caused by hyperemia and pain-related inactivity) may be observed. As

metaphysis with a surrounding hypointense rim indicative of sclerosis. b Coronal STIR image and c axial T2-weighted images exhibit fluid-like characteristics. d Rim enhancement is shown on the coronal T1-weighted contrast-enhanced image

1097

1098

8  Musculoskeletal System

Fig. 8.5.4a–c  14-year-old patient with chronic-sclerosing osteomyelitis. a The radiograph exhibits enlargement of the femur with cortical thickening, periosteal new bone formation and destruction. b Axial T1-weighted image shows thickening of the cortex and foci of increased signal. The marrow signal is

reduced, and ill-defined areas of intermediate signal with obliteration of the fat planes are present in the paraosseous region. c Contrast-enhanced T1-weighted image demonstrates contrast enhancement of the cortex and paraosseous tissue

Fig 8.5.5a–c  58-year-old patient with posttraumatic osteomyelitis and sinus tract to the skin. a Lateral radiograph of the knee and proximal lower leg. Extensive osteoarthritis and healed fracture of the fibula. Defect in the tibial metaphysis with peripheral sclerosis and sequestration. b Coronal T1-weighted im-

age. The defect in the proximal tibia and the fluid collection in the adjacent soft tissues exhibit low signal intensity. c Coronal T1-weighted fat-suppressed images with contrast enhancement show the intraosseous lesion and the soft tissue abscess, with a sinus tract to the skin

8.5  Inflammatory Diseases of Bone and Joints

Fig. 8.5.6a–f  CRMO in a 23-year-old female patient. a STIR image of the spine shows signal increase in T4, T5, and T6. Wedge-shaped compression fracture of T5. b Corresponding low signal intensity on T1-weighted image and c extensive contrast uptake in T1-weighted fat-suppressed images after administration of i.v. Gd-DTPA. 15 months later the patient complained

of pain in the right hip. d and e Bone marrow signal alterations (SI increase on STIR image, SI decrease on T1-w image) in the right proximal femur with edema in the adjacent soft tissues, f focal contrast uptake in the cortex and in the adjacent muscles. No frank bone destruction or discrete soft tissue mass is found

the disease progresses, erosions of the subarticular bone may be found. On MRI, intra-articular fluid collections can be identified with T2-weighted sequences. This is especially important in joints that are difficult to assess by physical exam and ultrasound. In T1-weighted images, joint effusion presents with low signal. On T2-weighted images, joint effusion exhibits high signal intensity, resulting in an “arthrographic effect” (Fig. 8.5.7). Intra-articular fluid, however, also can be identified in various other joint diseases, and the assessment of relax-

ation times does not allow for differentiation of pyarthrosis from other joint effusions. However, the destruction of cartilage due to bacterial arthritis is directly visible, especially in the presence of a joint effusion. In the initial phase of septic arthritis, cartilage edema may be identified on T2-weighted images. In later stages of the disease, a SI reduction is observed, which cannot be differentiated from osteoarthritis. Periarticular edema may indicate that a joint effusion is due to septic arthritis. An irregular SI reduction in the subarticular bone marrow in combination with a joint effusion is also an important clue to

1099

1100

8  Musculoskeletal System

Fig. 8.5.7a–e  Septic arthritis. Hematogenous fungal infection with geotricum capitatum in a 20-year-old patient with AML while undergoing chemotherapy. a T1-weighted image: low-intensity foci of hematopoietic bone marrow due to bone

marrow conversion. b,c T2-weighted images with fat suppression: extensive joint effusion. d,e T1-weighted SE images with fat suppression and i.v. gadolinium: extensive synovial contrast enhancement

the presence of septic arthritis. When septic arthritis is suspected, fluid from the joint should be aspirated for microbiological analysis.

Radiographic signs of spondylitis are not visible for days or weeks after onset of symptoms. Narrowing of the disk space is caused by destruction of the disk and portions of the adjacent vertebral body. CT allows for assessment of bone destruction and paravertebral fluid collections representing abscesses. MRI currently is the method of choice in the diagnosis of spondylitis (Tables 8.5.3, 8.5.4). In T1-weighted sequences, the signal intensity of the affected parts of the vertebrae and the disk is decreased, and well-defined vertebral endplates cannot be identified. In contrast to erosive osteochondritis, the signal alterations within the vertebral bodies extend beyond half of the height of the affected vertebral body. In T2-weighted sequences, signal intensity in the vertebral body and disk is increased. The nucleus pulposus loses its normal bandlike/biconvex shape and has an irregular configuration with very bright signal. The “intranuclear cleft” can no longer be identified. However, only approximately 50%

8.5.1.3  Spondylitis, Spondylodiskitis Spondylitis usually starts from hematogenous inoculation of bacteria (or other microorganisms) within one vertebra adjacent to the disk space, with subsequent spread to the disk and the adjacent vertebral body. In infants, the disk is vascularized, so that direct hematogenous infection of the disk is possible. After disk surgery or other spinal interventions, a primary infection of the disk also is possible. However, it should be noted that postoperative signal changes in the disk and adjacent vertebral bodies may have an appearance similar to that of spondylodiskitis.

8.5  Inflammatory Diseases of Bone and Joints Table 8.5.3  MRI in spondylodiskitis Sequence

Findings in spondylodiskitis

T1-weighted

Disk and adjacent vertebrae hypointense Poorly/not differentiable

T2-weighted/ STIR

Irregular increase of SI in the disk Intranuclear cleft not visible Hyperintense paravertebral abscess

T1-weighted CE/ fat-suppressed

Enhancement of vertebral body and disk Rim enhancement of paravertebral and epidural abscess

show increased signal intensity of the vertebral bodies on T2-weighted images. Therefore, the absence of increased signal on T2-weighted images should not result in misinterpretation if other signs of spondylitis are present. On contrast-enhanced T1-weighted fat-suppressed images, enhancement of the disk and the affected vertebrae is demonstrated. On non-enhanced T1-weighted images a reduction of signal intensity may be difficult to detect if the marrow has low signal, as may be present due to a high portion of hematopoietic cells (as found in children) or due to increased hematopoiesis or bone marrow infiltration. In spondylodiskitis, involvement of two adjacent vertebral bodies and the intervening disk is a typical constellation, whereas the posterior vertebral elements are usually not affected. Atypical manifestations of spondylitis, such as isolated involvement of one or of more than two

vertebrae with unremarkable disks, or the involvement of posterior elements, are more frequently found in tuberculosis and brucellosis. After 5–20 weeks of adequate antibiotic treatment, small T1-hyperintense spots in the periphery of the affected vertebrae are indicative of multifocal conversion to fatty marrow. Simultaneously, signal intensity on T2weighted and contrast-enhanced images decreases, with complete normalization occurring in 21–44 weeks. This is paralleled by a decrease of signal intensity in the disk space on T2-weighted images, until ultimately fibrosis with low signal intensity is found. It is most important to differentiate erosive osteochondritis from spondylodiskitis. On radiographs, these two diseases may present with similar findings. In spondylodiskitis, the signal increase in STIR images often reaches the opposed endplate. In erosive osteochondrosis, the bone marrow edema usually does not extend beyond the middle of the vertebral body. In both entities, the endplates have irregular margins. In the case of erosive osteochondritis, the endplates are continuous and clearly defined. In spondylodiskitis, on the other hand, the endplates are not identified as continuous bands. In acute spondylitis, the disk space has a high signal equivalent to that of fluid, while in erosive osteoarthritis the signal intensity is not as high. MRI is quite sensitive for detection of epidural abscesses. Epidural abscesses are typically isointense to the cord on T1-weighted and hyperintense on T2-weighted images. With MRI, impingement on the spinal cord and nerve roots can be precisely assessed. Strong contrast enhancement in the periphery is an important sign of epidural abscesses. On coronal and sagittal images the longitudinal extension of epidural abscesses is clearly identified (Figs. 8.5.8, 8.5.9).

Table 8.5.4  Differential diagnosis of spondylodiskitis Vertebral body

Disk

Vertebral body

Disk

T1-weighted

T1-weighted

T2-weighted

T2-weighted

Active spondylitis

2 adjacent vertebrae hypointense

Not definable

2 adjacent vertebrae hyper- or isointense

SI irregularly increased, no intranuclear cleft

Healed spondylitis /spondylodiskitis

2 adjacent vertebrae hypointense with multifocal fat deposition

Hypointense and reduced in height

2 adjacent vertebrae Normal or hypointense

Hypointense, no intranuclear cleft

Osteochondrosis

Normal or hyperintense adjacent to the disk

Reduced in height, hypointense with enhancement

Normal

Hypointense, no intranuclear cleft

Tumor

1 vertebra, centrally or excentrically hypointense

Normal

1 vertebra, centrally or excentrically hyperintense

Normal, intranuclear cleft preserved

1101

1102

8  Musculoskeletal System

Fig. 8.5.8a–d  Spondylodiskitis in a 60-year-old patient. a STIR image shows bone marrow edema and bright signal of the disk. b T1-weighted image: low signal intensity of the disk space and the affected portion of the involved vertebral bodies.

c T2-weighted image: bright signal of the disk is similar to fluid. d T1-weighted image after gadolinium shows contrast enhancement in the bone marrow and disk. Note the irregular endplates of the vertebrae

Fig. 8.5.9a–d  Spondylodiskitis with epidural abscess in an 83-year-old patient. Epidural abscess impinging on the cauda equina. a STIR image shows extensive bone marrow edema and bright signal in the disk space. b T1-weighted image: low signal of the involved vertebral bodies and disk space. c Bright signal

of the disk space on the T2-weighted image equivalent to fluid. d T1-weighted SE image after i.v. gadolinium shows contrast enhancement within the involved vertebral bodies and the epidural abscess

8.5.1.4  Soft-Tissue Infections

on T2-weighted images. In necrotizing fasciitis, signal increase is also visible in the deep fascias and muscles. T1-weighted images do not provide adequate contrast between the soft tissue infection and the surrounding muscles since both have low signal intensity. On T2weighted images, however, fluids have a very bright signal. T2-weighted fat-suppressed and contrast-enhanced T1-weighted fat-suppressed images are most useful for imaging of soft tissue infection (Figs. 8.5.10, 8.5.11).

Due to limited contrast resolution, CT does not allow for adequate assessment of soft tissue infection. However, for therapy planning the question of bone involvement and differentiation of superficial versus deep phlegmonous and abscess-like infections is crucial. Rahmoumi et al. (1994) demonstrated that MRI allows reliable differentiation of superficial infections (erysipelas) from necrotizing, deep, soft-tissue infections. This differentiation is especially important since an antibiotic therapy may be sufficient and successful in patients with superficial infections whereas in necrotizing fasciitis immediate surgical resection of necrotic tissue is mandatory. In erysipelas, thickening and signal increase of the skin with streaky extension into the subcutaneous fat is seen

8.5.1.5  Rheumatoid Diseases Rheumatoid diseases include a variety of conditions with different clinical and radiological characteristics. Proliferation of the synovial membrane is a common feature

8.5  Inflammatory Diseases of Bone and Joints

Fig. 8.5.10a–c  Soft tissue abscess in the anterior portion of the right deltoid muscle in a 71-year-old patient. a Axial T2weighted image. Fluid collection in the anterior portion of the deltoid muscle. b T1-weighted axial image. The fluid collection

is isointense with the surrounding muscle tissue. c Contrast-enhanced T1-weighted fat-suppressed axial image shows peripheral contrast enhancement of abscess

Fig. 8.5.11a–d  Soft tissue and bone infection after surgery. a The sagittal T1-w SE image shows a hypointense lesion in the distal portion of the patellar tendon and extending into Hoffa‘s fat pad and the tibial tuberosity. b In the STIR image high signal intensity is found. c Contrast-enhanced T1-weighted and d contrast-enhanced fat-suppressed T1-weighted images reveal a hypointense abscess with rim enhancement

of all. Clinical, serological, and immunological examinations play an important role in diagnosis and classification. Ultrasound and radiographic exams play a dominant role in imaging of rheumatoid diseases. However, MRI may contribute substantially in the early diagnosis

since unambiguous radiographic findings are frequently absent in the early stages of these disease entities. New treatment options, such as tumor necrosis factor (TNF)alpha inhibitors, require sensitive assessment of response and MRI might be useful for such follow-up.

1103

1104

8  Musculoskeletal System

Irreversible damage of the joints may take place in the early phase of the disease. Therefore, diagnosis needs to be established at an early stage so that effective therapy can be instituted. Sugimoto et al. (1996) showed that contrast-enhanced T1-w fat-suppressed imaging of the hands and wrists in patients without any findings in radiographs is very sensitive and specific for the diagnosis of early rheumatoid arthritis (RA), and has a higher sensitivity compared to clinical exams according to the revised criteria of the American Rheumatism Association of 1987. These criteria include (1) morning stiffness;

(2) soft-tissue swelling; (3) swelling of the proximal interphalangeal, metacarpophalangeal, or wrist joints; (4) symmetric swelling; (5) rheumatoid nodules; (6) presence of rheumatoid factor; and (7) radiographic erosions and/or periarticular osteopenia. Clinically, RA is defined by the presence of four or more criteria. MRI allows monitoring of therapeutic effects, assessment of surgical indications, and planning of synovectomies. The earliest recognizable pathology in rheumatoid arthritis is acute synovitis. Synovial proliferations present as villous, nodular, laminar, or even solid structures. Synovial proliferations can be directly visualized by MRI (Fig. 8.5.12). On T1-weighted images, they exhibit

Fig. 8.5.12a–d  Rheumatoid arthritis of the knee in a 35-yearold patient. a Coronal PD-weighted fat-suppressed image and b coronal T1-weighted image demonstrating destruction of the lateral condyle of the tibia. b Sagittal PD-weighted fat-suppressed image. Extensive joint effusion and solid synovial pro-

liferations in the posterior portion of the joint cavity, which exhibits an intermediate, inhomogeneous signal intensity. c Contrast-enhanced coronal T1-weighted and d contrast enhanced axial T1-w fat-suppressed images show substantial synovial enhancement

8.5.1.5.1  Rheumatoid Arthritis

8.5  Inflammatory Diseases of Bone and Joints

low signal intensity, which makes differentiation from joint effusion difficult. On T2-weighted images, they manifest with variable signal intensities; low signal intensity is due to fibrous, collagenous connective tissue or hemosiderin. Foci of high signal intensity are due to inflammatory tissue with edema. After intravenous administration of contrast agents, a strong signal increase is found within the synovial proliferations, reaching its maximum after 50 s and continuing until approximately 200 s (Adam 1991; Reiser 1988). On contrast-enhanced T1-weighted fat-suppressed images, synovial proliferations are clearly highlighted. Joint effusion experiences a delayed enhancement, so that high contrast of joint effusion and synovial proliferations is obtained. Approximately 6 min after i.v. contrast administration, signal enhancement is found in the joint effusions, starting in the periphery and gradually progressing to the center. After approximately 24 min, a complete filling of the effusion is visualized (Yamamoto 1993). MRI also allows differentiation between active and inactive pannus, since active hypervascularized pannus shows a significantly higher contrast uptake than inactive fibrous pannus. MRI allows for visualization of synovial proliferations in tendon sheaths. Inflammatory lesions and ruptures of tendons and ligaments are also clearly visualized. When using sequences adequate for cartilage imaging (e.g., 3D FLASH, DESS, trueFISP) structural damage and thinning of cartilage can be assessed. On radiographs, joint-related osteopenia is detected as a typical sign of rheumatoid arthritis; this cannot be seen on conventional MRI. Cysts

and erosions, on the other hand, are detected with higher sensitivity with MRI. In addition to the small joints of the hand and foot, the wrist, the knee, the elbow, the glenohumeral and acromioclavicular joints, the cervical spine, and the atlanto-occipital junction are frequently involved in rheumatoid arthritis. Synovial proliferations, pseudobasilar impression, and atlantodental subluxation may result in neurological complications; all can be clearly delineated in MRI. Anterior atlantodental, vertical (also known as atlantoaxial impaction), lateral and posterior subluxations are found in 40% to 85% of patients with RA (Fig. 8.5.13). With sagittal MRI, the foramen magnum and the distance to the medulla oblongata and the superior cervical cord can be readily appreciated. On T2- and PD-weighted images, myelomalacia due to pressure on the cord and medulla oblongata, respectively, manifests with signal increase. 8.5.1.5.2  Seronegative Spondyloarthropathies The three major types of seronegative spondyloarthropathy include ankylosing spondylitis, psoriatic arthritis, and Reiter’s disease. Like in RA, there is involvement of synovial joints, but with a different morphology and distribution of articular lesions. The key finding in spondyloarthropathies is sacroiliitis. Radiographs and CT are frequently negative in the early phase of the disease. As

Fig. 8.5.13a,b  Rheumatoid arthritis of the atlantodental joint. a Sagittal T1-weighted image shows erosion of the odontoid process with posterior subluxation relative to the anterior arch of the atlas. The odontoid process is surrounded by hypointense pannus

1105

1106

8  Musculoskeletal System Fig. 8.5.14a,b  Sacroiliac joints of two patients with psoriatic arthritis. a STIR image shows bone marrow edema in the right sacrum, indicating sacroiliitis. b Coronal T1-weighted images demonstrates ankylosis of both SI joints in advanced psoriatic arthritis

sacroiliitis progresses, erosions, subchondral sclerosis, transarticular bone bridges, and ankylosis are found. MRI allows detection of early stages of sacroiliitis and assessment of disease activity (Fig. 8.5.14). On STIR images, bone marrow edema in the subchondral zone of the iliac and sacral bones is found in active inflammation, and is associated with increased contrast enhancement. Active erosions are hyperintense on STIR and T2-weighted images and exhibit pronounced contrast enhancement. Histologically they represent invasive-destructive pannus at the cartilage-bone junction. On T1-weighted images, band-like hypointense signal is found in both active inflammation and sclerosis. Moreover, increased contrast enhancement is also found in periarticular bone marrow as well as in the anterior and posterior capsules of the sacroiliac joint. Another common finding in spondyloarthropathies is enthesitis, such as in the pelvis, the femoral trochanter, the iliac crest, the ischial tuberosity, the patella and the calcaneus. Bone marrow edema and contrast enhancement as well as inflammation of tendons, ligaments, and muscles may be found at these sites.

4.

5.

6. 7. 8.

9.

10.

11.

References 1. 2.

Rahmouni A, Chosidow O, Mathieu D et al (1994) MR imaging of acute infectious cellulitis. Radiology 192:493–496 Sugimoto H, Takeda A, Masuyama J, Furuse M (1996) Early-stage rheumatoid arthritis: diagnostic accuracy of MR imaging. Radiology 198:185–192

Suggested Reading 1.

2.

3.

Bennett DL, Ohashi K, El-Khoury GY (2004) Spondyloarthropathies: ankylosing spondylitis and psoriatic arthritis. Radiol Clin North Am 42:121–134 Buhne KH, Bohndorf K (2004) Imaging of posttraumatic osteomyelitis. Semin Musculoskelet Radiol 8:199–204. Review Chatha DS, Cunningham PM, Schweitzer ME (2005) MR imaging of the diabetic foot: diagnostic challenges. Radiol Clin North Am 43:747–759, ix. Review

12. 13.

Collins MS, Schaar MM, Wenger DE, Mandrekar JN (2005) T1-weighted MRI characteristics of pedal osteomyelitis. AJR Am J Roentgenol 185:386–393 Hermann KG, Bollow M (2004) Magnetic resonance imaging of the axial skeleton in rheumatoid disease. Best Pract Res Clin Rheumatol 18:881–907. Review Jevtic V (2004) Vertebral infection. Eur Radiol 14(Suppl 3): E43–52. Review Jurik AG (2004) Chronic recurrent multifocal osteomyelitis. Semin Musculoskelet Radiol 8:243–253. Review Ledermann HP, Schweitzer ME, Morrison WB, Carrino JA (2003) MR imaging findings in spinal infections: rules or myths? Radiology 228:506–514 Levine DS, Forbat SM, Saifuddin A (2004) MRI of the axial skeletal manifestations of ankylosing spondylitis. Clin Radiol 59:400–413. Review Puhakka KB, Jurik AG, Schiottz-Christensen B, Hansen GV, Egund N, Christiansen JV, Stengaard-Pedersen K (2004 Magnetic resonance imaging of sacroiliitis in early seronegative spondylarthropathy. Abnormalities correlated to clinical and laboratory findings. Rheumatology (Oxford) 43:234–237 Santiago Restrepo C, Gimenez CR, McCarthy K (2003) Imaging of osteomyelitis and musculoskeletal soft tissue infections: current concepts. Rheum Dis Clin North Am 29:89–109. Review Shih HN, Shih LY, Wong YC (2005) Diagnosis and treatment of subacute osteomyelitis. J Trauma 58:83–87 Tali ET (2004) Spinal infections. Eur J Radiol 50:120–33. Review

8.6  Avascular Necrosis The term avascular necrosis comprises a variety of disease entities and developmental variants. They may be localized in virtually any bone and may differ considerably in etiology and appear in patients of different age distributions. Ischemic episodes within the bone marrow are considered the common pathogenetic pathway of many of these entities. In radiography, usually there is increased bone density at some stage of the disease. With the interruption of bone perfusion there will be necrosis of the cellular components of the bone marrow

8.6  Avascular Necrosis

depending on the duration of hypoxia. Apoptosis occurs within hours in the hematopoeitic cell lines, and with osteocytes, osteoclasts, and osteoblasts, fat-cell apoptosis will occur after several days. In addition, there can be edema and venous congestion within the bone marrow. The signal intensity alterations of the bone marrow are the hallmark of the MRI-based diagnosis of avascular necrosis. Typically, there is (homogeneous or patchy) hypointensity of the fatty marrow (normally exhibiting high signal intensity) in T1-weighted sequences and high signal intensity in STIR/T2-weighted FS sequences. However, the time point of MRI-based diagnosis still is dependent on the one hand on the amount of what is believed to represent edematous change, on the amount of fat cell necrosis and on the degree of fatty marrow replacement by hemorrhage, fibrosis, or sclerosis. 8.6.1  Avascular Necrosis of the Hip and Transient Bone Marrow Edema Syndrome 8.6.1.1  Avascular Necrosis of the Hip The femoral head is a frequent site of AVN because of its large cartilaginous surface precluding vascular supply. Instead, the latter is achieved via the nutritional branches of the circumflex femoral artery. In the infant, there is additional blood supply by a small artery accompanying the ligamentum teres (capitis femoris), which is obliterated with increasing age. Direct disruption of the nutrient artery is evident in posttraumatic AVN of the hip. Nontraumatic AVN is caused by embolic disease (fat, gas) or elevated intramedullary pressure. The incidence of AVN

is increased in dysbaric conditions, with steroid therapy, Cushing’s syndrome, pancreatitis, alcoholism, Gaucher’s disease, and sickle cell anemia. Men are four times more frequently affected than women are. Patients usually present with groin pain of gradual onset present at rest and increased by weight bearing. Prognosis is related to the degree of MRI morphologic change, location, and local extension of the abnormalities (Beltran et al. 1990). If less than 25% of the weight bearing surface of the femoral head is affected the risk of infraction is very low. Currently the ARCO (Association Recherche Circulation Osseuse) classification system is widely used to describe the imaging findings used to stage the disease (Imhoff 1997) in correlation with histologic findings. In extension of the traditional X-ray based stages according to Ficat and Arlet (1968) five stages of imaging morphology are described comparing radiographic, CT, bone scan, and MRI appearances (Table 8.6.1) in conjunction with a grading of extent and location of the abnormalities. Initially no imaging findings reveal the (small) histologically defined foci of osteonecrosis (stage 0). Second, edemalike signal alterations are shown on MRI without a clear linear demarcation of the avascular area (stage I). Usually, these findings are more conspicuous on T2-weighted FS/STIR sequences than on T1-weighted sequences. The appearance of a reactive peripheral zone demarcating the lesion usually indicates transition from a reversible state into irreversible disease (stage II). However, when there is a small extension this kind of lesion may well respond to therapy (usually core decompression, drilling) and come to a stable situation. Often and visibly from this stage of disease onward, the virtually pathognomonic “double-line” sign is present in 65–80% of cases—

Table 8.6.1  ARCO (Association Recherche Circulation Osseuse) classification of AVN of the hip  Stage

0 Initial

I Reversible early

II Irreversible early

III Transient

IV Late

Radiogram

Normal

Normal

Sclerotic rim/foci

Crescent sign, flattening

Collapse, secondary changes

CT

Normal

Normal

Asterisks sclerotic rim

Subchondral Fracture

Collapse, secondary changes

Scintigraphy

Normal

Diffuse enhancement or cold spot

Cold in hot spot

Hot in hot spot

Hot spot

MRI

Edema (?)

Edema (?)

Necrotic area and reactive border

Crescent sign, flattening

Collapse, secondary changes

Localization

A medial, B central, C lateral

Size

A < 15%, B = 15–30%, C > 30%   (of femoral head circumference)

A < 2 mm, B = 2–4 mm, C > 4 mm

1107

1108

8  Musculoskeletal System

a high signal line toward the demarcated lesion and an adjacent low-signal line toward the surrounding bone on T2-weighted sequences. The former reflects vascularized granulation tissue, the latter sclerotic demarcation. On T1-weighted sequences, differentiation between vascularized and sclerotic portions of the demarcating zone is difficult, as both lines usually present with hypointense signal intensity. Intravenous application of gadolinium shows vivid uptake in the inner rim. Flattening of the femoral head and linear low signal intensity immediately underneath the subchondral bone plate (“crescent” sign) indicate (imminent or beginning) subchondral fracture (stage III). Collapse and findings of degenerative change define stage IV. Mitchell et al. (1987) correlated histopathological findings with four (A–D) MRI signal patterns in T1-weighted and T2-weighted sequences based on the (chronological) order, in which they appeared and reflecting the type of tissue predominating in the lesion

(Table 8.6.2). Signal patterns consistent with fat, hemorrhage, edema, and granulation tissue (Mitchell A, B) indicated better outcomes than patterns compatible with sclerosis, fibrosis, and fluid (Mitchell C, D). Most often, the AVN is located in the anterosuperior segments of the femoral head. Not rarely there is effusion accompanying already early stages of the disease. In up to 70% of cases, there is metachronous affection of the contralateral hip. Therefore, imaging should contain at least one (coronal) sequence covering both hips. Although depiction of subtle subchondral fractures is easier with CT, MRI demonstrates the extent and location of the lesions to much better advantage. Visualization of the necrotic area in the sagittal plane (Fig. 8.6.1) helps to plan therapy and may contribute to the differential indication of different treatment options. The sensitivity for the diagnosis of AVN is higher with MRI (Mitchell et al. 1986, 1987; van de Berg et al. 1992)

Fig. 8.6.1a–d  Bilateral AVN of the hip. Radiographically, only left-sided necrosis was notable. a Coronal T1-weighted SE image shows subchondral SI reduction (arrow) in the upper portion of the left femoral head delineated by a low-SI rim. Central to this rim, within the necrotic area, a band-shaped signal increase can be seen. There is flattening of the femoral head. On the right side, only a small, oval-shaped area is demarcated by a low-SI line, the demarcated area showing normal signal of fatty mar-

row. b Coronal T2-weighted SE image with fat saturation. The hypointense necrotic area is delineated by a band-shaped hyperintensity. There are extensive joint effusion and bone marrow edema in the femoral head and neck. c Coronal T1-weighted SE image after Gd application, with the typical “double-line sign.” d Axial T1-weighted image after Gd application with fat saturation demonstrates (reactive) synovial enhancement

8.6  Avascular Necrosis

than with any other imaging modality including radio­gra­ phy, CT, and scintigraphy (Table 8.6.3). Measurements of intramedullary pressure exhibit higher sensitivity but lower specificity than MRI. Moreover, MRI has the advantage of providing additional morphologic information, e.g., on joint effusion and cartilage. It is not surprising that compared to histopathological findings MRI may

fail to demonstrate small foci of very early AVN as it may take up to 5 days for fat cell necrosis (and thus relevant signal alteration) to occur. Nonetheless, MRI is to be considered the imaging modality of choice for early diagnosis of AVN as well as for assessing possible differential diagnoses, especially in the setting of clinical complaints and negative radiographs (Fig. 8.6.2).

Table 8.6.2  Signal intensities in avascular necrosis of the hip (central area within low SI demarcation (according to Jergesen et al. 1985)

Table 8.6.3  Detection of avascular necrosis of the hip (according to Beltran et al. 1986)

Type

Correlate

T1-weighted

T2-weighted

A

Fat

+

Intermediate

B

Blood

+

+

C

Fluid

–

+

D

Fibrosis

–

–

MRI

Bone scintigraphy

Bone marrow pressure measurement

Sensitivity (%)

89

78

92

Specificity (%)

100

75

57

Accuracy (%)

94

76

80

+ hyperintense, – hypointense Fig. 8.6.2a–d  72-year-old female patient with bilateral AVN of the hip. a Coronal T1-weighted SE image. In the upper (subchondral) portion of the right femoral head, there is demarcation of a hypointense zone with irregular margins (arrow). In the left femoral head a small annular subchondral SI reduction can be seen surrounding normal appearing marrow signal. b Sagittal T1-weighted SE image of the right femoral head. The necrotic area with low SI is found in the anterosuperior portion of the femoral head. c Coronal T1-weighted SE image after contrast media application. Distinct enhancement of the necrotic area as a sign of initial revascularization. d Sagittal T2-weighted SE image of the right femoral head. Small hyperintense cystic lesion within the necrotic area and in the subchondral acetabulum

1109

1110

8  Musculoskeletal System

8.6.1.2  Bone Marrow Edema Syndrome Bone marrow edema syndrome (BMES) is considered a self-limiting disorder responsive to conservative management with resolution of symptoms and imaging findings within approximately 9 to 12 months from onset (Watson 2004). MRI reveals findings consistent with bone marrow edema (BMEP) in the femoral head and neck with frequent extension into the intertrochanteric area. In contrast to AVN, there is no demarcated area of necrosis and the epicenter of the BMEP is rather in the femoral neck than in the femoral head. Concomitant joint effusion is a frequent finding. Due to the presence of BMEP in patients showing transient osteoporosis of the hip radiographically, these MRI findings have been termed transient bone marrow edema syndrome. Sometimes there is a migratory pattern of the BMEP involving hip, knee, and ankle at different time points. It has been suggested that the entity is a non-traumatic form of reflex sympathetic dystrophy or that it is due to a transient ischemic insult. It is unclear, if not unlikely, that BMES is a precursor of AVN (Bloem et al. 1988; Hayes et al. 1993; Neuhold et al. 1992, 1993; van de Berg et al. 1993). Although first described in pregnant women, most commonly middle-aged men have been found affected (Figs. 8.6.3, 8.6.4). Patients present with rapidly increasing pain (during one month), which reaches a plateau at about 2 months, with subsequent remission of symptoms spontaneously and/or with protected weight bearing. Most often, they lack risk factors for AVN. Diagnosis should be carefully established by exclusion of other potential entities such as AVN, stress fracture, and infection, all of them showing some overlap of imaging findings. Clinical history and assessment of risk factors can be helpful. 8.6.1.3  Subchondral Insufficiency Fracture of the Femoral Head Subchondral insufficiency fractures appear as linear, band-like, low-signal-intensity abnormalities immediately underneath the subchondral bone plate, accompanied by BMEP (Yamamoto and Bullough 2000a). Mostly

overweight female patients older than 65 years and renal-transplant recipients are affected. In contrast to subchondral fractures in AVN, histologically there are only small foci of necrosis, and they are located almost exclusively on the cortical side of the fracture line. Similar to the femoral head, subchondral insufficiency fractures in comparable patient groups have been described in the femoral condyles. Reports from Zanetti et al. (2003) and Sokoloff et al. (2001), support the etiology to be one of insufficiency fractures in a high percentage of patients but suggest that in some patients overuse may be the underlying cause of the fracture. 8.6.2  Perthes Disease and Coxitis Fugax 8.6.2.1  Perthes Disease Perthes disease is a disease of childhood with a peak incidence from ages 4 to 8 years (Fig. 8.6.5). Children affected limp and complain about pain in the groin, thigh, or knee. Contractions may develop. In about 15% of cases, metachronous bilateral affection can be seen. Although ischemia may well be the cause of Perthes disease, the etiology to date is not clear. Without timely treatment the disease will follow several distinct stages (see Table 8.6.4), potentially leading to deformity, broadening and flattening of the femoral head and of the proximal femur, functional impairment, and premature osteoarthritis. Therefore, early diagnosis and treatment is mandatory. In the healthy child, the femoral epiphysis exhibits the high signal intensity of fatty marrow in T1-weighted sequences. In the early stages of Perthes disease, there are signal inhomogeneities in the periphery of the femoral head with foci of globular or linear signal loss (Figs. 8.6.5, 8.6.6, 8.6.7) in the T1-weighted sequences. In T2-weighted sequences, these alterations usually are less conspicuous. Often there is joint effusion and a lateral position of the femoral head (Dillon et al. 1990; Heuck et al. 1988) with reduced size of the ossification centers compared to the contralateral side. Normal bone marrow extending from the metaphysis into the epiphysis is called a metaphyseal spur and considered to indicate regeneration. Depending on the outcome, there is reconversion to fatty marrow in Fig. 8.6.3a–c  Transient bone marrow syndrome in a woman after delivery. Coronal STIR-images a postpartum, b 2 months later, c 9 months later. The excessive bone marrow edema in the right femoral neck with sparing of the apophyses has significantly reduced after three months and has resolved completely after 9 months

8.6  Avascular Necrosis

Fig. 8.6.4a,b  Transient bone marrow edema syndrome. a Coronal STIR-image, b axial T1-weighted fat saturated image after i.v. Gd injection. Enhancement is noted in the femoral neck and in some portions of the femoral head

Fig. 8.6.5  Early stage of Perthes disease of the left hip. Coronal T1-weighted image with patchy SI reduction of the left femoral epiphysis (arrows). There is moderate excentricity of the femoral head due to cartilage thickening

Fig. 8.6.6a–c  Perthes disease of the left hip. a Radiograph of the pelvis. The femoral epiphysis is slightly condensed and flattened. There is minimal eccentricity of the left femoral head. b Coronal T1-weighted image with massive, slightly inhomogeneous SI reduction of the left femoral epiphysis. c STIR image. Inhomogeneous SI reduction of the femoral epiphysis with large effusion within the medial aspects of the hip joint. Edema of the lateral aspect of the femoral neck

1111

1112

8  Musculoskeletal System Table 8.6.4  Stages in Perthes disease (radiographic changes) Stage

Changes

I Initial stage

Soft-tissue swelling Enlarged joint space Delayed ossification Subchondral radiolucency

II Stage of fragmentation

Diffuse condensation Fragmentation Flattening Enlargement Broadening

III Stage of regeneration

Rebuilding of the femoral head

IV Late stage

Coxa vara greater trochanter raises epiphyseal deformity Femoral neck shortened and broadened

Fig. 8.6.7a–d  Perthes disease of the left hip. a Coronal T1weighted SE image. There is irregular signal intensity reduction, flattening and fragmentation of the femoral epiphysis and broadening and shortening of the femoral metaphysis. b,c GRE-

the epiphysis presumably due to revitalization (Fig. 8.6.7). Already early in the course of the disease a thickening of the cartilage can be found (Jamamillo et al. 1995). Since this occurs predominantly on the medial aspect of the femoral head (and since there may be additional synovitis), it results in loss of containment of the femoral head (Rush et al. 1988). Joint effusion is evident in T2-weighted sequences and usually is most prominent in the inferomedial aspect of the joint. Band-like low signal intensity abnormalities, i.e., filling defects inferomedially correspond to edematous and fibrous swelling of the joint capsule. The articular cartilage may show increased signal intensity on the T2-weighted sequences. As MRI has been described to be more sensitive for the diagnosis of Perthes disease than are radiography and scintigraphy, it is recommended when there are clinical complaints but unrevealing radiographs. In this connection, exclusion of Perthes disease is valuable also since it helps to avoid unnecessary followup examinations.

images, coronal and sagittal. Hyperintense signal of the hyaline cartilage. Lateralization of the left femoral epiphysis. d Coronal STIR image. There is joint effusion and extensive edema of the proximal femur

8.6  Avascular Necrosis

8.6.2.2  Coxitis Fugax In T1-weighted images, often there is but faint signal loss within the epiphysis. Signal increase may be more prominent in STIR or T2-weghted FS sequences. Because bacterial and tuberculous coxitis may present similarly to coxitis fugax in MRI, joint-fluid aspiration and microbiological workup should be performed if clinical findings do not allow a final diagnosis to be established. Local disturbances of ossification are characterized and are to be differentiated by the presence of a circumscribed bony defect in a subchondral location with otherwise-normal signal intensity in the remainder of the femoral head. Radiographically they exhibit a radiolucent well-defined subchondral area surrounded by sclerosis. 8.6.3  Bone Infarction Necrosis in the medullary cavity of the long and flat bones is observed with increased incidence in Caisson’s disease, with steroid therapy, sickle cell anemia, and alcoholism. However, often no definite etiology can be found. In contrast to osteonecrosis where the necrotic area mostly is located in the subchondral regions of the epiphysis, in bone infarction, the abnormality is located predominantly in the metaphysis and diaphysis. The typical radiographic appearance consists in a central radiolucency with peripheral serpiginous calcifications. In mature, long-standing cases, there is increasing central calcification. In the cases of typical presentation, a clear diagnosis can be established and differentiation from enchondroma and central chondrosarcoma is possible (Fig. 8.6.8). In MRI, the periphery of the infarct shows low signal intensity in T1-weighted images. According to the stage of disease (mature versus fresh), the central portions may

Table 8.6.5  Stages of AVN of the lunate (radiographic changes according to Decoulx) Stage

Change

I

Condensation

II

Mosaic-like pattern of sclerosis   and condensation

III

Fragmentation, dorsalization   of posterior horn components

IV

Complete destruction, osteoarthritis

show normal fatty marrow signal or a more or less pronounced signal loss. This pattern is virtually mirrored (Fig. 8.6.8) in STIR/T2-weighted FS sequences. Perifocal edema can be pronounced and may occasionally extend into the adjacent soft tissues. 8.6.4  Kienboeck’s Disease AVN of the lunate bone is associated with repetitive overloading/microtrauma (Table 8.6.5). Normal variants such as ulna minus or ulna plus variants are predisposing factors. Patients’ complaints may precede radiologic findings (increased bone density, reduced height, and fragmentation; compare Table 8.6.5) by months and even years. MRI enables early diagnosis before radiographic findings are present (Trumble and Irving 1990). In these early stages, there is signal reduction in T1-weighted and signal increase in T2-weighted FS/STIR sequences. Gadolinium uptake may be prominent, especially in T1-weighted FS sequences. With progressing disease, there is reduction Fig. 8.6.8a,b  Bone marrow infarction of the proximal tibia. a Coronal T1-w-SE image. The medullary infarction shows an irregular serpiginous marginal zone of low SI with normal bone marrow signal in the centre of the lesion. b Coronal STIR-image showing a clearly hyperintense marginal zone of the bone infarct with centrally suppressed fat signal mirroring the findings in the T1-weighted sequence

1113

1114

8  Musculoskeletal System Fig. 8.6.9a,b  Necrosis of the scaphoid bone in a 25-year-old patient after scaphoid fracture. T1-weighted GRE sequence a preand b post-contrast. a There is edema within the distal fragment of the scaphoid bone and a homogenous SI reduction of large parts of the scaphoid with an area of intermediate SI in the distal scaphoid pole (arrow). b Strong enhancement of the distal scaphoid pole and its surrounding soft tissue is observed after contrast injection, indicating the persistence of perfusion there in contrast to the proximal aspect of the scaphoid

of Gadolinium uptake and findings indicate necrosis, sclerosis, and cyst formation. This time course suggests that the initial event is stress-related edema in the bone marrow followed by fibrosis and sclerosis possibly contributing to reduced perfusion and lastly leading the way to AVN (Trumble and Irving 1990). 8.6.5  Necrosis of the Scaphoid Bone Subsequent to fracture of the scaphoid bone AVN of one or both fragments may occur. The proximal fragment is at higher risk since the blood supply enters from the distal portion of the scaphoid. Increased bone density may become apparent only years after the traumatic event, followed by cyst formation and reduction in height. Six weeks after AVN due to scaphoid fracture increased signal intensity in the STIR/T2-weighted FS and reduced signal intensities in T1-weighted sequences (re)appear or persist in contrast to a tendency to normalization of signal intensity in the unaffected fragment. Lack of enhancement after i.v. gadolinium is in favor of lost fragment vitality as well. This is especially valid if contrast uptake can be demonstrated in the unaffected fragment. AVN often leads to pseudarthrosis, and then conservative treatment holds little promise. The rare spontaneous osteonecrosis of the scaphoid bone (Preiser’s disease) has imaging findings similar to those of AVN (Fig. 8.6.9). 8.6.6  Osteochondritis Dissecans and Spontaneous Osteonecrosis of the Knee 8.6.6.1  Osteochondritis Dissecans This disease entity is observed in various joints. Most often affected are the femoral condyles, the trochlea tali, and the distal humerus. The etiology is not yet clear, but repetitive microtraumata, focal ischemia, and disturbances of ossification are discussed. Typically, children, adolescents, and young adults are affected. The potential

of healing, and thus the likelihood of a favorable outcome, is greater in younger patients. The stability of the fragment and the status of the overlaying cartilage are essential factors in choosing therapy. Radiography shows the dissected element separated by a rim of sclerosis from the surrounding bone. In MRI, the signal of the fragment may be heterogeneous and may show several small lowsignal-intensity foci corresponding to bony elements. These findings are not related to instability. Demarcation towards the surrounding bone shows low signal intensity in T1-weighted sequences. If there is linear high signal intensity in the interface between fragment and surrounding bone in T2-weighted sequences it has to be determined whether this pattern corresponds to loosening of the fragment, i.e., fluid entering the interface, or whether this pattern represents granulation tissue, i.e., repair. Intravenous gadolinium may help in such cases, clearly demonstrating vascularized granulation tissue by post-injection enhancement indicating an ongoing repair process. Continuity of the high signal intensity from underneath the fragment through the bone and a full-thickness cartilage defect to the joint space is indicative of loosening (DeSmet 1996). Therefore, it is strongly suggested that the presence of a high-signal-intensity line longer than 5 mm underneath the fragment on T2weighted sequences be interpreted as a predictor of instability only when there is a concomitant breach through the cartilage. Direct MR arthrography may be helpful. Large cyst-like areas underneath the fragment are suggestive of loosening as well. 8.6.6.2  Spontaneous Osteonecrosis of the Knee Traditionally described as Ahlbäck’s disease, spontaneous osteonecrosis of the knee (SONK) is characterized by sudden onset of severe, often-disabling pain without a remembered causative event in elderly female patients. In most of the cases, the weight-bearing portions of the medial femoral condyle are affected. Radiography reveals well-defined subchondral radiolucency, subsequent flat-

8.6  Avascular Necrosis

Fig. 8.6.10a,b  SONK (Ahlbäck’s disease) in a 63-year-old male patient. a Coronal PD fat-saturated image, b coronal T1weighted SE image and c contrast-enhanced T1-weighted image. There is an osteochondral lesion in the weight-bearing aspect of the medial femoral condyle with a low-SI contour in the T1-weighted image. In the PD fat-saturated image there is a small band-shaped area with strong hyperintensity defining

the core of the lesion accompanied by a larger area showing but mild perifocal BMEP. After i.v. Gadolinium there is contrast enhancement of the band-shaped area, indicating (possibly reactive) vascularization. Please note the focal area of low signal intensity immediately adjacent to the subchondral bone possibly reflecting condensed trabeculae/sclerosis

Fig. 8.6.11a,b  Osteonecrosis of the left humeral head. a Coronal T1-weighted image demonstrating subchondral SI reduction (triangles). b In the T2-weighted images there is a slightly hyperintense signal within this area. Still, the necrotic zone can easily be identified. Please note the small joint effusion (arrows)

tening of the femoral condyle, increasing sclerosis, and associated degenerative change. Often there is an accompanying meniscal lesion. MRI contributes to early diagnosis and to defining the extent of the lesion (Bjorken-

gren et al. 1990). In T1-weighted sequences, there is a band or oval-shaped area of subchondral low signal intensity (Fig. 8.6.10). STIR/T2-weighted sequences show surrounding, often extensive, edema, sometimes even

1115

1116

8  Musculoskeletal System

involving the soft tissues. Recently, the spontaneous nature of the lesion has been questioned and there is evidence that the initial event in Ahlbäck’s disease is instead a stress fracture (Zanetti 2003; Yao 2004; Sokoloff 2001; Yamamoto and Bullough 2000). 8.6.7  Osteonecrosis and Osteochondritis in Other Locations Besides “primary” osteonecroses, there is a variety of diseases is whose etiology has been related to traumatic or ischemic events. Some of them may be due to variants/disturbances of ossification. From these, secondary osteonecroses associated with systemic lupus erythematosus (SLE), renal transplantation, sickle cell anemia, steroid intake, or Cushing’s disease have to be differentiated (Fig. 8.6.11). The common morphologic element in this heterogenous group is circumscribed sclerosis in a subchondral location in a certain stage. References 1.

Beltran J, Noto AM, Mosure JC, Weiss KL, Zuelzer W, Christoforidis AJ (1986) The knee: surface coil MR imaging at 1.5 T. Radiology 159:747–751 2. Beltran J, Knight CT, Zuelzer WA et al (1990) Core decompression for avascular necrosis of the femoral head: correlation between long-term results and preoperative MR staging. Radiology 175:533–539 3. Berg BE van de, Malghelm J, La Baisse MA et al (1992) Avascular necrosis of the hip: comparison of contrast-enhanced and non-enhanced MR imaging with histologic correlation. Radiology 182:445–450 4. Berg BE van de, Malghelm J, La Baisse MA et al (1993) MR imaging of avascular necrosis and transient marrow edema of the femoral head. Radiographics 13:501–520 5. Bjorkengren AG, Alrowaih A, Lindstrand A, Wing-strand H, Thorngren KG, Pettersson H (1990) Spontaneous osteonecrosis of the knee: value of MR imaging in determining prognosis. AJR 154:331–336 6. Bloem JL (1988) Transient osteoporosis of the hip: MR imaging. Radiology 167:753–755 7. De Smet AA, Tuite MJ, Norris MA, Swan JS (1994) MR diagnosis of meniscal tears: analysis of causes of errors. AJR 163:1419–1423 8. Dillon EH, Pope EF, Jokl P, Lynch K (1990) The clinical significance of stage 2 meniscal abnormal­ities on Magnetic Resonance knee images. Magn Reson Imaging 8:411–415 9. Ficat RP, Ariel J (1968) Diagnostic de 1‘osteonecrose femorocapitale primitive au stade I. Rev Chir Orthop 54:637 10. Hayes CS, Conway WF, Daniel WW (1993) MR imaging of bone marrow edema pattern: transient osteoporosis, transient bone marrow edema syn­drome, or osteonecrosis. Radiographics 13:1001–1011

11. Heuck A, Lehner K, Schittich I, Reiser M (1988) Die Wertigkeit der MR fur Diagnostik, Differenzial-diagnostik und Therapiekontrolle des Morbus Perthes. RoFo 148:189–194 12. Imhof H, Breitenseher M, Trattnig S, Kramer J, Hofmann S, Plenk H, Schneider W, Engel A (1997) Imaging of avascular necrosis of bone. Eur Radiol 7:180–186 13. Jamamillo D, Kasser JR, Villegas-Medina OL, Garry E (1995) Cartilaginous abnormalities and growth disturbance in Legg-Calve-Perthes disease: evalua­tion with MR imaging. Radiology 187:767 14. Jergesen HE, Heller M, Genant HK (1985) Magnetic resonance imaging of the femoral head. Orthop Clin North Am 16:705–716 15. Mitchell DG, Kundel HL, Steinberg ME (1986) Avascular necrosis of the hip: comparison of MR, CTand scintigraphy. AJR 147:67–71 16. Mitchell DG, Rao VM, Dalinka MK, Spritzer CE, Alavi A, Steinberg ME, Fallon M, Kressel HY (1987) Femoral head avascular necrosis: correlation of MR imaging, radiographic staging, radionuclide imaging, and clinical findings. Radiology 162:709–715 17. Neuhold A, Hofmann S, Engel A (1993) Knochenmarkodem-Fruhform der Hiiftkopfnekrose. RoFo 159:120–125 18. Neuhold A, Hofmann S, Engel A et al. (1992) Bone marrow edema of the hip: MR findings after core decompression. J Comp Assist Tomogr 16:951–955 19. Rush BH, Bramson RT, Odgen JA (1988) Legg-CalvePerthes disease: detection of cartilaginous and synovial changes with MR imaging. Radiol­ogy 167:473 20. Sokoloff RM, Farooki S, Resnick D. Spontaneous osteonecrosis of the knee associated with ipsilateral tibial plateau stress fracture: report of two patients and review of the literature. Skeletal Radiol 30:53–56 21. Trumble TE, Irving J (1990) Histologic and mag­netic resonance imaging correlations in Kienbock’s disease. J Hand Surg 15:879–884 22. Watson RM, Roach NA, Dalinka MK (2004) Avascular necrosis and bone marrow edema syndrome. Radiol Clin North Am 42:207–219 23. Yamamoto T, Bullough PG (2000a) Subchondral insufficiency fracture of the femoral head and medial femoral condyle. Skeletal Radiol 29:40–44 24. Yamamoto T, Bullough PG (2000b) Spontaneous osteonecrosis of the knee: the result of subchondral insufficiency fracture. J Bone Joint Surg 82:858–866 25. Yao L, Stanczak J, Boutin RD (2004) Presumptive subarticular stress reactions of the knee: MRI detection and association with meniscal tear patterns. Skeletal Radiol 33:260–264 26. Zanetti M, Romero J, Dambacher MA, Hodler J (2003) Osteonecrosis diagnosed on MR images of the knee. Relationship to reduced bone mineral density determined by high resolution peripheral quantitative CT. Acta Radiol 44:525–531

8.7  Imaging of Internal Joint Derangement

8.7  Imaging of Internal Joint Derangement Imaging of internal derangements of the joints is among the leading indications of MRI, whose high soft-tissue contrast and multiplanar imaging capabilities enable depiction of the internal structures of the joints with the highest precision. In addition to technical improvements, a much more profound understanding of joint pathology and considerably increased experience with MRI of the joints facilitate the precise diagnosis of internal joint derangements. Currently, we recommend imaging all joints in all three major orthogonal planes using moderately T2weighted fat-suppressed TSE sequences as the basis of the MRI protocol, facilitating detection of most pathological entities. These can be substituted by STIR sequences in cases of problematic field homogeneity. A T1-weighted SE sequence would facilitate analysis of anatomy and bone marrow and thus complete a basic protocol. Depending on the clinical question, (in)direct MR-arthrography for the detection of capsule-labroligamentous lesions, i.v. contrast medium application for inflammation, tumors, and infection, and GRE sequences for the analysis of cartilage and the depiction of hemosiderin depositions may be applied, extending the basic protocol. Slice angulation is to be adapted to locoregional (joint) anatomy. 8.7.1  Imaging of Normal Joint Structures Cancellous bone, ligaments, and tendons exhibit low signal intensity independent of the pulse sequences applied. The same is true for fiber cartilage (menisci, triangular disc). Intermediate-to-low signal intensity is characteristic for muscles in all types of sequences. In contrast, effusion, bone marrow, and hyaline cartilage exhibit considerable differences with different pulse sequences. Effusion and edema show high SI in T2-weighted sequences and low SI in T1-weighted sequences. In general, cartilage SI decreases visibly with increasing echo time because of its short T2 relaxation time. In addition, this T2 relaxation time varies strongly within the cartilage and accordingly cartilage SI decreases from the surface towards the subchondral bone. 8.7.2  Knee The MRI protocol for the knee should comprise images in all three major orthogonal planes with T2-weighted or moderately T2-weighted/PD-weighted fat-saturated TSE sequences, which provide a better signal-to-noise (SNR) ratio than STIR sequences do. However, if field strength is low or field homogeneity is not optimal, then STIR sequences enable more robust fat suppression. One T1-weighted sequence should also be included in order

to obtain additional information on anatomy and bone marrow. Depending on specific clinical questions the protocol may be extended (contrast media application for tumors, inflammation, infection, T2*-weighted sequences for detection of hemosiderin deposition). The knee is affected by injury more often than any other joint. Not rarely, there is a complex lesion pattern that involves the cruciate ligaments, menisci, collateral ligaments, and the articular cartilage. Therefore, especially in the knee, MRI should be used to identify specific combinations of lesions rather than to focus on one lesion alone (Hayes 2000). Very helpful is the distribution of bone marrow edema-like signal alterations (BMEP). Although mostly faint in avulsive injury, this type of signal alteration usually is very prominent in contusional injury. The location of the BMEP area in the anterior tibia and eventually in the overlying anterior portions of the femoral condyle indicates hyperextension or dashboard injury, often associated with posterior cruciate ligament (PCL) injury. The location of BMEP in the posterolateral tibia and, in some cases, in the intermediate portions of the lateral femoral condyle is typical for anterior cruciate ligament (ACL) injury. Location in the lateral aspect of the tibia and the femoral condyle may indicate direct valgus stress, with disruption of the medial collateral ligament and eventually the ACL. BMEP in the inferomedial aspect of the patella and the anterior aspect of the lateral femoral condyle is typical for patellar dislocation and must initiate a detailed search for a cartilage or cartilage on bone flake fracture from the patella and in addition a disruption of the medial patellofemoral ligament (Elias 2002). Thus, the value of BMEP and its distribution is in alerting the radiologist to specific trauma mechanisms and concomitant lesion patterns and in directing attention to a targeted search for these lesions. 8.7.2.1  Menisci For the assessment of the menisci a combination of coronal and sagittal sections is appropriate (Beltran et al. 1990), enabling visualization of all parts of the menisci (Fig. 8.7.1). Transverse images may in some cases be helpful in depicting radial tears. Specific techniques such as radial reconstructions perpendicular to the curvature of the menisci from 3D data sets have not gained general acceptance. Lesions in the menisci are depicted as signal increase in sequences using short echo times (T1-weighted or PD-weighted fat saturated (turbo) spin-echo (TSE) sequences). Meniscal alterations can be classified in three grades with MRI using the guidelines developed by Stoller et al. (1987) and modified by Reicher et al. (1986) (Figs 8.7.1, 8.7.2). Alterations of grades I (globular) and II (linear) are arthroscopically occult because they do not reach the surfaces of the menisci. Underlying pathology

1117

1118

8  Musculoskeletal System

is degenerative change with mucoid transformation of meniscal tissue. In grade II alterations, damage to longitudinally aligned collagen fibers is thought to explain the course of the linear signal alterations. Follow-up observations showed that grade II alteration mostly remains stable and only rarely progresses to clinically relevant ruptures or; instead, it usually remains stable, or may even disappear (Dillon et al. 1990). Triangularly shaped signal alterations entering the meniscus from its base are to be differentiated from intrameniscal degenerations, as they correspond to the fibrovascular bundles in the peripheral third of the menisci. Their shape is uniform and they are mostly symmetric. In contrast to grade I and II meniscal alterations, signal abnormalities reaching at least one of the surfaces of the menisci are to be called meniscal tears (Figs. 8.7.3, 8.7.4). Such grade III lesions correspond to arthroscopically detectable ruptures of variable depth. The most common location is the posterior horn of the

medial meniscus. This is clinically relevant because arthroscopically the posterior horn of the menisci, especially the inferior surface of the medial meniscus, is difficult to visualize during arthroscopy. While in some cases these lesions may become evident only by probing at arthroscopy, they can be diagnosed with a high degree of confidence on MRI. On the other hand, horizontal tears and fibrillation at the free margin of the menisci are more difficult to depict with MRI. Often they occur subsequent to progressive degenerative change. Traumatic tears, however, exhibit a predominantly vertical orientation.

Fig. 8.7.1  Imaging planes for the menisci. Above: sagittal plane, below: coronal plane

Fig. 8.7.2  Classification of meniscal SI alterations (grades 0– IIIb)

Fig. 8.7.3a–c  Traumatic meniscal tear, T1-weighted sagittal (a) and coronal (b), and STIR coronal (c) images. The tear characterized by the linear signal increase reaching the surface of the meniscus (b,c) and shows a predominantly vertical component. The tear is located in the peripheral third of the meniscus (b,c) exhibiting a complex pattern running in more than one plane (a)

8.7  Imaging of Internal Joint Derangement

Characteristic imaging findings of bucket handle tears are the “fragment-in-notch” sign and the “truncated-meniscus” sign in coronal sections. These correspond to a longitudinally separated meniscal fragment that is displaced into the intercondylar notch. In the more common case of a bucket handle tear of the medial meniscus, this fragment is displaced underneath the posterior cruciate ligament against the barrier of the ACL or its sy-

Fig. 8.7.4  Complex tear of the posterior horn of the medial meniscus, indicative of a potentially unstable lesion. Sagittal moderately T2-weighted fat-saturated (FS) FS TSE

Fig. 8.7.5a–c  Bucket-handle tear of the medial meniscus. Sagittal T1 (a more peripheral section, b more central section) and coronal moderately T2-weighted FS (c). There is a disturbed volume relationship between the anterior and posterior horns of the medial meniscus (a) because a posterior portion of the meniscus is flipped anteriorly over the anterior horn (flippedmeniscus sign). The displaced inner fragment in this longitudinal tear (handle) is found in the femoral notch underneath

novial coverage preventing further dislocation. In central sagittal sections, this gives rise to the “double-PCL” sign (Weiss et al. 1991) (Fig. 8.7.5). Portions of the posterior meniscus are flipped anteriorly over the remaining anterior horn, the “flipped-meniscus” sign. This is a very conspicuous sign of serious meniscal pathology because it reflects the disturbed relationship of the volumes of the anterior and posterior horns of the menisci. In the lateral menisci, the anterior and posterior horns are of approximately the same size, whereas in the medial meniscus the posterior horn is of about twice the size of the anterior horn. Any disturbance of this volume relationship is indicative of an instable meniscal lesion or previous meniscus surgery. The shape and course of a meniscal tear have to be precisely described because this is relevant for therapy. Especially vertically oriented peripheral (peripheral third) tears can be sutured because healing response usually is good in the vascularized peripheral third of the menisci (red zone). According to DeSmet et al. (1994, 2006) diagnostic accuracy is considerably improved when linear surfacing meniscal abnormalities can be visualized in more than one section and more than one imaging plane. In addition to these linear signal alterations, any change in shape of the menisci has to be considered, such as shortening, truncation, or lack of visualization of a portion of the meniscus in its normal, expected anatomic position. In some arthroscopically correlated studies, the diagnostic

the PCL (double-PCL sign) (b). The ACL (or its fragments and synovial coverage) prevent further displacement. The remaining peripheral portion of the meniscus appears truncated c in coronal sections (truncated-meniscus sign), and the displaced fragment is clearly visible in the femoral notch underneath the PCL (c fragment-in-notch sign). Please note the ACL tear and its characteristic BMEP in the posterolateral tibia

1119

1120

8  Musculoskeletal System

accuracy of MRI for meniscal tears is reported to be as high as 88–98% (Boeree et al. 1991; De Smet et al. 1994; Fisher et al. 1991; Justice and Quinn 1995) (Fig. 8.7.6). Because of uncertainties in the clinical diagnosis (up to 30% false-positive findings), the MRI diagnosis of meniscal lesions is very important. In summary, MRI is able to confirm or exclude clinically relevant meniscal lesions with a high degree of confidence and thus contributes to avoiding purely diagnostic arthroscopy and to selecting patients for and planning arthroscopic procedures. The postsurgical MRI appearance of the menisci depends on the procedure that was carried out. After suturing, grade III signal alterations may persist in clinically asymptomatic patients. Only newly developing signs of meniscal tears in follow-up examinations may provide evidence of a recurrent tear. Partial resection of meniscal tissue may bring central scar tissue to the newly created

surface of the operated meniscus. Such findings must not be described as a rupture (Smith and Totty 1990). On the other hand, not infrequently after total meniscectomy, some meniscal tissue remnants may be observed. Usually they exhibit higher signal intensity than would be expected from normal meniscal tissue. All these aspects underline that for the assessment of the postoperative menisci previous MRI studies should be taken into account. Recent studies suggest that (McCauley 2005) in cases with more than 25% of meniscal tissue resection direct MR arthrography may be beneficial for detecting recurrent meniscal tears. Various anatomical structures may mimic meniscal tears (Watanabe et al. 1989). In the posterolateral corner, the popliteal tendon must not be misinterpreted as a vertical tear of the posterior horn of the lateral meniscus. Usually, the tendon is easily identified by following its Fig. 8.7.6  Degenerative meniscal tear with linear signal intensity increase clearly contacting the undersurface of the posterior horn of the medial meniscus. The oblique horizontal course of the tear is typical for its degenerative nature without association to major trauma

Fig. 8.7.7a–c  Discoid meniscus in a 6-year-old girl. a Coronal T1-weighted sequence and b,c coronal and sagittal PD-w FS sequence. Please note the thickening of the lateral meniscus with loss of the typical biconcave shape. There is diffuse SI increase in the central portions of the meniscus, a frequent finding in discoid menisci thought to represent either beginning degeneration or persistent vascularization

8.7  Imaging of Internal Joint Derangement

course from inferomedial to superolateral in transverse and sagittal images. Anteriorly, the transverse ligament connects the anterior horn of the medial and lateral menisci and may mimic a meniscal tear in central sagittal sections. Again, continuity of this ligamentous structure (running together with the inferior lateral geniculate artery) through several contiguous sections helps to avoid this pitfall. Similarly, the anterior and/or the posterior meniscofemoral ligaments should be identified and differentiated from free articular bodies or meniscal fragments. Often, central sagittal sections help to identify these structures because they appear as a characteristic dot anterior or posterior to the intermediate portion of the posterior cruciate ligament. MRI easily depicts intra- or parameniscal cysts and helps to determine their localization and extension. Their typical, loculated appearance with high signal intensity content in T2-weighted images as well as their topographic relationship to the menisci allow for a confident diagnosis. Usually, they are associated with (degenerative) meniscal tears (Fig. 8.7.7). Sometimes MRI allows visualization of the communication between the cysts and meniscal tear (Fig. 8.7.8). 8.7.2.2  Cruciate and Collateral Ligaments, Patellar Tendon, and Quadriceps Tendon The anterior cruciate ligament (ACL) has its origin at the inner aspect of the posterior portion of the lateral femoral condyle and inserts in the intercondylar region anterior to the intercondylar eminence. On average, it is 3.5-cm long and 11-mm thick. The posterior cruciate ligament (PCL) has its origin at the inner aspect of the

medial femoral condyle and inserts posterior inferior in the intercondylar area of the tibia. It is of note that the insertion site of the PCL is approximately 1 cm deeper than the articular surface of the tibia, close to the (former) physis. The PCL is about 3.8-cm long and stronger than the ACL is, with approximately 13-mm thickness. In the extended knee, the PCL shows a slight curvature, which is convex with respect to the cranial direction. The ACL runs straight and parallel to the roof of the femoral notch (line of Blumensaat). In experienced hands, ligamentous lesions in the knee can be clinically detected with a high degree of confidence. However, in acute trauma the clinical examination is limited. As already underlined, often in the knee there is a combination of several lesions that, if undetected, may lead to chronic instability and subsequent osteoarthritis even with adequate ACL repair. Therefore, the search for such associated lesions (menisci, posterior corner, cartilage) is important. With an external rotation of the knee of 10–20°, the ACL usually is depicted in sagittal sections in its full extension. Paracoronal angulation provides a second plane, where the ACL or ACL grafts are visualized in their complete length, and this may help to evaluate the status of the ligament or its repair especially in problematic cases. Often in sagittal sections, there is some partial volume averaging at the insertion of the ligament to the femoral condyle. In contrast to the other ligaments in the knee exhibiting low signal intensity, the ACL shows high signal intensity strands especially in its tibial portion. This is due to fat interspersed in between the fascicles of the ACL. With high-resolution MRI, it may be possible to differentiate the two main fiber bundles (anteromedial, posterolateral) of the ACL. This may be useful in centers performing ACL repair in the so-called double-bundle technique. Given the oblique course of the ACL with respect to the main magnetic field, the magic angle phenomenon, too, is considered to contribute to the higher signal intensity of the ACL. The PCL usually is depicted in one or two adjacent sections in neutral position as well as in slight external rotation of the knee with homogeneously low signal intensity (Fig. 8.7.9 and Table 8.7.1). 8.7.2.2.1  ACL Tears

Fig. 8.7.8  Parameniscal ganglion. Huge, lobulated mass of hypointense signal medial to the medial femoral condyle and in close contact with the posterior horn of the medial meniscus

Especially in ACL lesions, direct and indirect signs and their validity for the diagnosis of a tear are well investigated. Direct signs of ACL rupture are lack of continuity of the ligament, absence of the ligament in its anatomical position in the lateral intercondylar space, a wavy contour of the ligament, and displacement of tibial or femoral portions of the ligament from a course parallel to the line of Blumensaat. In femoral-sided ruptures, the

1121

1122

8  Musculoskeletal System Table 8.7.1  MRI of the cruciate ligaments ACL

PCL

Signal intensity

Higher, fibers notable

Homogenously low

Imaging plane

Sagittal, external rotation 10–20°

Sagittal

Bending

Unbent

Strained

Extension

Strained

Unbent

Femoral insertion

Lateral condyle

Medial condyle

Tibial insertion

Area intercondylaris anterior

Posterior aspect tibial epiphysis

tibial portion of the ACL takes a more horizontal course (Fig. 8.7.10). Especially in acute trauma, focal or diffuse signal intensity increase is visualized within the ACL and its immediate vicinity together with obtuse margins and swelling of the ligament (Fig. 8.7.11). These alterations are especially pronounced in fat-saturated T2-weighted or STIR images (Vahey et al. 1991). Indirect signs of an ACL tear are the increased curvature of the PCL (Fig. 8.7.12e) as well as the posterior displacement of the lateral meniscus relative to the posterior border of the lateral tibia. Both signs result from the anterior displacement of the tibia in relation to the femur resulting from anterior instability of the knee due to the ACL rupture. Anterior subluxation of the tibia relative to a vertical tangent to the dorsal boundary of the lateral femoral conFig. 8.7.9a,b  Intact a anterior and b posterior cruciate ligament on sagittal T1-weighted SE images. Straight course of the ACL parallel to the roof of the intercondylar notch. On its tibial portion, there is a signal intensity increase due to interpositioning of fat between the fiber bundles. The PCL has its origin at the inner aspect of the medial femoral condyle and its insertion at the posterior inferior intercondylar area of the tibia. The PCL shows a slightly convex curvature with respect to the cranial direction. Posterior to the PCL (arrow) a posterior meniscofemoral ligament (Wrisberg) is depicted

Fig. 8.7.10a,b  Acute ACL rupture close to the femoral insertion. a T1-weighted and b T2-weighted SE sequence. The residual tibialsided ligament shows an almost horizontal course (white arrow), the femoral portion cannot be detected. There is hematoma and effusion adjacent to the ligament

8.7  Imaging of Internal Joint Derangement

dyle of more than 7 mm is strongly predictive of a complete tear of the ACL (Chan et a. 1994). In acute ACL injury, subchondral contusional bone marrow edema-like signal alterations (BMEP) often are very conspicuous in the intermediate portion of the lateral femoral condyle and in the posterolateral tibia (Figs. 8.7.12, 8.7.13). They occur subsequent to transient subluxation and impaction of the lateral femur on the posterolateral tibia.

For complete ACL tears (Figs. 8.7.14, 8.7.15, 8.7.16), MRI is reported to have sensitivity of up to 95% and specificity of up to 98% (Fowler et al. 1995; Vellet et al. 1995). Partial tears of the ACL may progress to complete tears and subsequent instability. This can be avoided by early surgery. In partial ACL tears, focal or diffuse signal intensity increase on T2-weighted images are depicted along with intact portions of the ligament, which can be Fig. 8.7.11  Morphological findings in ACL lesions. Ligament discontinuity; absent in expected anatomical position, increased signal intensity within the intercondylar space, localized zone of increased signal intensity

Fig. 8.7.12a–f  Segond fracture. a Coronal PD FS image, b T1weighted SE sequence, c coronal PD FS image, d T1-weighted SE sequence, e,f T2-weighted sagittal image. a,b There is extensive BMEP in the posterolateral aspect of the tibia resulting from impaction of the femur on the tibia in the moment of ligament disruption. c,d Note the small, vertically oriented bony fragment

immediately posterior to the insertion of the iliotibial tract resulting from avulsion injury to the tibia showing very subtle (avulsive) BMEP. The menisci and the LCL are not injured in this patient. e As an indirect finding of ACL rupture there is hyperbuckling of the PCL. Less common in adults than in children, there is bony avulsion of the ACL from the tibia (d,f)

1123

1124

8  Musculoskeletal System

Fig. 8.7.13  Sagittal T1-weighted sequence in a patient with ACL rupture. Please note the very subtle vertical tear of the posterior horn of the lateral meniscus. Such subtle tears are not infrequently overlooked. There is only a small area of BMEP in the posterolateral tibia. The medial femoral condyle is shifted posteriorly and is located more dorsally than the posterior tibial border

followed from the femoral through to the tibial insertion site. Usually the contour of the ligament shows some bowing or a wavy configuration. Diagnostic accuracy is lower for partial ruptures than for complete ruptures (Umans et al. 1995) (Fig. 8.7.14). MRI is of value in following up ACL repair (McCauley 2005). It can depict the position of the transplant, rerupture, and impingement of the transplant, focal anterior fibrosis (“Cyclops lesion”) as well as patellar fractures and alterations in the patellar tendon (Recht et al. 1996) in case of bone patellar tendon bone reconstruction. The two main techniques for ACL surgery are bone patellar tendon bone grafting or hamstring grafting. Knowledge of the operational technique is important for the radiologist because in hamstring grafts the excised tendon is doubled, sutured, and redoubled again, so that the ACL is reconstructed by a bundle of up to 4 separate tendon strands which normally exhibit linear strands of fluid in between the separate tendon strands. In the case of a bone–patellar tendon bone graft, such a finding would alert the reader to a possible graft fissuring or even imminent graft failure. The position of the bone tunnels in the femur and tibia is important for the isometry of the transplant. In sagittal images, the femoral tunnel should be placed at the intersection of the line of Blumensaat, with the dorsal cortex of the femur. The tibial tunnel should be placed in any position of the knee (that is especially in the fully extended knee) posterior to the intersection of the line of Blumensaat with a tangent

Fig. 8.7.14a,b  ACL rupture with contusion zones. a Coronal PD FS sequence. There is the typical contusion area at the posterolateral tibial plateau. b Coronal PD FS sequence in the same patient with an additional contusion area of the intermediate medial femoral condyle

8.7  Imaging of Internal Joint Derangement

to the tibial plateau. The signal intensity of the transplant varies with time. In the first 3 to 4 months after the operation, it exhibits low signal intensity in all sequences. Within 4 to 8 months, and in some cases even 10 to 12 months, months SI is increased, presumably due to resynovialization and revascularization of the graft. After 1–2 years, the SI of the graft should become low in all sequences again. Usually, a complete rerupture of the graft can be confidently identified by depicting the loose ends of both remnants. Impingement of the grafts can occur against the roof of the femoral notch as well as against the inner aspect of the lateral femoral condyle and shows a variable degree of signal increase. Especially with a too-far-anterior placement of the tibial tunnel, impingement against the intercondylar roof may occur. In coronal sections, the entry of the femoral tunnel should be placed at the 10 or 11 o’ clock and/or 1 or 2 o’ clock position. Arthrofibrosis may lead to persistent pain after ACL repair. Its focal form in the anterior intercondylar compartment gives rise to the term Cyclops lesion, which exhibits low SI. In the first months after ACL repair, pronounced inflammatory reactions may be present in the patellar tendon or the hamstring insertion site. With the course of time (up to 3 years) defects in the patellar tendon may fill with repair tissue in cases of bone patellar tendon bone grafts, whereas a so-called neo-tendon may form in the hamstring grafts growing from proximal to distal within the former tendon track.

layer. The intermediate portion biomechanically is the strongest of the medial supporting structures. Usually there is fat or a bursa-like structure interposed between the capsule and the MCL. The ligament originates from the medial femoral condyle and inserts on the medial aspect of the tibia approximately 7-cm distal to the tibial joint surface. In contrast, the capsule, composed by the meniscofemoral, the meniscotibial, and the femorotibial ligaments, all blending with each other at the base of the meniscus, inserts closer to the joint surfaces of femur and tibia. The MCL blends with the posterior third of the inter­mediate portion of the medial meniscus, presumably leading to a reduced mobility of the medial meniscus. Damage to the MCL results from excessive valgus stress in the flexed knee. Usually the deepest layer (capsule) is ruptured first, as it is the weakest of the medial supporting structures. Mostly, lesions are located in the femoral portions of the ligament (Fig. 8.7.16). In some cases meniscocapsular separation may occur. The coronal plane, preferably with T2-weighted or stir sequences, is most appropriate for the visualization of MCL tears (Schweitzer et al. 1995; Jergesen et al. 1985). There are three grades of ligament affection, which correlate with specific findings on MRI. Grade I is microscopic fiber disruption with some internal hemorrhage in the ligament, local pain at palpation, and no instability. Grade II corresponds to partial macroscopic tears of the ligament, local pain, and increased knee mobility with valgus stress. Grade III is the complete disruption of the ligament and clear insta-

8.7.2.2.2  PCL Tears PCL tears are less common than ACL tears are. They classically occur in the flexed knee, subject to forces from anterior against the proximal tibia. Often the site of tear is in the intermediate portion of the ligament, and there is concomitant damage to the posterior joint capsule. In hyperextension injury, frequently there is a bony avulsion of the PCL from its tibial insertion site. Most often, lesions to the PCL are associated with lesions to the posterolateral corner of the knee. Usually PCL tears can be diagnosed with a high degree of confidence on MRI (Fig. 8.7.15) (Grover et al. 1990). Contusional bmep is located in the anterior portions of the tibial plateau and in the overlaying anterior portions of the femoral condyles. 8.7.2.2.3  Collateral Ligaments The medial collateral ligament (MCL) is important for the stability of the knee. It constitutes the intermediate portion of the three layers of the medial supporting structures, composed of the fascial planes most superficially, and followed by the so-called medial collateral ligament as the intermediate and the joint capsule as the deepest

Fig. 8.7.15  Sagittal T2-weighted sequence. Rupture of the anterior (ACL) and posterior (PCL) cruciate ligament. There are only irregular residuals of both cruciate ligaments visible. There is prominent dorsal subluxation of the tibial head

1125

1126

8  Musculoskeletal System

bility. MRI findings are obtuse margins of the ligament, focal signal intensity increase (edema, hemorrhage); disruption of ligament continuity, wavy contours, and eventually separation from the bony insertion site. There may be increased signal intensity in the periligamentous tissue, with loss of a clear delineation of the subcutaneous fat. BMEP in the lateral femorotibial compartment is a result from direct valgus stress and may be prominent. Note that faint bmep may occur as a result of bony avulsion at the insertion site(s), too. In contrast to the MCL, the LCL does not blend with the joint capsule. It has a more oval, cross-sectional shape running from the lateral aspect of the fibular head (note: not the very tip of the fibula) to the lateral femoral condyle. Slight angulation of the coronal plane may help to visualize the LCL in its entire length within one section.

Similar to the MCL, three grades of injury can be diagnosed for the LCL with corresponding MRI findings. However, whereas more and more often there is conservative treatment for MCL tears, complete LCL disruption constitutes a serious injury to the posterolateral corner of the knee, which if left untreated, results in continued posterolateral instability and a high rate of ACL or PCL repair failure and premature osteoarthritis OA. 8.2.2.2.4  Knee Extensor Apparatus The extensor apparatus of the knee is composed mainly of the patellar tendon, quadriceps tendon, the patella, and its retinacula. The quadriceps tendon inserts in part at the upper pole of the patella, in part continues over the anterior surface of the patella, and has its final insertion (as patellar tendon) at the tibial tuberosity. In MRI, it is mostly of low signal intensity with an anterior–posterior (a.p.) cross-sectional diameter of 5 mm. In asymptomatic persons, a V-shaped focal signal intensity increase may be present, which is most often located at the proximal, and sometimes at the distal end of the patellar tendon. In higher age groups and in overweight patients the ligament may take a wavier course. Chronic overuse can lead to repetitive episodes of inflammation and degeneration of the patellar tendon, termed “jumper’s knee” (Khan 1996). MRI shows an increase of the thickness of the patellar tendon to more than 7 mm in the a.p. direction together with signal intensity increase, obtuse (especially posterior) margins, and adjacent signal alterations in Hoffa’s fat pad (Sonin 1995). Rupture of the patellar tendon exhibits lack of continuity on MRI and the remaining fragments of the tendon show a wavy course. There is patella alta. In OsgoodSchlatter disease, there is thickening and signal intensity increase in T2-weighted sequences in the distal portion of the patellar tendon close to insertion at the tibial tuberosity. In Sinding-Larsen-Johansson disease, corresponding signal intensity alterations are present at the lower pole of the patella (Sonin 1995). 8.7.2.3  Cartilage Imaging

Fig. 8.7.16a,b  MCL tear. Coronal T1-weighted (a) and T2weighted (b) sequences. The MCL is displaced from the medial femoral condyle. There is concomitant effusion and soft-tissue swelling. In b, the meniscofemoral and meniscotibial ligaments (capsular attachments of meniscus) are well delineated by the effusion

Dedicated MRI of the cartilage is made possible by high field strength MRI. In the last decade, MRI of the cartilage has profited strongly from advances in scanner and coil technology as well as gradient and sequence design. In addition, because of the high impact of cartilage lesions for the pathogenesis of osteoarthritis (OA), increasing effort has been put into cartilage-dedicated therapeutic strategies such as mosaicplasty, drilling, and chondrocyte transplantation. Those techniques create a need for a non-invasive, accurate, valid, and widely available method to diagnosis cartilage lesions to aid in therapy-oriented staging, follow-up of cartilage damage, and

8.7  Imaging of Internal Joint Derangement

cartilage repair as well evaluation of the different therapeutic strategies. Because of the complex, zonally anisotropic internal structure of the cartilage resulting in a short T2 relaxation time and because of its low thickness (1 to at best 6 mm), MRI of the cartilage is still challenging. The highest anatomical resolution at acceptable SNR and contrast-tonoise ratio (CNR) and imaging times of around 5–8 min are required, motivating ongoing development of dedicated cartilage sequences. Improved spatial resolution reduces partial volume averaging and truncation effects. Gradient-echo sequences allow a reduction of imaging time, and 3D, as opposed to 2D, techniques contribute additional SNR (Smith and Totty 1990; Tervonen et al. 1993). Fat suppression improves contrast between cartilage and adjacent tissues and eliminates chemical shift artifacts. Currently, at 1.5T, optimized image resolution with T1-weighted 3D FS GRE sequences is about 0.3 × 0.3 × 1.5 mm3. Relevant signal arises only from uncalcified cartilage. Calcified cartilage and subchondral bone delineate the uncalcified cartilage as a line devoid of signal from the marrow. High-resolution T1-weighted 3D FS GRE sequences exhibit cartilage with homogeneous high signal intensity and show excellent delineation from the subchondral bone. Delineation from serous effusion is good. Delineation of synovial proliferations is only reliable subsequent to i.v. Gd application (Peterfy 1994). FS sequences with long repetition time and echo time show excellent contrast between cartilage and joint fluid. This so-called arthrographic effect is used to delineate superficial contour irregularities. However, due to the short T2 relaxation times of the cartilage, especially in its deep portions, SNR is low in long-echo-time sequences and leads to difficulties in delineating the deeper portions of cartilage from the subchondral bone. Therefore, currently, moderately T2-weighted sequences (TE ≤ 50 ms) and sometimes double-echo steady-state (DESS) sequences are used together with the T1-weighted 3D GRE sequences for cartilage evaluation (Broderick et al. 1994; Disler et al. 1995) (Fig. 8.7.17). In the early stages of cartilage degeneration and in the immediate neighborhood of cartilage defects proteoglycan depletion is observed. Local softening of cartilage at arthroscopic probing usually is ascribed to a local disruption of the collagenous fiber architecture, with subsequently altered water and proteoglycan content. Mechanically induced degeneration usually starts from a focal superficial lesion, leading to continuous erosion of cartilage tissue. Acute trauma results in flake fractures, impaction fractures, and subsequent joint surface incongruity as a pre-arthrotic condition (McCauley 2002). Diagnostic criteria of the various classification schemes for cartilage lesions are related to arthroscopic grading schemes. They comprise intracartilaginous signal intensity alterations, eventually a focal blistering of cartilage with intact cartilage surface, fissuring, and reduction of cartilage thick-

ness (focal as a defect or more widespread, diffuse). Such thickness reduction usually is given in intervals of half, a third, or a fourth of total cartilage thickness in addition to delineation or even eburnation of subchondral bone. In addition to lesion depth, the extent (area) and the affected compartments are described. Given the currently possible resolution between 4 × 0.6 × 0.6 to 1 × 0.3 × 0.3 mm3, it is clear that the diagnosis of fine superficial fibrillation, small defects and fissuring of cartilage is not reliable (Link et al. 1998). However, depiction of lesions

Fig. 8.7.17  Grading scheme for classification of different cartilage lesions by MRI (most of the MRI schemes are borrowed by arthroscopy). Intact surface, but focal inner signal alterations, local thickening, small surface lesions, gradual reduction of cartilage thickness (more or less than 25/50/75%), exposure and arrosion of subchondral bone

Fig. 8.7.18a,b  Chondromalacia patellae. Axial 3D DESS FS (left-hand side) and moderately T2-weighted FS TSE (right-hand side). Grade I lesions a show but intracartilaginous signal alterations with a still well-preserved, smooth articular surface. These lesions may be both hyperintense as well as hypointense. Grade II lesions (b) exhibit subtle surface irregularities in addition to intracartilaginous signal alterations. They are well delineated by the arthrogram-like effect of the effusion

1127

1128

8  Musculoskeletal System

Fig. 8.7.19a–c  Patellar flake fracture. Coronal STIR (a), axial T1-weighted FLASH FS (b) and DESS FS (c). The fragment is displaced in the medial recess (a–c) and well delineated by the extensive effusion (a,c) whose intermediate signal intensity in the T1-weighted sequence (b) is compatible with hemorrhagic components

Fig. 8.7.20a–c  Autologous osteochondral transplantation for osteochondritis dissecans (OCD) of the inner portion of the medial femoral condyle. Coronal STIR (a), T1-weighted SE (b) and sagittal T1-weighted SE (c) sequence. Alignment of the graft is challenging due to the highly variable degree of joint surface curvature in all planes, which has to be modeled with the

grafts (b,c). There is hypointensity of the graft within the first 3 months after operation due to surgical trauma and initial lack of vascularization (b,c). More anteriorly, the donor site is seen (a). Please note the lateral discoid meniscus, known to be associated with OCD of the medial femoral condyle

of the order of magnitude of one fourth to one third of the total thickness of the cartilage should be possible with reasonable diagnostic confidence. Diagnostic confidence is especially high for chondromalacia patellae as a specific disease entity of the younger adult (Ficat et al. 1999; Shahriaree 1985; McCauley et al. 1992; Recht et al. 1996) (Fig. 8.7.18). In addition to cartilage alterations in degenerative or inflammatory joint disease, subchondral cysts or syno-

vial proliferations are apparent. In case of hemosiderin depositions with hemophilic arthropathy or cases of pigmented villonodular synovitis, characteristic susceptibility induced signal voids can be visualized using T2*weighted GRE sequences. For early diagnosis and follow-up of degenerative disease, quantitative approaches are being developed. They comprise volumetric approaches, delayed Gd-enhanced MRI contrast imaging (dGEMRIC), and T2 relaxation

8.7  Imaging of Internal Joint Derangement

Fig. 8.7.21a–c  Cartilage lesions/defects on the medial femoral condyle and medial and lateral patellar facet. a Coronal PD FS sequence, b axial PD FS sequence, and c sagittal DESS sequence, showing reduced cartilage thickness and foci of hyperintense

signal within the remaining cartilage indicating chondromalacia and effusion delineation cartilage defects. In c, a Baker cyst is depicted. Note concomitant meniscal pathology (c)

time quantification (Mlynarik et al. 1999; Vahlensieck et al. 1994 Eckstein and Glaser 2004; Glaser 2005; Glaser et al., 2000; Gray et al. 2001; Goodwin 2001; Mosher and Dardzinski 2004) (Figs. 8.7.19, 8.7.20, 8.7.21).

9.

10.

References 1.

2.

3.

4.

5.

6.

7.

8.

Beltran J, Knight CT, Zuelzer WA et al (1990) Core decompression for avascular necrosis of the femoral head: correlation between long-term results and preoperative MR staging. Radiology 175:533–539 Boeree NR, Watkinson AF, Ackroyd CE, Johnson C (1991) Magnetic resonance imaging of meniscal and cruciate injuries of the knee. J Bone Joint Surg Br 73:452–457 Broderick LS, Turner LA, Renfrew D, Schnitzer TJ, Huff JP, Harris C (1994) Severity of articular cartilage abnormality in patients with osteoarthritis: evaluation with fast spinecho MR vs arthroscopy. Am J Roentgenol 162:99–103 Chan WP, Peterfy C, Fritz RC, Genant HK (1994) MR diagnosis of complete tears of the anterior cruciate ligament of the knee: importance of anterior subluxation of the tibia. AJR 162:355–360 De Smet AA, Tuite MJ, Norris MA, Swan JS (1994) MR diagnosis of meniscal tears: analysis of causes of errors. AJR Am J Roentgenol 163:1419–1423 De Smet AA, Tuite MJ (2006) Use of the “two-slice-touch” rule for the MRI diagnosis of meniscal tears. AJR Am J Roentgenol 187:911–914 Dillon EH, Pope EF, Jokl P, Lynch K (1990) The clinical significance of stage 2 meniscal abnormalities on Magnetic Resonance knee images. Magn Reson Imaging 8:411–415 Disler DG, McCauley TR, Wirth CR, Fuchs MD (1995) Detection of knee hyaline cartilage defects using fat-suppressed three-dimensional spoiled gra­dient-echo MR Imaging: Comparison with standard MR Imaging and correlation with arthroscopy. Am J Roentgenol 165:377–382

11. 12.

13.

14.

15.

16.

17.

18.

19.

Eckstein F, Glaser C (2004) Measuring cartilage morphology with quantitative magnetic resonance imaging. Semin Musculoskelet Radiol 8:329–353. Review Elias DA, White LM, Fithian DC (2002) Acute lateral patellar dislocation at MR imaging: injury patterns of medial patellar soft-tissue restraints and osteochondral injuries of the inferomedial patella. Radiology 225:736–743 Ficat RP, Philippe J, Hungerford DS (1979) Chon-dromalacia patellae: a system of classification. Clin Orthop 144:55–62 Fisher SP, Rox JM, Del Pizzo W, Friedman MJ, Snyder SJ, Ferkel RD (1991) Accuracy of diagnoses from magnetic resonance imaging of the knee: a multicenter analysis of one thousand and fourteen patients. J Bone Joint Surg Am 73:2–10 Fowler PJ, Miniasci A, Amendola A (1995) Anterior cruciate ligament tear: prospective evaluation of diagnostic accuracy of middle- and high-field strength MR imaging at 1.5 and 0.5 T. Radiology 197:826–830 Glaser C (2005) New techniques for cartilage imaging: T2 relaxation time and diffusion-weighted MR imaging. Radiol Clin North Am 43:641–653, vii. Review Goodwin DW (2001) Visualization of the macroscopic structure of hyaline cartilage with MR imaging. Semin Musculoskelet Radiol 5:305–312. Review Gray ML, Burstein D, Xia Y (2001) Biochemical (and functional) imaging of articular cartilage. Semin Musculoskelet Radiol 5:329–343. Review Grover JS, Bassett LW, Gross ML, Seeger LL, Finer-man GAM (1990) Posterior cruciate ligament: MR imaging. Radiology 174:527–530 Hayes CW, Brigido MK, Jamadar DA, Propeck (2000) Mechanism-based pattern approach to classification of complex injuries of the knee depicted at MR imaging. Radiographics 20(Spec. no.):S121–S134 Jergesen HE, Heller M, Genant HK (1985) Magnetic resonance imaging of the femoral head. Orthop Clin North Am 16:705–716

1129

1130

8  Musculoskeletal System 20. Justice WW, Quinn SF (1995) Error patterns in the MR imaging evaluation of menisci of the knee. Radiology 196:617–621 21. Khan KM, Bonar F, Desmond PM, Cook JL, Young DA, Visentini PJ, Fehrmann MW, Kiss ZS, O’Brien PA, Harcourt PR, Dowling RJ, O’Sullivan RM, Crichton KJ, Tress M, Wark JD (1996) Patellar tendinosis (jumper’s knee): findings at histopathologic examination, US, and MR imaging. Victorian Institute of Sport Tendon Study Group. Radiology 200:821–827 22. Link T, Majumdar S, Peterfy CG, Daldrup HE, Uffmann M, Dowling C, Steinbach L, Genant HK (1998) High resolution MRI of small joints: impact of spatial resolution on diagnostic performance and SNR. Magn Reson Imaging 16:147–155 23. McCauley TR (2002) MR imaging of chondral and osteochondral injuries of the knee. Radiol Clin North Am 40:1095–1107. Review 24. McCauley TR (2004) MR imaging evaluation of the postoperative knee. Radiology 234:53–61. Review 25. McCauley T, Kier R, Lynch KR, Jokl P (1992) Chon-dromalacia patellae: diagnosis with MR imaging. Am J Roentgenol 158:101–105 26. Mosher TJ, Dardzinski BJ (2004) Cartilage MRI T2 relaxation time mapping: overview and applications. Semin Musculoskelet Radiol 8:355–368. Review 27. Mlynarik V, Trattnig S, Huber M, Zembsch A, Imhof H (1999) The role of relaxation times in monitoring proteoglycan depletion in articular cartilage. JMRI 10:497–502 28. Peterfy CG, Majumdar S, Lang P, van Dijke CF, Sack K, Genant HK (1994) MR imaging of the arthritic knee: improved discrimination of cartilage, synovium, and effusion with pulsed saturation transfer and fat-suppressed T1weighted sequences. Radiology 191:413–419 29. Recht MP, Piraino DW, Applegate G, Richmond BJ, Yu J, Parker RD, Andrish JT (1996) Complications after anterior cruciate ligament reconstruction: radiographic and MR findings. AJR 167:705–710 30. Reicher MA, Hartzmann S, Duckwiler GR, Basset LW, Anderson LJ, Gold RH (1986) Meniscal injuries: detection using MR imaging. Radiology 159:753–757 31. Schweitzer ME, Tran D, Deely DM, Hume EL (1995) Medial collateral ligament injuries: evaluation of multiple signs, prevalence and location of associated bone bruises, and assessment with MR imaging. Radiology 194:825–829 32. Shahriaree H (1985) Chondromalacia. Contemp Orthop 11:27–32 33. Smith D, Totty WG (1990) The knee after partial meniscectomy: MR imaging features. Radiology 176:141–144 34. Sonin AH, Fitzgerald SW, Bresler ME, Kirsch MD, Hoff FL, Friedman H (1995) MR imaging appearance of the extensor mechanism of the knee: functional anatomy and injury patterns. Radiographics 15:367–82 35. Stoller DW, Martin C, Crues JV, Kaplan L, Mink JH (1987) Meniscal tears: pathologic correlation with MR imaging. Radiology 163:731–735

36. Tervonen O, Dietz MJ, Carmichael SW (1993) MR imaging of knee hyaline cartilage: evaluation of two- and three-dimensional sequences. J Magn Reson Imaging 3:663–668 37. Umans H, Wimpfheimer O, Haramati N, Applbaum YH, Adler M, Bosco J (1995) Diagnosis of partial tears of the anterior cruciate ligament of the knee: value of MR imaging. AJR 165:893–897 38. Vahey TN, Broome DR, Kayes KJ, Shelbourne KD (1991) Acute and chronic tears of the anterior cruciate ligament. Differential features at MR imaging. Radiology 181:251–253 39. Vahlensieck M, Dombrowski R, Leutner C, Wagner U, Reiser M (1994) Magnetization transfer contrast (MTC) and MTC-subtraction enhances cartilage lesions and intrasubstance degeneration in vitro. Skelet Radiol 23:535–539 40. Vellet AD, Le DH, Munk PL et al (1995) Anterior cruciate ligament tear: prospective evaluation of diagnostic accuracy of middle- and high-field strength MR imaging at 1.5 and 0.5 T. Radiology 197:826–830 41. Watanabe AT, Carter BC, Teitelbaum GP, Bradley WG (1989) Common pitfalls in Magnetic Resonance imaging of the knee. J Bone Joint Surg Am 71:857–862 42. Weiss KL, Morehouse HT, Levy IM (1991) Sagittal MR images of the knee: a low signal band parallel

8.7.3  Shoulder MRI is established for the workup of shoulder pathologies and provides information relevant for treatment. However, recommendations for the proper examination technique still are not uniform. Posttraumatic and degenerative changes in the shoulder are frequently associated with chronic pain and functional disability. Especially in the shoulder, capsuloligamentous and muscular elements are critical for stabilizing and guiding movement. This is because the size of the bony articular surface of the glenoid is but a third the size of the articulating surface of the humeral head. Principles of examination technique. In MRI of the shoulder, the arm should be positioned in a neutral position, and, if tolerated, even in slight external rotation in order to allow for adequate distension of the anterior capsule and rotator cuff. The hand should not be placed on the abdomen in order to prevent respiratory induced movement artifacts. For an adequate resolution dedicated surface coil arrays are mandatory. The field of view (FOV) should not exceed 140–180 mm2 when using a 2562 matrix, slice thickness should be not more than 3–4 mm. Again, imaging in all three major orthogonal planes is required and T2-weighted TSE or STIR sequences accompanied by T1-weighted and PD/moderately T2-weighted FS sequences. Vahlensieck and colleagues (1992) recommend a dual-gradient-echo sequence, providing good anatomical detail from the first echo (in-phase) and a T2*-weighting with the second echo (opposed-phase).

8.7  Imaging of Internal Joint Derangement

Especially in the shoulder, MR-arthrography is very useful (Jbara et al. 2005). Both indirect and direct MRarthrography increase diagnostic accuracy in rotator cuff and capsulolabral lesions (Flannigan et al. 1990; Palmer et al. 1993, 1994). Usually, for the evaluation of (partial) rotator cuff tears, indirect arthrography is sufficient. Some reports suggest that indirect arthrography has high accuracy for the evaluation of capsulolabral tears as well, but currently, more experience and very good results for the evaluation of capsulolabral tears are available from direct arthrography. Pulley lesions still are difficult to evaluate with either MRI-based technique. For direct MR arthrography, 10–15 ml NaCl with or without 2 µmol/ml Gd-DTPA is injected under fluoroscopic control. Fat-saturated sequences increase lesion conspicuity. Anatomy and section orientation. The three orthogonal planes applied in the shoulder are angulated out of the axial (transverse), sagittal, and coronal planes in such a way as to best correspond to an orientation perpendicular and parallel to the glenoid (compare Fig. 8.7.22, Table 8.7.2). Nonetheless, although strictly speaking oriented in the paracoronal, parasagittal, and para-axial planes, in day-to-day terminology they are called coronal, axial, and sagittal sections. From transverse sections, the coronal sections are angulated parallel to the long axis of the belly and tendon of the supraspinous muscle. This muscle and its tendinous insertion on the humeral tuberosity, the deltoid muscle, the subacromial–subdeltoid bursa, the subacromial fat, and the acromioclavicular Joint (ACJ) are depicted. The bursa usually is not depicted directly but is delineated by a thin layer of fat on the lateral (superficial) aspect of the bursa, visible in 70% of cases. Although sometimes (mistakenly) neglected, perpendicular to the coronal sections, the sagittal sections give important cross-sectional and topographic information about the complete rotator cuff with the infraspinous and teres minor muscles posteriorly, the subscapular muscle anteriorly, and the supraspinous muscle superiorly. Anterosuperiorly there is the rotator cuff interval (not to be confounded with a rotator cuff tear), a gap in the muscular coverage of the glenohumeral joint between the anterior border of the supraspinous tendon and the upper

Fig. 8.7.22  Imaging planes for the MRI assessment of the shoulder

border of the subscapularis tendon, containing the biceps tendon when passing from intra-articular to extra-articular (Bigoni and Chung 2006). More superficially, the coracoacromial arch with the acromion, the coracoid process, and the coracoacromial ligament are visualized (Fig. 8.7.23). The shape and angle of the acromion are to be evaluated when it comes to diagnosis of, and pre-operative planning for, impingement (Gagey et al. 1993). A low angle and a hooked shape are predisposing factors for impingement and rotator cuff tears (Morrison and Bigliani 1987). Capsular recesses can be detected with adequate capsular distension (effusion, direct filling) anteriorly underneath the coracoid process and inferiorly in the axilla. Not infrequently, there is communication between the glenohumeral joint and the subscapular bursa. Transverse sections allow the evaluation the anterior and posterior portions of the glenoid labrum and the joint capsule, together with the long biceps tendon and its tendon sheath passing through the intertubercular sulcus. The middle glenohumeral ligament regularly is identified as a linear structure of low signal intensity deep to the subscapularis tendon (Fig. 8.7.24). The evaluation of the glenoid labrum is challenging due to both, its small anatomic extension and its high variability in shape (Bencardino and Beltran 2006). Especially anteriorly, it is delineated in only 45% (as opposed to 70% posteriorly in its “typical” triangular shape with low signal intensity. Irregularities and central signal inhomogeneities without association to trauma have to be differentiated from traumatic tears. In the upper anterior quadrant (12 to 3 o’ clock position in sagittal sections) the attachment of the labrum to the bony glenoid, the labrum’s relationship to the articular cartilage, and to the joint capsule with the capsule’s three reinforcing (superior, middle, and inferior) glenohumeral ligaments is highly variable. The first two of these ligaments arise from the upper half of the glenoid rim running (infero-)

Table 8.7.2  Imaging planes for the MRI assessment of the shoulder Axial

Labral lesions Hill-Sachs defect Anterior capsule space Long biceps tendon and tendon sheath

Oblique coronal

Supraspinous muscle and tendon AC joint Subacromial space Subacromial and subdeltoid bursa

Oblique sagittal

Acromioclavicular arch Shape of the acromion Rotator cuff components

1131

1132

8  Musculoskeletal System

Fig. 8.7.23a–c  Normal anatomy. T1-weighted (a) paracoronal and (b) axial and T2-weighted parasagittal (c) sequences. The paracoronal plane (a) displays the supraspinous tendon and muscle with its insertion on the humeral head passing underneath the acromion. Some fat signal should be visible underneath the acromion and the acromioclavicular joint (ACJ). Note the fat plane outlining the subdeltoid bursa. The axial plane illustrates the subscapular and infraspinous muscles and tendons as well as the deltoid muscle, all of them protecting and controlling mobility of the shoulder. A small amount of fluid around the biceps tendon running in the intertubercular groove is a normal finding. The parasagittal plane (c) provides important information on potential atrophy of the components of the rotator cuff as well as on the coracoacromial arch, the rotator interval, and the superior border of the subscapularis tendon

Fig. 8.7.24  Schematic visualization of the anatomical structures in the oblique-sagittal plane. 1 coracoid process, 2 acromion, 3 supraspinous muscle, 4 infraspinatus muscle, 5 teres minor muscle, 6 subscapularis muscle, 7 humeral head, 8 coracoacromial ligament, 9 long biceps tendon, 10 superior glenohumeral ligament, 11 intermediate glenohumeral ligament, 12 inferior glenohumeral ligament

8.7  Imaging of Internal Joint Derangement Table 8.7.3  Rotator cuff lesions: SIGNAL pattern (modified according to Rafii et al. 1990) Morphology

PD-weighted

T2-weighted

Normal

Normal

–

–

Tendinitis

Normal, swollen

0

(+)

Degeneration

Thinned, irregular

(+)

+

Partial tear

Focal SI increase - Intraligamentous - Synovial side - Bursal side

+

+

Complete tear

Broad SI increase - < 1 cm - > 1 cm - Retraction

+

+

– low signal intensity, + high signal intensity (fluid-like), 0 intermediate signal intensity

laterally to their insertion sites on the lesser tubercle. The biomechanically important inferior glenohumeral ligament is a hammock-like structure reinforcing the inferior and the postero- and anteroinferior portions of the capsule. The labrum may not be attached to the glenoid for several millimeters and consecutively is separated from it by fluid, leading to the diagnosis of a sublabral hole/foramen. If it is absent (as it is in up to 8% of patients) and the middle glenohumeral ligament is thickened (sometimes described as “cord-like”) there is a Buford complex. Sometimes, proximal portions of the labrum are but loosely overlying the articular cartilage and must not be mistaken for a labral tear. Likewise, fluid between the base of the labrum and the attachment of the capsule immediately medial to the labrum’s attachment may mimic a labral tear. This anterior capsular attachment can be highly variable, too (see Fig. 8.7.30). The three principle variants include a capsular attachment immediately adjacent to the base of the labrum (type I), an attachment up to one centimeter more medial on the glenoid (type II), and an attachment far medial on the neck of the scapula (type III). Types II and III must not be mistaken for capsular separation, but are considered potentially predisposing factors for anterior dislocation of the shoulder. 8.7.3.1  Lesions to the Rotator Cuff The range of rotator cuff defects comprises subtle degenerative changes as well as complete tears (Table 8.7.3). Most of the lesions to the rotator cuff are attributed to chronic impingement rather than to trauma. The former is related to repetitive impingement and consecutive overuse of the peripheral portions of the rotator cuff, the

glenohumeral joint capsule, the subacromial-subdeltoid bursa and in part, the long biceps tendon between the humeral head and the coracoacromial arch. The two main components are abnormal movement, and narrowed anatomic subacromial space. According to Neer (1972, 1982) three stages are differentiated, including pathologic findings, age of the patient and therapeutic concepts. Stage I is edema and hemorrhage in the subacromial tissues. Stage II is continued repetitive episodes of inflammation and fibrosis with or without partial tear. Stage III is a complete tear of the rotator cuff (Table 8.7.4). Most often by far, the tendon of the supraspinous muscle is affected, followed by the subscapularis tendon (coracoidal impingement) and the posteroinferior portion of the supraspinous muscle (posterosuperior impingement in overhead sports-related activities). Alterations immediately at the tendon insertion site and at the musculotendinous junction are to be differentiated from impingement. Factors associated with impingement are (Fig. 8.7.25): flat acromion, hooked shape of acromion, osteophytes at the tip of the acromion (insertion of coracoacromial ligament), capsular hypertrophy and osteoTable 8.7.4  Stages of the impingement syndrome according to Neer (1972) Stage

Finding(s)

I

Edema and bleeding within the subacromial soft tissue

II

Progressive inflammation and fibrosis with or without partial tear of the rotator cuff

III

Complete tear of the rotator cuff

1133

1134

8  Musculoskeletal System

Fig. 8.7.25a,b  Two patients with impingement syndrome. a Oblique coronal T1-weighted sequence. Marked hypertrophy of the capsule in AC joint osteoarthritis. The subacromial fat has almost disappeared and the supraspinatus muscle is impinged. b Axial image: Os acromiale in a professional female javelin thrower. (Courtesy of Dr. H. Bonel)

phytes in osteoarthritis (OA) of the ACJ, thinning of the fat interposed between ACJ/acromion and supraspinous muscle, os acromiale, a thickened coracoacromial ligament and (rarely) hypertrophy of the supraspinous muscle (Seeger et al. 1988), and a prominent coracoid process that may show degenerative changes on its undersurface (Figs. 8.7.26, 8.7.27; Table 8.7.5). 8.7.3.1.1  Degeneration, Tendinitis, Bursitis There is no general agreement on the exact etiology and terminology of structural changes within the rotator cuff. Inflammatory change, steroid medication, aging, reduced vascularization close to the insertion site, and inadequate loading may contribute. In histology, findings of degeneration and of inflammation are both described, and according to the prevalence of one component or the other the terms tendinosis or tendinitis are applied. On MRI, an intact tendon is of low signal intensity in all sequences and exhibits smooth contours. Predominantly

Fig. 8.7.26a,b  58-year-old patient with tear of the supraspinous tendon. a Oblique coronal T1-weighted sequence. There is a diffuse signal intensity increase at the insertion site of the supraspinous tendon. b Oblique coronal T2-weighted sequence. There is fluid-like signal intensity adjacent to the major tubercle (straight arrow). Thickening and slight signal increase on the T1weighted and T2-weighted sequences in the more medial portion of the supraspinous tendon (curved arrow) indicate chronic degenerative change

degenerative changes lead to increased signal intensity in T1-weighted and PD-weighted sequences, with only slightly irregular contours. In predominantly inflammatory changes, there is high signal intensity in T2-weighted or STIR sequences as well. Thickening of the tendon may be observed in both cases. Calcifications show low signal intensity in all sequences and are quite prominent in T2*weighted sequences. Interpretation of signal intensity increase in the dis-

8.7  Imaging of Internal Joint Derangement

Fig. 8.7.27a–c  Complete supraspinous tear. a Oblique coronal PD FS sequence and b oblique sagittal T2-weighted sequence exhibiting discontinuity of the tendon with fluid-like signal intensity (↑) representing a complete tear of the supraspinous ten-

don with retraction of the tendon. c Indirect MR-arthrography, T1-weighted SE sequence in another patient: partial, joint-sided tear (↑) of the supraspinous tendon close to the insertion site

Table 8.7.5  Relevant findings to report in rotator cuffs lesions

painful. Acute inflammation in bursae is characterized by fluid filling the lumen, prominent in T2-weighted FS sequences. In T1-weighted sequences, the peribursal fat is displaced laterally and inferiorly (depending on the amount of fluid present, e.g., if there is a tear-drop-like appearance in rheumatoid arthritis [RA]). Inflammation and fibrosis may lead to obliteration of the peribursal fat, visible in T1-weighted sequences. Such findings are nonspecific but direct the attention of the reader to search for associated pathologies. In chronic bursitis, fat may be present on the inner aspect of the bursa as well, causing a doubling of the high-signal-intensity line delineating the bursa in T1-weighted sequences.

Localization of lesion

- Insertion site - Intratendinous - Myotendinous junction

Type of lesion

- Partial - Complete

Affected muscles Size of lesion

In centimeters, in two planes

Extension of retraction

In centimeters

Extension of fatty degeneration

In percentage of muscle cross section

Status of long biceps tendon

- - - -

Intact Degenerative Torn Dislocated

Osseous changes

- - - -

Secondary degenerative changes Acromion AC joint Calcifications

tal portion of the supraspinous tendon, approximately 1 cm medial to its insertion site may be difficult when short TE sequences (up to roughly 30–40 ms) are used. In such cases, it is not possible to differentiate between real pathology and artifactual signal increase due to the magic angle effect. Elimination of this area of increased signal intensity in T2-weighted sequences (TE > 50–60 ms) or after a change in arm position (e.g., ABER [abduction–external rotation] position) and lack of swelling and contour irregularities is helpful. Often, there is associated bursitis, which may be very

8.7.3.1.2  Rotator Cuff Tear Usually, the terms “rupture,” “defect,” and “tear” are used simultaneously without implying a specific etiology. A purely traumatic tear of the supraspinous muscle without preexisting degenerative change is rare, requires a severe trauma, and presents with extensive hematoma, (hemorrhagic) effusion and eventually bone bruising at the greater tuberosity. A tear of the subscapularis muscle may more often be of traumatic origin. Associated not infrequently with a concomitant tear of the supraspinous muscle, it must not be missed. Partial tears present a circumscribed defect that does not involve the complete cross-sectional diameter of the tendon/muscle. Localization may be bursa sided, joint sided, or intratendinous. The defect exhibits high, fluid-like signal intensity in T2-weighted/STIR sequences (Singson et al. 1995; Sonin et al. 1996). Sometimes the differentiation between small partial tears and degeneration is difficult due to partial volume averaging and because usually there is concomitant tendon degeneration around the (partial) tear. MR arthrography may be valuable in such cases (Tirman et al. 1944). In our institution,

1135

1136

8  Musculoskeletal System

when using indirect MR-arthrography, immediate postinjection data are acquired as well, because they help to differentiate between vascularity-related and diffusionrelated enhancement. Complete tears by definition are full thickness tears irrespective of the cross-sectional extent of the defect. They create a communication between the joint cavity and the subacromial–subdeltoid bursa (Farley et al. 1992; Reimus et al. 1955; Reiser et al. 1986), clearly visible with joint distension by effusion (or direct MR arthrography). In extensive lesions, even a communication with the ACJ may be evident. In recent tears, edema and effusion may be prominent. However, in more chronic cases when there is a delay of several months before MRI, they may be absent and the former defect may be masked by (more or less strongly vascularized) scar tissue. Muscle belly retraction and fatty degeneration as well as atrophy are associated findings of chronic rotator cuff tears (see Fig. 8.7.29). Operative therapy aims at repairing the defect (suture, reinsertion) and at improving the space available for movement (acromioplasty, lateral resection of the clavicle, resection of osteophytes). Outcome, technique (arthroscopic, open), and approach (size, location) depend on the size and the localization of the defect, the extent of muscle retraction and atrophy and on the accompanying osseous change. An efficient MR imaging sign to evaluate the degree of fatty atrophy of the supraspinous muscle belly is the so-called tangent sign. It refers to drawing a tangent from the tip of the base of the coracoid to the most cranial aspect of the spina scapulae in a central sagittal section (Fig. 8.7.28). Normally, this tangent should intersect with the muscle belly. It is important to report on the long biceps tendon, because it acts as a (secondary) depressor of the humeral head. Table 8.7.6 summarizes relevant MRI findings that should be addressed in the report (Morag 2006). 8.7.3.1.3  Lesions of the Biceps Tendon Tendinitis, rupture. With effusion in the glenohumeral joint, often, there is a considerable amount of fluid in the tendon sheath of the long biceps tendon. A small amount of fluid is not necessarily to be described as pathologic. In tendinitis, effusion may be accompanied by swelling, splitting, and increased signal intensity in T2-weighted sequences. A rupture of the biceps tendon with consecutive retraction of the fragments may cause the appearance of an “empty sulcus.” The tendon is then visible in more caudal sections only. The biceps tendon is secured in its sulcus by the transverse ligament, fibers continuing from the superficial portions of the subscapular muscle, thus bridging the sulcus from the lesser tuberosity to the greater tuberosity. Fibers of the subscapular tendon in part reach the in-

Fig. 8.7.28a,b  Atrophy of the supraspinous muscle in a patient after complete tear of the supraspinous tendon. a Oblique sagittal T1-weighted SE image and b oblique coronal T1-weighted SE image with severe fatty degeneration and atrophy of the supraspinous muscle. Positive tangent sign: the tangent (*) from coracoid to spina scapulae does not intersect with the supraspinous muscle belly indicating severe atrophy of the muscle (a)

ner aspect of the medial wall of the intertubercular sulcus. Therefore, in case of severe trauma with rupture of the transverse ligament and/or the subscapular tendon the biceps tendon may dislocate medially (Figs. 8.7.28, 8.7.29). More often, this will be articular-sided and may

8.7  Imaging of Internal Joint Derangement Table 8.7.6  Lesions of the joint capsule and the glenoid labrum Type/acronym

Characterization

Bankart lesion

- Most commonly anteroinferior capsulolabral lesion - Complete detachment of the labrum with detachment of the capsule and the periosteum - In some cases: avulsed bony fragment

APLSA lesion

Anterior labroligamentous periosteal sleeve avulsion

- Labrum is completely detached - Capsule and periost remain continuous, but avulsed from bone - Complex is dislocated inferomedially

GLAD lesion

Glenolabral articular disruption

- Rare - Tear at labral basis (no complete detachment) with lesion to hyaline cartilage

Bennet lesion HAGL lesion

- Often seen in throwers - Posteroinferior labral lesion Humeral avulsion glenohumeral ligament

- Humeral Sided Rupture Of The Glenohumeral ligaments with   (HAGL-B) or without avulsion of a bony fragment - Less common than glenoid sided ligament lesions

Fig. 8.7.29a,b  Subscapular tear and dislocation of the biceps tendon. Coronal T1-weighted (a) and T2-weighted (b) sequences. Medial displacement of the biceps tendon is appreciated in the coronal plane. More commonly, there is displacement underneath the subscapular muscle and tendon, but medial displacement superficial to the muscle and tendon may be seen as well. In any case, medial displacement of the biceps tendon is indicative of subscapular muscle or tendon tear

hinder reposition of the humeral head in case of dislocation. Sometimes the displaced tendon will be positioned superficially to the subscapular tendon/muscle. 8.7.3.2  Capsulolabral Tears A lesion to the capsule or the labroligamentous complex results from (1) traumatic dislocation with subsequent glenohumeral (Fig. 8.7.30) instability and recurrent dislocation or (2) is observed in chronic instability. Bony lesions such as the Hill-Sachs lesion in anterior dislocation or the “through” lesion in posterior dislocation as well as traumatic rotator cuff tears usually can be depicted with

high diagnostic confidence. Diagnosing capsulolabral lesions is more challenging but highly relevant to outcome and therapeutic (conservative versus operation) approach. Grading of capsulolabral lesions is performed according to their position relative to the glenoid rim (anterior, posterior, superior, inferior quadrants) and according to their local extension (partial tear, complete avulsion, biceps tendon or inferior/middle glenohumeral ligament involvement) (Fig. 8.7.31). The most frequent location is the anterior-inferior quadrant (two thirds of cases: Bankart, ALPSA, GLAD lesions; compare Fig. 8.7.32 and Table 8.7.6). Findings in the area of the incisura glenoidalis are normal variants in most cases. They are to be dif-

1137

1138

8  Musculoskeletal System

Fig. 8.7.30  Anatomy of the glenohumeral joint capsule. Anterior attachment to the scapula varies from close to the glenoid rim (type I), over insertion at the level of the scapular neck (type II) to an attachment far medially at the body of the scapula (III)

Fig. 8.7.31a,b  Schematic visualization of the labral topography with normal variants and pathologic findings. View from lateral on the glenoid fossa. a Normal findings in the framed field: Usually (left) the labrum is attached to the surface of the bony glenoid. In the area of the incisura glenoidalis (middle), the labrum may not be attached to the bone for a distance of some millimeters, leading to the image of the sublabral foramen in the axial plane. Right Buford complex with partial absence of the superior anterior labrum and (possibly) thickening of the medial glenohumeral ligament. b SLAP lesions type 1–7: longitudinal

ferentiated from lesions located in the superior glenoid rim near the insertion site of the biceps tendon (superior labral tear with anterior or posterior extension [SLAP] lesions), the second most common location. Especially when there is suspicion of a capsulolabral tear, spatial resolution and CNR should be as high as possible. The anterior and posterior portions of the labrum are best depicted in axial sequences; the superior and inferior portions are best depicted in paracoronal sections. Parasagittal sections may help in identifying longitudinal tears and lesions to the glenohumeral ligaments. In view of depiction of the anatomy, it is helpful to orient coronal and sagittal sections as exactly as possible parallel and vertical to the plane of the glenoid rim. The primary aim of MRI is in establishing the diagnosis of a labroligamentous lesion and in directing the orthopedic surgeon to the problematic location rather than performing a perfect grading of the lesion in every case (Fig. 8.7.33; Table 8.7.7). In contrast to native MRI (accuracy 44–67%) the diagnostic accuracy is improved by direct arthrography to up to 90% sensitivity and specificity (Tirman et al. 1993). The ABER (arm in abduction and external rotation) position puts tensile stress on the anterior and inferior portions of the capsule facilitating

tear of the labrum, which is either limited to the insertion site of the biceps tendon (type 1), to slightly anterior and/or posterior within the labrum (type 2) or extending far anterior-inferiorly (type 5) or even beyond the labrum to the medial glenohumeral ligament (type 7). The torn labral portion may dislocate into the joint space similar to a bucket handle tear with (type 4) or without (type 3) affection of the (undersurface of the) biceps tendon or may hang into the joint with a free end similar to a flap (type 6)

8.7  Imaging of Internal Joint Derangement

Fig. 8.7.32a–e  Direct MR arthrography. Upper row (a,b) Bankart lesion, lower row (c–e) SLAP type 3 lesion. a,b Oblique axial, moderately T2-weighted FS TSE sequences. There is complete detachment of the labrum (*) in combination with periosteal stripping (↑) (c–e) Oblique coronal moderately T2-weighted FS TSE sequence. The labrum is torn at the superior glenoid cir-

cumference with intra-articular dislocation (↑) of the detached portion (comparable to a bucket-handle tear), consistent with a SLAP type 3 lesion. Please note that the fluid collection between the two portions of the torn labrum points superolaterally in SLAP (types 2–4) lesions (in contrast to the sublabral recess indicated by fluid pointing superomedially)

Fig. 8.7.33a,b  Shoulder dislocation. Inferomedial dislocation resulting in post-contusional BMEP at the posterosuperolateral aspect of the humeral head with impression fracture (Hill-Sachs lesion) and anteroinferior labral tear (Bankart lesion). Low-magnetic-field-strength MRI (0.2 T). T1-weighted (a) and STIR (b) transverse sections show the impression fracture (arrows). Please note the subtle linear low signal intensity abnormalities adjacent

to the fracture presumably representing additional fracture lines and trabecular condensation (arrowheads). The anterior labrum appears fragmented and there is stripping of the capsule from the anterior aspect of the scapula (b triangles). The intermediate signal intensity of the effusion indicates its hemorrhagic nature (a)

1139

1140

8  Musculoskeletal System Table 8.7.7  SLAP (superior labral tear with anterior or posterior extension) lesions are injuries of the superior glenoid circumference and the insertion of the biceps tendon. According to their extension, SLAP lesions are divided into four to seven subtypes

References 1.

Type

Finding(s)

2.

1

Tear is limited to the insertion site of the biceps tendon, doubtful clinical significance

3.

2

Longitudinal anterior or posterior extension of the tear beyond the bicipital insertion, parallel to the glenoid circumference

4.

3

Intra-articular dislocation of the detached labrum (cf. type 2), comparable to a bucket handle tear

5.

4

Cf. type 2 with additional tear of the long biceps tendon with additional dislocation of (portions of) the long biceps tendon into the joint space

6.

5

Longitudinal tear reaching from superior over anterior to inferior

7.

6

Longitudinal tear with a loose end floating in the joint space (“flap tear”)

8.

7

Longitudinal tear reaching from superior to anterior with extension onto the middle glenohumeral ligament

9.

10. Table 8.7.8  Lesions of the glenoid labrum

11.

Diffuse SI increase Linear SI increase

12.

Detachment of a part of the labrum Rounding and increased mobility of the labrum

diagnosis. Especially in older lesions (most patients do not consult the physician immediately after trauma but only after months of incomplete recovery) when (partial) scarring has occurred, the ABER position may help in lesion detection. External rotation can be advantageous for the detection of SLAP lesions. Cyst-like lesions of the labrum usually are associated with labral tears (Table 8.7.8). They may occur within or adjacent to the labrum and may lead to labral thickening or may show a small connection to a concomitant labral tear. Typically, they appear lobulated with thin septa and exhibit high signal intensity in T2-weighted sequences. When localized in close vicinity to the supraspinous fossa or the incisura scapulae they can cause fatty atrophy of the supra- and infraspinous muscles due to compression of branches of the suprascapular nerve.

13.

14.

15.

16.

17.

18.

Bencardino JT, Beltran J (2006) MR imaging of the glenohumeral ligaments. Radiol Clin North Am 44:489–502, vii. Review Bigoni BJ, Chung CB (2006) MR imaging of the rotator cuff interval. Radiol Clin North Am 44:525–536, viii. Review Farley TE, Neumann CH, Steinbach LS (1992) Full-thickness tear of the rotator cuff of the shoulder: diagnosis with MR imaging. AJR 158:347–351 Flannigan B, Kursunoglu-Brahme S, Snyder S, Kar-zel R, Del Pizzo W, Resnick D (1990) MR-arthrography of the shoulder. AJR 155:829–832 Gagey N, Ravaud E, Lassau JP (1993) Anatomy of the acromial arch: correlation of anatomy and Magnetic Resonance imaging. Surg Radiol Anat 15:63–70 Jbara M, Chen Q, Marten P, Morcos M, Beltran J. Shoulder MR arthrography: how, why, when. Radiol Clin North Am 43:683–692, viii. Review Morag Y, Jacobson JA, Miller B, De Maeseneer M, Girish G, Jamadar D. MR imaging of rotator cuff injury: what the clinician needs to know. Radiographics :1045–1065. Review Morrison DS, Bigliani LU (1987) The clinical significance of variations in acromial morphology. Orthop Trans 11:234–244 Neer CS (1972) Anterior acromioplasty for the chronic impingement syndrome in the shoulder. J Bone Joint Surg Am 54:41 Neer CS (1982) Impingement lesions. Clin Orthop 173:70–77 Palmer WE, Brown JH, Rosenthal DI (1993) Rotator cuff: evaluation with fat-suppressed MR arthrography. Radiology 188:683–687 Palmer WE, Brown JH, Rosenthal DI (1994) Labral-ligamentous complex of the shoulder: evaluation with MR-arthrography. Radiology 190:645–651 Rafii M, Firooznia H, Sherman O et al (1990) Rotator cuff lesions: signal patterns at MR imaging. Radiology 177:817–823 Reimus WR, Shady KL, Mirowitz SA, Totty WG (1995) MR diagnosis of rotator cuff tears of the shoulder: value of using T2-weighted fat-saturated images. AJR 164:1451–1455 Reiser M, Kahn T, Rupp N, Allgayer B (1986) Er-gebnisse der MR-Tomographie in der Diagnostik der Osteomyelitis und Arthritis. RoFo 145/6:661–666 Seeger LL, Gold RH, Bassett LW, Ellmann H (1988) Shoulder impingement syndrome: MR findings in 53 shoulders. AJR 150:343–347 Singson RD, Hoang T, Dan S, Friedman M (1995) MR evaluation of rotator cuff pathology using T2-weighted fast spin-echo technique with and without fat suppression. AJR 166:1061–1065 Sonin AH, Peduto AJ, Fitzgerald CM, Callahan CM, Bresler ME (1996) MR imaging of the rotator cuff mechanism: comparison of spin-echo and turbo spin-echo sequences. AJR 167:333–338

8.7  Imaging of Internal Joint Derangement 19. Tirman PFJ, Stauffer AE, Crues JV et al. (1993) Saline magnetic resonance arthrography in the evaluation of glenohumeral instability. Arthroscopy 9:550–559 20. Tirman PFJ, Bost FW, Steinbach LS et al. (1994) MR arthrographic depiction of tear/ of the rotator cuff: benefit of abduction and external rotation of the arm. Radiology 192:851–856 21. Vahlensieck M, Majumdar S, Lang P, Genant HK (1992) Shoulder MRI: routine examinations using gradient recalled and fat-saturated sequences. Eur J Radiol 2142–147

changes under therapy. Small wrap-around coils may be advantageous for imaging the fingers. Direct MR arthrography is very helpful for the assessment of ligamentous and triangular fibrocartilage complex (TFCC) lesions (Schmitt et al. 2003; Cerezal et al. 2005). 8.7.4.1  Intrinsic Ligaments: Scapholunate and Lunotriquetral Ligaments

MRI of the wrist has become an important indication of MRI. Coronal STIR/T2-weighted FS sequences are the foundation of the MRI protocol of the wrist. In addition, a T1-weighted sequence in the same plane should be performed (Fig. 8.7.34). In principle, the wrist should be imaged in all three orthogonal planes, just like any other joint. Due to the small anatomic structures to be depicted, (3D) gradient-echo sequences may be advantageous in view of obtaining the best possible spatial resolution at sufficient SNR. Dedicated phased-array coils are extremely helpful, either as dedicated wrist coils or as small-extremity coils able to cover both hand and wrist together. Using such coils also facilitates rapid imaging necessary to follow the Gd enhancement dynamics in inflammatory joint diseases, in which they are needed to establish an early diagnosis and to evaluate potential

The current concept to approach an understanding of the mechanical functioning of the wrist consists in recognizing the distal carpal row on the one hand, and the distal radius with the TFCC on the other hand as functional carpal segments. Between these two, the proximal carpal row is interposed as an additional, intercalated segment mediating the movement between forearm and hand. In all segments, the carpal bones are interconnected by (intrinsic) ligaments allowing only for a small range of well coordinated movement between the individual bones. Disturbance of the intercalated segment’s kinematics is key to carpal instability, which usually can be attributed to ligamentous damage because no muscles or tendons insert on the intercalated segment. Clinically, lesions to the scapholunate (SL) or lunotriquetral (LT) ligaments account for the vast majority of carpal instabilities and the consecutive dissociation of the formerly interconnected proximal row’s bones facilitates proximal migration of the distal fixed unit (Miller 2001). Questions to be answered are whether there is disruption of the dorsal or palmar strands of the SL or LT ligaments, whether there

Fig. 8.7.34a,b  Fracture of the scaphoid with subsequent osteonecrosis. a Coronal T1-weighted sequence with hypointense signal of most of the scaphoid. b Coronal PD FS sequence with ex-

tensive bone marrow edema. Note the low-signal-intensity line perpendicular to the long axis of the scaphoid in its midportion representing the fracture line in a typical location

8.7.4  Wrist

1141

1142

8  Musculoskeletal System

Fig. 8.7.35a,b  Direct MR arthrogram of the wrist. a,b Partial tear of the scapholunate ligament exhibiting a swollen aspect, with marked signal inhomogeneities in coronal and axial sections. Note, the palmar strand of the ligament (b) as well as the

lunotriquetral ligament (a,b) are intact. Axial, moderately T2weighted FS TSE in another patient (c, for direct comparison) shows intact scapholunate and lunotriquetral ligaments

native MRI, the LT ligament sometimes cannot be visualized because of its small dimensions and high anatomic variability. In addition to these direct signs of intrinsic ligament lesions, indirect signs such as fluid collections, the dissociation, and the malrotation (DISI, PISI) of the affected bones with respect to the carpus. 8.7.4.2  Triangular Fibrocartilage Complex

Fig. 8.7.36  TFCC lesion. Direct MR arthrogram, no recent wrist trauma. 3D DESS FS sequence. There is a subtle split in the radial aspect of the TFC consistent with a full-thickness defect (Palmer Ia) in the disk. Please note the chondromalacia of the corresponding surface of the lunate with underlying intraosseous ganglion formation. The (dorsal portion of the) scapholunate ligament is intact

are stumps of ligament on the lunate allowing for repair, and whether the cartilage in the radiocarpal and intercarpal articulations is preserved. Care should be taken not to overcall small perforations in the central portions of these ligaments as lesions, as such perforations become frequent with increasing age and are of no mechanical significance. Both ligaments can be identified as low-signal-intensity structures (Smitz and Snearly 1994) at the proximal aspect of the proximal row bones. SI increase (Fig. 8.7.35) and disruption indicate lesions (Romiger et al. 1993). In

Functionally, the TFCC stabilizes the rotation of distal radius and ulna around each other, and it serves as a cushion between the ulnar head and the carpal bones. It consists of the biconcave fibrocartilaginous disk (TFC) itself (Totterman and Miller 1995), the meniscus homologue, the ulnar collateral ligament, the extensor carpi ulnaris tendon and the posterior and volar radioulnar ligaments. The TFC attaches to the lunate fossa of the radius and, with a variable extent, to the distal ulna (from the fovea up to the tip of the styloid). Within the TFCC it is connected to the ulnar-sided extrinsic posterior and volar ligaments. Lesions of the TFCC are commonly classified according to Palmer (1989) by their traumatic (Palmer I) or chronic (Palmer II) nature and their location within the TFC and their extension to or involvement of other structures (Fig. 8.7.36). Traumatic ruptures frequently are located in the radial aspect, whereas degenerative fenestration may often be encountered in the central portions of the TFC, increases in frequency with age and is not infrequently asymptomatic. Signal intensity is increased in TFCC lesions and fluid may communicate through them between the distal radioulnar and the radiocarpal joints (Zlatkin 2006). In contrast to these asymptomatic defects, Zanetti et al. (2000) describe non-communicating proximal-sided TFCC defects as coinciding more often with ulnar-sided wrist pain.

8.7  Imaging of Internal Joint Derangement Fig. 8.7.37a,b  Ulnocarpal impaction syndrome. Persistent pain 20 months after distal radial fracture. Coronal T1-weighted (a) and STIR (b) sequences. The former fracture appears consolidated but there is marked effusion in the distal radioulnar and radiocarpal joint as well as BMEP in the ulnar aspect of the lunate and the adjacent (proximal) portion of the triquetrum. There is subtle BMEP in the distal ulna as well. Tear of the TFC in its radial portion. Please note the avulsed ulnar styloid

8.7.4.3  Ulnocarpal Impaction

8.7.4.4  Ganglia

Ulnocarpal impaction is characterized by excessive ulnarsided load bearing, transmitted by the TFCC from the distal ulna to the ulnar aspect of the lunate and the radial aspect of the triquetrum. Usually there is a (congenital or acquired) increased length of the ulna with respect to the radius. BMEP, subchondral cyst-like formations and sclerosis in the proximal aspect of the lunate and/or triquetrum, altered signal and/or reduced thickness in the ulnar, and carpal bone cartilage as well as degenerative lesions to the TFCC may be seen (Fig. 8.7.37). Usually, shortening of the ulna is the therapeutic approach. This condition is differentiated from ulnar styloid impaction on the triquetrum and from the impingement of a short ulna on the distal portion of the radius.

Ganglia often can be seen in the wrist and may explain symptoms by their local mass effect. They exhibit a jelly-like consistency and arise from joints, ligaments, or bone. Often they are found arising from the SL ligament. Ganglia have low SI in T1-weighted and high SI in T2-weighted sequences and not infrequently show septations. Sometimes a indication to their origin can be depicted (Fig. 8.7.38). 8.7.4.5  Nerve Compression Syndromes: Carpal Tunnel Syndrome and Loge de Guyon Median or ulnar nerve compression syndromes are characterized by pain and paresthesia, weakness, and atrophy in the radial two-and-a-half fingers and thenar muscles

1143

1144

8  Musculoskeletal System

(median nerve) or in the hypothenar muscles and fourth to fifth fingers (ulnar nerve). The median nerve runs superficial to the flexor tendons underneath the flexor retinaculum in the carpal tunnel. When compressed it may exhibit increased SI in T2-weighted sequences and increased volume, and it may appear more flattened than usually. Palmar bowing of the flexor retinaculum may be best identified at the level of the hamate bone. In chronic disease, fibrosis may develop leading to low SI in all sequences and occasionally muscle atrophy can be observed. Usually clinical diagnosis is based on typical symptoms and electrophysiological findings. MRI should be used in clinically equivocal cases, to identify the precise cause of the syndrome, to ascertain the position of the median nerve prior to operation and to assess recurrent postoperative disease. Typical underlying diseases are tumors, ganglia, lumbrical muscle hy-

pertrophy, (teno)synovial proliferations in rheumatoid disease, or degenerative change. Carpal dislocations, displaced bony fragments, or abundant callus from fracture healing may be causative as well. Scarring, neuroma (in iatrogenic nerve lesions) and incomplete retinaculum dissection may be responsible for recurrent disease. Guyon’s canal is a fascial tunnel into which the ulnar nerve accompanied by the ulnar artery enters on its course from the forearm to the hand. It is delineated by the pisiform proximally and medially, the hook of the hamate as well as the hypothenar laterally and distally, the flexor retinaculum dorsally and the palmar aponeurosis ventrally. Compression can be induced by ganglia, lipoma, anomalous muscles, aneurysm of the ulnar artery, pisotriquetral OA and fracture of the hook of the hamate. Visualization of the nerve is facilitated by the abundant fat surrounding it.

Fig. 8.7.38a–e  Ganglion on the posteroradial aspect of the wrist arising from underneath the common extensor tendon with broad contact to the wrist joint capsule leading to extensive surgical revision. The axial sections help to appreciate the displacement of the extensor tendons (a T1-weighted SE, b T2weighted TSE, e T1-weighted FLASH FS). Coronal moderately

T2-weighted FS TSE shows the typical lobulated aspect of this well-demarcated lesion of fluid-like signal intensity. The sagittal T1-weighted FLASH FS image illustrates the close relationship with the dorsal aspect of the capsule (d). Please note the median nerve in the carpal tunnel and Guyon’s canal with the branches of the ulnar nerve

8.7  Imaging of Internal Joint Derangement

8.7.4.6  Tendinopathy In the wrist the extensor tendons topographically are located in six compartments (Fig. 8.7.39). They contain (from radial to ulnar) the abductor pollicis longus and extensor pollicis brevis tendons (I), the extensor carpi radialis longus et brevis tendons (II), medial to the Lister’s tubercle the extensor pollicis longus tendon (III), followed by the extensor indicis and digitorum tendons (IV) and the extensor digiti minimi tendon (V), and lastly the extensor carpi ulnaris tendon in its small groove medial to the ulnar styloid (VI). The flexor digitorum prof. et superf. and the flexor pollicis longus tendon are grouped together in the carpal tunnel between the palmar aspect of the carpal bones and the flexor retinaculum. Tendinopathy is characterized by fluid in the tendon sheaths

Fig. 8.7.39a–d  De Quervain’s chronic stenosing tendinitis. Transverse T1-weighted (a) and T2-weighted (b) images at the level of the distal radioulnar joint and T1-weighted coronal images prior to (c) and after i.v. Gd (d). There is marked soft tissue swelling around the extensor and abductor pollicis ten-

(tenosynovitis), thickening of the tendon with or without increased tendinous signal and eventually partial or full thickness tear. De Quervain’s chronic stenosing tendinitis of the first extensor compartment at the level of the radial styloid as well as extensor carpi ulnaris tenosynovitis with or without subluxation out of its shallow groove at the base of the ulnar styloid can be confidently diagnosed (Fig. 8.7.39). References 1.

Cerezal L, Abascal F, Garcia-Valtuille R, Del Pinal F (2005) Wrist MR arthrography: how, why, when. Radiol Clin North Am 43:709–731, viii. Review

don showing intermediate signal intensity in the T1-weighted sequence (a,c black arrows), increased signal intensity in the T2weighted sequence and strong Gd uptake (d open arrows). The tendon itself is showing low signal intensity (arrowheads a,c)

1145

1146

8  Musculoskeletal System 2.

3. 4.

5.

6.

7.

8.

9.

Miller RJ (2001) Wrist MRI and carpal instability: what the surgeon needs to know, and the case for dynamic imaging. Semin Musculoskelet Radiol 5:235–240. Review Palmer AK (1989) Triangular fibrocartilage complex lesions: a classification. J Hand Surg 14:594 Romiger MB, Bernreater WK, Kenney PJ, Lee DH (1993) MR imaging of anatomy and tears of wrist ligaments. Radiographics 13:1233–1246 Schmitt R, Christopoulos G, Meier R, Coblenz G, Frohner S, Lanz U, Krimmer H (2003) [Direct MR arthrography of the wrist in comparison with arthroscopy: a prospective study on 125 patients.] RoFo 175:911–919 Smitz DK, Snearly WN (1994) Lunotriquetral in-terosseous ligaments of the wrist: MR appearances in asymptomatic volunteers and wrists. Radiology 191:199–202 Totterman SMS, Miller RJ (1995) Triangular fibro-cartilage complex: normal appearance on coronal three-dimensional gradient-recalled-echo MR images. Radiology 195:521–527 Zanetti M, Linkous MD, Gilula LA, Hodler J (2000) Characteristics of triangular fibrocartilage defects in symptomatic and contralateral asymptomatic wrists. Radiology 216:840–845 Zlatkin MB, Rosner J (2006) MR imaging of ligaments and triangular fibrocartilage complex of the wrist. Radiol Clin North Am 44:595–623, ix. Review

8.7.5  Temporomandibular Joint MRI of the temporomandibular joint is performed to assess diskoligamentous pathology. The exam is performed both with the mouth closed in usual occlusion and with the mouth opened (approximately 30-mm wide). Silicon splints are useful to achieve a constant width of mouth

opening. Surface coils with a diameter between 6 and 12 cm should be used, preferentially in a design allowing for simultaneous bilateral examination. The exact positioning of the high-resolution parasagittal and paracoronal sequences is accomplished on axial sections. The parasagittal sections are oriented perpendicular to the transverse axis of the temporomandibular joint, the paracoronal section parallel to this axis. Usually 2D or 3D GRE sequences are applied providing good spatial resolution and SNR. The articular disk subdivides the temporomandibular joint in a superior meniscotemporal and an inferior meniscocondylar compartment. The disk itself shows a biconcave shape with thickened anterior and posterior portions and a thinner intermediate portion. Continuous with the posterior portion is the bilaminar area. Disk displacement can be followed by MRI to reasonable detail. In anterior displacement, which is more frequent, complete and partial displacement are differentiated (Fig. 8.7.40). In partial displacement, the posterior portion of the disk is found anterior to the 11 o’ clock position, but contact to the condyle is preserved. This contact is lost in complete displacement. In partial displacement, usually, the disk reduces with opening of the mouth. This, too, is not found in complete displacement. Anterior displacement can be combined with medial or lateral displacement. Posterior disk displacement is rare. Although disk perforation (Fig. 8.7.41) can be depicted with MRI, its diagnostic confidence is lower than for disk displacement. The former can be diagnosed reliably only when there is definite lack of continuity in several consecutive sections. Mostly, in MRI there is anterior displacement of the disk with pronounced intradisk signal heterogeneity and concomitant condylar degeneration.

Fig. 8.7.40a,b  Oblique sagittal FLASH 2D. Closed mouth (a) and opened mouth (b). There is partial anterior disk displacement (a) with the mouth closed. The disk reduces with opening of the mouth (b)

8.7  Imaging of Internal Joint Derangement

8.7.6  Ankle and Foot Similar to the wrist, MRI of the ankle and foot is an established field of musculoskeletal radiology, especially with respect to treatment planning (Rosenberg 2000). Except for the depiction of Morton’s neuromas, imaging of the foot and ankle should be performed with the patient in a supine position, with neutral rotation and good fixation of the foot in the extremity or wrap-around coils. The latter are mandatory for high-resolution imaging of the toes. Any imaging protocol should include STIR/PDweighted FS/T2-weighted FS sequences and T1-weighted sequences, the FOV should not exceed 162 × 162 cm2 for the ankle and 12 × 12 cm2 for the forefoot; the acquisition matrix should not be below 256 × 256. For both, ankle and foot, imaging should be performed in the three major orthogonal planes, for the ankle imaging, in an oblique plane angulated 45° from the coronal plane may be beneficial for the diagnosis of ligamentous and tendon pathologies. Direct MR arthrography may be helpful in the assessment of doubtful capsuloligamentous, especially syndesmotic lesions, posttraumatic impingement and in detecting cartilage abnormalities (Cerezal 2005).

overlaying chondromalacia, chondral fractures, fluid (to be differentiated from reactive granulation tissue: therefore in difficult cases we tend to rely on i.v. Gd) extending between the demarcated element and surrounding bone, especially when there is continuity through the cartilage to the joint space (Fig. 8.7.42). 8.7.6.2  Osteonecrosis A variety of aseptic osteonecroses are described in the foot. They may occur in the navicular bone (Köhler’s disease), the heads of the second and/or third metatarsal bones (Freiburg’s disease), the calcaneal apophysis (Sever’s disease), the base of the fifth metatarsal bone (Iselin’s disease), and the bases of the phalanges (Thiemann’s disease). Hallmarks of diagnosis are signal alterations of the bone marrow in a typical location and a demarcation line towards the unaffected bone. Time course is characterized by BMEP, in the very beginning without, then with a demarcation line, followed by signal loss in all sequences due to sclerosing processes, and last by reconstitution of normal fatty marrow signal indicating successful repair.

8.7.6.1  Osteochondral Lesion of the Talus: OLT

8.7.6.3  Stress Fractures

The term osteochondral lesion of the talus (OLT) comprises both osteochondral fractures due to an acute traumatic event and those lesions ascribed to chronic overuse (repetitive microtrauma) and ischemia. It thus tends to replace the term “osteochondritis dissecans.” Nonetheless, assessment of the stability of the demarcated bony element is mandatory (for classification please refer to Sect. 8.6). Findings indicating potential instability are

Inadequate loading on normal bone or a reduction of bone strength under normal loading can lead to stress reactions, i.e., repetitive (trabecular) microfractures and reduced repair capacity of the affected bone in relation to the ongoing microscopic damage. When this imbalance between bony destruction and bony repair continues, the initial microfractures may progress to open fractures within the trabecular and lastly even within the corti-

Fig. 8.7.41a,b  Oblique sagittal FLASH 2D. Closed mouth (a) and opened mouth (b). Advanced osteoarthritis of the temporomandibular joint with disk perforation showing a broad based defect. Arrows indicate erosions in the mandibular head

1147

1148

8  Musculoskeletal System

Fig. 8.7.42a–e  Osteochondral lesion of the talus (OLT) in the medial talar dome. Coronal STIR (a), T1-weighted SE pre- (b) and post- (c) Gd and coronal and sagittal moderately T2-weighted FS in another patient (d,e). a–c The demarcated area shows Gd enhancement including the demarcation line and therefore is indicative of a potential healing reaction. There is surrounding

edema and some joint effusion indicative of chondromalacia. Note the intact posterior talotibial and talocalcaneal ligaments. In another patient (d,e), there is marked degeneration, flattening of the talus, pronounced chondromalacia and osseous destructions. Effusion is more prominent presumably due to (reactive) synovitis, most notably in the distended posterior recess (e)

cal bone, becoming apparent as stress fractures. Typical locations are the shafts of the central metatarsal bones, the neck of the talus, and the calcaneus. The typical and conspicuous finding is BMEP around the fracture line. The fracture line itself is either of linear low signal intensity (ascribed to compacted trabeculae along the fracture line) or, in case of distension, there may be linear high signal intensity, thought to represent fluid in the distended fracture. Very often, the fracture line is oriented (near to) perpendicular to the adjacent cortex.

8.7.6.4  Transient Edema-Like Bone Marrow Abnormalities An area of BMEP without associated findings of other pathologies (e.g., subchondral necrosis, osteoarthritis, and stress fractures) is characteristic of the so-called transient bone marrow edema-like abnormalities. Please note that the diagnosis of transient bone marrow edema-like abnormalities should only be established after exclusion of such other entities (Zanetti 2002). Reduced weight bear-

8.7  Imaging of Internal Joint Derangement

ing usually leads to resolution of the MRI findings within 3–6 months. 8.7.6.5  Tendon Pathologies Tendon pathologies are common in the foot and ankle. Complete rupture, partial rupture, as well as (mostly concomitant) degenerative and reactive inflammatory change with preserved continuity have to be differentiated. Tendon sheaths (tendovaginitis) as well as the peritendinous soft tissues (peritendinitis) in locations where there are not tendon sheaths may be affected. Rupture is indicated by complete discontinuity in the tendon; often the two fragments are distracted, and the space between them is filled with hematoma and fluid. In partial ruptures, some remaining fiber continuity can be identified. Initially, there is thickening and signal increase of the tendon (Marcus et al. 1989) itself. Tendovaginitis/ peritendinitis is characterized by fluid accumulation and gadolinium uptake in the tendon sheath and surrounding soft tissues, respectively. Yet, some little amount of fluid within the tendon sheaths, especially in the flexor hallucis longus tendon sheath at the level of the ankle, may well be a normal finding and must not be overcalled. Degeneration and rupture of the Achilles tendon mostly occur within the least vascularized area 2–6 cm above its insertion to the calcaneus. Rupture of the tibialis posterior tendon usually is localized at its insertion to the navicular bone. Increase of the cross-sectional area of the tibialis posterior tendon to more than double the size of the two accompanying (flexor digitorum and flexor hallucis longus) tendons is indicative of degeneration. Peroneal tendon pathology is not as frequent as alterations of the Achilles or tibialis posterior tendon. Usually complete rupture (mostly peroneus longus tendon) occurs at the level of the cuboid bone, whereas partial rupture (mostly peroneus brevis tendon) occurs at the level of the tip of the fibula. Underlying subtendinous BMEP may indicate otherwise less conspicuous tendon pathologies (Morrison 2000). 8.7.6.6  Capsuloligamentous Pathologies In more than 75% of cases, the lateral compartment of the ankle is subject to traumatic injury. Usually, the anterior talofibular ligament ruptures first, followed by the calcaneofibular and only rarely the posterior talofibular ligament. Para-axial sections angulated anteroinferiorly by 20° from the axial plane in slight plantar flexion may be helpful for depicting the anterior talofibular ligament in its entire length. Complete rupture can be diagnosed by visualizing complete discontinuity (Fig. 8.7.43); a wavy contour indicates biomechanical insufficiency (Schneck et al. 1992). When there are extensive soft tissue signal

Fig. 8.7.43  Complete tear of the anterior talofibular ligament, for better demonstration in comparison with the contralateral ankle. T2-weighted para-axial section. Note the complete discontinuity in contrast to the straight course of the intact contralateral ligament

alterations, the exact differentiation between partial and complete rupture may become difficult. Chronic tears often manifest as thickened, sometimes thinned, elongated bands with irregular contours. Evaluation of syndesmotic injury is clinically important, but not always a straightforward diagnosis in MRI, especially in chronic injury. Direct MR arthrography may be helpful in such cases. Acute rupture of the anterior tibiofibular ligament usually requires substantial trauma, is accompanied by edema-like alterations and in almost all cases by rupture of the anterior talofibular ligament (Brown 2004). Additionally, a prominent tibiofibular recess in coronal sections and distal tibiofibular joint incongruity may be found. 8.7.6.7  Sinus-Tarsi Syndrome In 70% of cases, sinus tarsi syndrome is due to traumatic ankle sprain. Other causes are inflammatory disease, ganglion cysts, and foot deformities. The anatomy is complex (for review see: Lektrakul 2001) but information on pathology in the sinus tarsi is of value for the clinician. High signal intensity in T2-weighted FS sequences or absence of the high signal of fat in T1-weighted sequences is the diagnostic landmark, usually evident in sagittal sections. The morphologic correlate is chronic inflammation of synovial recesses, scarring or hemorrhage and lesions to the cervical or interosseous ligaments (Erickson et al. 1990; Klein and Spreitzer 1993), although the latter may be difficult to establish.

1149

1150

8  Musculoskeletal System

8.7.6.8  Plantar Fasciitis, Plantar Fibromatosis Plantar fasciitis refers to mixed inflammatory and degenerative changes of the calcaneal aspect of the plantar fascia. MRI depicts signal alterations within the fascia and its immediate surroundings in T2-weighted FS sequences as well as a thickening of the fascia exceeding 5 mm (Steinborn 1999). Often, there is concomitant BMEP in the calcaneus (Fig. 8.7.44). Plantar fibromatosis (Ledderhose’s disease) consists of circumscribed nodular fibrous proliferations within the plantar fascia (Fig. 8.7.45). Most often, they are located in the medial portion of the fascia and exhibit low signal intensity in T2-weighted sequences.

tal nerves (Fig. 8.7.46). It is a common cause of metatarsalgia. Most often, it is located in the second and third metatarsal space. Depending on their content of collagen, they appear as foci of more or less low signal intensity on T1-weighted and T2-weighted images located in the plantar aspect of the intermetatarsal space. Correct evaluation of the proper location and differentiation against intermetatarsal bursitis, stress fractures, synovitis, sesamoid bone necrosis, and infection is essential for treatment (Zanetti 1999) Imaging in the prone position may improve MRI-based diagnosis (Weishaupt 2003).

8.7.6.9  Morton Neuroma Morton neuroma is a perineural fibrosis leading to chronic irritation of the plantar branches of the interdigi-

Fig. 8.7.44a,b  Plantar fasciitis and plantar spur. Sagittal T1weighted (a) and STIR (b) sequences. There is moderate edema around the plantar fascia. Edema in the adjacent portions of the calcaneus, however, is quite pronounced. It is much more conspicuous in the STIR than in the T1-weighted sequence. (Courtesy Dr. M. Steinborn, Pediatric Radiology, Munich-Schwabing)

Fig. 8.7.45a–c  Plantar fibromatosis. Axial T2-weighted (a), axial T1-weighted FS after Gd (c) and sagittal T1-weighted after Gd (b). There is a well-circumscribed nodular mass of heterogeneous, albeit predominantly low signal intensity in the distal portion of the medial fascicle of the plantar aponeurosis (arrows). The lesion shows clear Gd uptake in the FS sequence (c)

8.7  Imaging of Internal Joint Derangement

Fig. 8.7.46a,b  Morton’s neuroma. Axial a T1-weighted image, b T2-weighted image, and c FS T1-weighted image after contrast. There is a small roundish mass between the heads of the third

and fourth metatarsals exhibiting hypointense (T2-weighted) and intermediate (T1-weighted) signal intensity and in this case showing prominent contrast enhancement

References

12. Zanetti M, Strehle JK, Kundert HP, Zollinger H, Hodler J (1989) Morton neuroma: effect of MR imaging findings on diagnostic thinking and therapeutic decisions. Radiology 212: 583–588 13. Zanetti M, Linkous MD, Gilula LA, Hodler J (2000) Characteristics of triangular fibrocartilage defects in symptomatic and contralateral asymptomatic wrists. Radiology 216:840–845

1.

Brown KW, Morrison WB, Schweitzer ME, Parellada JA, Nothnagel H (2004) MRI findings associated with distal tibiofibular syndesmosis injury. AJR Am J Roentgenol 182:131–136 2. Cerezal L, Abascal F, Garcia-Valtuille R, Del Pinal F (2005) Wrist MR arthrography: how, why, when. Radiol Clin North Am 43:709–731, viii. Review 3. Erickson SJ, Quinn SR, Kneeland JB et al (1990) MR imaging of the tarsal tunnel and related spaces: normal and abnormal findings with anatomic correlation. AJR 155:323–328 4. Klein MA, Spreitzer AM (1993) MR imaging of the tarsal sinus and canal: normal anatomy, pathologic findings, and features of the sinus tarsi syndrome. Radiology 186:233–240 5. Lektrakul N, Chung CB, Lai Ym , Theodorou DJ, Yu J, Haghighi P, Trudell D, Resnick D (2001) Tarsal sinus: arthrographic, MR imaging, MR arthrographic, and pathologic findings in cadavers and retrospective study data in patients with sinus tarsi syndrome. Radiology 219:802–810 6. Marcus DS, Reicher MA, Kellerhouse LE (1989) Achilles tendon injuries: the role of MR imaging. J Comput Assist Tomogr 13:480–486 7. Morrison DS, Greenbaum BS, Einhorn A (2000) Shoulder impingement. Orthop Clin North Am 31:285–293 8. Rosenberg ZS, Beltran J, Bencardino JT (2000) From the RSNA Refresher Courses. Radiological Society of North America. MR imaging of the ankle and foot. Radiographics 20(Spec no.):S153–S179 9. Schneck CD, Mesgarzadeh M, Bonakdarpour A, Ross GJ (1992) MR imaging of the most commonly injured ankle ligaments. Radiology 184:499–512 10. Steinborn M, Heuck A, Maier M, Schnarkowski P, Scheid­ ler J, Reiser M (1999) [MRI of plantar fasciitis.] RoFo 170:41–46 11. Weishaupt D, Treiber K, Kundert HP, Zollinger H, Vienne P, Hodler J, Willmann JK, Marincek B, Zanetti M (2003) Morton neuroma: MR imaging in prone, supine, and upright weight-bearing body positions. Radiology 226:849–56

8.7.7  Elbow The elbow consists of three articular components, the humeroulnar, the humeroradial, and the proximal radioulnar joints. Elbow pain and impaired function more often are related to chronic overuse than to acute trauma. After plain films, MRI usually is the procedure of choice for the imaging workup of elbow pain. It demonstrates a variety of diseases ranging from injuries to ligaments and tendons, compressive or entrapment neuropathy, bony injury to inflammatory conditions, and soft-tissue masses. The patient should be positioned supine with the extended arm comfortably at his or her side in neutral position. Correct positioning of the coronal and sagittal scans greatly facilitates the image evaluation and therefore should be carefully performed. Imaging is recommended in all three major orthogonal planes on the base of (moderately) T2-weighted FS sequences. Assessment of the cartilaginous surfaces may profit from high-resolution 3D GRE sequences. Collateral ligaments and epicondyles usually are depicted best in coronal sections; the joint surfaces, biceps, and triceps tendon in sagittal and nerve compression in axial sections (Fowler and Chung 2006). Direct MR arthrography is useful for the evaluation of intra-articular loose bodies and for the evaluation of cartilage defects. However, lack of anatomical resolution in combination with the small thickness of the cartilage in the elbow limits its value to the diagnosis of lesions higher than grade 2 according to Shahriaree (Waldt et al. 2005).

1151

1152

8  Musculoskeletal System

8.7.7.1  Tendon Pathologies Epicondylitis. Insertional tendinopathy at the medial/lateral epicondyles is due to repetitive valgus (medial) or extensor and rotational stress (lateral). Lateral epicondylitis (“tennis elbow,” common extensor tendon) is much more common than is medial epicondylitis (common flexor tendon) (Coel et al. 1993) (Fig. 8.7.47). Insertion tendinopathy manifests as thickening of the insertional portion of the tendon and increased SI in T1-weighted

and PD-weighted (FS) sequences usually without signal increase in T2-weighted sequences. Subsequent to local steroid injection, increased signal may be seen up to 1 month post injection. Partial or complete ruptures result in thinning and more or less pronounced defects in the tendon delineated by fluid like SI in T2-weighted sequences. Biceps tendon. Whereas lesions to the triceps tendon tend to be rare, tendinopathy of the biceps tendon is more common. Running superficial to the brachialis tendon, it

Fig. 8.7.47a–e  Lateral epicondylitis. Coronal T1-weighted SE pre- (a) and post- (b) Gd, and coronal T2-weighted SE (c). Axial STIR (d) and T1-weighted FS after Gd (e). There is thickening and signal intensity increase in the common extensor tendon, most pronounced in the FS sequences in the proximal portions of the common extensor tendon. Gd uptake is due to reactive fibrovascular proliferations

8.7  Imaging of Internal Joint Derangement

inserts on the radial tuberosity approximately 3 cm distal to the articular surface of the radial head. A complete tear usually results from tendon failure in a region of mucoid degeneration. Sometimes this chronic process (Fig. 8.7.48) with repetitive episodes of reactive inflammatory change may lead to large, cyst-like structures, causing a pronounced mass effect. Complete rupture with retraction can be demonstrated on sagittal images and by the lack of the tendon in its normal anatomic position on axial sections, which should be acquired from the myotendinous junction through to the distal aspect of the radial tuberosity. Acute tear of the biceps tendon results from forced extension of the forearm and is more often located at the radial tuberosity than at the myotendinous junction. 8.7.7.2  Collateral Ligaments

Fig. 8.7.48a–c  Chronic partial tear of the distal biceps tendon. Axial T1-weighted SE pre- and post- Gd (a,c) and axial T2weighted SE (b). Thinning of the tendon (arrow) can be appreciated. Please note the extensive concomitant bursitis (Courtesy Dr. M. Steinborn, Pediatric Radiology, Munich-Schwabing)

Deep to the common flexor/extensor tendons, the collateral ligaments reinforce the joint capsule being tightly interwoven with the capsule. The ulnar (medial) collateral ligament (UCL) protects the elbow against valgus stress. It originates from the inferior aspect of the medial epicondyle distal to the common flexor tendons. It inserts with an anterior bundle (showing a deep and a superficial layer) anteriorly at or up to 3mm distal to the articular margin of the ulnar coronoid, and with a posterior bundle more posteriorly at the posteromedial margin of the trochlear notch/olecranon. The lateral collateral ligament complex (LCLC) limits varus stress. It originates from the inferior aspect of the lateral epicondyle distal to the common extensor tendons. According to their insertions on the forearm, the LCLC consists of a radial collateral ligament (RCL), a lateral ulnar collateral ligament (LUCL), and a third portion, the annular ligament (AL). The latter encircles the radial head originating from the anterior and posterior margins of the radial notch on the ulna acting as the primary stabilizer of the proximal radioulnar joint. The RCL inserts broadly on the annular ligament. The LUCL runs more posteriorly, blending with the annular ligament at the posterolateral margin of the radial head but then continues posteromedially to finally insert on the ulna providing posterolateral stability to the elbow. The two components most important for joint stability are the anterior bundle of the ulnar (medial) collateral ligament anteromedially and the ulnar portion of the lateral collateral ligament posterolaterally. Lesions to the UCL often occur in throwing athletes, while lesions to the LUCL result from a fall on the outstretched arm. Degeneration and tears of the UCL can be associated with lesions of the common extensor tendon. Incomplete tears usually affect the deep layer of the UCL. Direct MR arthrography may be helpful in evaluating lesions to the UCL (Steinbach et al. 2002).

1153

1154

8  Musculoskeletal System

8.7.7.3  Neuropathies The aim of MRI is to exactly depict the location and to narrow down the differential diagnosis of the underlying pathology of nerve compression, such as ganglia, lipoma, hematoma, inflammatory disease, osteophytes, ectopic calcification, other tumors, and accessory muscles. Among the ulnar, radial, and median nerves, the ulnar nerve is most commonly subject to injury at the elbow. This is due to its location posterior to the medial humeral epicondyle, where it runs through a fibro-osseous tunnel clearly depicted in axial MR images. Lack of the arcuate ligament may allow for anterior subluxation of the ulnar nerve during flexion, with consecutive friction against the medial epicondyle. MRI shows thickening of the nerve and edema may be seen in T2-weighted FS sequences. Compression of the median and/or radial nerves usually is due to anatomic variants, bursae, or tumors. Associated with neuropathy there may be edema in the early time course (within 1 month), fatty degeneration and atrophy (when there is ongoing neuropathy for several months) of the dependent muscles. 8.7.7.4  Osteochondritis Dissecans and Osteonecrosis Associated with repetitive microtrauma due to valgus stress, osteochondritis dissecans (OD) usually occurs between the ages of 13 and 16 years. The typical location is the capitulum humeri or the radial head. Fluid-like signal intensity entering the interface between a demarcated fragment and the underlying bone from the joint space in T2-weighted (FS) sequences is indicative of instability

(Fig. 8.7.49). In contrast to OD, in Panner’s disease, the osteonecrosis of the capitulum humeri predominantly affects the age group between 5 and 10 years. There is fragmentation within the capitulum humeri and sclerosis but no evidence of loosening or fragment detachment is seen. In most cases, the outcome is favorable, with good reconstitution of the articular surface. 8.7.7.5  Synovial Pathologies and Bursitis Synovitis is a frequent finding in the elbow associated with septic arthritis, crystal deposition disease, synovial osteochondromatosis, pigmented villonodular synovitis (PVNS), and neuropathic osteoarthropathy. In case of primary rheumatoid disease, it is usually quite pronounced and synovial cyst formation may occur. Olecranon and biceps tendon bursitis are common. Imaging profits from i.v. gadolinium and fat saturation. 8.7.7.6  Intra-Articular Loose Bodies Intra-articular loose bodies may arise from trauma, osteochondrosis dissecans, synovial chondromatosis, and osteoarthritis, causing locking and painful limitation of the range of motion in the elbow. Concomitant effusion facilitates their diagnosis in T2-weighted FS sequences. Profiting from joint distension, direct arthrography is useful in the dedicated search for such loose bodies. As the joint capsule comprises all three articulations loose bodies may displace into all five recesses (anterior humeral, olecranon, annular and the two recesses of the collateral ligaments attaching to the ulna/radius) of the capsule. Fig. 8.7.49a,b  Osteochondritis dissecans of the capitulum humeri. Coronal (a) and sagittal (b) T2-weighted TSE FS sequences. A small osteochondral fragment has become detached from the underlying bone (arrow). It is displaced as intra-articular loose body in the anterior recess ventrolaterally to the humeral metaphysis. There is prominent effusion distending the capsule with its anterior and posterior recesses (Courtesy Dr. M. Steinborn, Pediatric Radiology, MunichSchwabing)

8.8  Bone and Soft Tissue Tumors

References 1.

2.

3.

4.

5.

Coel M, Yamada CY, Ko J (1993) MR imaging of patients with lateral epicondylitis of the elbow (tennis elbow): importance of increased signal of the anconeus muscle. AJR 161:1019 Fowler KA, Chung CB (2006) Normal MR imaging anatomy of the elbow. Radiol Clin North Am :553–567, viii. Review Steinbach LS, Palmer WE, Schweitzer ME (2002) Special focus session. MR arthrography. Radiographics 22:1223–1246 Steinbach LS, Palmer WE, Schweitzer ME (2002) Special focus session. MR arthrography. Radiographics 22:1223– 1246 Review Waldt S, Bruegel M, Ganter K, Kuhn V, Link TM, Rummeny EJ, Woertler K (2005) Comparison of multislice CT arthrography and MR arthrography for the detection of articular cartilage lesions of the elbow. Eur Radiol 15:784–791

8.8  Bone and Soft Tissue Tumors The number of tumors affecting the skeleton is large (Table 8.8.1) and in many cases, a specific diagnosis is not possible. Therefore, it is the radiologist’s task to select the most likely diagnosis. Conventional radiographs are always the fundamental imaging technique. CT and MRI then provide additional information. When interpreting images additional clinical information is very important. Although radiographic and MR findings do not always allow an exact diagnosis of tumors and tumor-like lesions, they often provide information regarding the nature of the tumor and its aggressiveness. By considering the site of the lesion and the age of the patient (in combination with good clinical information), a reasonable differential diagnosis seems possible in many cases. Many primary bone tumors and tumor-like lesions are observed in children, teenagers, and young adults, while metastases and multiple myeloma predominate in older adults. For soft tissue tumors, such an age dependency is not observed. Although the radiological analysis of the destruction pattern, the tumor matrix, and periosteal reactions frequently allows a specific diagnostic assignment of bone tumors, a histological diagnosis is usually indispensable. Only in some entities, like nonossifying fibroma, enchondroma of the small bones, bone cysts, and fibrous dysplasia, an observation is justified due to the characteristic radiological image. For lipomas, the characteristic density values in CT and the typical signal pattern in MRI, respectively, allow a distinct diagnosis. The other soft tissue tumors usually cannot definitely be determined by imaging findings.

The most important role of MRI in the diagnostic process of bone and soft tissue tumors is the evaluation of the exact tumor size, the visualization of any infiltration of vessels and nerves, the assessment of response to chemotherapy, and the detection or exclusion of tumor recurrence. 8.8.1  Intraosseous Tumor Extension The majority of bone tumors show low signal intensity (SI) in T1-weighted and high SI in T2-weighted spin-echo (SE) images. T1-weighted SE and short inversion recovery (STIR) sequences allow an excellent evaluation of the tumor extension within the bone marrow since they provide high contrast to the normal bone marrow. Also, T2-weighted and contrast-enhanced T1-weighted sequences with fat saturation allow a precise visualization of intramedullary tumor infiltration. In proton density (PD)-weighted images and mildly T2-weighted images as well as after contrast injection without fat saturation, intramedullary tumor masses often appear isointense to the normal bone marrow. To allow an exact evaluation of the longitudinal tumor infiltration into the long bones sagittal and coronal images and, if necessary, oblique images along the longitudinal axis of the particular part of the extremity/body are recommended (Fig. 8.8.1). Osteosarcoma is second in incidence only to multiple myeloma as a primary malignant bone neoplasm. It is usually seen in the second or third decade of life, with men more frequently being affected than are women (2:1). The most typical sites are the long bones of the appendicular skeleton (80%), particularly the femur (40%). A metaphyseal location predominates. MRI is helpful to define the extent of the neoplasm and its relation to surrounding structures of the neurovascular bundles. Large field-of-view T1-weighted imaging of the affected bone is important to identify additional lesions, termed “skip metastases” (Fig. 8.8.2). It is still not known whether skip lesions are metastases or a simultaneous or delayed de novo emergence of the same tumor. Generally, the absence of a joint effusion in MRI has high predictive value for the absence of any joint involvement. With regard to the sensitivity for the detection of intramedullary tumor extension, MRI is superior to conventional radiography and CT, since these procedures imply that there is a destruction of the bone trabeculae. Approximately 30% of bone substance has to be destructed before a loss of bone substance can be recognized in conventional radiography. Merely in the diaphysis, a density increase due to the replacement of fat marrow may be noted on CT. MRI, however, can also show tumor infiltrations of the bone marrow that have not yet led to a destruction of the spongious bone or to endosteal bone destruction so far.

1155

1156

8  Musculoskeletal System Table 8.8.1  Summary of bone tumor entities (according to Resnick et al. 2005) Bone-forming tumors Benign

Malignant

Osteoma, osteoid osteoma, osteoblastoma, ossifying fibroma

Aggressive osteoblastoma, osteosarcoma

Cartilage-forming tumors Benign

Malignant

Enchondroma, enchondromatosis (Ollier’s disease), hereditary multiple exostoses (HME), Maffucci’s syndrome, periosteal chondroma, chondroblastoma, chondroymyxoid fibroma

Chondrosarcoma

Histiocytic or fibrohistiocytic tumors Benign

Malignant

Histiocytoma, giant cell tumor

Malignant fibrous histiocytoma

Tumors of fatty differentiation Benign

Malignant

(Intraosseous) lipoma, liposclerosing myxofibrous tumor of bone

Liposarcoma

Tumors of muscle differentiation Benign

Malignant

Leiomyoma

Leiomyosarcoma

Tumors of vascular differentiation Benign

Malignant

Hemangioma, cystic angiomatosis, lymphangioma, lymphangiomatosis, glomus tumor

Angiosarcoma, hemangiopericytoma

Tumors of neural differentiation Benign

Malignant

Neurofibroma

Malignant peripheral nerve sheath tumor (MPNST)

Tumors of notochord origin Benign

Malignant

Chordoma

Chordoma (locally aggressive)

Tumors of miscellaneous or unknown origin Benign

Malignant

Simple (solitary or unicameral bone cyst), aneurysmal bone cyst, intraosseous ganglion cyst

Adamantinoma, Ewing’s sarcoma (neuroectodermal origin suggested by immunohistochemical and cytogenic studies)

8.8  Bone and Soft Tissue Tumors

Fig. 8.8.1a–d  Osteosarcoma of the left distal femur. a Coronal PD-weighted fat-saturated image showing an ill-defined hyperintense mass of the distal left femur reaching the joint space. b Coronal T1-weighted image with correspondingly hypointense signal. c Coronal T1-weighted image after i.v. Gd, with pe-

ripheral enhancement and central necrosis. d Axial T1-weighted fat-saturated image after i.v. Gd shows that there is no infiltration of the neurovascular bundle; however, reactive synovial enhancement is seen

1157

1158

8  Musculoskeletal System Table 8.8.2  Age dependency of bone tumors and tumor-like lesions Preferred age (years) Bone tumors Ewing’s sarcoma

5–20

Osteosarcoma

10–20

Chondroblastoma

10–20

Osteoid osteoma

10–20

Enchondroma

20–50

Chondrosarcoma

20–40

Osteochondroma

10–50

Giant cell tumor

20–40

8.8.3  Extraosseous Tumor Extension

Tumor-like lesions Solitary bone cyst

The measurement of signal in compact bone is indicative of an infiltration of cortical bone, which may not be notable in radiography and CT. Previously it was assumed that MRI was less suitable for the detection of destructive and infiltrative processes of compact bone than was computed tomography or radiography. With appropriate pulse sequences and coil systems, however, minor infiltrations of cortical bone are also noted that might not be shown by CT and conventional radiography. Periosteal bone reactions may also frequently be shown by MRI, although they are not visualized with the same morphological precision as on radiographs and CT. Calcifications and ossifications lead to focal signal extinctions in T2-weighted SE and T2*-weighted gradient-echo (GRE) sequences. Therefore, sclerotic or periosteal ossifications are clearly noted.

5–20

Nonossifying fibroma

10–20

Aneurysmal bone cyst

10–30

8.8.2  Compact Bone Compact bone and as solid calcifications are represented by the contrast to the bone marrow and to the paraosseous soft tissue structures. However, a direct visualization of calcified bone structures is not given. Streak artifacts as watched on CT in the long bones—particularly in the peripheral skeleton segments—which aggravate the differentiation of periosteal bone reactions, are missing on MRI.

Visualization of extraosseous extension of bone tumors and the extension of soft tissue tumors are considerably better on MRI than on CT (Figs. 8.8.3, 8.8.4). Precise differentiation from muscle is possible with T2-weighted SE, with STIR, or with contrast-enhanced T1-weighted sequences, whereas confident differentiation from fat tissue is achieved with native T1-weighted sequences. Moreover, STIR sequences, fat-saturated T1-weighted, and contrast enhanced T1-weighted sequences have been found valuable for the assessment of any extraosseous tumor infiltration. Since individual muscle groups can be identified and visualized in their anatomical course, MRI allows better differentiation of extra- and intracompartmental tumor spread, which is relevant for surgery planning. A pseudocapsule surrounding the tumor is more often shown

Fig. 8.8.2a,b  Osteosarcoma of the left distal femur of a 75-year-old patient with evidence of skip lesions in the left proximal femur shaft (a) and the left caudal portion of the sacrum (b)

8.8  Bone and Soft Tissue Tumors

by MRI than by CT. This capsule seems to arise from the compression of surrounding muscles and/or the bundling of perifocal vessels. This pseudocapsule has also been found and operatively confirmed in malignant bone and soft tissue tumors; so it does not seem to result from chemical shift or susceptibility artifacts.

MRI is also superior to CT when analyzing infiltration of the neurovascular bundle, which is critically important for surgery planning. Even a discrete SI reduction in the perivascular fatty tissue (T1-weighted image) or a perivascular edema (T2-weighted SE or STIR image) may indicate tumor invasion into the neurovascular bundle.

Fig. 8.8.3a–e  A 22-year-old female with Ewing’s sarcoma in the left distal femur. a,b Radiographs depict an osteolytic lesion in the metaphysis of the left distal femur with moth-eaten cortical destruction and periosteal reactions. c Coronal T1-weighted

image. d,e Coronal spread into surrounding muscles (vastus medialis, intermedius and lateralis muscles) and into the suprapatellar recess. There is contact to the neurovascular bundle without any infiltration

1159

1160

8  Musculoskeletal System

In the case of vessel compression or arteriovenous shunting, flow velocity changes, resulting in a change in SI within the vessel lumen that becomes particularly obvious upon comparison to the healthy side. Also, intraluminal tumor infiltration can be easily detected by MRI

due to isointensity with the other tumor tissue. Moreover, the displacement or occlusion of vessels by the tumor as well as tumor feeders and the pathological vascular architecture are readily depicted using high-resolution MRA techniques (Fig. 8.8.5).

Fig. 8.8.4a–c  Large cell tumor of the distal femur epiphysis and metaphysis. a Lateral radiograph with extensive geographic destruction and cortical distension and thinning. There are trabeculae within the tumor. b Coronal T2-weighted image with fat saturation. High, inhomogeneous SI of the tumor (arrows) with

extensive bone marrow edema of the distal femur. The tumor reaches the joint space (tip of arrow). c Axial T2-weighted image. Inhomogeneous, partially low SI of the tumor with initial infiltration into the joint space (arrow)

Fig. 8.8.5a,b  Osteosarcoma of the lower leg with infiltration of the soleus muscle. a Contrast-enhanced MRA demonstrating increased vascularity of the tumor, with lacunae and cork screw

vessels. b The T1-weighted fat-saturated image after contrast shows a large area of strong, inhomogeneous contrast enhancement and central necrosis

8.8  Bone and Soft Tissue Tumors

However, differentiation of tumor and perifocal edema which both have high SI in T2-weighted and contrast-enhanced images may be problematic. In dynamic contrast-enhanced studies, the SI increase of perifocal edema is slower and reaches a lower maximum than the SI of the tumor itself. However, within in the edematous zone micrometastases are frequently found, which must be included in the resection. 8.8.4  Soft Tissue Tumors In soft tissue tumors, which are frequently malignant, MRI is obviously the method of choice. It is superior to

other modalities such as CT and sonography with regard to sensitivity and the evaluation of the tumor extension. However, findings in most lesions are nonspecific. In some cases, soft tissue tumor(s) may not be easily differentiated from soft tissue infection (Table 8.8.3). Hemangioma is the most frequent soft tissue tumor and the most common tumor in infants and children (Fig. 8.8.6). Malignant fibrous histiocytoma and liposarcoma are the two most frequent malignancies of soft tissue in adults. A variety of synovial disorders producing intra- or periarticular masses may also be summarized under soft tissue tumors or tumor-like lesions classifications. These

Table 8.8.3  Summary of soft tissue tumor entities (according to Resnick et al. 2005) Tumors of fat content Benign

Malignant

Lipoma, lipoblastoma

Liposarcoma

Tumors of fibrous tissue Benign

Malignant

Fibroma, fibromatoses

Fibrosarcoma

Tumors of muscle tissue Benign

Malignant

Leiomyoma, rhabdomyoma, myxoma

Leiomyosarcoma, rhabdomyosarcoma

Fibrohistiocytic tumors Benign

Malignant

Benign fibrous histiocytoma, xanthomatoses

Malignant fibrous histiocytoma,   dermatofibrosarcoma protuberans

Vascular/lymphatic tumors Benign

Malignant

Hemangioma, lymphangioma, glomus tumor

Angiosarcoma

Cartilaginous and osseous tumors (rare) Benign

Malignant

Extraskeletal chondroma, extra-articular chondromatosis, synovial chondromatosis, osteoma

Chondrosarcoma, osteosarcoma

Neurogenic tumors Benign

Malignant

Neurilemoma (schwannoma), neurofibroma,   Morton’s neuroma, fibrolipomatous hamartoma

Malignant peripheral nerve sheath tumor (MPNST)

1161

1162

8  Musculoskeletal System

Fig. 8.8.6a–c  A 31-year-old patient with hemangiomatosis of bones and soft tissues. a Axial T2-weighted image demonstrating cortical thickening, with SI increase of the compact bone of the tibia and fibula. In addition, there is a SI increase within the tibial anterior muscle, parts of the tibial posterior muscle, and

the extensor digitorum longus muscle. b Axial T1-weighted fat saturated image after contrast with prominent contrast of the hemangiomatosis compared with muscle and compact bone. c T2-weighted image with high signal intensity of the hemangiomatosis

processes include inflammatory joint diseases and septic arthritis (which are described elsewhere in the text) and pigmented villonodular synovitis. Pigmented villonodular synovitis (PVNS) (Fig. 8.8.7) typically occurs in adults in the third to fourth decade of life and is seen, less frequently, in adolescents and even children. The most common location is the knee, and monoarticular disease seems to be the rule. In MRI, the presence of hemosiderin deposition (T2-weighted and T2*-weighted images) is consistent with the diagnosis of this synovial proliferative disorder, but it may also be seen in hemophilia and other bleeding disorders. Contrast enhancement after intravenous (i.v.) gadolinium is usually found, but not observed as a rule. The radiographic features are different in cases of tendon sheath involvement, including the nodular tendon sheath disease known as giant cell tumor. These giant cell tumors of tendon sheaths are the most common cause of soft tissue masses in the fingers.

Moreover, similar SI may also be observed in benign lesions such as inflammatory processes. Even after contrast enhancement, a clear differentiation is not possible, however. The SI increase after injection of contrast media is usually more rapid and higher in malignant neoplasms than it is in benign lesions. The rate of enhancement usually correlates with the grade of perfusion, vascularity, and permeability and therefore indirectly with the aggressivity of bone tumors. In malignant neoplasms, tumor inhomogeneities are more frequently observed than in benign tumors. These inhomogeneities are attributed to calcification, edema, bleeding, necrosis, and fatty transformation. However, in some cases benign lesions may also show inhomogeneous signal distribution. After contrast application, differentiation of viable necrotic and regressive tumor portions is definitely easier (Table 8.8.4). A pseudocapsule, more frequently seen on MRI than on CT, cannot be interpreted as a sign that a lesion is benign, since it is also observed in malignant tumors. However, an edematous peritumoral marginal zone provides an important clue. In benign tumors, an edema at the tumor margins is rare, and the zone of edema is narrow, whereas in malignant tumors extensive edema is frequently seen. However, this sign does not apply to all cases, since, for example, eosinophilic granulomas and osteoid osteomas may show extensive perifocal edema. The edematous zone is hyperintense on T2-weighted SE and STIR images as well as on contrast-enhanced T1-weighted images, whereas it often cannot be distinguished from normal muscle on native T1-weighted images.

8.8.5  Evaluation of Tumor Nature Still, radiographs are definitely most important for the evaluation of the destruction pattern (“geographic,” “moth-eaten,” and “permeative” according to the Lodwick classification). The analysis of the destruction pattern permits differentiation between lesions of different levels of aggressiveness. In benign and malignant bone and soft tissue tumors, similar SI are found on native MRI, making differentiation between benign and malignant lesions impossible.

8.8  Bone and Soft Tissue Tumors

Fig. 8.8.7a–d  PVNS in a 12-year-old male patient who reports swelling and pain of the ankle. a Lateral radiograph. Extensive soft tissue swelling of the posterior portion of the ankle. b T1-weighted image and c T1-weighted image after contrast showing synovial proliferations in the posterior portion of the ankle, with hypointense signal and contrast enhancement. d T2*-weighted image. There is signal intensity loss of synovial proliferations due to susceptibility artifacts of hemosiderin Table 8.8.4  SI of bone and soft tissue tumors T1-weighted SE

T2-weighted TSE

T1-weighted CE

Tumor

_

+

+

Necrosis

_

+

_

Fat

+

(+)

+ (FS)

Muscle

_

_

_

Calcification, ossification, cortical structures

_

_

_

TSE turbo spin echo, CE contrast enhanced, + hyperintense, – hypointense, FS fat saturated

1163

1164

8  Musculoskeletal System

8.8.6  Tumor Classification With regard to the assignment of a bone or soft tissue tumor to a specific entity, decisive information can be derived from MRI only in a few cases. The majority of the bone and soft tissue tumors shows low SI in T1-weighted images and high SI in T2-weighted images. Only a few tumors show a different signal behavior (Table 8.8.5), which is then, however, useful for differential diagnosis. Chondroid (cartilage-forming) tumors usually show intermediate SI on T1-weighted and high SI on T2-weighted images. Often chondroid tumors exhibit a

microlobulation and an annular contrast enhancement of septae (Fig. 8.8.8). Enchondromas are benign cartilaginous neoplasms (Fig. 8.8.9). While enchondromas usually have no clinical significance, they may cause complications, such as pathologic fractures and malignant transformation. Enchondromas are mostly solitary benign lesions located centrally in small bones of the hands and feet (50% in the proximal phalanx). Progressive calcification over a period of years is not unusual. Focal loss of calcification suggests malignant degeneration with destruction of the underlying enchondroma by sarcomatous tissue.

Table 8.8.5  Signal behavior of selected bone and soft tissue tumors High SI in T1- images

Macroscopic appearance

Lipoma, liposarcoma

High fat content

Hemangioma

High fat content, blood flow

Giant cell tumor, aneurysmal bone cyst, teleangiectatic osteosarcoma, hemophilic pseudotumor

Subacute bleeding

Solitary bone cyst

High protein content, subacute bleeding

Low SI in T1-weighted images

Macroscopic appearance

Fibrous dysplasia, nonossifying fibroma, fibroblastic osteosarcoma

High content of collagen fibers

Osteoblastic osteosarcoma

Ossifications, calcifications

Giant cell tumor, aneurysmal bone cyst, teleangiectatic osteosarcoma

Hemosiderin deposition

PVNS

Hemosiderin deposition

Fig. 8.8.8a,b  A 53-year-old patient with chondrosarcoma arising from an exostosis of the pubic bone. a Radiograph of the pelvis showing an exostosis with peripheral sclerosis (arrow) and a soft tissue component (arrow) with calcifications adjacent to the

ischial bone. b Axial PD-weighted image with high-signal-intensity tumor with small lobulations. The adjacent muscles are displaced (triangles)

8.8  Bone and Soft Tissue Tumors

When multiple enchondromas coexist, the diagnosis of enchondromatosis should be considered. Multiple enchondromas may occur in three distinct entities: 1. Ollier’s disease is a nonhereditary disorder characterized by multiple enchondromas with a predilection for unilateral distribution. The enchondromas can become large and be disfiguring (Fig. 8.8.10). 2. Maffucci syndrome is nonhereditary and less common than is Ollier’s disease. It consists of multiple hemangiomas in addition to enchondromas.

3. Metachondromatosis consists of multiple enchondromas and osteochondromas and is the only hereditary (autosomal dominant) disorder among these three entities. Bone cysts are characterized by signal intensities paralleling that of fluid. The rim may exhibit a thin layer with contrast enhancement. Usually, the nidus in osteoid osteomas is hyperintense in T2-weighted images and is readily differentiated from the surrounding sclerotic bone. Osteoid osteomas Fig. 8.8.9a–e  Enchondroma of the proximal phalanx of the second finger in a 35-year-old female patient. a,b Anterior–posterior radiograph. There is an oval, well-defined large radiolucency of the proximal phalanx of digit II with thinned cortical bone, endosteal scalloping, and subtle calcifications within the osteolysis. In addition, there are multiple small circular sclerotic opacities representing osteopoikilosis. c Coronal STIR image and d axial T2-weighted image with homogenous strong hyperintensity, e coronal T1-weighted image with correspondingly hypointense signal, and f coronal T1-weighted contrast enhanced image with inhomogeneous enhancement with sparing of the calcifications

1165

1166

8  Musculoskeletal System

Fig. 8.8.10a,b  Ollier’s disease with multiple enchondromas of the right humerus shaft. a STIR image showing numerous small lobulated hyperintense lesions, b T1-weighted image with hypointense signal of the enchondromas and c Contrast-enhanced T1weighted image demonstrating inhomogeneous contrast uptake Fig. 8.8.11a–d  Osteoidosteoma of the right acetabulum in a 13-yearold male patient. a Coronal STIR image. Extensive edema within the acetabulum, the obturator internus muscle, and effusion in the inferior joint space. b Axial T2-weighted image with a small hyperintense lesion and a hypointense central spot, representing a calcification within the nidus. c Coronal T1-weighted image and d coronal T1-weighted image after contrast shows a hypointense enhancing intraosseous lesion within the right acetabulum. Note the enhancement within the joint space indicating synovial proliferation

are often surrounded by an extensive zone of edema in the marrow and adjacent soft tissue (Fig. 8.8.11). Osteochondromas represent the most common bone tumor (20–50% of all benign bone tumors and 10–15% of all bone tumors). Their radiologic features are often pathognomonic. They are composed of cortical and medullary bone with an overlying hyaline cartilage cap and are continuous with the underlying parent bone cortex and medullary cavity. Osteochondromas may be solitary or multiple, the latter being associated with an autosomal dominant syndrome (hereditary multiple exostoses [HME]). Complications associated with osteochondro-

mas are more frequent in HME and include deformity, fractures, neurologic and vascular sequelae, and malignant transformation. Malignant transformation is found in 1% of solitary osteochondromas and in 3–5% of patients with HME. Continued lesion growth and a hyaline cartilage cap greater than 1.5 cm in thickness, after skeletal maturity, suggest malignant transformation. The exact thickness of the cartilage cap of osteochondromas can only by evaluated by MRI, while it is regularly underestimated on CT. Fluid-fluid levels were first detected in aneurysmal bone cysts (ABC), and it has been assumed that they are

8.8  Bone and Soft Tissue Tumors

Fig. 8.8.12a–d  A 48-year-old patient with osteosarcoma of the proximal tibia. Status post-surgery, with incomplete operative resection. Large residual tumor/tumor recurrence. a Axial T1weighted image. Bony defect in the ventral portion of the tibia with low-signal areas within the bone and soft tissues (black triangles). b Axial T1-weighted image with fat saturation. The high SI within the posterior tibial marrow represents a subacute hematoma. c Axial T1-weighted image with fat saturation after

contrast. Strong enhancement of the surgical defect and the adjacent soft tissue edema with hypointense areas, which may correspond to hematomas (black arrows). Differentiation between the residual tumor/tumor recurrence to the fatty tissue is difficult. d Axial contrast enhanced T1-weighted image with fat saturation. In this sequence technique, the enhancing tumor tissue and the adjacent edema are readily differentiated from fatty tissue and muscles

a specific sign for diagnosis of ABC. Fluid-fluid levels derive from sedimentation of corpuscular components and debris in cystic cavities. However, for differential diagnosis it has to be noted that fluid-fluid levels were also described in various other bone tumors or tumor-like lesions, such as giant cell tumor, chondroblastoma, osteoblastoma, malignant fibrous histiocytoma, and teleangiectatic osteosarcoma.

8.8.7  Therapy Control Morphological signs, such as the reduction of extraosseous tumor components, an increasing ossification of bony destructions, the reduction of peritumoral edema, and the delineation of the intramuscular fat layers indicate good response to chemotherapy, surgery, radiation therapy, and combined regimens.

1167

1168

8  Musculoskeletal System

Fig. 8.8.13 a-c  Recurrence of a giant cell tumor of the calcaneus. Bone cement was implanted (open arrow) which is free of signal in all pulse sequences. a) T1-w image. There is a formation of intermediate signal intensity well demarked from normal

bone marrow. b) Contrast-enhanced T1-w SE image with mild enhancement within this region. c) Contrast-enhanced T1-w SE image with fat saturation. With fat suppression technique the (surgically proven) tumor recurrence is well identified

Scar tissue (older than 6 months) is usually hypointense on T1-weighted and T2-weighted images and does not show contrast enhancement, allowing a differentiation from tumor recurrence or residual tumor tissue, which usually exhibits high SI on T2-weighted images and high contrast uptake. However, reactive changes in the operation field may lead to misjudgments, especially in the early postoperative phase, but even after longer postoperative intervals. Differentiation between tumor and hematoma, edema, or inflammatory changes may not be possible. Moreover, MRI is not fully reliable after radiation therapy since hyperintense signal on T2-weighted images may persist for some months without tumor recurrence. In these cases, comparison of pre- and postoperative images is mandatory and may give valuable information. After radiation therapy with 40–60 Gy, signal intensity increase within the bone marrow in T1-weighted and T2-weighted images due to transformation of a hematopoietic marrow into fat marrow is seen. Dynamic, contrast-enhanced images are helpful both for the detection of tumor recurrence and for the evaluation of response to chemotherapy. In therapy responders, the rapid and high SI increase, which is characteristic for untreated malignant tumors, is no longer observable. In nonresponders, however, no significant difference to the contrast kinetics before chemotherapy can be observed. After surgery, image quality and therefore the ability to detect small tumor recurrence may be limited due to metal artifacts. Non-ferromagnetic implants lead to local SI loss and to oval zones with increased signal intensity at the ends of the implant. Susceptibility artifacts, particularly of GRE sequences, even for the smallest metal

particle, must be taken into account. Radiologically not provable metal abrasion, which arises, for example, from drilling might lead to obvious artifacts. However, larger tumor recurrences close to endoprostheses and osteosynthesis plates or screws are usually detectable by MRI, especially in axial T1-weighted fat-saturated images after contrast enhancement with slice thicknesses of 2–3 mm. Please see Figs. 8.8.12, 8.8.13. References 1.

2.

3.

4.

5.

6.

Bancroft LW, Peterson JJ, Kransdorf MJ (2005) MR imaging of tumors and tumor-like lesions of the hip. Magn Reson Imaging Clin N Am 13:757–774 Doi H, Ono A, Kawai A, Morimoto Y, Kunisada T, Nakata E, Ozaki T (2006) Magnetic resonance angiography without contrast enhancement medium in bone and soft tissue tumors. Oncol Rep 15:681–685 Erlemann R (2006) Imaging and differential diagnosis of primary bone tumors and tumor-like lesions of the spine. Eur J Radiol 58:48–67 Feydy A, Anract P, Tomeno B, Chevrot A, Drape JL (2006) Assessment of vascular invasion by musculoskeletal tumors of the limbs: use of contrast-enhanced MR angiography. Radiology 238:611–621 Fuchs B, Spinner RJ, Rock MG (2005) Malignant peripheral nerve sheath tumors: an update. J Surg Orthop Adv 14:168–174 Gehanne C, Delpierre I, Damry N, Devroede B, Brihaye P, Christophe C (2005) Skull base chordoma: CT and MRI features. JBR-BTR 88:325–327

8.9  Posttraumatic Alterations 7.

8.

9.

10.

11. 12.

13.

14.

15.

16.

17. 18. 19.

20.

21.

22.

Hanna SL, Fletcher BD (1995) MR imaging of malignant soft-tissue tumors.Magn Reson Imaging Clin N Am 3:629–650 Hernandez JA, Camacho A, Palacio D, Swischuk LE (2005) Low-grade (often multifocal) osteomyelitis (a diagnostic problem and/or a mimicker of tumor).Emerg Radiol 11:322–327 Hudson TM, Hamlin DJ, Enneking WF, Pettersson H (1985) Magnetic resonance imaging of bone and soft tissue tumors: early experience in 31 patients compared with computed tomography. Skeletal Radiol 13:134–146 Keenan S, Bui-Mansfield LT (2006) Musculoskeletal lesions with fluid-fluid level: a pictorial essay. J Comput Assist Tomogr 30:517–524 Kransdorf MJ, Murphey MD (2006) Soft tissue tumors: posttreatment imaging. Radiol Clin North Am 44:463–472 Leichtle C, Leichtle U, Rudert M (2005) Juvenile bone cyst, osteochondroma and non-ossifying fibroma in a male patient. A case report with description of entities [in German]. RoFo 177:1580–1582 Meissner SA, Vieth V, August C, Winkelmann W (2006) Radiology-pathology conference: osteosarcoma in a cartilaginous exostosis of the femur. Clin Imaging 30:206–209 Murphey MD, Choi JJ, Kransdorf MJ, Flemming DJ, Gannon FH (2000) Imaging of osteochondroma: variants and complications with radiologic-pathologic correlation. Radiographics 20:1407–1434 Oudenhoven LF, Dhondt E, Kahn S, Nieborg A, Kroon HM, Hogendoorn PC, Gielen JL,Bloem JL, De Schepper A (2006) Accuracy of radiography in grading and tissue-specific diagnosis—a study of 200 consecutive bone tumors of the hand. Skeletal Radiol 35:78–87 Resnick D, Kyriakos M, Greenway GD (2005) Tumors and tumor-like lesions of bone: imaging and pathology of specific lesions. In: Resnick D, Kransdorf MJ (ed) Bone and joint imaging, 3rd edn. Saunders, New York Turcotte RE (2006) Giant cell tumor of bone. Orthop Clin North Am 37:35–51 Vahlensieck M (2006) Synovial lesions around the knee joint [in German]. Radiologe 46:65–70 Weatherall PT (1995) Benign and malignant masses. MR imaging differentiation. Magn Reson Imaging Clin N Am 3:669–694 Weekes RG, Berquist TH, McLeod RA, Zimmer WD (1985) Magnetic resonance imaging of soft-tissue tumors: comparison with computed tomography. Magn Reson Imaging 3:345–352 Woertler K (2005) Soft tissue masses in the foot and ankle: characteristics on MR imaging. Semin Musculoskelet Radiol 9:227–242 Woude HJ van der, Verstraete KL, Hogendoorn PC, Taminiau AH, Hermans J, Bloem JL (1998) Musculoskeletal tumors: does fast dynamic contrast-enhanced subtraction MR imaging contribute to the characterization? Radiology 208:821–828

8.9  Posttraumatic Alterations With MRI, posttraumatic changes can be visualized that could not previously be seen with any imaging modality. This applies particularly to contusional injuries to the cancellous bone and the bone marrow (“bone bruises”), occult fractures, not visible on projection radiography, various soft tissue injuries (Fleckenstein et al. 1989; Gregg et al. 1995), and early stages of stress fractures. STIR and fat-saturated (moderately) T2-weighted TSE sequences are particularly sensitive for the detection of such changes. 8.9.1  Bone Injuries 8.9.1.1  Stress Fractures Stress fractures comprise an ongoing process of continued structural transformation in bone. They develop in either weakened (e.g., irradiation, Paget’s disease, osteomalacia, osteoporosis, corticosteroid therapy, rheumatoid arthritis) bone, under what is considered normal loading conditions (“insufficiency fracture”) or in normal bone due to continued inadequate loading (“fatigue fracture”). Stress fractures can be identified in the cortical or cancellous bone. MRI and bone scintigraphy show stress fractures well before radiographic changes are visible. Only 2–3 weeks after the onset of symptoms peri- and endosteal new bone formation, and a fracture line become recognizable in cortical stress fractures. Stress fractures in cancellous bone show diffuse sclerosis with obtuse margins (endosteal callus, trabecular condensation). The diagnosis of a stress fracture is supported by clinical data: a slow onset of symptoms in connection with unusual physical activity, and worsening with physical effort and improvement with rest. Preferred localizations of stress fractures are the tibia, the second and third metatarsal bones and the pelvis. In the pelvis, they are observed mostly as bilateral stress fractures in the sacral bone (Fig. 8.9.1), in the femoral neck, in the supra-acetabular ileum and in the pubic bones. Not rarely, insufficiency fractures of the sacrum arise in women who underwent radiotherapy for pelvic malignancies and may be misinterpreted as metastases (Blomlie et al. 1993). In the T1-weighted image, usually there is an area of—often only moderately—reduced SI in the bone marrow, which corresponds to an area of increased SI in T2weighted images. These signal alterations are best visualized with STIR or fat-saturated (moderately) T2-weighted TSE sequences, and margins often are obtuse. A linear component of low SI, which corresponds to the fracture line, can be found in this area in all sequences. Usually, it runs vertical to the neighboring cortical bone, but there are stress fractures running longitudinally in the tibia. If

1169

1170

8  Musculoskeletal System

there is distension, SI can be high due to water and hemorrhage in the fracture plane in T2-weighted sequences. In the long bones, stress fractures commonly are localized in the metaphysis or epiphysis.

Extensive edema and inflammatory reaction can be found in the soft tissues adjacent to the stress fractures. With high-resolution MRI periosteal and endosteal ossifications may be depicted. Occasionally, subperiosteal hemorrhage and the elevated periosteum itself are visualized.

Fig. 8.9.1a–c  Sacral insufficiency fracture. Minor sclerosis in the left portion of the sacrum (a). MRI reveals marked concomitant signal loss in T1-weighted sequences in a corresponding location extending parallel and adjacent to the iliosacral joint (b). T1-weighted FS sequences show marked Gd uptake and central in this area a linear low-signal-intensity line, corresponding to the fracture line (c). Both MRI scans are angulated in a paracoronal plane parallel to the long axis of the sacrum

Fig. 8.9.2a–c  Occult bilateral femoral neck fractures. In projection radiography (only left side shown here), no evidence of a fracture is detectable (a). T2-weighted axial MRI reveals a lowsignal-intensity line perpendicular to the cortices involved in both proximal femora (b). Please note the extensive Gd uptake in the T1-weighted FS sequence (c), which contrasts with the discrete findings in the T2-weighted non-FS sequence. There is soft tissue enhancement reflecting periosteal and synovial reaction and there is concomitant effusion

8.9  Posttraumatic Alterations

Before a stress fracture develops, pathologic changes, which are already accompanied by pain, occur and are described as stress reactions. Under inadequate loading, a progressive bone resorption takes place. This is based on focal hyperemia and high osteoclast activity. With MRI signal alterations in the bone marrow, a so-called bone marrow edema-like signal pattern (BMEP) are then recognized, particularly in the STIR and (moderately) T2-weighted TSE sequences. 8.9.1.2  Occult Fractures

ment of post-traumatic complaints are easily recognizable, too (Bogost et al. 1995). Especially after traumatic injury to the wrist, occult fractures, which are recognizable in projection radio­gra­ phy only weeks after the trauma, can easily be proven with MRI. However, if, in the early posttraumatic phase, one cannot distinguish between a bone contusion and an occult fracture, then high-resolution computed tomography should be performed. With MRI, fractures of the epiphyses can be visualized more exactly and with better sensitivity than with radiographic techniques. If there are disturbances of growth after physeal fractures, osseous or fibrous bridging that may have occurred can be visualized with MRI.

Fractures, usually non-displaced fractures, not visible on radiography studies that have been correctly carried out and read, are described as occult fractures. Such radiographically occult fractures can reliably be proven or excluded by MRI. Especially in elderly patients, bone scintigraphy is positive only after approximately 3 days, which makes MRI the preferred modality (Haramati et al. 1994) for the reliable and early detection or exclusion of such a fracture. As demonstrated by Quinn and McCarthy (1993), T1-weighted SE sequences can show occult femoral neck fractures as reliably as radiographic follow-up, bone scintigraphy, conventional tomography, and computed tomography, and MRI is also cost effective. Occult fractures appear as linear or band-shaped SI reduction in the bone marrow, which usually connects to the cortical bone (Fig. 8.9.2) in T1-weighted images. Extensive perifocal edema is shown in the T2-weighted or STIR images, while the fracture line itself is of low SI (trabecular compression), or, in the case of distension, of high SI (hematoma, edema) (Fig. 8.9.3). In GRE sequences (susceptibility effects), more or less extensive SI reductions are found. If additionally STIR sequences are used, soft tissue injuries (hematoma, contusion of the musculature), which can be meaningful for the assign-

Bone contusions (bone bruises) are caused by a direct blow or an axial, eccentric, compressive force. They often are associated with ligamentous injuries (e.g., anterior cruciate ligament or collateral ligament) (Mink and Deutsch (1989). The pathological alterations are limited to the bone marrow and are thought to correspond to bleeding, granulation tissue, trabecular microfractures, and edema. They manifest as inhomogeneous areas of low SI in T1-weighted and high SI in T2-weighted sequences and follow a geographic or reticular pattern, which is distinctive for acute bone contusion. The cortical bone usually is not affected unless an open fracture has occurred. Bone contusions tend to normalize within 6 weeks to 4 months, but persisting signal abnormalities up to 2 years have been described, especially following extensive trauma or continued instability. Proper recognition of bone contusions is important since they may require reduced weight bearing of the extremities affected. A subchondral area of BMEP in the lateral femo-

Fig. 8.9.3a,b  Sports injury in a 35-year-old female, with partial tear of the plantar aponeurosis and occult fracture of the cuboid bone. The T1-weighted coronal sequence shows thickening and central signal inhomogeneities in the medial fascicle of

the plantar aponeurosis (a arrow). The sagittal STIR sequence exhibits extensive, ill-defined BMEP in the lower aspect of the cuboid, together with a linear low-signal-intensity abnormality indicating the fracture line (b arrow)

8.9.1.3  Bone Contusions

1171

1172

8  Musculoskeletal System

Fig. 8.9.4a–c  Patellar dislocation with flake fracture and tear of the medial patellofemoral ligament. a BMEP at the inferomedial aspect of the patella and at the lateral aspect of the femur. b,c Osteocartilaginous flake in the suprapatellar recess. a,b The medial retinaculum, especially the patellofemoral ligament, is torn at the patellar insertion side. (Courtesy Dr. M. Steinborn, Pediatric Radiology, Munich-Schwabing)

ral condyle and in the posterolateral aspect of the tibia is characteristic for recent tears of the anterior cruciate ligament. After patellar dislocation, contusions are found in the lateral femoral condyle and in the (infero)medial patellar facet (Virolainen et al. 1993) (Fig. 8.9.4). In ruptures of the medial collateral ligament, contusional areas can be observed in corresponding subchondral regions of the lateral femoral condyle and the lateral aspect of the tibial plateau. Posterosuperior location of BMEP in the humeral head is frequent in anterior dislocation of the shoulder. Chronic overuse in ulna impaction syndrome is reflected by BMEP in the ulnar aspect of the lunate bone and the adjacent portions of the triquetral bone. A subchondral BMEP is found in osteochondral and chondral injuries, too. It is of note that the bone abnor-

mality may not exactly overlie, but may be seen in close vicinity to the associated cartilage lesion. While geographic lesions are deemed to often lead to early osteoarthritis (Vellet et al. 1991), a reticular lesion pattern is considered to have a better prognosis without persisting chronic damage. References 1.

Blomlie V, Lien HH, Iversen T et al (1993) Radiation-induced insufficiency fractures of the sacrum: evaluation with MR imaging. Radiology 180:241–244

8.11  Diagnostic Value of MRI and Comparison with Other Imaging Modalities 2.

3.

4.

5.

6.

7.

8.

9.

Bogost GA, Lizerbram EK, Crues JV (1995) MR imaging in evaluation of suspected hip fracture: frequency of unsuspected bone and soft-tissue injury. Radiology 197:263–267 Fleckenstein JL, Weatherall PT, Parkey RW, Payne JA, Pshock RM (1989) Sports-related muscle injuries: evaluation with MR imaging. Radiology 172:793–798 Gregg A, Bogost MD, Lizerbram EK et al (1995) MR imaging in evaluation of suspected bone and soft-tissue injury. Radiology 197:263–267 Haramati N, Staron RB, Barax C et al (1994) Magnetic resonance imaging of occult fractures of the proximal femur. Skelet Radiol 23:19–22 Mink JH, Deutsch AL (1989) Occult cartilage and bone injuries of the knee: detection, classification and assessment with MR imaging. Radiology 170:823–829 Quinn SF, McCarthy JL (1993) Prospective evaluation of patients with suspected hip fracture and indeterminate radiographs: use of T1 weighted MR images. Radiology 156:77–82 Vellet AD, Marks P, Fowler P, Munro PH (1991) Occult posttraumatic osteochondral lesions of the knee: prevalence, classification, and short-term sequelae evaluated with MR-imaging. Radiology 178:271–276 Virolainen H, Visur T, Kuusela T (1993) Acute dislocation of the patella: MR findings. Radiology 197:826–830

8.10  Differential Diagnosis Differential diagnosis is very important in cases of unclear pathology and especially when the patient’s history, clinical findings, and radiography do not yield conclusive results or, at worst, are contradictory. Fortunately, however, radiologists’ ever-increasing experience with MRI is allowing more and more typical lesion patterns to be identified, thus facilitating specific interpretations that help establish definite diagnoses in many cases. Mostly, signal intensities, relaxation times, and contrast enhancement do not allow specific (in terms of histology) diagnoses. Rather, they may reveal otherwise undetected pathology and reflect disease activity. Therefore, combining MRI morphology with clinical, laboratory, and radiologic data is the key to narrowing down the often-broad spectrum of differential diagnostic options to a few clinically relevant diagnoses. Pain, fever, erythrocyte sedimentation rate (ESR), and leukocytosis often lead the way to diagnosis in infections of the musculoskeletal system. “Low-grade” and chronic infections as well as atypical presentation of disease, however, may be problematic in particular cases. Diffuse signal alterations in T1-weighted and T2-weighted sequences with indistinct margins often are present in the immediate vicinity of the bones and joints. Such findings are much more obvious and thus sensitivity for detection is higher in STIR, fat-presaturated T2-weighted/PD-weighted, as well as in T1-weighted fat-saturated contrast-enhanced

sequences. Contrast enhancement in dynamic sequences generally is slower and weaker in edema than in, especially malignant, tumors. Aseptic osteonecroses are frequently localized in the subchondral bone (epiphyses of the long bones, flat bones). In typical cases, they are demarcated by a low-SI line. Recent evidence suggests a relationship to subchondral stress reactions/stress fractures in those cases with linear low-SI lines close to the subchondral bone plate. Calcifications may appear similar in enchondroma and bone infarcts. However, in bone infarcts, calcifications are typically in the periphery of the lesion and have a polylobular or serpentine configuration, whereas enchondromas are characterized by stippled central calcifications. Endosteal cortical thinning is observed in both enchondroma and chondrosarcoma; erosion of more than two thirds of the cross-sectional cortical diameter is indicative of malignant transformation within an enchondroma. MRI shows the lobulated chondroid matrix with high signal intensity in T2-weighted images. In chondroid tumors, a variable amount of enhancement may be present. Although not allowing for a clear differentiation between enchondroma and chondrosarcoma, contrast enhancement usually is absent within (mature) bone infarcts. Bone infarcts, on the other hand, exhibit linear (band-like), high signal intensity in the periphery (STIR, T2-FS), which is a valuable diagnostic sign. The differential diagnosis of osteoporotic versus tumor-associated vertebral body fractures may be difficult or even impossible. In tumor-associated fractures, the complete vertebral body may show low signal intensity in T1-weighted sequences while in osteoporosis signal abnormalities mostly exhibit a band-like configuration adjacent to the vertebral endplates. Increased signal intensity in opposed-phase GRE images as well as a mass effect are indicative of tumor-associated fractures. Diffusion-weighted imaging may further improve the specificity of MRI (Baur 2003). 8.11  Diagnostic Value of MRI and Comparison with Other Imaging Modalities Due to its unique advantages, such as excellent spatial as well as contrast resolution and multi-planar imaging capabilities, MRI has become the modality of choice for most diagnostic problems in the field of musculoskeletal imaging. Previously, use of MRI was limited by the long examination times. With the development of fast pulse sequences, hardware such as multi-array coils, and more ergonomic user interfaces, considerable improvement has taken place and imaging times could be reduced down to even 15–20 min for, for example, a complete knee examination.

1173

1174

8  Musculoskeletal System

8.11.1  Bone Marrow With MRI, the bone marrow can be investigated with superior accuracy. Apart from a variety of pathological changes, the distribution and conversion of red marrow into fatty marrow and vice versa (reconversion) may be observed. Both diffuse infiltration (leukemia, osteomyelofibrosis, aplastic anemia, and lipidoses) as well as focal invasion by neoplastic disease (metastatic disease, lymphoma, myeloma) as well as inflammatory and posttraumatic (bone marrow edema-like alterations) changes are readily depicted. Often, this allows a whole cascade of examinations (radiography, scintigraphy, CT, and sometimes biopsy) to be bypassed. 8.11.2  Bone Tumors and Bone Metastases For the detection of bone tumors and metastases, MRI can be considered the modality of choice when radiography is not conclusive. In bone lesions with an aggressive pattern of growth and unclear periosteal reactions, MRI adds important information. Clearly benign tumors and tumor-like lesions, on the other hand, do not need further investigation. Depiction of the complete spine, pelvis, and proximal femora using phased-array coils enables visualization of a large proportion of the hematopoietic marrow, the preferred localization of bone metastases. Today, especially for bone metastases of tumor entities that do not exhibit unequivocal scintigraphic enhancement, e.g., bronchial carcinoma, renal cell cancer, and multiple myeloma, MRI has become a valuable alternative to scintigraphy in clinical routine. Results are more specific and the higher spatial resolution provides important morphological detail. With whole-body MRI, the complete skeletal system can be imaged within acceptable examination times. Detection of metastases is superior to PET/CT (Schmidt G 2006). With multi-detector row CT (MDCT), large body parts can be imaged within short acquisition times and with very high spatial resolution. Isotropic voxels can be acquired, so that high-quality, multiplanar reformats can be generated, which greatly enhance anatomical orientation. As shown in multiple myeloma, CT is inferior to MRI, especially in cases with diffuse myeloma infiltration of the bone marrow. Fracture risk, on the other hand, can be better assessed with MDCT. 8.11.3  Infections MRI enables early detection of bone, joint, and soft-tissue infections. The extension of the inflammation can be exactly determined in the medullary cavity, in the cortical bone, and in the soft tissues. This has great importance for the treatment of patients, particularly with regard to the question of whether an antibiotic treatment may be

sufficient or whether an operative treatment, eventually including bone surgery, is required. In difficult clinical situations such as diabetic osteoarthropathy and chronic infections, MRI usually enables the differentiation between acute infection and remission. In ambiguous cases, contrast-enhanced scans with fat saturation proved highly useful. The detection of an abscess with fluid in the center and peripheral contrast enhancement is highly indicative of an active infection. In radiography, osseous destructions usually are detected only late (up to 3 weeks from onset) in the time course of acute osteomyelitis. In chronic osteomyelitis and posttraumatic or postoperative osteitis, irregular sclerosis may render evaluation even more difficult. Scintigraphy, in such cases, often shows only unspecific tracer accumulation. MRI may solve the diagnostic dilemma. 8.11.4  Aseptic Osteonecroses MRI is considered the modality of choice for the detection and classification of aseptic osteonecrosis and osteochondrosis dissecans. Early findings are signal-intensity changes in the bone marrow while radiography is still negative. MRI may reveal bilateral affection in osteonecrosis of the hip and in Perthes disease, while radiography and clinical presentation are innocuous. By precisely assessing the extent and site of the necrosis, MRI is helpful for preoperative planning. Fragment loosening in osteochondrosis dissecans can be diagnosed with high confidence. 8.11.5  Joints Assessment of internal derangement of the joints has become one of the mainstays of musculoskeletal MRI. Stress views, scintigraphy, and conventional arthrography have but an ancillary role and, clearly, conventional arthrography has been replaced by direct MR arthrography. Especially in the knee, MRI may effectively replace diagnostic arthroscopy. 8.11.6  Bone and Soft Tissue Tumors For the diagnosis and staging of bone and soft-tissue tumors—especially malignant and aggressive entities— MRI is superior to all other imaging modalities. However, for reasons of cost effectiveness, it should be avoided in the examination of patients in whom diagnosis can be established from radiography alone, for example, in patients with non-ossifying fibroma and enchondroma of the short tubular bones. Assessment of the nature and especially follow-up of chemotherapy of bone and softtissue tumors profits from dynamic contrast-enhanced

8.12  Diagnostic Process

studies including first-pass enhancement. These procedures are also helpful for the differentiation of fibrosis and tumor recurrence. The intramedullary extension, soft-tissue infiltration, the invasion of the neurovascular bundles as well as the extension into the epiphyses and the joint space can be evaluated more precisely with MRI than with all other imaging procedures. This significantly improves the staging of bone tumors, particularly malignant bone tumors, and focally aggressive entities. 8.11.7  Traumatology Detection of bone contusions, osteochondral lesions, and occult fractures are important contributions of MRI in traumatology, and unclear postoperative conditions may be classified. 8.12  Diagnostic Process Still, projection radiography is the indispensable basis of any imaging-based diagnostic approach to the musculo-

skeletal system. When radiography is neglected, dangerous errors may result, and not only in rare cases. MRI most often is the next appropriate diagnostic imaging procedure after radiography and often yields a definite diagnosis. Depending on the clinical problem, MRI may be warranted when there is no pathological finding in radiography, e.g., with early osteonecrosis, or in cases of suspected metastases without radiographically detectable bone destruction. The need to use evidence-based criteria in the application of MRI, as in the application of other imaging modalities, is gaining broad recognition. From this standpoint, the reduction of ionizing radiation is an important aspect in choosing the right study, as are availability and cost effectiveness. Guidelines designed to facilitate the adequate use of imaging modalities are becoming increasingly available and their normative power is growing. The guidelines published by the European Commission (in 2000) for the use of imaging modalities emphasize the issue of radiation protection. They further underline that MRI should be preferred to CT when both procedures can be expected to yield comparable diagnostic information.

1175

Chapter 9

Skeletal Muscle

9

9.7.1.1  Acquired Myopathies  .. . . . . . . . . . . . . . . .   1188

9.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1177 C. Born

9.2 

Examination Techniques  .. . . . . . . . . . . .   1178

9.2.1 

Patient Preparation   . . . . . . . . . . . . . . . . . .   1178

9.2.2 

Patient Positioning   and Selection of Coils  . . . . . . . . . . . . . . . .   1178

9.2.3 

Examination Sequences and Imaging Planes  . . . . . . . . . . . . . . . . . .   1178

9.2.4 

Whole-Body MRI  .. . . . . . . . . . . . . . . . . . .   1179

9.7.3.1  General Features of Denervation  . . . . . .   1206

9.2.5 

Use of Contrast Media  .. . . . . . . . . . . . . . .   1180

9.7.4 

9.3 

Value of MRI  .. . . . . . . . . . . . . . . . . . . . . . .   1181

9.4 

Normal Anatomy  .. . . . . . . . . . . . . . . . . . .   1181

9.5 

MR Imaging Findings  .. . . . . . . . . . . . . . .   1184

9.5.1 

Muscle Edema  .. . . . . . . . . . . . . . . . . . . . . .   1184

9.5.2 

Fatty Infiltration  . . . . . . . . . . . . . . . . . . . . .   1184

9.5.3 

Mass Lesions and Calcification  .. . . . . . .   1184

9.5.4 

Contrast Enhancement  . . . . . . . . . . . . . . .   1184

9.5.5 

Lesion Patterns  . . . . . . . . . . . . . . . . . . . . . .   1185

9.5.6 

Distribution  . . . . . . . . . . . . . . . . . . . . . . . . .   1186

9.6 

Muscle Biopsy  .. . . . . . . . . . . . . . . . . . . . . .   1188

9.7 

Skeletal Muscle Diseases  . . . . . . . . . . . . .   1188

9.9.2.1  Proton (1H) MR Spectroscopy of Human Muscle  .. . . . . . . . . . . . . . . . . . .   1210

9.7.1 

Muscle  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1188

9.9.2.2  31Phosphorus MRS  .. . . . . . . . . . . . . . . . . .   1210

9.7.1.2  Inherited Myopathies  .. . . . . . . . . . . . . . . .   1196 9.7.2 

Neuromuscular Junction  .. . . . . . . . . . . . .   1204

9.7.2.1  Myasthenia Gravis  . . . . . . . . . . . . . . . . . . .   1204 9.7.2.2  Lambert-Eaton Syndrome  . . . . . . . . . . . .   1205 9.7.2.3  Botulism  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1205 9.7.3 

Nerves  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1206

Central Nervous System: Motor Neuron Diseases  .. . . . . . . . . . . . . .   1208

9.7.4.1  Primary Lateral Sclerosis   .. . . . . . . . . . . .   1208 9.7.4.2  Amyotrophic Lateral Sclerosis  .. . . . . . . .   1208 9.7.4.3  Spinal Muscular Atrophies   . . . . . . . . . . .   1208 9.7.4.4  Poliomyelitis  .. . . . . . . . . . . . . . . . . . . . . . . .   1208 9.8 

Follow-Up  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1209

9.9 

Future Developments  .. . . . . . . . . . . . . . .   1209

9.9.1 

Diffusion-Weighted Imaging and BloodOxygen Level-Dependent Imaging  . . . .   1209

9.9.2 

Spectroscopy  . . . . . . . . . . . . . . . . . . . . . . . .   1210

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1210

9.1  Introduction C. Born The diagnosis of muscle disease (myopathy) has traditionally relied on clinical examination combined with the microscopic examination of a muscle biopsy specimen. While histological examination of a muscle biopsy remains the cornerstone of diagnosis, macroscopic analysis of muscle using a variety of imaging techniques can be useful.

A diagnosis of myopathy is suspected when patients complain of difficulty performing tasks that require muscle strength or when they developed various types of rash or respiratory problems. To establish the diagnosis the following considerations are important: • Clinical history • Clinical examination of muscle strength, tonus, sensibility, trophy, reflexes. • Laboratory examination including the muscle enzymes creatine kinase (CK), lactate dehydrogenase (LDH)

1178

9  Skeletal Muscle

and aldolase, as well as erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) and blood cell counts • Electrophysiological examination, e.g., electromyogram (EMG) and electronstagmography (ENG) • MRI Magnetic resonance imaging (MRI) has an important role in the detection and characterization of pathological conditions of skeletal muscle that cause changes in muscle signal intensity. There are many conditions that may affect muscle signal intensity, such as inflammatory, infectious, traumatic, neurologic, neoplastic, and iatrogenic conditions. Although the MR imaging findings are similar in many conditions, distinct patterns of signal intensity abnormalities may be recognized. Additional clues to the diagnosis may also be present on the MR images, thus allowing the differential diagnosis to be further narrowed. Finally, correlation with the patient’s clinical history and the results of other direct differential diagnosis help to achieve a single correct diagnosis. MRI of the muscle may help to select the appropriate site of biopsy, enable evaluation of the distribution of muscle wasting, and prove useful in monitoring disease progression or the response to treatment. MRI is proven useful in evaluating patients with muscle disorders. In time, MRI may replace EMG and muscle biopsy in the evaluation of these conditions. 9.2  Examination Techniques 9.2.1  Patient Preparation No special patient preparation and positioning are necessary. The patient is placed in the supine position. The musculoskeletal system, especially in the extremities, is not prone to involuntary motion and as a consequence, motions artifacts are rare. However, the patient should be positioned very comfortably to avoid motion artifacts, pain, and compression of different muscle groups. Sometimes it is useful to mark the mostly affected muscle group with a marker, for example a nitroglycerin capsule, which exhibits high signal intensity on T1-weighted images and T2-weighted images. 9.2.2  Patient Positioning and Selection of Coils Patient positioning considerations include patient size, body part, and anatomical structures to be examined and expected examination time. MRI is more sensitive in the detection of pathologic findings in the lower extremities than in the shoulder girdle and the upper extremities. If the symptoms or the neurophysiologic findings are clearly limited to the upper extremities, it is useful to begin the MRI examination in this mostly affected body part. Most of the muscle diseases affect the muscles of the

legs. Therefore, the examination of the muscle of the femur and the lower leg are mostly mandatory. Muscles of the lower extremities are easily accessible for a muscle biopsy. The patient should be studied with the most closely coupled coil possible (i.e., the smallest coil that covers the anatomy) to achieve the maximum signal-to-noise ratio (SNR) and the best spatial resolution. Many musculoskeletal examinations in the extremities are performed using surface coils. The SNR is four to six times greater when surface coils are used than when body coils are used (Fisher et al. 1985). Multiple coil arrays will facilitate examinations and patient throughput. These multiple coil arrays allow the use of multiple coils simultaneously to increase the SNR. The examination time can also be reduced (Berquist 1991). A new approach especially useful for the assessment of muscle diseases is whole-body MRI. Whole-body MRI represents a new method of comprehensive imaging instead of the stepwise concept for the detection of muscle diseases, especially concerning the distribution and the extent of the disease as well as the identification of the best site of a planned muscle biopsy. 9.2.3  Examination Sequences and Imaging Planes Current experiences indicate that in most situations selection of a T1-weighted and a STIR sequence (short-tau inversion recovery sequence), will provide the necessary diagnostic information. Suspected muscle hemorrhages are best detected using T2*-weighted gradient-echo (GRE) sequences, which should be performed additionally. The signal intensity of normal skeletal muscle is generally slightly higher than that of water and much lower than that of fat in T1-weighted images, and much lower than that of both fat and water on T2-weighted images. On inversion recovery and fat-suppressed T2-weighted images, normal muscle intensity is much lower than that of water but higher than that of fat. Pathological conditions affecting skeletal muscle may cause alterations in muscle size, shape, and signal intensity. Abnormalities in muscle size and shape are detected with virtually any MR imaging sequence. Alterations in muscle signal intensity may include alterations in muscle T1 and T2 relaxation times. Alterations in muscle T1, notably T1 shortening caused by fatty infiltration, methemoglobin, or proteinaceous material, are detected with T1-weighted SE images. Alterations in muscle T2, notably T2 prolongation due to increased intracellular or extra cellular free water (muscle edema), are best detected with STIR images or, as an alternative, with fat-suppressed T2weighted images (Hernandez 1993). The slice thickness has to be selected according the size of the suspected pathology and the type of coil employed. When small anatomical areas are scanned and high spatial resolution is required, a slice thickness of

9.2  Examination Techniques

> 5 m should be employed. Larger areas can be studied with thick (1 cm) slices. This slice thickness is especially useful for screening examinations. It is also important to select an optimal field of view (FOV). This is especially important when using surface coils on the peripheral extremities. A smaller FOV improves spatial resolution significantly. However, SNRs decrease with smaller FOV. Enlarging the field of view to encompass both sides of the body allows assessment of muscle asymmetry and may assist in the detection of subtle signal intensity abnormalities, particularly in focal diseases. Evaluation of the extent of muscle disease is most easily accomplished using two planes. The use of axial planes allows evaluation of the distribution of affected muscles. Axial slices are mandatory to identify those muscles that are most affected and mass lesions. Coronal images also allow coverage of larger areas of the body, so that abscesses in infectious myositis can be detected.

A useful imaging protocol is suggested in Table 9.2.1, using a 1.5-T scanner (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany) which combines 76 coil elements (matrix coils) and 32 receiver channels. It allows parallel imaging techniques (PAT) in all three dimensions with free table movement at a total field of view of 205 cm. STIR and T1-weighted images can be obtained with an acceleration factor of 2 due to reduced acquisition time and to achieve an adequate SNR. 9.2.4  Whole-Body MRI Originally, whole-body MRI of skeletal muscle was performed in sequential scanning of five body levels, with time consuming coil rearrangement and repositioning of the patient.

Table 9.2.1  MRI protocol for neuromuscular whole-body imaging on a 32-channel whole-body scanner (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany) with the use of PAT Whole-body MRI protocol FOV

Scanning time

 

 

In acute inflammatory conditions

STIR coronal

T1-w coronal

 

TR 2,680, TE 101, TI 150

TR 540, TE 13

 

SL 8 mm

SL 8 mm

 

STIR coronal

T1-w coronal, transverse

T1-w FS coronal, transverse

TR 2,680, TE 101, TI 150

TR 540, TE 13

TR 179, TE 3.33

SL 8 mm

SL 8 mm

SL 8 mm

STIR coronal

T1-w coronal, transverse

T1-w FS coronal

TR 2,680, TE 101, TI 150

TR 540, TE 13

TR 179, TE 3.33

SL 8 mm

SL 8 mm

SL 8 mm

STIR coronal

T1-w coronal, transverse

T1-w FS coronal, transverse

TR 2,680, TE 101, TI 150

TR 540, TE 13

TR 179, TE 3.33

SL 8 mm

SL 8 mm

SL 8 mm

STIR coronal

T1-w coronal, transverse

T1-w FS coronal, transverse

TR 2,680, TE 101, TI 150

TR 540, TE 13, SL 8 mm

TR 179, TE 3.33

STIR coronal

T1-w coronal, transverse

T1-w coronal, transverse

TR 2,680, TE 101, TI 150

TR 540, TE 13

TR 179, TE 3.33

SL 8 mm

SL 8 mm

SL 8 mm

50 min

T1-w T1-weighted, SL slice thickness, T2-w T2-weighted

Additionally, 20 min

1179

1180

9  Skeletal Muscle

The past 5 years have witnessed the development of multiple receiver coil systems that allow simultaneous recording of MR data from large body regions, with high sensitivity (Fig. 9.2.1). Even systems with up to 32 parallel receiver channels and more than 70 single coils covering the entire body are now offered commercially, which enable fast whole-body examinations without requiring manual repositioning of the patient in the scanner. Additionally, the presence of multiple receiver coils was found to be useful for improvement of spatial resolution or shortening of measuring times. Parallel imaging strategies can clearly reduce the number of phase-encoding steps at the expense of a slightly decreased signal-tonoise ratio (SNR). The introduction of a rolling platform mounted on top of a conventional MRI examination table facilitated whole-body MR imaging and, with the use of fast gradient echo, T1-weighted and STIR-imaging techniques, for the first time allowed whole-body imaging within less than one hour. With the development of PAT in combination with global matrix coil concepts (Fig. 9.2.2), acquisition time could be reduced substantially without compromises in spatial resolution, enabling the implementation of more complex and flexible examination protocols (Table 9.2.1). 9.2.5  Use of Contrast Media

Fig. 9.2.1a,b  Whole-body MRI of a 25-year-old woman with a limb girdle dystrophy. a Composed T1-weighted image in coronal orientation: Slight atrophy of thigh muscles is seen bilaterally. b Composed STIR images in coronal orientation: Muscle edema, a sign of active inflammation of skeletal muscles, is absent

The use of intravenous injection of gadolinium chelates in muscle diseases continues to be explored. Increased contrast enhancement is found in infectious myositis, acute rhabdomyolysis, necrotic processes, and focal myositis as well as in mass lesions. If fatty degeneration in the muscle predominates or if neuropathic lesions result in muscle edema, contrast enhancement is not increased. Therefore, contrast agents should be applied only in those cases in which added value can be anticipated.

Fig. 9.2.2  Scheme of a global matrix coil concept for wholebody imaging   (Magnetom Avanto, Siemens, Erlangen, Germany)

9.4  Normal Anatomy

9.3  Value of MRI Imaging techniques, such as CT, ultrasound, and MRI, have assumed increasing importance in the diagnosis of patients with muscle disease. For most muscle disorders, ultrasound and MRI are more useful than CT is. Advantages of ultrasound include accessibility at the bedside and lower cost. However, MRI remains the gold standard for detecting changes in muscle tissue. In some cases, MRI examinations can take the place of muscle biopsy for diagnosis. New advances in MRI include diffusionweighted imaging and blood-oxygen level–dependent imaging to evaluate tissue oxygenation. It has been suggested that MRI should be included along with serologic testing and tissue biopsy as part of the routine approach to evaluating patients with myositis (Sayers et al. 1992). Unlike muscle biopsy or electromyogram, MRI is noninvasive, making it possible to perform repeat studies for longitudinal analyses of outcome (Park and Olsen 2002).

For long-term studies, quantitative data are generated, which allow for accurate comparisons over time. 9.4  Normal Anatomy The cross-sectional anatomy of the muscle system is rather complex. The following illustrations (Figs. 9.4.1 through 9.4.8) demonstrate the most important muscle groups concerning the diseases of the skeletal muscle systems. The signal intensity of normal skeletal muscle is generally slightly higher than that of water and much lower than that of fat (Fig. 9.47) on T1-weighted images, and much lower than that of both fat and water on T2weighted images. On inversion-recovery and fat-suppressed T2-weighted images, normal muscle signal intensity is much lower than that of water but higher than that of fat (Fig. 9.4.8).

Fig. 9.4.1  Left pelvic girdle. 1 Gluteus maximus, 2 gluteus minimus, 3 gluteus medius, 4 iliopsoas. Top ventral view, right lateral view, bottom dorsal view, left medial view

Fig. 9.4.2  Right thigh. 1 Vastus lateralis, 2 vastus intermedius, 3 vastus medialis, 4 sartorius, 5 rectus femoris, 6 gracilis, 7 adductor muscles, 8 biceps femoris, 9 semimenranosus, 10 semitendinosus, F Femur. Top ventral view, right lateral view, bottom dorsal view, left medial view

1181

1182

9  Skeletal Muscle

Fig. 9.4.3  Right lower leg. 1 Gastrocnemius mediale, 2 gastrocnemius laterale, 3 soleus, 4 tibialis posterior, 5 tibialis anterior, 6 peroneii muscles, 7 extensor digitorum longus, F fibula, T Tibia. Top ventral view, right lateral view, bottom dorsal view, left medial view

Fig. 9.4.4  Left shoulder girdle. 1 Deltoideus, 2 infraspinatus, 3 subscapularis, 4 serratus anterior, 5 pectoralis major, 6 pectoralis minor, 7 coracobrachialis. Top ventral view, right lateral view, bottom dorsal view, left medial view

Fig. 9.4.5  Right upper arm. 1 Biceps brachii (caput breve), 2 biceps brachii (caput longum), 3 brachialis, 4 triceps brachii (caput mediale), 5 triceps brachii (caput laterale), 6 triceps brachii (caput longum), H Humerus. Top ventral view, right lateral view, bottom dorsal view, left medial view

9.4  Normal Anatomy

Fig. 9.4.6  Right forearm. 1 Extensor muscle group, 2 flexor digitorum profundus, 3 flexor carpi ulnaris, 4 flexor digitorum superficialis, 5 flexor carpi radialis, R radius, U ulna. Top ventral view, right lateral view, bottom dorsal view, left medial view

Fig. 9.4.7  Composite T1-weighted coronal image. The signal intensity of the muscles (thin arrow) is much lower than that of subcutaneous or bone marrow fat (thick arrows), but much higher than that of water

Fig.9.4.8  Composite short-tau inversion recovery (STIR) weighted coronal image. The signal intensity of muscle (thin arrow) is higher than that of fat (thick arrows), shown as subcutaneous fat or bone marrow fat, but much lower than that of water (see bladder, asterisk)

1183

1184

9  Skeletal Muscle

9.5  MR Imaging Findings Abnormal signal intensity within skeletal muscle is frequently encountered at MR imaging. Potential causes are diverse, including trauma, infectious, autoimmune, inflammatory, neoplastic, neurologic, and iatrogenic conditions. Alterations in muscle signal intensity seen in pathologic conditions usually fall into one of three recognizable patterns: muscle edema, fatty infiltration, and mass lesions. These conditions have to be differentiated, and biopsy may be required to establish the correct diagnosis. Clues to the correct diagnosis and whether biopsy is necessary or appropriate are frequently present on the MR images, especially when they are correlated with clinical features. Although the MR imaging findings of many disease entities are similar, distinct patterns of signal abnormalities may be recognized. The best approach for MR image interpretation is to analyze patterns of the following features: edema, fatty infiltration, mass lesions, contrast enhancement, lesion pattern, and distribution. 9.5.1  Muscle Edema Muscle edema is almost always due to increased intracellular or extracellular free water, that is, true muscle edema (Fleckenstein et al. 1993). However the accumulation of abnormal metabolites that may occur in some metabolic conditions may also contribute to T2 prolongation and thus to increased signal intensity on T2-weighted images (Fig. 9.5.1). Muscle edema can be quite subtle and detectable only with inversion recovery or fat suppressed T2-weighted images (Fleckenstein et al. 1993; Plotz et al. 1989). Muscle edema may be seen in autoimmune diseases such as polymyositis and dermatomyositis (Fleckenstein and Reimers 1996; Fraser et al. 1991; Lamminen 1990; Hernandez et al. 1993), mild injuries (Scott 2003), infectious myositis without cellulitis or abscesses (Resnick et al. 1995a), radiation therapy (May et al. 1997), subacute denervation (Beltram et al. 1991), compartment syndrome (Shintani and Shiigai 1993), early myositis ossificans (DeSmet et al. 1992), rhabdomyolysis (Shellock and Fleckenstein 1997), and sickle cell crisis (Feldman et al. 1993) as well as in transient, physiological conditions (e.g., after muscle exercises) (Disler et al. 1995).

destroyed muscle fibers and subsequent fibro-fatty replacement. The cause of muscular atrophy is variable and may be due to chronic inflammation and the consecutive failure of membrane proteins, e.g., Na+/K+-ATPase, reduction of physical activity during disease activity and long-term, high-dose corticotherapy (Askari et al 1976). MRI reveals increased quantities of fat with its characteristic signal intensity within the involved muscle, usually associated with a decreased volume of muscle tissue (Fig. 9.5.2). Paradoxically, fatty infiltration may also result in apparent hypertrophy of a chronically denervated muscle, but this condition is rare. Fatty infiltration due to chronic denervation is usually accompanied by muscle atrophy and represents an irreversible type of muscle injury (Fleckenstein et al. 1993). Various non-neurological disorders may mimic chronic denervation. Chronic disuse leads to atrophy with fatty infiltration (e.g., following a tendon tear) as well as corticosteroid use (Resnick 1995). Corticosteroids, especially when used in high doses for a long period, are associated with truncal muscle atrophy and fatty replacement of muscle tissue, which should be considered in the interpretation of images in patients suffering from myo­ pathies who are undergoing therapy with corticosteroids (Resnick 1995). 9.5.3  Mass Lesions and Calcification A localized mass lesion with morphological characteristics and signal intensities different from those of normal muscle is found with all sequences. This pattern may be seen in soft-tissue neoplasm (May et al. 1997), intramuscular abscess (Resnick et al. 1995a), any condition associated with myonecrosis (Stoller and Ferkel 1997), traumatic injuries (DeSmet et al. 1992; Scott 2003), myositis ossificans, muscular sarcoidosis (Matsuo et al. 1995), and parasitic infection (Resnick et al. 1995a). Careful attention to the signal intensity characteristics of the mass may reveal its nature. If a fluid-fluid level is seen, necrosis, blood, or pus is likely to be present. Uniform high signal intensity in T1-weighted images suggests the presence of the blood breakdown product methemoglobin, proteinaceous material, or fat. Intramuscular hematoma and myositis ossificans may demonstrate peripheral calcifications on radiographs and CT.

9.5.2  Fatty Infiltration

9.5.4  Contrast Enhancement

Pathological fatty infiltration of muscle tissues due to an abnormal deposition of fat within a muscle is usually found in association with muscle atrophy. In T1-weighted images, fat exhibits high signal intensity, whereas water exhibits low signal intensity (Fig. 9.5.1). Fatty degeneration occurs in the late stages of various pathological conditions involving the skeletal muscle and is due to the

Areas with increased vascularity (neoplasm, inflammation, etc.) enhance rapidly and retain the contrast longer than normal tissue does (Beltran et al. 1991). The administration of contrast material may be useful to differentiate benign and malignant mass lesions. If enhancement is observed within the central portions of a mass, myonecrosis or liquefaction is unlikely. On the

9.5  MR Imaging Findings Fig. 9.5.1  26-year-old woman with limb girdle dystrophy (calpain deficiency). The T1-weighted image of the composed whole-body (left) shows fatty atrophy of the gastrocnemius muscles, left > right (white arrows). The increased signal intensity of the involved muscle (lower leg, left > right) in the corresponding STIR image on the right side indicated an active inflammatory process (note the extension, white box). The muscles of the thighs are normal

other hand, absence of central enhancement does not allow exclusion of a neoplasm because central necrosis or decreased vascularity may be present (May et al. 1997). An abscess presents as a fluid-filled mass lesion with a surrounding membrane, which strongly enhances after gadolinium. However, gadolinium may slowly diffuse into a fluid-filled space such as an abscess or a hematoma. Therefore, adequate timing of the scan after contrast material injection is mandatory in order to avoid enhancement within a mass that would falsely suggest it to be solid. Myonecrosis may simulate an abscess at both clinical evaluation and MR imaging. The differentiation of focal myositis and a soft-tissue tumor may be difficult. Focal myositis (a benign inflammatory pseudo-tumor) may simulate a neoplasm. MRI reveals a mass-like lesion with high signal

intensity on T2-weighted images and strong enhancement. Active phases of several myopathies demonstrate minor or moderate signal increase after intravenous gadolinium. Rhabdomyolysis, on the other hand, is characterized by strong enhancement of the involved muscle groups. 9.5.5  Lesion Patterns Fatty infiltration may present on T1-weighted images as homogenous or reticular. Muscle dystrophy and metabolic myopathy as well as chronic myopathies are often characterized by homogenous fatty infiltration of the involved muscle groups. Reticular patterns are found in neuropathic, endocrine, or toxic myopathies.

1185

1186

9  Skeletal Muscle

Muscle edema is spotted when it is present in parts of an involved muscle or homogenous when the whole muscle is affected. Homogenous edema is a relatively non-specific finding, whereas the spotted pattern is characteristic for myositis, vasculitis, or rhabdomyolysis. 9.5.6  Distribution Principally primary skeletal muscle diseases, muscular dystrophy, congenital myopathy, and polymyositis are often characterized by distinct patterns and the grade of involvement of selective muscles. In muscular dystrophy and polymyositis, the overall involvement is substantially more severe than in patients with congenital myopathy. Peculiarly, particular muscles are selectively spared from

involvement. For example, the gracilis muscle is significantly less affected than the other muscles of the thigh in all primary skeletal muscle diseases. This also applies to the rectus femoris muscle and the sartorius muscle as well as the tibialis posterior muscle. Common anatomical and functional characteristics of muscles may be related to distribution of muscular disease (Lamminen 1990). Edema and atrophy with fatty replacement should be considered separately. The finding of edema indicates pronounced disease activity with 80% sensitivity and 75% specificity (Lamminen 1990). Fatty replacement and atrophy, on the other hand, are findings associated with chronic diseases (Fig. 9.5.2). Table 9.5.1 gives an overview of the patterns of the most important muscle disorders.

Fig. 9.5.2  32-year-old woman suffering from a limb girdle dystrophy (calpain deficiency). T1-weighted images from the whole-body MRI clearly show the distribution of fatty muscle degeneration, predominantly in the proximal muscles of the extremities. The insets show fatty muscle degeneration in the thighs bilaterally (vastus muscles) and the lower legs (note the extensions on the left side)

+++

–

++

+++

+++

–

+++

+

+

+

–

+++

+++

–

–

+++

+

–

+++

+

+

+

Acute dermatomyositis

Chronic dermato–/ polymyositis

Inclusion body myositis

Focal myositis

Infectious myositis

Muscle dystrophy

Toxic/endocrine/ metabolic m.

Rhabdomyoslysis

Spinal muscle atrophy

Radiculopathy

Polyneuropathy

Traumatic neuropathy

+

–

–

–

–

++

–

–

+++

++

–

–

+

+

CM

R

R

R

H/r

–

R

H

–

–

H/r

H/r

–

H

H

H

H

H/spots

–

H

H/epimysial

H/epimysial

H

H

H/spots/epimysial

H

(STIR, FS T2)

(T1) –

Edema

Fat

Pattern

+

+

+

++

–

++

+

–

–

+++

++

+

+

Atrophy

Variable

+++

Variable

+

–

+++

++

–

–

++

+++

+++

+++

Symmetry

Distribution

–

Caudal

Variable

–

Variable

Cranial>caudal

Cranial>caudal

–

Caudal

Cranial

Cranial

Cranial

Cranial

Predominantly cranial or caudal

Fat fatty infiltration, CM+ contrast material enhancement, + less, ++ moderate, +++ severe, – non, h homogenous, r reticular, spot spotted

+

–

+++

–

Acute polymyositis

Edema

Fat

Disease

Table 9.5.1  MR imaging findings of the most important muscle diseases, overview (modified, according to Beese et al.)

Topic

Topic

Topic

Topic

Deep muscle groups

Not selective

Not selective

Single muscles or muscle groups

Single muscles, often lower leg

Thigh anterior > posterior

Thigh anterior > posterior

Thigh anterior > posterior

Thigh anterior > posterior

Special remarks

9.5  MR Imaging Findings 1187

1188

9  Skeletal Muscle

9.6  Muscle Biopsy In clinical practice, the distinction between inflammatory muscle disease and neuropathic disorders is often difficult because both types of processes can cause weakness and decreased muscle mass. MRI enables excellent soft-tissue differentiation and multiplanar capability, making it very useful in muscle disease diagnosis. MRI is very sensitive in detecting inflammation and edema, especially with fat-suppression sequences. MRI enables early detection of various muscle diseases, allows assessment of the extent, number, and distribution of affected muscles, and thereby facilitates direct biopsy of the area of active disease. Moreover, MRI is useful for monitoring response to therapy (Garcia 2000). T2-weighted images help to distinguish areas of acute and chronic inflammation, normal muscle, and fat. In myositis, areas with inflammatory changes can be identified. Patients with diabetes mellitus (n = 7) and polymyositis (PM) (n = 4) were examined with STIR sequences in order to determine the sites for biopsies. Samples from muscles with abnormalities on MRI had significantly more inflammation as measured by cellular infiltrates than samples from areas that had normal MRI appearance (p = 0.008). Although messenger RNA specific for the inflammation-associated cyclo-oxygenases and lipoxygenases were detected in both inflamed and non-inflamed areas at histology, expression was greater in areas with MRI abnormalities. However, the findings of MRI are not specific for a given muscle disorder. Polymyositis may be difficult to distinguish from inclusion body myositis or other myopathies (Dion et al. 2002). Indications for a muscle biopsy are the presence of some evidence of muscle disease, e.g., weakness, muscle symptoms such as discomfort, cramps, fatigue with activity (rule out myasthenia) as well as elevated (very high or high) CK and myopathic EMG. However, in cases of suspected neurop-

athy a nerve biopsy is indicated. Biopsy is not generally indicated in certain disorders that are better diagnosed by electrodiagnostic methods, such as myasthenia gravis or myotonia. An episode of rhabdomyolysis is a relative indication for a muscle biopsy after a period of 1 month. In chronic disease, muscles with moderate but not severe abnormalities should be the site of the biopsy. In acute diseases, the muscles with severe or moderate weakness have to be defined. The best muscles for biopsy are the deltoid, the biceps, or the quadriceps. In conclusion, MRI with its high sensitivity for muscle abnormalities is an excellent procedure for locating the biopsy site. 9.7  Skeletal Muscle Diseases 9.7.1  Muscle Skeletal muscle diseases include inherited syndromes and acquired entities. Table 9.7.1 demonstrates the classifications. 9.7.1.1  Acquired Myopathies 9.7.1.1.1  Inflammatory Myopathy Inflammatory myopathies are frequently caused by immune processes or other inflammatory processes and can occur in a large number of very different conditions (Table 9.7.2). Idiopathic inflammatory myopathy syndrome is constituted by a heterogeneous group of diseases of unknown etiology that have an inflammatory muscle process in common. It is characterized by proximal muscular weakness, elevation of serum enzymes related mainly to striated musculature (e.g., CK-MB),

Table 9.7.1  Classification of muscle diseases Inherited clinical syndrome

Molecular syndromes

Acquired

Congenital weakness

AChRs

Inflammatory myopathies

Distal myopathies

DNA repeat

Serum autoantibodies

Limb-girdle syndrome

Dystrophin

Infections

Myoglobinuria

Glycogen metabolism

Paraneoplastic

Myotonia

Intermediate filaments

Renal, respiratory, skin, toxic, gastrointestinal, endocrine

Ion channels Mitochondria Structural proteins

9.7  Skeletal Muscle Diseases Table 9.7.2  Some of the most important inflammatory conditions and their typical symptoms Syndrome

Distinctive features

Idiopathic

Proximal weakness, high CK, inflammatory myopathy

Collagen vascular disease

Myalgia, younger onset, scleroderma & mixed connective disease

Jo-1 antibodies

Interstitial pneumonitis, Raynaud’s phenomena, arthritis, perimysial pathology

Signal recognition particle antibodies

Acute onset, severe weakness, capillary pathology, fibrosis

MAS antibody

Acute onset, rhabdomyolysis

Drug induced

d-Penicillamine

Familial

Homozygosity at HLA-DQA1 locus

Graft-versus-host disease

7 –24 months post–bone marrow transplantation

Granulomatous

Sarcoid, infections, autoimmune diseases

Sarcoidosis

Myopathy, neuropathy, lung disease

Malignancy (necrotic)

Rapid onset, older patients, necrotic myopathy

Mitochondrial (p-Cox)

Quadriceps weakness, older onset, steroid resistant

Other systemic disorders

HIV, fasciitis

alterations in electromyography, and the presence of mononuclear infiltrates within the muscles (Pichiecchio et al. 2004) . This group of diseases presents with a large range of histopathological findings, responses to treatment, and clinical courses. Dermatomyositis, polymyositis, and inclusion body myositis are the most important types of idiopathic inflammatory myopathy. Myopathies may also be associated with other connective tissue diseases, especially systemic lupus erythematosus, scleroderma, and rheumatoid arthritis. 9.7.1.1.2  Polymyositis Polymyositis and dermatomyositis are autoimmune inflammatory conditions of skeletal muscles characterized by gradual onset of muscle weakness in the thighs and pelvic girdle, which typically progresses to involvement of the upper extremities, neck flexors, and pharyngeal musculature (Resnick and Niwayama 1995). These conditions are caused by cell-mediated (type IV) autoimmune attack on striated musculature. Polymyositis (PM) involves only skeletal muscles; dermatomyositis (DM) both skeletal muscles and skin. However, these conditions may overlap in clinical and imaging features. Polymyositis frequently manifests during the fourth decade of life. Dermatomyositis has a dual distribution of manifestation, with peaks during childhood and the fifth decade of life. Childhood-onset dermatomyositis tends to be more severe than the adult-onset form (Resnick and

Niwayama 1995). However, the adult-onset form is associated with increased prevalence of a variety of malignancies, including those of the breast, prostate, lung, and gastrointestinal tract. Several serological abnormalities have been identified and may be helpful in the classification of subtypes and assessment of prognosis, but they are not used for routine diagnosis. As a group, these antibodies have been termed myositis-specific antibodies (MSAs). These autoantibodies occur in about 30% of all DM and PM patients. A positive antinuclear antibody (ANA) titer is common in patients with DM. Anti–Mi-2 antibodies are highly specific for DM but lack sensitivity because only 25% of the patients with DM are positive for this antibody. The presence of the autoantibodies is associated with acuteonset classic DM with the V-shaped and shawl rash (poikiloderma) and a relatively good prognosis. Anti-Jo-1 (antihistidyl transfer RNA [t-RNA] synthetase) is more frequent in patients with PM than in patients with DM. It is associated with pulmonary involvement (interstitial lung disease), Raynaud’s phenomenon, and arthritis. Other MSAs include anti–signal-recognition protein (anti-SRP), associated with severe PM. One of the major challenges related to the diagnosis and classification of myopathies is the search for highly sensitive and specific non-invasive methods of diagnosis as well as for treatment follow-up. Among the imaging methods, MRI has shown itself to be superior to any other modality. Muscle weakness occurs predominantly in the proximal portions of the extremities and in a symmetric manner. The involvement of the esophagus leads to dys-

1189

1190

9  Skeletal Muscle

phagia; further, the muscles of the posterior neck and the quadriceps muscle are also affected. The age of onset of the disease is usually >20 years. There is a slow progression over several months. Associated disorders are: • Cardiac: arrhythmias; inflammatory cardiomyopathy • Pulmonary: respiratory muscle weakness; interstitial lung disease • Esophageal paresis • Malignancy: mildly increased risk • Autoimmune activity • Lupus • Sjögren’s syndrome • Anti-phospholipid antibodies and syndrome 9: 5–8% • Thyrotoxicosis: may resolve with anti-thyroid medication alone

ease, steroids as first-line therapy and immunosuppressive agents in steroid–non-responders are used. Pathologically the perimysial connective tissue is affected in Jo-1 syndrome. Typical is a myopathy with perifascicular atrophy, degeneration, and regeneration, most pronounced in close proximity to pathological alterations of the connective tissue. These findings are most prominent in patients with more severe weakness. The inflammation in endomysial and perivascular regions is less common than in other types of polymyositis. The prognosis is severe with a high mortality due to systemic (lung) involvement (Fig. 9.7.2). The best treatment option is corticosteroids.

Early in the course of polymyositis and dermatomyositis, bilateral and symmetric edema in pelvic and thigh musculature are found in MRI, especially in the vastus lateralis and vastus intermedius muscles (Fraser et al. 1991; Lamminen 1990). After a course of months to years, the muscles involved by PM or DM exhibit an increasing content of fat. 9.7.1.1.3  Jo-1 Syndrome Antisynthetase antibodies like anti–Jo-1 antibodies are known to be highly specific for inflammatory myopathies. Patients with this antibody often exhibit a combination of symptoms, including interstitial lung disease, fever, polyarthritis, inflammatory myositis (Fig. 9.7.1), and Raynaud’s syndrome. In the management of this dis-

Fig. 9.7.1  MR images of a patient suffering from a Jo-1-syndrome. a Coronal STIR-image of the distal thigh and proximal lower leg shows an increase of signal intensity in the sartorius muscle (arrow), left > right, due to extensive edema. b The coronal native T1-weighted image, however, demonstrates a normal signal intensity of the involved muscle, with blurring of its contours. c,d The fat suppressed T1-weighted images after Gd-DTPA i.v. coronal c and axial d show an increased enhancement of the involved muscles (sartorius, semimembranosus, biceps femoris), suggestive of an active phase of the inflammatory muscle disease without a fatty replacement

Fig. 9.7.2  Computed tomography of the lung in a patient with Jo-1-syndrome. Sixty percent of these patients show involvement of the lung, which is responsible for the poor prognosis. Fibrosis of the lung with predominant affection of the lung bases, which manifests as fine to coarse interstitial patterns, honey-combing, and reduction of lung volume

9.7  Skeletal Muscle Diseases

9.7.1.1.4  Myositis and Collagen Vascular Diseases These diseases are most frequently found in young individuals, predominantly in females. Muscle pain and arthralgia are common symptoms. They are mediated by myositis-overlap antibodies and respond to lower doses of corticosteroids. The incidence of this muscle disease in scleroderma is 14–17%. The heart is frequently involved, and many patients suffer from pulmonary fibrosis and finger contractures. Mixed connective tissue disease (U1-RNP antibody) is characterized by systemic lupus erythematosus; systemic sclerosis, and inflammatory myopathy. The muscle disease is often mild and non-progressive. 9.7.1.1.5  Signal-Recognition Particle (SRP) Antibody Syndrome Females are affected more often than males are, and the age of onset is usually in the fourth and fifth decade. The acute phase of the disease lasts from weeks to months and shows a peak from July to January. The patients suffer from severe muscle weakness and myalgia. The proximal muscle groups are more frequently affected. Symptoms of systemic disease are palpitations, dyspnea on exertion and Raynaud’s phenomenon (30%), but no skin changes. The courses of the disease are severe or fulminate with frequent recurrences and a high rate of mortality. Serum CK is high to very high. There is an association with human leukocyte antigen (HLA)-DRw52, -DR5 in 57% of cases. The antibody target is a protein, the anti–signalrecognition particle. 9.7.1.1.6  McCune-Albright Syndrome The average age of onset is 36 years. The illness is characterized by short duration with an acute onset. Clinical signs are rhabdomyolysis (alcoholic), myalgia (50%), and palpitations. 9.7.1.1.7

Myositis with Antibodies to U1 Small Nuclear Ribonucleoprotein (U1 snRNP) 28

This disease predominantly affects females aged 36–48 years. Clinical signs are swollen hands, Raynaud’s phenomenon, and mild-to-severe weakness of the proximal muscles. Other organ systems may be also involved: • Lung: interstitial pneumonitis, vital capacity reduced due to lung disease • Arthritis: symmetrical, non-destructive, distal joints • Raynaud’s syndrome: some patients

• • • •

Sicca syndrome: some patients Lymphadenopathy (20%) Neurological: lymphocytic meningitis (20%) Systemic: fever, weight loss

9.7.1.1.8  Familial Idiopathic Myopathy The clinical signs are similar to sporadic polymyositis. A known genetic risk factor is homozygosis at HLA-DQA1 locus. 9.7.1.1.9  Malignancy-Associated Necrotic Myopathy These entities are more frequent in higher age groups. Patients suffer from weakness of proximal muscles, which are involved symmetrically. Usually the course of the diseases is characterized by rapid determination and disability within 2–3 months. The survival rate is related to the response to treatment of the underlying lung cancer, cancer of the stomach, colon (adenocarcinoma), or prostate. Histology demonstrates muscle fiber necrosis with small and large foci of macrophages and C5b-9 complement deposition in muscle fibers. The therapy with corticosteroids is effective in treatment of this type of myopathy, if the neoplasm is under control. 9.7.1.1.10  Drug-Induced Myositis d-Penicillamine may result in polymyositis in about 40% of treated patients and can occur at any time during treatment. Muscle weakness affects predominantly proximal muscles. Other symptoms may be dysphagia in 40% and myocarditis or conduction block. Eighty percent of patients have a positive ANA titer. d-Penicillamine may also be responsible for myasthenia gravis, neuromyotonia, and pemphigus. Phenytoin may result in inflammatory myopathy with fever, rash, lymphadenopathy, and eosinophilia. Other drugs possibly related to myositis are penicillin, sulfonamide, levodopa, cimetidine, leuprolide, propylthiouracil, and carbimazole. 9.7.1.1.11  Graft-Versus-Host Disease Graft-versus-host disease (GVHD) is defined as a cellmediated immune reaction due to donor cytotoxic T cells (CD8+) and lymphokine-secreting helper T cells, resulting in a delayed-type hypersensitivity reaction versus recipient cells. The disease is caused by a donor-recipient HLA major histocompatibility complex disparity and involves different end organs:

1191

1192

9  Skeletal Muscle

Acute phase (<100 days after bone marrow transplantation): Skin: maculopapular exanthem, epidermal necrolysis Liver: similar to primary biliary cirrhosis Bowel: microangiopathy Chronic phase (>80 days after bone marrow transplantation): Skin: Lichen planus-like, scleroderma, vitiligo Lichenoid lesions: oral & esophageal mucosa Keratoconjunctivitis sicca Bronchiolitis: obstructive pulmonary failure Myositis Myasthenia gravis Chronic immune demyelinating polyneuropathy (CIDP) Guillain-Barré syndrome Fasciitis CNS (rare): myelitis, otic neuritis

9.7.1.1.12  Multinodular Polymyositis The multinodular polymyositis is a very rare condition; recently some case reports were published. The age of onset is 40–76 years. Clinical signs are predominant proximal muscle weakness and symmetric involvement with a progression over months. In 80% of patients dysphagia is a major problem. Characteristic are masses of nodules in multiple locations in the muscles of limbs and trunk. The disease is associated with HIV infection and hepatitis C infection. Differential diagnoses are focal myositis, myositis ossificans, and granulomatous myositis. Spontaneous improvement may occur. The nodular enlargements present with increased signal intensity in T2-weighted images and a strong enhancement with gadolinium.

Fig. 9.7.3a–c  MRI of the lower leg in a 41-year-old man with dermatomyositis. Duration: 8 months, CK: 650 IU/l, moderate weakness and localized involvement of the muscle of the lower leg. a The coronal STIR image shows increased signal intensity longitudinal diffuse high-intensity area in the medial head of the left gastrocnemius muscle (arrow). b Correspond-

9.7.1.1.13  Dermatomyositis Dermatomyositis is an idiopathic inflammatory myopathy (IIM) with characteristic cutaneous findings. In 1975, Bohan and Peter suggested a set of criteria to aid in the diagnosis and classification of DM and polymyositis. Four of the five criteria are related to the muscle disease: progressive proximal symmetrical weakness, elevated muscle enzymes, an abnormal finding on electromyogram, and an abnormal finding on muscle biopsy. The fifth criterion was compatible cutaneous disease. The association between DM (and possibly PM) and malignancy has been recognized for a long time. DM is a systemic disorder that frequently affects the esophagus and lungs. Less commonly, the disease may affect the heart. Calcinosis is a complication that is observed most often in children or adolescents. Therapy of the muscle disease involves the use of corticosteroids with or without an immunosuppressive agent. The skin disease is treated with sun avoidance, sunscreens, topical corticosteroids, antimalarial agents, and/or methotrexate. The prognosis depends on the severity of the myopathy, the presence of a malignancy, and/or the presence of cardiopulmonary involvement. Bohan and Peter (1975) suggested five subsets of myositis: • DM (Fig. 9.7.3) • PM • Myositis with malignancy • Childhood DM/PM • Myositis overlapping with another collagen–vascular disorder A skin biopsy reveals an interface dermatitis that is difficult to differentiate from systemic lupus erythematosus.

ing T1-weighted image of the lower leg in the same patient exhibits a moderate reticular hyperintensity of the muscle due to beginning fatty replacement. In c, the involved muscles present a moderate contrast enhancement in the fat-suppressed T1weighted image

9.7  Skeletal Muscle Diseases

Vacuolar changes of the columnar epithelium and lymphocytic inflammatory infiltrates at the dermal–epidermal junction basement membrane may be present. Muscle biopsy in DM reveals perivascular and interfascicular inflammatory infiltrates with adjoining groups of muscle fiber degeneration/regeneration. This contrasts with PM infiltrates, which are mainly intrafascicular (endomysial inflammation) with scattered individual muscle fiber necroses. DM may cause death because of muscle weakness, cardiopulmonary involvement or due to the underlying malignancy. Most patients with DM may develop residual weakness and disability. In children with severe disease, contractures can develop. 9.7.1.1.14  Inclusion Body Myositis Inclusion body myositis (IBM) is a very rare disease. In contrast to the other myopathies men are affected more commonly than women (male-to-female ratio is 3:1), the patients tend to be older (an overwhelming majority of affected individuals are 50 years or older). Sporadic inclusion body myositis (s-IBM) and inherited inclusion body myopathies (i-IBM) encompass a group of disorders that have the common pathological finding of inclusion bodies on muscle biopsy. They demonstrate a wide variation in clinical expression, age of onset, associated diseases, and prognosis. The term inclusion body myositis was originally used by Yunis and Samaha in 1971 for a case of myopathy that phenotypically suggested a case of chronic polymyositis but showed cytoplasmic vacuoles and inclusions on muscle biopsy (Fig. 9.7.4). Expression of s-IBM is variable, but all cases eventually evolve into a syndrome of diffuse, progressive weakness that usually does not respond to immunosuppressive treatment. The distribution of weakness in s-IBM is variable, but both proximal and distal weakness can be present. Characteristic changes are predominant knee extensor weakness in the legs and wrist/finger flexor weakness in the arms. A common and distinct characteristic of s-IBM is

Fig. 9.7.4  Chronic course of inclusion body myositis (IBM) in a 56-year-old man. Axial STIR image of the distal thigh. The mostly fatty infiltrated extensor muscle group shows no edema; the ischiocrural muscle groups exhibit a moderate and bilateral

its severe involvement of the wrist and finger flexors, out of proportion to their extensor counterparts. Hence, loss of finger dexterity and grip strength may be a presenting or prominent symptom. The disproportionate and marked weakness in the wrist and finger flexors relative to her corresponding extensors prompted the treating physician to consider superimposed cervical radiculopathy. While the distribution of weakness is typically symmetric in polymyositis, it is often asymmetric in s-IBM. Because of slow progression, any treatment trial should last for at least 6 months (possibly 12–18 months) to evaluate benefit (Askanas et al. 1998). Viral infections by paramyxovirus are the likely cause of inclusion body myositis (Sanders 1995), if secondary IBM is present. Unfortunately there is no causal effective treatment for inclusion body myositis. Patients may be given a trial course of prednisone, followed by methotrexate or azathioprine. However, if there is no improvement within 2–3 months, all drugs should be discontinued. Physical therapy and exercise are important in the treatment of inflammatory myopathies. 9.7.1.1.15  Other Inflammatory Myopathies A large number of conditions may induce inflammatory processes in skeletal muscles, the most important are in Table 9.7.3.

Table 9.7.3  Further causes of inflammatory myopathies Other types of inflammatory myopathies Infection Focal myositis Granulomatous Necrotizing myonecrosis Fasciitis Perimyositis Myositis ossificans

increase of signal intensity. On the axial T1-weighted image there is pronounced fatty replacement of the knee extensors and vastus muscles, which is asymmetric in this particular case

1193

1194

9  Skeletal Muscle

9.7.1.1.16  Infections (Myositis)

9.7.1.1.17  Granulomatous Myopathy

Muscle infection without abscess formation or necrosis may produce edema as the sole abnormality on MR images. Bacterial myositis may result from direct extension of infection in tissues adjacent to a muscle, such as osteomyelitis or a subcutaneous abscess (Resnick et a. 1995a). Bacterial myositis frequently progresses to abscess formation and thus often has a mass-like appearance on MR images (Fig. 9.7.5). An intramuscular abscess may be caused by various infectious agents, frequently Staphylococcus aureus. Bacterial infections cause fever, pain, swelling, and marked tenderness. If MR imaging is performed before a phlegmon or abscess is formed, the only finding may be muscle edema. Fungal infections are less common but may also cause abscesses. Necrotizing fasciitis is often associated with myonecrosis and abscess formation in adjacent muscles. Infectious agents which affect skeletal muscles include: • Viral • Benign acute myositis • Acute: influenza A and B • Parainfluenza; adenovirus • Coxsackie B (Bornholm disease, epidemic pleurodynia) • Retrovirus: HIV, HTLV-1 • Bacterial • Legionnaire’s disease • Leptospirosis • Lyme • Pyomyositis (S. aureus) • Clostridia perfringens • Tuberculosis • Parasitic • Protozoa: toxoplasmosis, trypanosomiasis, microsporidosis, malaria • Cestode: cysticercosis, echinococcosis • Nematode: trichinosis, larva migrans • Other: Candida, Cryptococcus, Cunninghamella

Muscular sarcoidosis, as a typical representative of a granulomatous inflammatory systemic disease with muscle involvement, results in distinctive, well-circumscribed intramuscular nodules that contain a central starshaped area of fibrosis surrounded by an inflammatory granuloma (Matsuo et al. 1995). The central fibrosis has low signal intensity in all sequences, and the surrounding granuloma exhibits high signal intensity that is slightly higher than that of normal muscle on T1-weighted images and much higher on T2-weighted images. Frequently, surrounding edema is present (Fig. 9.7.6). Occasionally the inflammatory process occurs with an increase of muscle volume. Biopsy is required to confirm the diagnosis. Muscular sarcoidosis is more likely to respond to immunotherapy than granulomatous myopathy without sarcoidosis. Other rare causes of a granulomatous myopathy are rheumatoid arthritis, mixed CTD, Wegener’s disease, myasthenia gravis, Crohn’s disease and other infections. 9.7.1.1.18  Focal Myositis Focal myositis is a rare idiopathic inflammatory condition involving a single muscle, which frequently presents as a soft-tissue pseudotumor. MR images reveal a small mass-like lesion with high signal intensity on T2-weighted images and intense enhancement after intravenous administration of Gd chelates. Resection is often curative (Fleckenstein and Reimers 1996). Clinical signs are pain and local enlargement. In principle, all muscles may be involved, including the arm, forearm, neck, tongue, or psoas muscle. However, focal myositis typically occurs in the legs, involving the quadriceps or gastrocnemius muscles. Focal myositis may simulate a neoplasm, so that soft sarcomas have to be excluded. Biopsy reveals a T-cell–predominant inflammation with perimysial connective tissue thickening. Fig. 9.7.5a,b  Intramuscular abscess, caused by Staphylococcus aureus. Both images show a mass-like lesion in the thigh muscles of the left side, with central hypointensity and strong enhancement of the abscess membranes and adjacent tissues. a Axial fat-suppressed T1-weighted image of the proximal thigh. b Coronal fat-suppressed T1-weighted image

9.7  Skeletal Muscle Diseases Fig. 9.7.6a,b  Muscular sarcoidosis. Intramuscular nodules that contained several central starshaped areas of fibrosis (arrows) surrounded by an inflammatory granuloma. a Axial T1-weighted image. b Axial T1-weighted image with fat saturation after GD-DTPA

Fig. 9.7.7a–c  Diabetic myonecrosis of the left quadriceps muscles in a 43-year-old man with poorly controlled diabetes mellitus and an extremely painful and tender left thigh. a Axial T1-weighted MR image. b Axial T2-weighted fast spin-echo MR image obtained with fat suppression. c Axial fat-suppressed T1-weighted MR image obtained after intravenous administra-

tion of gadolinium contrast material. The images show an enlargement of the left quadriceps muscles, especially the vastus medialis (arrows), as well as diffuse edema throughout the left quadriceps muscles with enhancement. There is a region of absent enhancement (× in c) in the left vastus medialis muscle, a finding consistent with myonecrosis

9.7.1.1.19  Myonecrosis

Necrotizing fasciitis has also been referred to as hemolytic streptococcal gangrene, Meleney ulcer, acute dermal gangrene, hospital gangrene, suppurative fasciitis, and synergistic necrotizing cellulitis. Fournier gangrene is a form of necrotizing fasciitis that is localized to the scrotum and perineal area. Necrotizing fasciitis is a progressive, rapidly spreading, inflammatory infection located in the deep fascia, with secondary necrosis of the subcutaneous tissues. If gas-forming organisms are present, subcutaneous air is classically described in necrotizing fasciitis. This may be seen only on radiographs and is not found in all cases. The speed of spread is directly proportional to the thickness of the subcutaneous layer. Necrotizing fasciitis moves along the deep fascial plane. These infections can be difficult to diagnose in their early stages, but they rapidly progress. They require aggressive treatment to combat the associated high morbidity and mortality (70–80%). The causative bacteria may be aerobic, anaerobic, or mixed flora. Surgical procedures may cause local tissue injury and bacterial invasion, resulting in necrotizing fasciitis. These procedures include surgery for intraperito-

Myonecrosis can occur in sickle cell crisis (Feldman et al. 1993), diabetes (Van Slyke and Ostrov 1995), severe ischemia, crush injury (Resnick et al. 1995a), and rhabdomyolysis (Shellock and Fleckenstein 1997). Any condition that can cause myonecrosis can simulate an abscess at both clinical evaluation and MR imaging. Diabetic myonecrosis is a distinctive myopathy associated with poorly controlled diabetes mellitus (Van Slyke and Ostrov 1995). Clinical features include severe pain, most frequently in the thigh muscles, with comparatively mild physical examination findings and occasionally a low-grade fever. MRI demonstrates marked edema and enhancement around an often irregular mass-like lesion. (Fig. 9.7.7). 9.7.1.1.20  Fasciitis and Perimyositis Necrotizing fasciitis can occur after trauma or around foreign bodies in surgical wounds, or it can be idiopathic, as in scrotal or penile necrotizing fasciitis.

1195

1196

9  Skeletal Muscle

neal infections and drainage of ischiorectal and perianal abscesses. Intramuscular (i.m.) injections and intravenous (i.v.) infusions may also lead to necrotizing fasciitis. Minor insect bites may set the stage for necrotizing infections. Streptococci can be introduced into the wounds, but the bacteriologic pattern may change due to hypoxiainduced proliferation of anaerobes. Local ischemia and hypoxia can occur in patients with systemic illnesses (e.g., diabetes). Host defenses can be compromised by underlying systemic diseases favoring the development of these infections. Illnesses such as diabetes or cancer have been described in over 90% of cases of progressive bacterial gangrene. The number of diabetic patients has been reported to be 20–40%. As many as 80% of Fournier gangrene cases occur in individuals with diabetes. In some series more than 35% of the patients were alcoholics. Recent studies have shown a possible relationship between the use of non-steroidal anti-inflammatory agents (NSAIDs), such as ibuprofen, and the development of necrotizing fasciitis during varicella infections. MRI and CT have been utilized recently in the diagnosis of necrotizing fasciitis. Absence of contrast enhancement on T1-weighted imaging reliably indicates fascial necrosis in those requiring operative debridement. Combined with clinical assessment, MRI allows determination of the presence of necrosis and the need for surgical debridement (Brothers et al. 1988). A synonym of eosinophilic fasciitis is eosinophilia–myalgia syndrome. The age of onset is mostly 30–60 years. More men are affected than females are. Prodromal symptoms may consist of fever, myalgia, and arthralgia as well as fatigue. Clinical features involved the skin (peau d’orange, edema, thickening, but no Raynaud’s syndrome), the muscle (pain, weakness, distal > proximal) and joints as well as the bone marrow (aplastic anemia, lymphoma, leukemia). Serum CK is normal or mildly increased. Histopathology is characterized by lymphocytes, eosinophils, and macrophages. Muscles are affected in areas close to inflamed fascia with perimysial inflammation (perimyositis) and perifascicular atrophy. 9.7.1.1.21  Myositis Ossificans Myositis ossificans is an aberrant reparative process that causes benign heterotopic (extraskeletal) ossification in soft tissues. Von Dusch first suggested the term myositis ossificans in 1868. The term, however, is a misnomer because the condition involves no muscle inflammation, and the process is not limited to muscle. Myositis ossificans manifests in two forms. 9.7.1.1.22  Myositis Ossificans Circumscripta Myositis ossificans circumscripta can develop either in response to soft tissue injury (e.g., blunt trauma, stab

wound, fracture/dislocation, surgical procedures) or can occur without known injury. Potential mechanisms of atraumatic myositis ossificans include non-documented trauma, repeated minor mechanical injuries, and nonmechanical injuries caused by ischemia or inflammation. Clinical symptoms include pain, tenderness, focal swelling, and a restricted range of motion reduction. Rarely, the condition is asymptomatic and may be diagnosed from radiographs obtained for unrelated problems. However, the history of trauma may be difficult to elicit. Most (80%) ossifications arise in the thigh or arm. Other sites include intercostal spaces, erector spinae, pectoralis muscles, glutei, and the chest. 9.7.1.1.23  Hereditary Myositis Ossificans Progressiva Myositis ossificans progressiva is an autosomal dominant genetic disorder with complete penetrance and variable expression. Over-expression of bone morphogenetic protein 4 and its messenger RNA occurs; this protein has been mapped to chromosome band 14q22–q23. The condition produces painful lumps and stiffness in the neighboring joint. Lumps decrease in a few weeks, but joint mobility reduction persists. Exacerbating factors for ossifications at new sites include venous puncture, biopsy of lumps, i.m. injections, dental treatments, and excision of masses. The most common sites are the sternocleidomastoid muscle, paraspinal muscles, the jaw’s masticatory muscles, as well as shoulder and pelvic girdle muscles. The abdominal muscles, extraocular muscles, and gastrointestinal tract and tongue muscles are spared. Ossification progresses from proximal to distal and cranial to caudal. MRI findings depend on the age of the lesion (Resnick 1995). In immature lesions, T2-weighted spin-echo images show a homogeneous hyperintense soft-tissue mass. Surrounding edema may be seen in lesions less than a few months old. In T1-weighted images, only mass effect may be noted with displacement of fascial planes. Mature lesions appear as inhomogeneous masses with central, fatlike signal intensity on both T1- and T2-weighted images. 9.7.1.2  Inherited Myopathies 9.7.2.1.1  Congenital Myopathy Several myopathies have been defined as congenital myopathies. Hypotonia is the clinical hallmark of congenital myopathies. It presents in the neonatal period as head lag, inability to bend the hips, knees, and elbows and to rotate the hips externally, diffuse weakness in facial, limb, and axial muscles and reduced muscle mass. The following types can be differentiated: • Nemaline rod myopathy (20%) • Central core disease (16%)

9.7  Skeletal Muscle Diseases

Fig. 9.7.8a,b  Distal thigh in a patient with a multiminicore disease. a The T1-weighted image illustrates the extensive fatty degeneration of the quadriceps muscles bilaterally with a mod-

erate sparing of the ischiocrural muscles; b the corresponding STIR image reveals no signs of intramuscular edema

• Centronuclear myopathy (14%) • Minimulticore myopathy (10%) (Fig. 9.7.8) • Congenital fiber-type disproportion or type 1 fiber predominance (21%) • Six other miscellaneous congenital myopathies (19%)

T1-weighted MR images reveal hyperintense fatty infiltration interspersed between the diseased muscles. The mean fat mass is significantly higher in diseased than in normal muscle (Leroy-Willig et al. 1997). MR imaging enables determination of the severity of fatty infiltration, which parallels the decrease in muscle strength. The muscle in the thigh that is most resistant to disease is the gracilis, followed by the sartorius, semitendinosus, and semimembranosus muscles (Liu et al. 1993). Asymmetric involvement of the thigh muscles is not unusual. The most characteristic histological features are the presence of hyaline fibers, adjacent parts of focal fiber necrosis, and phagocytosis.

In all myopathies, CK level is either in the reference range or mildly elevated. It can be elevated moderately in central-core disease (CCD) and may also be elevated in asymptomatic carriers of the ryanodine receptor mutation in CCD. If the CK level is very high, other disorders such as Duchenne, Becker, or limb-girdle muscular dystrophy should be considered. In congenital myopathy, nerve conduction study findings are normal and EMG findings are either normal or show the typical small-amplitude, narrow-duration motor unit potentials (MUPs) that are seen in myopathies. Fibrillations and positive sharp waves are rare. Muscle biopsy has to be performed in all patients in whom congenital myopathy is suspected (Goebel 2003). MR imaging (Fig. 9.7.8) enables determination of the severity of fatty infiltration, which parallels the decrease in muscle strength (Liu et al. 1993). 9.7.1.2.2  Inherited Progressive Muscular Dystrophy Congenital myotonia (CM) and congenital muscular dystrophy (CMD) are separate and distinct entities. However, some of the symptoms are similar. The muscular dystrophies (MDs) are a group of genetic diseases characterized by progressive weakness and degeneration of the skeletal muscles. There are many forms of muscular dystrophy, some noticeable at birth (congenital muscular dystrophy), others in adolescence (Becker MD), but the three most common types are Duchenne MD, facioscapulohumeral MD, and Becker MD. These three types differ in patterns of inheritance, age of onset, rate of progression, and distribution of weakness.

9.7.1.2.3  Congenital Muscular Dystrophy The congenital dystrophies are a group of conditions which share early presentation and a similar appearance of the muscle. Congenital means “from birth” and in congenital muscular dystrophy the initial symptoms are present at birth or in the first few months. Congenital muscular dystrophy is a very heterogeneous group of conditions, and in the last few years major efforts have been made in identifying the separate entities and in locating the responsible genes. Babies with congenital muscular dystrophy often have hypotonia (low muscle tone or floppiness), and may have reduced movements. Other common signs are contractures (tightness) in the ankles, hips, knees, and elbows. The contractures can sometimes be severe and affect several joints (known as arthrogryposis). They develop because of reduced muscle strength, so that the babies are unable to move freely in the womb. Some of these babies may also have respiratory problems because of weakness of breathing muscles. In some children who do not have contractures the first problems are only noted after a few months because of difficulties in holding the head or delay in learning how to sit unaided, stand or walk. The gene responsible for merosin-deficient congenital muscular dystrophy was the first gene to be identified in a

1197

1198

9  Skeletal Muscle

Fig. 9.7.9a,b  A 9-year-old girl with a merosin-deficient muscular dystrophy. a Axial T1-weighted image of the thigh demonstrates the extensive fatty replacement of almost all muscle groups except the rectus femoris and the sartorius muscles. The changes are symmetrical. b On the axial STIR image, the involved muscle does not show increased signal intensity, and edema can therefore be excluded

subpopulation of patients with congenital muscular dystrophy and lies on chromosome 6 (1p35–p36). This gene is responsible for the production of laminin 2, which contributes to the protein called merosin (Fig. 9.7.9). In general, CMDs are autosomal recessive diseases resulting in severe proximal weakness at birth (or within the first 12 months of life) that is either slowly progressive or non-progressive. Contractures are common, and CNS abnormalities can occur. Muscle biopsy is indicated in all cases of suspected CMD to help confirm the diagnosis and exclude other causes of weakness. Muscle biopsy shows signs of dystrophy: marked increase in endomysial and perimysial connective tissue; variability in fiber size with small, round fibers; immature muscle fibers, and (uncommonly) necrotic muscle fibers. No distinguishing features are present on muscle biopsy as in congenital myopathies.

Duchenne muscular dystrophy primarily affects boys and is the result of mutations in the gene that regulates dystrophin, a protein involved in maintaining the integrity of muscle fiber. Onset is between 3 and 5 years, and the disease progresses rapidly. Most boys become unable to walk at 12, and by 20 have to use a respirator to breathe. Early signs of Duchenne, which usually occur between the ages of 2 and 6, include frequent falling, difficulty getting up from a sitting or lying position, and a waddling gait. Another hallmark is the apparent enlargement of the calf and sometimes other muscles, which is really due to an accumulation of fat and connective tissue in the muscle (Figs. 9.7.10, 9.7.11). A blood sample shows a very high level of CK. Progression varies somewhat from child to child. The use of orthopedic devices and physical therapy can prolong the ability to walk. Frequently, however, a wheelchair will be needed by age 12. Mild mental retardation has been noted in some (but by no means all) boys with Duchenne dystrophy (Lamminen 1990; Liu et al. 1993). Fig. 9.7.10  Coronal T1weighted image of the thighs in a 9-year-old patient with Duchenne muscular dystrophy. The predominant findings are found symmetrically in the pelvis and the upper legs

9.7.1.2.4  Duchenne Muscular Dystrophy Age of onset:

2–6 years

Inheritance/gender affected:

X-linked/males

Muscles first affected:

Pelvis, upper arms, upper legs

Progression:

Slow, sometimes with rapid spurts

Fig. 9.7.11a,b  Carrier type of Duchenne muscular dystrophy. a T1-weighted image of the thigh demonstrates the typical predominant involvement of the adductor and the flexors muscles with sparing of the quadriceps muscles and a symmetric distribution. b More distally the fatty atrophy of the involved muscles is pronounced

9.7  Skeletal Muscle Diseases

9.7.1.2.5  Becker Muscular Dystrophy

9.7.1.2.7  Limb-Girdle Muscular Dystrophy

Age of onset:

2–16 years

Age of onset:

Teens or early adulthood

Inheritance/gender affected:

X-linked/males

Inheritance/gender affected:

Muscles first affected:

Pelvis, upper arms, upper legs

Autosomal recessive and dominant forms/ males and females

Progression:

Slow

Muscles first affected:

Hips, shoulders

Progression:

Usually slow

Becker muscular dystrophy is a slowly progressive primary degeneration of skeletal muscle, with an onset age of approximately 11 years and age at death up to the fourth decade. The disease is inherited as an X-linked recessive disorder, predominantly in boys. Duchenne and Becker muscular dystrophies affect the same dystrophin genetic system and have the same muscular involvement, but the Becker type is less pronounced. Symptoms appear in the lower limbs 5 to 10 years before they occur in the upper limbs. Serum CK in Becker muscular dystrophy is raised to a degree similar to that found in the Duchenne type. The selective muscular involvement in Becker muscular dystrophy is apparent on T1-weighted MR images. The rectus femoris, adductor longus, gracilis, sartorius, semitendinosus, and semimembranosus muscles are relatively spared and even hypertrophied in the thighs. Enlargement of the calves due to fatty infiltration (i.e., pseudohypertrophy) of the bilateral gastrocnemius and soleus muscles occurs in the Becker type, but it is less severe than in Duchenne muscular dystrophy. The histology of Becker muscular dystrophy resembles that of the Duchenne type, except that in the Becker type hyaline fibers are relatively uncommon and regenerative fiber clusters are often seen (Lamminen 1990). 9.7.1.2.6  Facioscapular Muscular Dystrophy Age of onset:

Teens or early adulthood

Inheritance/gender affected:

Autosomal dominant/ males and females

Muscles first affected:

Face, shoulders

Progression:

Slow, sometimes with rapid spurts

Facioscapulohumeral muscular dystrophy becomes apparent in adolescence and causes progressive weakness in facial muscles and various muscles in the arms and legs. It progresses slowly and can vary in symptoms from mild to disabling (Lamminen 1990).

Most often, the onset of limb-girdle muscular dystrophy (LGMD) is in adolescence or early adulthood. In the most common forms, the disease causes progressive weakness that starts in the hips and moves to the shoulders. The weakness progresses to include the arms and legs. Within 20 years after the onset, walking becomes difficult, if not impossible (Lamminen 1990). The autosomal recessive inherited limb-girdle dystrophy can result from gene defects on chromosomes 2, 13, 15, and 17. On the other hand, the autosomal dominant form can result from gene defects on chromosome 5. A gene on chromosome 15 that codes for the enzyme calpain 3 may play a role in some cases of limb-girdle dystrophy (Fig. 9.7.12). A flawed gene on chromosome 17 for the muscle protein adhalin is also known to cause other cases. Further genes have not yet been identified (Laval et al. 2004). 9.7.1.2.8  Oculopharyngeal Muscular Dystrophy Age of onset:

40s, 50s, 60s

Inheritance/gender affected:

Autosomal dominant/ males and females

Muscles first affected:

Eyelids, throat

Progression:

Slow

Oculopharyngeal muscular dystrophy, (OPMD) usually starts with drooping of the eyelids, most often in the 40s or 50s. This is followed by other signs of eye and facial muscle weakness, as well as by difficulty in swallowing. The later stages of this slowly progressive disease may include weakness in the pelvic and shoulder muscles. Swallowing problems can lead to choking and recurrent pneumonia. Special glasses to prop up the eyelids may be useful, and surgery can be done to alleviate both the drooping eyelids and the swallowing difficulties. The disease is linked to a gene defect on chromosome 14 (Lamminen 1990).

1199

1200

9  Skeletal Muscle

Fig. 9.7.12  Achilles tendon contractures and calf hypertrophy had been noticed since the late teens. First symptoms of muscle weakness occurred at the age of 43 with problems in climbing stairs, mild asymmetric proximal lower limb atrophy, and mild scoliosis. Mild weakness of proximal leg and arm muscles as well as foot extensors progressed. At examination at age 56, walking on tiptoes was only mildly impaired, whereas walking on heels was not possible. Now, at the age of 63, the patient is still able to walk without assistance and to climb stairs, but the ability to get up from a lying or sitting position is markedly impaired. CK levels are elevated with 830 U/l (normal <80). Coronal T1-weighted images (left) and axial STIR images (right)

show bilateral, but asymmetric muscle atrophy combined with fatty degeneration of the lower limb muscles. Hip abductors, hip adductors, rectus femoris muscle, hamstring muscles, and tibialis anterior muscle are strongly affected. Mild changes are seen in medial gastrocnemius muscles. The axial STIR sequence demonstrates no increase of the signal intensity in the involved muscles, so edema can be excluded. No increased enhancement was seen after administration of gadolinium (not shown). Biopsy detected a novel homozygous intronic mutation (IVS1 + 2T > C) of the CAV3 gene. This is the first splicing mutation reported for CAV3. These findings add to our knowledge of the clinical and genetic variability of CAV3 mutations

9.7.1.2.9  Distal Muscular Dystrophy

and involve fewer muscles than the other dystrophies, although spread to other muscles can occur. Walking can be improved with orthopedic devices that support the foot (Lamminen 1990).

Age of onset:

Adulthood

Inheritance/gender affected:

Autosomal recessive and dominant forms/ males and females

Muscles first affected:

Hands or lower legs

Progression:

Variable

Distal muscular dystrophy (DD) is actually a group of rare muscle diseases, which have in common weakness and wasting of the distal muscles of the forearms, hands, lower legs, and feet. A type of distal dystrophy called Welander is inherited in an autosomal dominant pattern and affects the hands first. Another type, known as Markesbery-Griggs, is autosomal dominant in its inheritance and affects the anterior portion of the lower legs first, as does Nonaka dystrophy. Miyoshi dystrophy, caused by a gene defect on chromosome 2, is autosomal recessive and affects the posterior portion of the lower legs first. In general, the distal dystrophies are less severe, progress more slowly

9.7.1.2.10  Emery-Dreifuss Muscular Dystrophy Age of onset:

Childhood to early teens

Inheritance/gender affected:

X-linked recessive/males

Muscles first affected:

Upper arms, lower legs

Progression:

Slow

Emery-Dreifuss muscular dystrophy (EDMD) is a rare form of muscular dystrophy. Muscle weakness and wasting generally start in the shoulders, upper arms, and lower legs (Fig. 9.7.13). Weakness may later spread to involve the muscles of the chest and pelvis. Contractures are found early in the disease, usually involving the ankle and elbow. Unlike other forms of muscular dystrophy, contractures in Emery-Dreifuss dystrophy often appear before the person experiences significant muscle weak-

9.7  Skeletal Muscle Diseases

Fig. 9.7.13a–e  60-year-old patient with long-lasting Emery-Dreifuss MD. a Coronal T1-weighted image of the shoulders shows the fatty replacement of the suprascapularis and the deltoideus muscles bilaterally. b Coronal T1-weighted images of the proximal femur demonstrate extensive fatty infiltration of all muscles, less involvement of the adductor and ischiocrural muscle. d,e The axial T1-weighted images depict almost no normal muscle tissue, which is replaced by fat

ness. Physical therapy is beneficial in minimizing the contractures. Life-threatening heart problems are a common complication of this disease. The heart problems are due to a conduction defect and can be treated with a cardiac pacemaker. These problems can even occur in females who do not suffer from the disease but are carriers, and therefore sisters and mothers of boys with Emery-Dreifuss should be examined. The skeletal muscle weakness is less severe than it is in some other dystrophies, such as Duchenne. Emery-Dreifuss dystrophy is caused by a defect in the gene on the X chromosome that codes for the protein called emerin. The function of this protein has not yet been identified (Lamminen 1990). 9.7.1.2.11  Metabolic Myopathy Metabolic myopathies refer to a group of hereditary muscle disorders caused by enzymatic defects due to defective genes. Metabolic myopathies are heterogeneous conditions that have in common abnormalities of muscle energy metabolism that result in skeletal muscle dysfunction. Most recognized metabolic myopathies are considered primary inborn errors of metabolism and are associated with known or postulated enzymatic defects that affect the ability of muscle fibers to maintain adequate adenosine triphosphate (ATP) concentrations. Traditionally, these diseases are grouped into abnormalities of glycogen, lipid, purine, or mitochondrial biochemistry. Metabolic myopathies are rare but potentially treatable disorders. Metabolic myopathies are the most clearly defined and etiologically understood muscle disorders, because their fundamental biochemical defects have be-

come known through recent developments in molecular biology and biochemistry. Also, many of the genetic defects have been characterized, and genetic counseling is now possible. An understanding of energy metabolism in exercising muscles is indispensable to an understanding of metabolic myopathies. Muscle contraction depends on the chemical energy of ATP. Many processes within the muscle cell maintain a supply of ATP to support muscle contraction. The 3 major pathways that supply ATP to meet the energy demands of exercising muscle are as follows: Glycogen metabolism: Aerobic exercise is essential for intermittent or sub-maximal contraction. Anaerobic exercise may be substituted for high-intensity muscular activity, particularly when blood flow is reduced and oxygen availability is limited. Lipid metabolism is an important source of energy in sustained sub-maximal exercise (i.e., exercise lasting longer than 40 min). Phosphocreatine stores, consumed in the purine nucleotide cycle, are vital for very high-intensity exercise of short duration, as other stores are depleted early. Metabolic myopathies have a wide age range of symptom onset. Most patients, however, present early in life (i.e., infancy, childhood, young adulthood). Metabolic myopathies presenting with exercise intolerance, cramps, and myoglobinuria. Cramps and muscle discomfort may occur after brief exercise or after prolonged physical activity. Glycogen is the main source of energy during brief exercise, while free fatty acids are the most important source of fuel during prolonged exercise. Accordingly, cramps are the hallmark of glycogen storage diseases (e.g., McArdle’s disease). However, in lipid storage disease (e.g., carnitine palmitoyltransferase [CPT] deficiency), muscle cramps and exercise intolerance oc-

1201

1202

9  Skeletal Muscle

cur only after prolonged exercise and are worse during fasting. Metabolic myopathies presenting with progressive muscle weakness may mimic limb-girdle muscular dystrophy or polymyositis and are a common manifestation of deficiencies of acid maltase, debrancher enzyme, and carnitine. Serum CK may be elevated modestly in metabolic myopathies, and the level may fluctuate significantly in patients prone to rhabdomyolysis. CBC and reticulocyte count may reveal signs of hemolytic anemia in patients with phosphofructokinase deficiency. Needle EMG may show myotonic discharges in acid maltase deficiency (AMD), a relatively specific finding in patients with suspected metabolic myopathy. In general, needle EMG may reveal short-duration, low-amplitude motor unit action potentials. However, EMG readings may be normal in many metabolic myopathies. McArdle introduced the Forearm ischemic exercise test in 1951. It is a useful screen to detect a possible enzymatic defect in the glycogenolytic and glycolysis pathways (DiMauro et al. 1992). MRI is useful in recognizing muscle involvement, which is characterized by increased signal intensity compared to normal muscle on T1-weighted images, presumably caused by excessive deposition of glycogen and lysosomal membranes. In patients with adult-onset AMD, muscle MRI shows a selective progressive pattern of muscle involvement with a constant involvement of the adductor magnus and semimembranosus at the early stage of the disease and later fatty infiltration of the long head of the biceps femoris, semitendinosus, and of the anterior thigh muscles. In the advanced phases a selective sparing of sartorius, rectus, and gracilis muscles and peripheral portions of the vastus lateralis has also been seen. Muscle strength and MRI findings have been positively correlated (Penn et al. 1992). 9.7.1.2.12  Glycogen Storage Diseases Muscle cells transport glucose from the circulating blood, synthesize glycogen, and then degrade it when energy demands increase. Muscle cell membrane (i.e., sarcolemma) is not freely permeable to glucose; therefore, utilization of circulating glucose is limited by its rate of transport through the sarcolemma. Glycogen is the main form of carbohydrate storage in the muscle. When energy is required for muscle contraction, glycogen is degraded to glucose and provides the energy required for muscle work. Any disturbance in either the synthesis or the degradation of glycogen could result in glycogen storage disease (i.e., glycogenoses). Glycogen storage diseases (glycogenoses) are named according to their specific defective enzyme function, an eponym, or by Roman numerals that correlate to the order of their discovery (glycogenosis types I–XII).

9.7.1.2.13  Lipid Storage Diseases Long-chain fatty acids are the major source of energy for the skeletal muscle during sustained exercise and fasting. The passage of these fatty acids through the mitochondrial membrane, for β-oxidation, requires their binding with carnitine. Carnitine is synthesized mainly in the liver and actively transported into the muscle against a concentration gradient. Free fatty acids are first converted to acyl-coenzyme A (CoA) compounds by the action of fatty acyl-CoA synthetases. Then, the long-chain acylCoA is bound to carnitine by acylcarnitine transferases, such as CPT I. This occurs on the outer mitochondrial membrane. The new compound passes through the inner mitochondrial membrane by the action of acylcarnitine translocase. Within the mitochondrial matrix, CPT II splits the transferred compound to free fatty acids and carnitine. In the mitochondria, β-oxidation of the longchain fatty acids is then carried out. Carnitine deficiency, deficiency of CPTs, or a defect in β-oxidation of these fatty acids may lead to myopathies. 9.7.1.2.14  Disorders of Purine Nucleotide Metabolism Adenylate deaminase is an enzyme that catalyses transformation of adenosine monophosphate (AMP) to inosine monophosphate (IMP) and ammonia. This reaction mainly occurs during anaerobic exercise to replenish ATP, which is an essential source of energy for the muscles. Whether myoadenylate deaminase deficiency causes muscle disease remains controversial. 9.7.1.2.15  Mitochondrial Disorders Mitochondrial disorders encompass a group of disorders resulting from abnormalities of the respiratory chain. The incidence is unknown; however, metabolic myopathies are rare and much less common than most of the muscular dystrophies. Mortality and morbidity rates vary depending on the specific disorder and the extent of enzymatic defect (i.e., complete or partial). Mortality rate is high in infantile acid maltase deficiency (AMD, Pompe disease). Invariably, the disease is progressive, leading to death within 1–2 years. The childhood form of AMD is less severe, and most children die by the end of the second decade of life from respiratory complications. In contrast, patients with the adult form of AMD present with slowly progressive limb girdle weakness, but some develop early respiratory failure secondary to involvement of intercostal muscles. The mortality rate of the adult form of AMD is much lower and the morbidity much less severe than those of the other two forms, likely due to the only partial deficiency of the enzyme.

9.7  Skeletal Muscle Diseases

9.7.1.2.16  Endocrine Myopathy Diseases of the endocrine system, including the thyroid, parathyroid, suprarenal, and pituitary glands, the ovaries, the testes, and the islands of Langerhans of the pancreas, usually result in multisystem signs and symptoms. A myopathy is very often present, and it rarely may be the leading symptom. Major categories of endocrine myopathy include those associated with (1) adrenal dysfunction (as in Cushing’s disease or steroid myopathy), (2) thyroid dysfunction (as in myxedema coma or thyrotoxic myopathy), (3) parathyroid dysfunction (as in multiple endocrine neoplasms), (4) pituitary dysfunction, and (5) islands of Langerhans dysfunction (as in diabetic myopathy). Although abnormal endocrine states frequently present with muscle weakness—most often proximal weakness—the exact pathophysiology remains incompletely understood. Even histological analysis and electromyographic testing may not show consistent, reproducible abnormalities in all cases. Consistent patterns of MR imaging findings do not exist. MRI in the diagnostic workup of endocrine myopathies is indicated to exclude other underlying muscle diseases when the findings of laboratory tests and biopsy do not correlate consistently with the clinical signs.

Compartment syndrome results from increased pressure within a restricted space. Trauma, burns, heavy exercise, extrinsic pressure, or intramuscular hemorrhage may initiate a vicious cycle of increasing pressure within confining fascia that leads to venous occlusion, muscle and nerve ischemia, arterial occlusion, and tissue necrosis. Acute and less severe chronic subtypes are seen. Clinical findings include severe pain and dysfunction of sensory and motor nerves that pass through the affected compartment. Early MR imaging findings include extremity swelling and diffuse edema within the affected compartment (Shintani and Shiigai 1993). If an acute compartment syndrome is clinically suspected, immediate therapeutic interventions in order to reduce intracompartmental pressure are required.

9.7.1.2.17  Trauma Muscle contusions are caused by a direct blow. MRI reveals high signal intensity in T2-weighted images at the site of impact, frequently due to interstitial hemorrhage as well as edema (Fig. 9.7.14). More severe contusions may result in a hemorrhage and thus reveal a mass-like lesion in addition to edema (Fig. 9.7.15). Muscle strains are injuries of the musculotendinous junction caused by overly forceful muscle contraction. Muscle strains occur most frequently in muscles that cross two joints, contain fast-twitch fibers, and contract during elongation. The hamstring, gastrocnemius, and biceps brachii muscles have these features and are the most frequently strained. MRI of a mild muscle strain may reveal edema centered along the musculotendinous junction. More severe muscle strains result in fluid collections and hematoma; they may contain grossly interrupted muscle fibers and a mass-like appearance may be found. Other causes of muscle trauma may mimic acute mild muscle strain, including delayed-onset muscle soreness (DOMS). DOMS is a type of overuse injury that does not become symptomatic until hours or days after the overuse episode, in contrast to muscle strain or contusion, which usually are immediately painful. Mild DOMS is frequently seen in recreational athletes. Severe forms may progress to rhabdomyolysis (Scott 2003).

Fig. 9.7.14a–c  Muscle strain of the right quadriceps and adductor muscles in 23-year-old man after playing soccer. a Axial STIR image demonstrates the extensive intramuscular edema. b Axial T1-weighted image shows a slightly increased signal intensity of the strained muscle as a sign of hemorrhage. c After administration of Gd-DTPA the fat-suppressed T1-weighted image shows enhancement

1203

1204

9  Skeletal Muscle

Fig. 9.7.15a,b  42-year-old man in whom a direct blow led to an intramuscular hematoma in the vastus intermedius muscle. a Axial T1-weighted image shows circumscribed increase of the signal intensity due to the paramagnetic effect of methemoglobin in the right vastus intermedius muscle. b The coronal T1weighted image gives a better overview of the longitudinal extension of the hematoma

The etiologies of rhabdomyolysis may be subdivided into traumatic, exercise induced, toxicological, environmental, metabolic, infectious, immunologic, and inherited. Definitive diagnosis is made by laboratory evaluation. The most useful test is serum CK. This assay is widely available and 100% sensitive. Total CK elevation is a sensitive but non-specific marker for rhabdomyolysis. If rhabdomyolysis is suspected in patients with serum CK levels in excess of two to three times the normal range and if risk factors for rhabdomyolysis are present, more extensive laboratory workup is required. A urine dipstick test for blood has positive results in the presence of hemoglobin or myoglobin. Aldolase, LDH, and serum glutamic-oxaloacetic transaminase (SGOT) are non-specific markers that are elevated in patients with rhabdomyolysis. Hyperkalemia, an immediate threat to life in the hours immediately after injury, occurs in 10–40% of cases. Acute renal failure develops in 30–40% of patients and is the most serious complication in the days after initial presentation (Gabow et al. 1982). Imaging studies generally play no role in the initial diagnosis of rhabdomyolysis. MRI may be useful in distinguishing various etiologies of rhabdomyolysis. One study suggests that bacterial myositis, focal myositis, and idiopathic rhabdomyolysis show a characteristic gadolinium enhancement on MRI. Abscesses were found only in bacterial myositis. Polymyositis and dermatomyositis have a characteristic uniform distribution pattern with emphasis on the quadriceps muscles. MRI initially reveals edema throughout the involved muscles, which may progress to findings of myonecrosis. The severity of the signal intensity alterations correlates with the severity of injury (Shellock and Fleckenstein 1997). The treatment has to be directed to the treatment of underlying conditions, such as trauma, infection, or toxins. General recommendations for the treatment of rhabdomyolysis include fluid resuscitation and prevention of end-organ complications.

9.7.1.2.18  Rhabdomyolysis

9.7.2  Neuromuscular Junction

Rhabdomyolysis is defined as the breakdown of muscle fibers with leakage of potentially toxic cellular contents into the systemic circulation. The final common pathway of rhabdomyolysis may be a disturbance in myocyte calcium homeostasis. Clinical consequences of rhabdomyolysis include: • Hypovolemia (sequestration of plasma water within injured myocytes) • Hyperkalemia (release of cellular potassium into the systemic circulation) • Metabolic acidosis (release of cellular phosphate and sulfate) • Acute renal failure (nephrotoxic effects of liberated myocyte components) • Disseminated intravascular coagulation (DIC)

Disorders of the neuromuscular junction are rare; they are mostly acquired and rarely congenital or familial. Pre-synaptic and post-synaptic dysfunctions have to be differentiated. 9.7.2.1  Myasthenia Gravis Myasthenia gravis (MG) is the most common primary disorder of neuromuscular transmission. The usual cause is an acquired immunological abnormality, but some cases result from genetic abnormalities at the neuromuscular junction. Much has been learned about the pathophysiology and immunopathology of MG during the past 20 years. A wide range of potentially effective treatments

9.7  Skeletal Muscle Diseases

are available, many of which have implications for the treatment of other autoimmune disorders. The prevalence of MG in the United States is estimated at 14/100,000 population, approximately 36,000 cases in the United States. However, MG is probably under diagnosed and the prevalence is probably higher. Previous studies showed that women are more often affected than men. The most common age at onset is the second and third decades in women and the seventh and eighth decades in men. As the population ages, the average age at onset has increased correspondingly, and now males are more often affected than females, and the onset of symptoms is usually after age 50. MG is an autoimmune disorder of peripheral nerves in which antibodies form against acetylcholine (ACh) nicotinic postsynaptic receptors at the myoneural junction. A reduction in the number of ACh receptors results in a characteristic pattern of progressively reduced muscle strength with repeated use of the muscle and recovery of muscle strength following a period of rest. The bulbar muscles are affected most commonly and most severely, but most patients also develop some degree of intermittent generalized weakness. The most important aspect of MG for emergency physicians is the detection and management of the myasthenic crisis. For unknown reasons autoantibodies against ACh nicotinic postsynaptic receptors develop in MG, although certain genotypes are more susceptible. Cholinergic nerve conduction to striated muscle is impaired by a mechanical blockage of the binding site by antibodies and, ultimately, by destruction of the postsynaptic receptor. Patients with MG can present with a wide range of signs and symptoms, depending on the severity of the disease (Waragai 1997). Mild presentations of MG may be associated with subtle findings, such as ptosis, that are limited to bulbar muscles. Findings may not be apparent unless muscle weakness is provoked by repetitive or sustained use of the muscles involved. Recovery of strength is seen after a period of rest or with application of ice to the affected muscle. Conversely, increased ambient or core temperature may worsen muscle weakness. Severe exacerbations of MG may present dramatically—facial muscles may be slack, and the face may look expressionless. The patient may be unable to support the head, which will fall onto the chest while the patient is seated. Thymic abnormalities are clearly associated with MG, but the nature of the association is uncertain. Ten percent of patients with myasthenia gravis have a thymic tumor and 70% have hyperplastic changes (germinal centers) that indicate an active immune response. These are areas within lymphoid tissue where B cells interact with helper T cells to produce antibodies. Because the thymus is the central organ for immunological self-tolerance, it is reasonable to suspect that thymic abnormalities cause the breakdown in tolerance that causes an immune-mediated attack on AChR in MG. The thymus

contains all the necessary elements for the pathogenesis of myasthenia gravis: myoid cells that express the AChR antigen, antigen-presenting cells, and immunocompetent T cells. Thymus tissue from patients with MG produces AChR antibodies when implanted into immunodeficient mice. However, it is still uncertain whether the role of the thymus in the pathogenesis of MG is primary or secondary. Most thymic tumors in patients with myasthenia gravis are benign, well-differentiated, and encapsulated, and can be removed completely at surgery. It is unlikely that thymomas result from chronic thymic hyperactivity because myasthenia gravis can develop years after thymoma removal and the HLA haplotypes that predominate in patients with thymic hyperplasia are different from those with thymomas. Patients with thymoma usually have more severe disease, higher levels of AChR antibodies, and more severe EMG abnormalities than patients without thymoma. Almost 20% of patients with myasthenia gravis whose symptoms began between the ages of 30 and 60 years have thymoma; the frequency is much lower when symptom onset is after age 60. CT scan or MRI of the chest is highly accurate in detecting thymoma. Every patient with MG should be screened for thymoma. Muscle edema is not typical in MG. The chronic disuse of the involved muscles may result in minor fatty atrophy (Waragai 1997). 9.7.2.2  Lambert-Eaton Syndrome Lambert-Eaton myasthenic syndrome (LEMS) is a rare autoimmune disorder of neuromuscular transmission. It was first described in association with lung cancer. The initial presentation can be similar to that of myasthenia gravis, but the progressions of the two diseases have some important differences. LEMS results from an autoimmune attack directed against the voltage-gated calcium channels on the pre-synaptic motor nerve terminal. Ptosis or ocular muscle weakness may be present. Proximal muscle weakness is noted, especially in the thighs and the hips (Sanders 1995). 9.7.2.3  Botulism Botulism is a paralytic disease caused by the neurotoxins of Clostridium botulinum and in rare cases, Clostridium butyricum and Clostridium baratii. These gram-positive, spore-forming anaerobes can be found in soil samples and marine sediments throughout the world. Wound botulism, caused by systemic spread of toxin produced by organisms inhabiting wounds, may results from trauma, surgery, subcutaneous heroin injection, and sinusitis from intranasal cocaine abuse. Infant botulism is due to intestinal colonization of organisms in infants younger than 1 year. In this age group, normal intestinal flora may

1205

1206

9  Skeletal Muscle

not have developed to the degree that prevents colonization by these organisms in healthy adults. The effects of the toxin are limited to blockade of peripheral cholinergic nerve terminals, including those at neuromuscular junctions, postganglionic parasympathetic nerve endings, and peripheral ganglia. This blockade produces a characteristic bilateral descending paralysis of the muscles innervated by cranial, spinal, and cholinergic autonomic nerves but no impairment of adrenergic or sensory nerves. Therapy consists in approximately 10,000 IU of antibodies against toxin types A, B, and E to neutralize serum toxin concentrations (Hatheway 1995). MRI has limited value in the diagnosis because of the rapid progression of the disease. In suspected cases immediate therapy with antibodies is required. 9.7.3  Nerves Potential mechanisms of muscle denervation are numerous and include spinal cord diseases, poliomyelitis, peripheral nerve injury, neuropathy, or nerve compression and neuritis. 9.7.3.1  General Features of Denervation Subacute muscle denervation causes edema uniformly throughout an involved muscle (Fleckenstein et al. 1993). This finding usually does not become evident in MRI until approximately 2–4 weeks after denervation has occurred, although such signal intensity changes have been observed as early as 2–4 days after denervation. The mechanism of this finding is not completely understood but appears likely to reflect shifting of water from intracellular to extracellular spaces. If normal innervation is restored, the MRI findings eventually return to normal. However, if innervation is not restored, atrophy with fatty infiltration develops after a period of months, indicating

Fig. 9.7.17  Chronic partial tear of the subscapularis tendon in a 70-year-old man. Oblique sagittal proton density–weighted fast spin-echo MR image shows atrophy and fatty infiltration of the superior two thirds of the subscapularis muscle (arrowheads) with relative sparing of the inferior one third (arrows)

irreversible changes in the muscle. Acutely denervated muscle does not demonstrate signal intensity alterations at MRI (Fleckenstein et al. 1993). This observation may allow distinction of denervation from acute traumatic muscle injury, which tends to have abnormal signal intensity at MRI that can be seen within hours or days after the injury. Fatty infiltration may be seen in the chronic stages of muscle denervation (Fleckenstein et al. 1993) and in chronic disuse, as a late finding after a severe muscle injury or chronic tendon tear, and as a consequence of corticosteroid use (Resnick 1995). Fatty infiltration due to chronic denervation is usually accompanied by muscle atrophy and represents irreversible muscle injury (Fleckenstein et al. 1993). Paradoxically, fatty infiltration may contribute to apparent hypertrophy of a chronically denervated muscle, but this association is rare. T2-weighted and inversion-recovery images show variable findings in chronic denervation (Fleckenstein et al. 1993) and thus are less reliable than T1-weighted images in revealing changes of chronic muscle denervation. Muscle atrophy with fatty infiltration may be seen in non-neurological conditions that may mimic chronic denervation on MRI. Chronic muscle disuse leads to atrophy with fatty infiltration (Figs. 9.7.16, 9.7.17.). 9.7.3.1.1  Diseases Resulting in Acute or Chronic Peripheral Denervation As mentioned before, numerous diseases my result in muscle atrophies due to acute or chronic denervation.

Fig. 9.7.16  Axial T1-weighted image of the thighs in a patient immobilized due to stroke. Reticular fatty replacement of all visualized muscle groups and an extensive atrophy of the muscles. The correlation with EMG is useful to differentiate these changes from chronic denervation

Inflammatory Demyelinating Polyneuropathy Acute Inflammatory Demyelinating Polyneuropathy Acute inflammatory demyelinating polyneuropathy (AIDP) is an autoimmune process that is characterized by progressive weakness and mild sensory changes. Many variants exist. In the West, the most common presentation

9.7  Skeletal Muscle Diseases

is a subacute ascending paralysis. This is associated with distal paresthesias and loss of deep tendon reflexes. The condition usually plateaus after about 2–3 weeks before slowly improving. AIDP is believed to be caused by an immunologic attack that is directed against myelin components. This results in a demyelinating polyneuropathy. Two thirds of patients with AIDP recall an antecedent upper respiratory or gastrointestinal infection or syndrome from 1–6 weeks prior to the onset of weakness. Increased CSF protein without an increased white blood cell count (albuminocytologic dissociation) is observed classically in AIDP. However, this finding is not specific to AIDP. More than 75% of patients have complete or near-complete recovery with no deficit or only mild residual fatigue and distal weakness. Other patients, almost all of whom required ventilation, report severe dysesthesias or moderately severe distal weakness as residual symptoms. Imaging is seldom necessary for diagnosing AIDP, but it may be necessary to exclude alternative diagnoses. However, electrodiagnostic testing is always necessary to confirm the diagnosis of AIDP. Chronic Inflammatory Demyelinating Polyneuropathy Chronic inflammatory demyelinating polyneuropathy (CIDP) is presumed to occur because of immunologic antibody-mediated reaction along with interstitial and perivascular infiltration of the endoneurium with inflammatory T cells and macrophages. The consequence is a segmental demyelination of peripheral nerves. Metabolic Polyneuropathy The term metabolic neuropathy includes a wide spectrum of peripheral nerve disorders associated with systemic diseases of metabolic origin. These diseases include diabetes mellitus, hypoglycemia, uremia, hypothyroidism, hepatic failure, polycythemia, amyloidosis, acromegaly, porphyria, disorders of lipid/glycolipid metabolism, nutritional/vitamin deficiencies, and mitochondrial disorders, among others. The hallmark of these diseases is involvement of peripheral nerves by alteration of the structure or function of myelin and axons due to metabolic pathway dysregulation. Diabetes mellitus is the most common cause of metabolic neuropathy, followed by uremia. Recognizing that some disorders involving peripheral nerves also affect muscles is important. Different levels of peripheral nerve involvement are found in type 1and type 2 diabetes, with milder compromise in type 2. Toxic Polyneuropathy Patients with toxic etiologies for neuropathy are less common than patients with other neuropathies such as those due to hereditary, metabolic, or inflammatory causes. Drug-related neuropathies are among the most common toxic neuropathies. Neuropathies from industrial agents (either from occupational or environmental sources),

presenting after either limited or long-term exposure, are insidious. Patients may present with subtle pain or weakness (Berger and Schaumberg 1994). Alcohol Exposure Ethanol intercalates into cell membranes, increasing membrane permeability. Alcohol also affects many signal-transduction proteins, including ion channels, secondary messengers, neurotransmitters, neurotransmitter receptors, G proteins, chaperonins, and regulators of genetic expression. Peripheral neuropathy is often the earliest symptom of chronic alcohol abuse. Peripheral nerve damage results from three processes, which are intensively debated: • Nutritional deficiency, especially thiamine deficiency; other nutritional deficiencies may involve niacin, folate, or protein • Direct toxicity from abnormal products (e.g., phosphatidyl ethanol, fatty acid ethyl esters) and from metabolites (e.g., acetaldehyde that reacts with proteins to form adducts) • Indirect toxicity (i.e., neuropathy from hepatic dysfunction) Mononeuropathy Mononeuropathies can occur secondary to direct trauma, compression, stretch injury, ischemia, infection, or inflammatory disease. Nerve entrapments are due to compression of the nerve by either normal structures or an external source. The most common nerve entrapments are at the median nerve of the wrist (i.e., carpal tunnel syndrome) and ulnar nerve of the elbow (i.e., cubital tunnel syndrome). In the lower extremity, peroneal neuropathy is the most common isolated mononeuropathy. In patients of our electrodiagnostic laboratory, it is the third most common mononeuropathy overall. Femoral neuropathies can occur secondary to direct trauma, compression, stretch injury, or ischemia. Femoral neuropathy causes weakness of the quadriceps, which results in difficulty with ambulation (al Hakim and Katirji 1993). Complete injuries disrupt all the neurons traversing the injured segment, causing total loss of distal motor or sensory function. Incomplete lesions disrupt some neurons but leave others unaffected, with some sparing of distal motor or sensory function. An incomplete nerve injury implies that at least part of the nerve remains in continuity; this has important therapeutic implications. One role of MRI in the setting of suspected peripheral nerve injury is detection of a surgically correctable cause of nerve compression such as a bone spur or ganglion cyst. Correlation with clinical and electromyelographic findings may be helpful in such cases because the level of nerve dysfunction can be localized to a specific anatomic region, thus allowing a focused MRI to asses definitely involved muscle and the corresponding peripheral nerve, the extent of fatty replacement and the atrophy.

1207

1208

9  Skeletal Muscle

9.7.4  Central Nervous System: Motor Neuron Diseases Motor neuron diseases (MNDs) are progressive degenerative diseases in which death of the cell bodies of motor neurons is the primary process. These should be distinguished from diseases in which primarily the axons of motor neurons are affected. The traditional classification of MNDs is according to the affected cell types, as follows: • Upper motor neurons alone: primary lateral sclerosis (PLS) • Lower motor neurons alone: progressive muscular atrophy (PMA) and the spinal muscular atrophies (SMAs) • Upper and lower motor neurons: amyotrophic lateral sclerosis (ALS) The muscle atrophy results from chronic denervation. However, MRI of muscles is seldom necessary. So these diseases are only briefly mentioned. 9.7.4.1  Primary Lateral Sclerosis PLS is a progressive, degenerative disease of upper motor neurons characterized by progressive spasticity (i.e., stiffness). It affects the lower extremities, trunk, upper extremities, and bulbar muscles (usually in that order). The cause of PLS is unknown. The terms pathophysiology and pathogenesis refer at this time to histological consequences of unknown etiologic factors, which result, in turn, in the clinical manifestation of PLS. In summary, PLS is a clinical diagnosis, and most reports (combining imaging and autopsy series) indicate neuronal loss in the precentral gyrus. 9.7.4.2  Amyotrophic Lateral Sclerosis ALS is a devastating disorder of the anterior horn cells of the spinal cord and the motor cranial nuclei that leads to progressive muscle weakness and atrophy. Although major recent advances have shed light on its etiology, the key mechanisms in both familial and sporadic ALS remain unknown. No cure is known. ALS primarily involves anterior horn cells in the spinal cord and cranial motor nerves. Patients may have weakness of bulbar muscles or of single or multiple limb muscle groups. Presentation is not always bilateral or symmetrical. A predominantly bulbar form usually leads to more rapid deterioration and death. Limb weakness is predominantly distal. Weakness and atrophy of the intrinsic hand muscles are prominent. Weakness progresses to involve the forearms and shoulder girdle muscles and the lower extremities. ALS leads to death within a decade. In most cases, death occurs within 5 years. Respiratory insufficiency is usually a late event.

Needle EMG and nerve conduction studies are the tests of choice for confirming the diagnosis of ALS. Muscle biopsy should be done if the presentation is atypical (e.g., very early onset, prominent lower extremity weakness with or without hand muscle involvement). Brain or cervical spine MRI should be done to rule out dysmyelinative lesions (e.g., in family history of TaySachs disease) or to rule out cervical myelopathy (Waragai 1997). MRI of the atrophic muscles obtains images presenting the atrophy with topic selectivity and fatty replacement of a reticular appearance. 9.7.4.3  Spinal Muscular Atrophies The SMAs are a clinically and genetically heterogeneous group of disorders. They are characterized by primary degeneration of the anterior horn cells of the spinal cord and often of the bulbar motor nuclei without evidence of primary peripheral nerve or long-tract involvement. Because bulbar features are often present, the term SMA does not technically describe the disorder. Consequently, alternative designations, such as bulbospinal muscular atrophy, hereditary motor neuronopathy (HMN), and progressive muscular atrophy, have been used. The SMAs present with a diversity of symptoms and differ in age of onset, mode of inheritance, distribution of muscle weakness, and progression of symptoms. Additionally, atypical forms of the disease have been described, including those with associated sensory deficits, hearing loss, or arthrogryposis (Resnick 1995). Patients with disorders of the motor unit present predominantly lower motor neuron signs, that include hypotonia (i.e., loss of postural tone), flaccid weakness, decreased or absent deep tendon reflexes, fasciculations, and atrophy. Imaging is seldom necessary. 9.7.4.4  Poliomyelitis Poliomyelitis is an enteroviral infection that can manifest in four different forms: inapparent infection, abortive disease, non-paralytic poliomyelitis, and paralytic disease. Before the 19th century, poliomyelitis occurred sporadically. During the 19th and 20th centuries, epidemic poliomyelitis was more frequently observed, reaching its peak in the mid 1950s. The worldwide prevalence of this infection has decreased significantly since then because of aggressive immunization programs. Eradication of this disease during the present decade is a top priority for the World Health Organization (WHO). Poliovirus is an RNA virus that is transmitted through the oral–fecal route or by ingestion of contaminated water. The destruction of motor neurons leads to the development of flaccid paralysis, which may be bulbar or spinal in distribution.

9.9  Future Developments

Fig. 9.7.18  Chronic muscle denervation in a 52-year-old man who contracted poliomyelitis as a child from the vaccine. Coronal T1-weighted MR image shows nearly complete fatty replacement of the left pelvic and thigh muscles (*)

Most patients infected with poliovirus develop inapparent infections and are frequently asymptomatic. When non-paralytic poliomyelitis develops, symptoms usually are those observed in abortive disease in addition to meningeal irritation. Paralytic poliomyelitis involves systemic manifestation, such as respiratory failure, in addition to symptoms observed in non-paralytic poliomyelitis. Patients who have recovered from poliomyelitis occasionally develop a post-poliomyelitis syndrome in which recurrences of weakness or fatigue are observed and usually involve groups of muscles that were initially affected. This post-polio syndrome may develop 20–40 years after infection with poliovirus (Cashman et al. 1987). MRI in patients with post-poliomyelitis syndrome shows asymmetric involvement of the muscle groups, with homogenous fatty replacement and no edema (Fig. 9.7.18). 9.8  Follow-Up MRI has various in diagnostics of muscle diseases over other methods. Unlike muscle biopsy or electromyogram, MRI is non-invasive and allows performing repeat studies for longitudinal analyses of outcome (Park and Olsen 2002). Long-term studies are facilitated and quantitative data needed for making accurate comparisons over time can be obtained. There is controversy whether of CK, LDH, and aspartate aminotransferase are most accurate to assess active muscle inflammation. CK-l may be normal in dermatomyositis and may not correlate with disease flares, even in patients who initially present with a raised CK level.

Some data suggest that lactate dehydrogenase may correlate better with disease activity, especially in the course of the disease and that aldolase and aspartate aminotransferase are also useful. Many centers use MRI to assess muscle inflammation. Signs of muscle inflammation and edema are readily visualized on T2-weighted MR images. Muscle inflammation in children with juvenile dermatomyositis can be assessed quantitatively from T2 relaxation times, which correlate well with clinical scores (Maillard et al. 2004). Moreover, active and inactive disease could be differentiated, suggesting that this might be a useful way to follow disease activity if a quantification system could be established. In contrast to most studies in the literature, which generally use four MRI parameters (edema, fatty replacement, distribution, and contrast enhancement), a study by Hilario et al. (2000) analyzed six types of findings for qualitative analysis: • Increase in muscle signal intensity in T2-weighted images • Increase of chemical shift • Perimuscular edema • Increase in the intensity of the subcutaneous fat signal • Muscular atrophy and • Muscle fat replacement “Increase in muscle signal intensity” was used to identify patients with greater disease activity with 80% sensitivity and 75% specificity. “Increase of chemical shift” had a sensitivity of 85.7% and a specificity of 83.3% in the detection of disease activity (Hilario et al. 2000). Muscle atrophy and fat replacement are associated with chronic muscle processes and are due to chronic inflammation with major muscle impairment, reduction of physical activity and long-term corticotherapy (Hilario et al. 2000). 9.9  Future Developments 9.9.1  Diffusion-Weighted Imaging and Blood-Oxygen Level-Dependent Imaging Diffusion-weighted imaging (DWI) utilizes MRI to detect molecular motion. In muscle tissues, DWI can be used to detect the motion of water, which relates to the delivery of metabolites and oxygen (Morvan and Leroy-Willig 1995). The calculated diffusion coefficient increases in the presence of muscle inflammation and decreases with fatty infiltration of muscle fibers. DWI therefore offers pathophysiological information regarding muscle status in disease states. In normal and inflamed muscles, anisotropic motion of water (z-direction, head to foot) is observed. Therefore, diffusion tensor imaging (DTI) and a fiber tracking algorithm can be used to detect inflammatory muscle disease (Basser et al. 1994). DTI has the po-

1209

1210

9  Skeletal Muscle

tential to detect segmentation of muscle fibers in disease states (Damon et al. 2002). Blood-oxygen level–dependent (BOLD) imaging was developed for evaluating brain activation. BOLD signal is generated by the changes in the ratio of oxyhemoglobin to deoxyhemoglobin in the activated region of the brain and has recently been applied to evaluate the tissue oxygenation in normal and diseased muscle. Slow- and fasttwitch muscles show different signals in BOLD imaging because they have different microvascular density. Noseworthy et al. (2003) has recently reported that BOLD signal-based muscle functional MRI could be beneficial in understanding microvascular-related disease such as muscular dystrophy, ischemia, and chronic/peripheral venous insufficiency. 9.9.2  Spectroscopy MR spectroscopy (MRS) of human skeletal muscle can play a significant role in (1) understanding healthy muscle metabolism and the mechanisms of muscle fatigue, (2) understanding the effects of disease on muscle metabolism and function, (3) monitoring the efficacy of therapeutic intervention, and (4) the confirmation of diagnoses. 9.9.2.1  Proton (1H) MR Spectroscopy of Human Muscle Using in vivo and in vitro 1H MRS, metabolites in the muscles such as creatine (CRN), taurine (TAU), anserine (ANS), and carnosine (CAR) can be simultaneously analyzed. These metabolites are rich in mammalian skeletal muscle tissues and may play an important role in several functions of the skeletal muscles such as energy metabolism and contraction. 1 H MRS of muscle provides a tool to quantitate intramyocellular lipids (i.e., lipid droplets within the muscle cell, which are directly available for energy production), a capability that became particularly important when intramyocellular lipid content was linked to insulin sensitivity. Moreover, various metabolites are detectable with MRS, including creatine (Cr) and phosphocreatine (PCr), and trimethylammonium (TMA), which in muscle includes carnitines, acetylcarnitine, carnosine, lactate, deoxymyoglobin, and tentatively taurine. Some of these metabolites are only visible after exercise or in ischemia (Fig. 9.7.15) (Boesch and Kreis 2000).

muscle contain seven resonances: three arise from the phosphate groups of ATP, one from PCr and one from inorganic phosphates (Pi); two additional smaller resonances can sometimes be observed from phosphomonoesters (PME) and phosphodiesters (PDE). The area under each resonance (i.e., the signal intensity) is proportional to the amount of the corresponding metabolite. The spectral distance between Pi and PCr provides information about intracellular pH (pHi) while the chemical shift of ATP can be used to calculate intracellular magnesium ion concentration. In living human muscle, only metabolites that are unbound and present at concentrations of at least 1 mM give rise to peaks that have sufficient SNR to be easily visible. The concentration of the metabolically active adenosine phosphate (ADP), which is important in the regulation of rates of mitochondrial ATP synthesis, does not produce a signal visible in the spectra, but can be calculated indirectly using the CK equilibrium equation. One of the most important features of 31P MRS is its ability to monitor time-dependent changes of metabolites non-invasively. 9.9.2.2.1  31P MRS as a Diagnostic Tool 31P MRS is well suited for gathering data on skeletal muscle energetics in vivo. The technique has evolved to the point where it has become an important tool in the study of the pathophysiology both of rare primary disorders of muscle such as the mitochondrial myopathies and of more common systemic diseases such as renal failure, which also influence muscle metabolism. 31P MRS is used for providing information about the biochemical composition of tissue without invasive sampling, and it has the unique ability to measure intracellular pH. In some conditions it can be used as an aid to diagnosis. Magnetic resonance spectroscopy is being increasingly used in conjunction with other noninvasive technologies to investigate effects of gene function on metabolism. 9.9.2.2.2  31P MRS as a Tool in Therapeutic Trials More recently, 31P MRS has been used as a measure of treatment response in therapeutic trials. It can be used to repeatedly assess the response to therapy over long periods. It has been used to supply evidence of therapeutic efficiency (or failure) in mitochondrial disorders treated by riboflavin and nicotinamide for example (Pearn 1980).

9.9.2.2  31Phosphorus MRS

References

31Phosphorus (31P) MRS records signals from high-energy phosphate compounds which are central to energy metabolism in vivo. Phosphorus MR spectra from

1.

al Hakim M, Katirji B (1993) Femoral mononeuropathy induced by the lithotomy position: a report of 5 cases with a review of literature. Muscle Nerve 16:891–895

9.9  Future Developments 2.

3. 4. 5.

6.

7.

8.

9. 10.

11.

12.

13. 14.

15.

16.

17.

Askanas V, Serratrice G, Engel WK (1998) Overview of pathologic and pathogenic comparisons between sporadic inclusion-body myositis and hereditary inclusion-body myopathies. In: Askanas V, Serratrice G, Engel WK (eds) Inclusion body myositis and myopathies. Cambridge University Press, Cambridge, pp 3–81 Askari A, Vignos PJ, Moskowitz RW (1976) Steroid myopathy in connective tissue disease. Am J Med 61:485–492 Basser PJ, Mattiello J, LeBihan D (1994) MR diffusion tensor spectroscopy and imaging. Biophys J 66:259–267 Beltran J, Chandnani V, McGhee RA, Kursunoglu-Brahme S (1991) Gadopentate dimeglumine-enhanced MR imaging of the musculoskeletal system. AJR Am J. Roentgenol 156:457–466 Berger AR, Schaumberg HH (1994) Disorders of the peripheral nervous system. In: Rosenstock L, Cullen MR (eds) Textbook of clinical occupational and environmental medicine. Saunders, Philadelphia, pp 482–513 Berquist TH (1991) Magnetic Resonance imaging techniques in musculoskeletal diseases. Rheum Dis Clin North Am 17:599–615 Boesch C, Kreis R (2000) Imaging and Spectroscopy of Muscle. In: Young IR, Grant DM, Harris RK (eds) Methods in biomedical magnetic resonance imaging and spectroscopy. Wiley, Sussex Bohan A, Peter JB (1975) Polymyositis and dermatomyositis (first of two parts). N Engl J Med 292:344–347 Brothers TE, Tagge DU, Stutley JE et al (1998) Magnetic resonance differentiates between necrotizing and nonnecrotizing fasciitis of the lower extremity. J Am Coll Surg 187:416–421 Cashman NR, Maselli R, Wollmann RL (1987) Late denervation in patients with antecedent paralytic poliomyelitis. N Engl J Med 317:7–12 Damon BM, Ding Z, Anderson AW et al (2002) Validation of diffusion tensor MRI-based muscle fiber tracking. Magn Reson Med 48:97–104 De Smet AA (1993) Magnetic resonance findings in skeletal muscle tears. Skeletal Radiol 22:479–484 DeSmet AA, Norris MA, Fisher DR (1992) Magnetic resonance imaging of myositis ossificans: analysis of seven cases. Skeletal Radiol 21:503–507 DiMauro S, Bonilla E, Hays AP (1992) Skeletal muscle storage diseases: Myopathies resulting from errors in carbohydrate and fatty acid metabolism. In: Mastalgia FL, Walton JN (eds) Skeletal muscle pathology, 2nd edn. Churchill Livingstone, Edinburgh Dion E, Cherin P, Payan C et al (2002) Magnetic resonance imaging criteria for distinguishing between inclusion body myositis and polymyositis. J Rheumatol 29:1897–1906 Disler DG, Cohen MS, Krebs DE, Roy SH, Rosenthal DI (1995) Dynamic evaluation of exercising leg muscle in healthy subjects with echo planar MR imaging: work rate and total work determine rate of T2 change (1993) J Magn Reson Imaging 5:588–593

18. Feldman F, Zwass A, Staron RB, Haramati MRI of soft tissue abnormalities: a primary cause of sickle cell crisis. Skeletal Radiol 22:506 19. Fisher MR, Barker B, Amparo EG, Brandt G, Brandt-Zawadski M, Hricak H, Higgens CB (1985) MR Imaging using specialized coils. Radiology 157:443–447 20. Fleckenstein JL, Reimers CD (1996) Inflammatory myopathies: radiologic evaluation. Radiol Clin North Am 34:427–439 21. Fleckenstein JL, Weatherall PT, Parkey RW, Payne JA, Peshock RM (1989) Sports-related muscle injuries: evaluation with MR imaging. Radiology 172:793–798 22. Fleckenstein JL, Watumull D, Conner KE et al (1993) Denervated human skeletal muscle: MR imaging evaluation. Radiology 187:213–218 23. Fraser DD, Frank JA, Dalakas M, Miller FW, Hicks JE, Plotz P (1991) Magnetic resonance imaging in the idiopathic inflammatory myopathies J Rheumatol 18:1693–1700 24. Gabow PA, Kaehny WD, Kelleher SP (1982) The spectrum of rhabdomyolysis. Medicine 61:141–152 25. Garcia J (2000) MRI in inflammatory myopathies. Skeletal Radiol 29:425–438 26. Goebel HH (2003) Congenital myopathies at their molecular dawning. Muscle Nerve 27:527–548 27. Hatheway CL (1005) Botulism: the present status of the disease. Curr Top Microbiol Immunol 195:55–75 28. Hernandez RJ, Sullivan DB, Chenevert TL, Keim DR (1993) MR imaging in children with dermatomyositis: musculoskeletal findings and correlation with clinical and laboratory findings. AJR Am J Roentgenol 161:359–366 29. Hilario MOE, Yamashita H, Lutti D et al (2000) Juvenile idio­pathic inflammatory myopathies: the value of magnetic resonance imaging in the detection of muscle involvement. Sao Paulo Med J 118:35–40 30. Krendel DA, Zacharias A, Younger DS (1997) Autoimmune diabetic neuropathy. Neurol Clin15:959–971 31. Lamminen AE (1990) Magnetic resonance imaging of primary skeletal muscle diseases: patterns of distribution and severity of involvement. BJR; 63:946–950 32. Laval SH, Bushby KM (2004) Limb-girdle muscular dystrophies—from genetics to molecular pathology. Neuropathol Appl Neurobiol 30:91–105 33. Leroy-Willig A, Willig TN, Henry-Feugeas MC et al (1997) Body composition determined with MR in patients with Duchenne muscular dystrophy, spinal muscular atrophy, and normal subjects. Magn Reson Imaging 15:737–744 34. Liu GC, Jong YJ, Chiang CH, Jaw TS (1993) Duchenne muscular dystrophy: MR grading system with functional correlation. Radiology 186:475–480 35. Maillard SM, Jones R, Owens C et al (2004) Quantitative assessment MRI T2 relaxation time and thigh muscles in children with dermatomyositis. Rheumatology 43:603–608 36. Matsuo M, Ehara S, Tamakawa Y, Chida E, Nishida J, Sugai T (1995) Muscular sarcoidosis. Skeletal Radiol 24:535–537

1211

1212

9  Skeletal Muscle 37. May DA, Good RB, Smith DK, Parsons TW (1997) MR imaging of musculoskeletal tumors and tumor mimickers with intravenous gadolinium: experience with 242 patients. Skeletal Radiol 26:2–15 38. Morvan D, Leroy-Willig A (1995) Simultaneous measurements of diffusion and transverse relaxation in exercising skeletal muscle. Magn Reson Imaging 13:943–948 39. Noseworthy MD, Bulte DP, Alfonsi J (2003) BOLD magnetic resonance imaging of skeletal muscle. Semin Musculoskelet Radiol 7:307–315 40. Palmer WE, Kuong SJ, Elmadbouh HM (1999) MR imaging of myotendinous strain. AJR Am J Roentgenol 173:703–709 41. Park JH, Olsen NJ (2002) Skeletal muscle imaging for the evalua-tion of myopathies. In: Wortmann RL (ed) Diseases of skeletal muscle. Lippincott Williams & Wilkins, Philadelphia, pp 293–312 42. Pearn J (1980) Classification of spinal muscular atrophies. Lancet 1:919–22 43. Penn AM, Lee JW, Thuillier P, Wagner M, Maclure KM, Menard MR et al (1992) MELAS syndrome with mitochondrial tRNA (Leu) (UUR) mutation: correlation of clinical state, nerve conduction, and muscle 31P magnetic resonance spectroscopy during treatment with nicotinamide and riboflavin. Neurology 42:2147–52 44. Pichiecchio A, Uggetti C, Ravaglia S, Egitto MG, Rossi M, Sandrini G, Danesino C (2004) Muscle MRI in adult-onset acid maltase deficiency. Neuromuscul Disord 14:51–5 45. Plotz PH, Dalakas M, Leff RL, Love AA, Miller FW, Cronin ME (1989) Current concepts in idiopathic inflammatory myopathies: polymyositis, dermatomyositis, and related disorders. Ann Intern Med 111:143–57 46. Reimers CD, Schedel H, Fleckenstein JL et al (1994) Magnetic resonance imaging of skeletal muscles in idiopathic inflammatory myopathies of adults. J Neurol 241:306–314

47. Resnick D (1995) Dermatomyositis and polymyositis. In: Resnick D (ed) Diagnosis of bone and joint disorders, 3rd edn. Saunders, Philadelphia, pp 1218–1231; 3309–3342; 4577–4584 48. Resnick D, Niwayama G (1995) Osteomyelitis, septic arthritis, and soft tissue infection: mechanisms and situations. In: Resnick D (ed) Diagnosis of bone and joint disorders, 3rd edn. Saunders, Philadelphia, pp 2325–2418 49. Resnick D, Georgen TG, Niwayama G (1995a) Physical injury: concepts and terminology. In: Resnick D (ed) Diagnosis of bone and joint disorders, 3rd edn. Saunders, Philadelphia, pp 2561–2692 50. Sanders DB: Lambert-Eaton myasthenic syndrome: clinical diagnosis, immune-mediated mechanisms, and update on therapies. Ann Neurol 1995 May; 37 Suppl 1: S63–73 51. Sayers ME, Chou SM, Calabrese LH (1992) Inclusion body myositis: analysis of 32 cases. J Rheumatol 19:1385–1389 52. Scott DL (2003) Imaging in muscle disease. In: Isenberg Da (ed) Imaging in rheumatology, 1st edn. Oxford University Press, Oxford, pp 188–212 53. Shellock FG, Fleckenstein JL (1997) Magnetic resonance imaging of muscle injuries. In: Stoller DW (ed) Magnetic resonance imaging in orthopaedics and sports medicine, 2nd edn. Lippincott-Raven, Philadelphia, pp 1341–1362 54. Shintani S, Shiigai T. Repeat MRI in acute rhabdomyolysis: correlation with clinicopathological findings. J Comput Assist Tomogr 1993; 17:786 791 55. Stoller DW, Ferkel RD (1997) The ankle and foot. In: Stoller DW (ed) Magnetic resonance imaging in orthopaedics and sports medicine, 2nd edn. Lippincott-Raven, Philadelphia, pp 443–595 56. Van Slyke MA, Ostrov BE (1995) MRI evaluation of diabetic muscle infarction. Magn Reson Imaging 13:325–329 57. Vincent A, Palace J, Hilton-Jones D (2001) Myasthenia gravis. Lancet 357:2122–2128 58. Waragai M (1997) MRI and clinical features in amyotrophic lateral sclerosis. Neuroradiology 39:847–851

Chapter 10

MRI of the Fetal Body

10

10.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1213 A.D. McKenna and F.V. Coakley

10.6.3 

Congenital Cystic Adenomatoid Malformation  . . . . . . . . . .   1221

10.2 

Historical Development of Fetal MRI    1213

10.6.4 

MRI for Fetal Lung Maturity  .. . . . . . . . .   1222

10.3 

Safety of MRI in Pregnancy  . . . . . . . . . .   1213

10.6.5 

Congenital Diaphragmatic Hernia  .. . . .   1222

10.4 

Fetal MRI Technique  .. . . . . . . . . . . . . . . .   1214

10.6.6 

Abdominal Masses  .. . . . . . . . . . . . . . . . . .   1226

10.5 

Normal MRI Findings   of the Fetal Body  . . . . . . . . . . . . . . . . . . . .   1215

10.6.6.1  Upper-Quadrant Masses   . . . . . . . . . . . . .   1226

10.6 

MRI Findings in Pathologic Conditions of the Fetal Body  . . . . . . . . . . . . . . . . . . . .   1218

10.6.1 

Airway Compromise  . . . . . . . . . . . . . . . . .   1218

10.6.2 

Pulmonary Sequestration  .. . . . . . . . . . . .   1219

10.6.7 

Gastrointestinal Tract Anomalies  .. . . . .   1227

10.6.8 

Genitourinary Abnormalities  . . . . . . . . .   1227

10.7 

Summary: Indications   for MRI for the Fetal Body  . . . . . . . . . . .   1228 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1229

10.1  Introduction

body, several limitations of the modality remain. These include a small field of view, limited soft-tissue contrast, A.D. McKenna and F.V. Coakley beam attenuation in maternal adiposity, poor access to the deeply engaged fetal head, poor image quality in oliThis chapter discusses the performance and interpreta- gohydramnios, and poor evaluation of the fetal brain subtion of MRI studies of the fetal body (MRI of fetal neuro- sequent to calvarial calcification after 33 weeks’ gestation. logical abnormalities is discussed separately in the neu- In contrast, MRI has a number of advantages including a roradiology sections of this text). The aim of this chapter large field of view, excellent multi-parameter soft-tissue is to allow the reader to recognize normal and abnormal contrast, absence of the same limiting physical factors of MRI findings in the fetal body, and to understand the ultrasound, and easy performance of volumetric measurecurrent indications for such studies. It should be noted ments. Despite these theoretical advantages and while that MRI of the fetal body is a less mature application MRI in pregnancy was first described in 1983 (Smith et al. than MRI of fetal neurological abnormalities is, and that 1983), it has only been in the past few years that fetal MRI the indications remain relatively anecdotal or based on has become more widely practiced. The primary reason small case series. Despite this, the management impact for this change is that fetal MRI was greatly limited for with respect to parental counseling and discussion of many years by fetal motion, and it is only the more recent treatment options can be substantial. Accordingly, such availability of new high quality rapid imaging sequences cases can be both clinically challenging and clinically that has allowed more widespread implementation of rewarding as far as the role of the radiologist may be of MRI for fetal evaluation. It is hardly an exaggeration to critical importance in pregnancy outcome. say that these sequences have revolutionized the ability to perform routine high quality fetal MRI.

10.2  Historical Development of Fetal MRI Despite the established primacy of ultrasound as the preferred imaging modality for screening and evaluation of fetal abnormalities, including abnormalities of the fetal

10.3  Safety of MRI in Pregnancy Safety is one of the key issues to consider when performing MRI in pregnancy. There are three potential risks

1214

10  MRI of the Fetal Body

that have to be considered, namely teratogenesis, acoustic damage, and gadolinium toxicity. It should be noted that the vast majority of studies investigating the safety of MRI in pregnancy with respect to teratogenesis have shown no ill effect (Schwartz and Crooks 1982; Wolff et al. 1980). However, a small number of studies have suggested at least the possibility of harm early in pregnancy, and caution should reasonably be exercised in the first trimester. For example, a study of 304 chick embryos that were divided into two groups (subject group placed in a 1.5-T scanner for 6 h and subjected to varying radiofrequency pulses and gradients and a control group not exposed to MR) showed a significantly higher number of dead or abnormal embryos in the subject group, when the embryos were examined on day 6 of a 21-day gestation and after MRI exposure on day 2 (the study demonstrated 19.5% abnormal or dead embryos in the subject group compared to 10.7% in the control group, p < 0.05) (Yip et al. 1994). From a practical point of view, elective studies in the first trimester would therefore seem at least relatively contraindicated, although clearly, if such studies had a compelling clinical indication, MRI would still be preferable to any study involving ionizing radiation. Acoustic damage may seem an unusual concern, but MRI scanners are noisy and the question has been raised as to whether the developing ear might be harmed by such acoustic exposure. Two studies from the group at Nottingham would suggest that this is unlikely (Baker et al. 1994; Gover et al. 1995). A follow-up study of 20 children who have undergone echo-planar imaging while in utero showed that 16 of 18 with follow-up information had passed their hearing test at 8 months postnatally, compared to the 16.7 expected (Baker et al. 1994a). In a second study, the same group showed that the sound intensity within a fluid-distended stomach of a volunteer (designed to simulate the gravid uterus) would suggest that the sound intensity to the fetus is lower than that to the mother, again indirectly suggesting that acoustic damage is unlikely (Gover et al. 1995). While the risks of teratogenesis and acoustic damage are very small or nonexistent, the risk of gadolinium administration during pregnancy is substantive. In particular, gadolinium administered to pregnant rabbits has been shown to result in fetal skeletal malformations, albeit administered at doses well above the physiologic range (0.5 millimoles per kilogram per day for 13 days; package insert, Nycomed, Princeton, New Jersey). While teratogenic effects have not been observed in a small number of human studies where gadolinium has been given in pregnancy (Marcos et al. 1997; Spencer et al. 2000), it is clear that gadolinium should not be administered in pregnancy unless there is an essential clinical indication, particularly during the period of organogenesis. Administration of gadolinium later in pregnancy may be reasonable, although such indication would likely be for a maternal or obstetric indication rather than for evaluation of a fetal

abnormality. An example would include the evaluation of placenta accreta with gadolinium-enhanced images. 10.4  Fetal MRI Technique It is sensible to have the mother avoid fluids and caffeinated beverages prior to scanning to reduce the chance that the scan may be interrupted by the need for patient urination. This is a particular concern in pregnancy, particularly late pregnancy, when bladder capacity is reduced by the gravid uterus. Perhaps counterintuitively, it is probably better not to have the mother fast prior to imaging, because most mothers report that fetal motion is reduced by a recent meal. At our institution, no other maternal preparation is used, although some centers recommend routine administration of a small dose of a benzodiazepine (specifically flunitrazepam 1 h prior to the exam) (Revel et al. 1993). In our experience, the frequency of non-diagnostic studies is so low (under 5%) that we do not consider the routine use of a sedating agent that crosses the placenta and reduces fetal motion to be appropriate. In the first and second trimester, most patients can be scanned in the supine position. In later pregnancy, or in patients with a larger than normal uterus (e.g,. multiple gestation) the left lateral decubitus position may be preferred, in order to avoid syncopal episodes related to compression of the inferior vena cava by the gravid uterus in the supine position. A surface coil should be used in order to increase signal to noise, since many of the ultrafast sequences used with fetal imaging suffer from inherent limitations in signal to noise ratios. However, occasionally the indwelling body coil of the scanner may be all that can be used, because patient size may preclude placement of a surface coil. After the patient has been appropriately positioned in the scanner, a localizer sequence is obtained. The particular sequence used and the planar section chosen is relatively unimportant and may vary from institution to institution. As a minor point, we have found the coronal plane of section to be helpful in assessing fetal lateralization (the use of an internal fetal landmark such as the heart or stomach to distinguish left from right is unreliable because of the potential presence of situs inversus or other heterotopic disorders). That is, evaluation of the fetal position relative to the mother allows the absolute distinction of left and right sides of the fetus, rather than depending on the internal fetal anatomy. After an appropriate localizer sequence has been obtained, ultrafast T1- and T2-weighted sequences are acquired in planes anatomic to the fetus. Image acquisition and the precise choice of sequences and planes of section must be closely supervised and tailored to the specific indication. T1-weighted images can be obtained with a two-dimensional interleafed spoiled gradient echo technique (e.g., FMPSPGR using Gen-

10.5  Normal MRI Findings of the Fetal Body

eral Electric equipment, or multiplanar FLASH using Siemens equipment). Three-dimensional T1-weighted sequences are generally not appropriate because of poor signal to noise. Sequential slice acquisition is also usually not appropriate because of fetal motion between slice acquisitions resulting in noncontiguous images. T2-weighted images can be obtained using a single-shot rapid acquisition with refocused echo (SS-RARE) sequence (SSFSE using General Electric equipment, and HASTE using Siemens equipment). All of these ultrafast sequences “freeze” fetal motion, and generally result in images of diagnostic quality. Despite rapid acquisition times, these sequences may still be degraded by fetal motion, in which case the sequences can simply be repeated or returned to later during the study. Because all of the sequences are rapid enough to be performed during a maternal breath hold, repeating the sequences does not present a large time penalty to the overall examination. In general, the T2-weighted images tend to be of better diagnostic quality than the T1-weighted images, although the latter can provide useful information and are worth obtaining (see below). As another general point, diagnostic quality images require a reasonably high signal-to-noise ratio. As a result, obtaining images with very small voxels, while increasing spatial resolution, can ultimately result in images that are too grainy and may work against the endpoint of obtaining useful clinical information. Accordingly, the field of view should not be reduced excessively; generally, 20 cm is a practical lower limit. Similarly, a matrix size of approximately 192 × 128 is usually sufficient, and a slice thickness of 4–7 mm is the approximate lower limit before images become signal starved. The use of minimum bandwidth techniques will also increase signal to noise. Finally, it should be noted that some groups have found steady state free procession gradient echo sequences result in high quality T2-weighted images (the FIESTA sequences using General Electric equipment, or the trueFISP sequence using Siemens equipment) (Chung et al. 2000). While such images are certainly comparable to SS-RARE based acquisitions, there seems to be little to choose between the two approaches, and this choice can reasonably be based on local practice and preference. The acquisition of multiple contiguous slices at MRI allows for easy and accurate measurement of fetal volumes, both of the entire fetus and of individual fetal organs (Baker et al. 1994b, c 1995; Coakley et al. 2000). Assessment of fetal liver volume by prenatal MRI may facilitate recognition of intrauterine growth retardation, which is difficult to diagnose accurately using clinical or sonographic criteria. In a study of 32 high-risk pregnancies, 11 resulted in the birth of a fetus with intrauterine growth retardation (Baker et al. 1995). Ten of these 11 fetuses had an abnormally small liver volume at prenatal MRI.

10.5  Normal MRI Findings of the Fetal Body Two major points should be kept in mind when considering normal findings at MRI of the fetal body. First, the fetus is a rapidly developing growing structure so that gestational age is a major factor in the ability to adequately delineate normal anatomy. From a practical point of view, studies conducted before 18 weeks’ gestation may be of limited diagnostic yield. This tends not to be a problem as most fetal MRI studies will be performed after this date, as they are generally requested subsequent to an abnormal or suspected abnormal detailed obstetric sonogram (which is generally performed at 18 weeks). The difference in organ visualization in early gestation compared to late gestation was nicely illustrated in a study of 54 fetuses examined by the SS-RARE technique where individual organ depiction was graded on a scale from 1 (organ not seen) to 4 (all of organ seen) by two readers (Levine et al. 1998). These results are summarized in Table 10.1. It cannot be overemphasized that results obtained for example in a fetus at 24 weeks’ gestation may be hugely superior to results obtained in a fetus at 16 weeks’ gestation. Table 10.1  Organ visualization scores in a study of 54 fetuses examined by SS-RARE technique (Baker et al. 1995), where the score is the mean of grading on a four point scale by two readers (1 = organ not seen, 2 = less than 50% of organ seen, 3 = more than 50% of organ seen, and 4 = all of organ seen) Organ

<20 weeks

>20 weeks

Lungs

3.8

4.0

Heart

3.8

4.0

Heart: four chambers

1.0

1.0

Diaphragm

3.4

3.9

Stomach

3.8

4.0

Liver

3.9

4.0

Gallbladder

2.9

3.7

Spleen

2.2

3.0

Small bowel

3.5

3.8

Large bowel

1.8

3.1

Umbilical cord insertion

3.2

3.7

Kidneys

3.7

3.9

Urinary bladder

3.5

4.0

Gender

3.1

3.7

Note the generally lower scores before 20 weeks’ gestation, and the low scores for evaluation of internal cardiac anatomy

1215

1216

10  MRI of the Fetal Body

Fig. 10.1  a Spoiled gradient echo T1 coronal image through the maternal abdomen at 28 weeks of gestation. Note the fetal lungs (white arrows) are of low T1 signal intensity while the fetal liver (black arrow) is of relatively high T1 signal intensity. b Singleshot RARE T2 coronal image through the maternal abdomen in the same case shows that the fetal lungs (white arrows) are of high T2 signal intensity, due to the contained amniotic fluid, while the fetal liver (black arrow) is of low T2 signal intensity

Fig. 10.2  Bilateral congenital diaphragmatic hernia at 22 weeks’ gestation. Diagnosis was suggested at ultrasound and confirmed by MRI. a Coronal ultrasound of fetal chest showing echogenic structures (white arrows) in both hemithoraces. Vascular pattern (black arrow) is atypical for pulmonary vessels, and raises the consideration of bilateral congenital diaphragmatic hernia. b Coronal fast spoiled gradient-echo T1-weighted (TR/TE of 140/4.2 ms, 70° flip angle) image shows tissue of liver signal

The second major concept that should be grasped when assessing normal anatomic findings is that in many cases, MRI and ultrasound are complimentary and structures that are well assessed by ultrasound may be poorly assessed by MRI and vice versa. For example, bony structures are poorly seen at prenatal MRI and evaluation of conditions such as spina bifida or radial hypoplasia can be extremely difficult at MRI, even though such conditions can be diagnosed with relative ease at ultrasound. Conversely, organs that are of similar echogenicity may be difficult to distinguish at ultrasound, but may be easily distinguished at MRI. Important examples include the signal intensity of lung, liver, and bowel, and therefore these will be discussed in greater detail. Unlike ultrasound, MRI allows the lung and liver to be very easily distinguished, because lung is of low T1 and of high T2 signal intensity, whereas the fetal liver is of high T1 and low T2 signal intensity (Fig. 10.1). The signal intensity of the fetal lung can be easily explained by remembering that in utero the fetal lung and airways are filled with fluid, not air, and are accordingly of high T2 intensity. These signal characteristics are not simply useful in assessing normal findings, but can also be of greater utility in the assessment of pathologic findings. Examples include the evaluation of left hepatic lobe position in left congenital diaphragmatic hernia (i.e., whether or not the left hepatic lobe has herniated up into the chest) and the evaluation of suspected bilateral congenital diaphragmatic hernia.

intensity (arrows) in both hemithoraces. c Corresponding T2weighted single-shot rapid acquisition with relaxation enhancement MR image (TR/effective TE, infinity/100) also shows tissue of liver signal intensity (arrows) in both hemithoraces. The T1 and T2 signal intensity of the tissue in the hemithoraces is that of liver, not lung, and therefore indicates bilateral CDH (after counseling, the parents opted for termination, and the diagnosis of bilateral CDH was confirmed at autopsy)

10.5  Normal MRI Findings of the Fetal Body

Whereas the latter condition could be extremely difficult to diagnose at ultrasound because there is no mediastinal shift, it can be easily diagnosed on MRI because the tissue within the chest will demonstrate the similar characteristics of liver rather than lung, that is, the tissue will be bright on T1 and low on T2-weighted images (Fig. 10.2). The signal intensity of bowel is often bright on T1-weighted images presumably due to the presence of meconium. This could be helpful for example in the identification of the normal rectum (Fig. 10.3). In the study of 40 fetuses with normal gastrointestinal tracts at 20–32 weeks’ gestation, the rectum was identified in 36 cases (90%) (Saguintaah et al. 2002). Bright T1 signal may also help identify otherwise-obscure abdominal masses at ultrasound, as being of gastrointestinal origin. Finally, another limitation of MRI that we have noted in prenatal evaluation is the assessment of the internal genitalia. While the external genitalia can be frequently identified and allow relatively confident distinction of male and female (Fig. 10.4), the internal genitalia are much more difficult to visualize. We have occasionally been asked to image the fetal pelvis in cases of ambiguous external genitalia in order to determine if a uterus is present or not (Fig. 10.5). Our experience is that this is an extremely difficult evaluation to perform with confidence, and is probably not an appropriate indication for prenatal MRI. It should be noted that the protocol outlined above is simply an outline that may need to be modified and

adapted to the particular indications. While it would be impossible to predict and enumerate every variation that might be required, the need for flexibility can be illustrated by the example of congenital hemochromatosis. In cases of suspected congenital hemochromatosis, a T2*-weighted sequence can be used to demonstrate low signal in the liver and hence confirm the diagnosis (Fig. 10.6). A T2*-weighted gradient-echo sequence with long repetition and echo times and a small flip angle (130/20, 20° flip angle) can be used, as this has been shown to optimize MR quantification of iron overload in adult hemochromatosis (Gandon et al. 1994).

Fig. 10.4  a Sagittal T2-weighted single-shot RARE image of a male fetus at 30 weeks’ gestation. The fetus was referred for a further evaluation of a cystic abdominal mass seen at prenatal ultrasound. The presence of a penis and scrotum (arrow) is easily appreciated, allowing identification of gender. b Axial T2weighted single-shot RARE image of a female fetus at 30 weeks’ gestation. The labia majora (arrows) are visible allowing for gender determination

Fig. 10.3  Spoiled gradient-echo T1 coronal image of the fetal torso at 30 weeks’ gestation. The fetus was being assessed for a cystic abdominal mass of indeterminate nature on prenatal ultrasound. The presence of high T1 signal intensity meconium in the rectum (arrow) is suggestive of a normal gastrointestinal tract, and the mass was ultimately found to be of genitourinary tract origin

Fig. 10.5  a Axial T2-weighted single- shot RARE image in a fetus referred for ambiguous external genitalia at 28 weeks’ gestation. The bladder (b) can be easily identified and a soft tissue structure (arrow) posterior to the bladder was tentatively identified as the uterus. Female gender was confirmed on amniocentesis and by postnatal examination. This case illustrates the difficulty of identifying the internal genitalia with confidence at prenatal MRI

1217

1218

10  MRI of the Fetal Body

10.6  MRI Findings in Pathologic Conditions of the Fetal Body 10.6.1  Airway Compromise

Fig. 10.6  a T2*-weighted spoiled gradient-echo coronal image through the maternal abdomen in a normal fetus using an echo sequence with long repetition and echo times and a small flip angle (130/20, 20° flip angle). The maternal liver (white arrow) and the fetal liver (black arrow) are seen to be of similar signal intensity (case included for comparison with b). b Sagittal T2*weighted spoiled gradient-echo sequence through the maternal abdomen. The fetal liver (horizontal arrow) seems to be of lower signal than the maternal liver (vertical arrow). The diagnosis of congenital hemochromatosis was suspected based on maternal history (two prior fetal demises due to hemochromatosis) and the finding of oligohydramnios in the current pregnancy. The low signal on T2*-weighted imaging was considered indicative of congenital hemochromatosis, which was confirmed postnatally. The MRI findings led to early delivery with intensive postnatal treatment for severe liver failure, which slowly resolved spontaneously without liver transplantation (the standard treatment for congenital hemochromatosis)

Fig. 10.7  Sagittal T2 single-shot RARE sequence through a fetus with a giant neck mass. The primary considerations for such a mass are cystic hydroma and teratoma. In this case, the mass is seen to be replacing the structures of the lower face and resulting in fetal hydrops with gross integumentary edema (black arrow) and intra-abdominal ascites (white arrow). After parental counseling, the parents opted for termination and a histopathological diagnosis was not established

Upper airway obstruction at birth is life threatening. Congenital obstruction of the upper airway is usually extrinsic due to a large neck mass (Fig. 10.7). The common congenital neck masses are cystic hygroma and teratoma (Donaldson et al. 1990). Cystic hygromas are composed of cystic lymphatic spaces, possibly secondary to local failure of lymphatic connections during development. Cystic hygromas are often complicated by hydrops, likely due to compression of neck vessels. Chromosomal anomalies are present in 30–70% of fetuses. Cervical teratomas are usually benign isolated tumors that can be cured by surgery, provided the airway can be maintained during and after delivery. Cystic hygromas and teratomas may both appear solid or cystic on prenatal imaging. The finding of a predominantly solid tumor or a cystic tumor with solid nodules favors the diagnosis of teratoma, while intrathoracic extension favors the diagnosis of cystic hygroma (Donaldson et al. 1990). Congenital high-airway obstruction syndrome (CHAOS) is a very rare intrinsic form of obstruction of the larynx or upper trachea (turner et al. 1986). The exact pathology has not been well described, because of the rarity of the condition, and because previously reported cases have resulted in fetal or neonatal demise. Upper-airway obstruction results in retention of bronchial secretions and pulmonary distension by the retained fluid. Overinflation of the lungs, with flattening or eversion of the diaphragm, is thought to impair venous return to the heart, resulting in fetal hydrops and ascites. This results in a characteristic constellation of ultrasound findings, including large bilateral echogenic fetal lungs, flattening, or eversion of the diaphragm, dilated fluid-filled airways below the level of obstruction, and fetal hydrops or ascites. These findings are also seen at MRI (Fig. 10.8). The ex utero intrapartum treatment (EXIT) procedure is a form of surgical delivery that can be used for fetuses with a prenatal diagnosis of upper airway obstruction (Quinn and Adzick 1997). During the EXIT procedure, the mother is anesthetized with inhaled halogenated agents to promote uterine relaxation, and the fetal head and neck are delivered through a hysterotomy. The fetus remains on placental circulation during this period, during which the airway can be secured in a controlled fashion. Bronchoscopy, endotracheal intubation, and tracheostomy can be performed as appropriate. The EXIT procedure was developed to deliver fetuses with congenital diaphragmatic hernia after therapeutic tracheal occlusion, but the technique has also been successfully applied to fetuses with large neck masses (Donaldson et al. 1990; Hubbard et al. 1988). CHAOS has been successfully managed by a combination of fetal tracheostomy and delivery using the EXIT procedure (Angtuaco et al. 1992).

10.6  MRI Findings in Pathologic Conditions of the Fetal Body

Fig. 10.8  a Coronal ultrasound image in a fetus undergoing screening examination at 24 weeks’ gestation. The lungs (L) are enlarged and echogenic with inversion of the diaphragm (arrow). The liver is outlined by a large volume of ascites (asterisk). The constellation of findings was considered suggestive of congenital high airway obstruction syndrome (CHAOS). b Oblique sagittal T2-weighted single-shot RARE sequence through the same fetus confirms diaphragmatic inversion (thick white arrow) due to pulmonary over-expansion, large-volume ascites

(asterisk), integumentary edema (black arrow), and a fluid-filled, blind-ending structure (thin white arrow) in the neck in the expected location of the trachea. c Repeated sagittal T2-weighted single-shot RARE images were obtained through the fetal neck and all demonstrated a blind-ending, fluid-filled structure (arrow) in the expected location of the trachea. These findings were considered to confirm the diagnosis of CHAOS, and the fetus successfully underwent in utero tracheostomy. The diagnosis was confirmed postnatally

10.6.2  Pulmonary Sequestration

through the pulmonary veins. The lesion is believed to result from abnormal budding of the primitive foregut (resulting in an “accessory lung”), although there is usually no visible communication with the tracheobronchial tree. Sometimes a fibrous pedicle is seen accompanying the feeding and draining vessels and may represent the involuted foregut bud. Intralobar sequestration is characterized pathologically by chronic inflammation, fibrosis, and cystic changes. The developmental nature of intralobar sequestration is controversial. It is possible that at least some cases of intralobar sequestration are acquired rather than congenital, and are due to bronchial obstruction followed by distal infection and recruitment of a systemic arterial supply through pleural granulation tissue. This hypothesis would explain why intralobar sequestration is frequently diagnosed after adolescence, the rarity of associated additional congenital anomalies, the occasional presence of a communication with the bronchial tree, and the characteristic pattern of venous drainage through the pulmonary veins. At prenatal imaging, extralobar sequestration typically appears as a solid, well-defined triangular mass that is echogenic on ultrasound or hyperintense on T2weighted MRI (Fig. 10.9) and that can be seen as early as 16 weeks’ gestation (Hubbard et al. 1999; Goldstein 2000). Extralobar sequestration is usually left-sided and can be supradiaphragmatic (90%) or subdiaphragmatic (10%). Cystic changes can sometimes be seen, particularly in hy-

Pulmonary sequestration is a developmental mass of non-functioning bronchopulmonary tissue that is separated from the tracheobronchial tree and receives arterial blood from the systemic circulation. Pulmonary sequestration accounts for 6% of congenital thoracic lesions and more than 75% of prenatally detected lung lesions (Adzick et al. 1998). It is sometimes difficult to distinguish sequestration from other congenital pulmonary abnormalities, and accordingly, MRI is increasingly used as a supplement to obstetric ultrasound in complex fetal anomalies, including thoracic lesions such as sequestration (Hubbard et al. 1999). Pulmonary sequestration is classified as either extralobar (15–25%) or intralobar (75–85%), depending on whether the sequestration has a separate pleural investment or is within the pleura of the lung, respectively. Most, if not all, prenatally diagnosed sequestrations are extralobar. Extralobar sequestration is more common in males (4:1). Extralobar sequestration is characterized pathologically by diffuse dilation of bronchioles, alveoli, and subpleural lymphatic vessels. The arterial supply is nearly always from the descending aorta, and likely represents persistence of primitive splanchnic arteries that supply the early foregut. Venous drainage is usually through the azygos system or the inferior vena cava; although in 25% of cases the venous drainage is

1219

1220

10  MRI of the Fetal Body

Fig. 10.9  a Single-shot RARE T2-weighted coronal image through the fetal chest at 22 weeks’ gestation, demonstrating a triangular hyperintense mass (arrow) in the left chest. b Axial single-shot RARE image demonstrates the mass (M) with cardiomediastinal displacement of the left (L) and right (R) lungs and the heart (H) to the right side. The diagnosis of sequestration was confirmed postnatally

Fig. 10.10  Coronal T2-weighted single-shot RARE sequence in a fetus at 22 weeks’ gestation, with sequestration. The presence of a feeding artery (arrow) arising off the descending aorta is well demonstrated. The presence of such a feeding artery supports the diagnosis of sequestration over a diagnosis of CCAM

brid lesions that combine elements of sequestration and congenital cystic adenomatoid malformation, although generally the presence of intralesional cysts would favor the diagnosis of congenital cystic adenomatoid malformation. Visualization of a systemic feeding artery (Fig. 10.10) arising from the thoracic or abdominal aorta is a useful finding that distinguishes sequestration from other masses such as congenital cystic adenomatoid malformation, lobar emphysema and bronchial atresia. Color and spectral Doppler can be helpful in visualizing the feeding artery, but visualization may still be difficult. In one study, a feeding systemic artery was identified by Doppler ultrasound in only 4 out of 10 cases of pathologically proven sequestration (Becmeur et al. 1998). The frequency with which MR imaging identifies feeding vessels has not been systematically established. Hubbard et al. (1999) did not visualize a feeding artery in both their cases, but we have often seen such vessels in cases of sequestration undergoing MRI at our institution (Fig. 10.10). An ipsilateral pleural effusion is seen in 6–10% of fetuses with extralobar sequestration, and may be related to the common pathological finding of dilated subpleural lymphatics or to torsion around the connecting vasculature and fibrous pedicle. The finding of a unilateral pleural effusion in association with a prenatal thoracic mass is suggestive of extralobar sequestration. Extralobar sequestration can occupy between one and two thirds of the hemithorax, and may cause mediastinal shift and even fetal hydrops. Detailed ultrasound evaluation of the entire fetus is important when an extralobar sequestra-

tion is suspected, because associated anomalies are common. The reported incidence of such anomalies ranges between 11 and 65% (Curtis et al. 1997; Stocker 1986). Associated anomalies that have been described include congenital diaphragmatic hernia (Fig. 10.11), congenital cystic adenomatoid malformation, diaphragmatic eventration or paralysis, bronchogenic cyst, pericardial defect, foregut duplication or diverticulum, ectopic pancreas, vertebral anomalies, and pectus excavatum. The incremental benefit of MRI over ultrasound remains under investigation. In our experience, MRI has been helpful in complex cases where there is an associated anomaly such as congenital diaphragmatic hernia and in the distinction of subdiaphragmatic sequestration from congenital neuroblastoma; sequestration is characterized by very high and uniform T2 signal intensity (Fig. 10.12). Extralobar sequestration has an excellent prognosis and frequently regresses spontaneously in utero. Rarely, the volume of fluid or lymph secreted by the mass may cause a tension hydrothorax. Tension hydrothorax may result in hydrops secondary to vena caval obstruction and cardiac compression. The detection of hydrops in a fetus with sequestration may be an indication for in utero drainage by thoracocentesis or thoracoamniotic shunting. In one series, substantial regression was seen in 28 of 41 fetuses with sequestration on serial prenatal ultrasound (Adzick et al. 1998), all 28 of whom were asymptomatic after birth. The sequestration did not regress and required postnatal resection for respiratory symptoms in seven cases. Four fetuses developed tension hydrothorax

10.6  MRI Findings in Pathologic Conditions of the Fetal Body

Fig. 10.11  a Axial T2-weighted single-shot RARE image of a fetus at 26 weeks’ gestation, with a left congenital diaphragmatic hernia and associated CCAM. Herniated stomach (white arrow) and liver (L) are visible at the left hemithorax, with displacement of the heart (H) and lungs to the right side. In addition, a large sequestration (asterisk) is seen within the left hemithorax and is of hyperintense T2 signal. b Sagittal T2-weighted singleshot RARE image demonstrating the herniated liver (black arrow), herniated stomach (S), and sequestration (asterisk) in the left hemithorax

Fig. 10.12  Coronal T2-weighted single-shot RARE image in a fetus at 28 weeks’ gestation, with left-sided subdiaphragmatic sequestration. Note that the subdiaphragmatic mass (arrow) is of markedly and uniformly hyperintense T2 signal, which helps distinguish this lesion from congenital neuroblastoma. This finding assisted parental counseling, allowing reassurance that the lesion was benign. The diagnosis was confirmed at postnatal resection

with secondary hydrops in utero; three were successfully treated prenatally by thoracoamniotic shunt placement (n = 2) or serial thoracocenteses (n = 1), and one died despite postnatal resection and extracorporeal membrane oxygenation. The two remaining pregnancies underwent elective termination. In a second series of 13 fetuses with sequestration, all fetuses survived and only one fetus required in utero placement of a thoracoamniotic shunt for tension hydrothorax (Lopoo et al. 1999). 10.6.3  Congenital Cystic Adenomatoid Malformation Congenital cystic adenomatoid malformation (CCAM) refers to a developmental mass of proliferated terminal bronchioles that may communicate with the airways or gastrointestinal tract. The lesion is typically supplied from the pulmonary artery. At prenatal MRI, CCAM appears as a well-defined chest mass of higher T2 signal intensity than the normal adjacent lung (Fig. 10.13). This may be indistinguishable from sequestration; however, the presence of internal cysts suggests a CCAM. Conversely, the identification of the feeding artery from the aorta would favor pulmonary sequestration. CCAMs may consist of a few large or medium sized cystic spaces (macrocystic type) or of multiple tiny cysts (microcystic type). Microcystic CCAMs may appear solid to the naked eye and at prenatal ultrasound. Macroscopic CCAMs usually have a benign course, and are treated by postnatal resection.

Fig. 10.13  Coronal T2 SS RARE weighted single-shot RARE image through a fetus at 21 weeks’ gestation, demonstrating a large hyperintense mass (M) in the right chest, with displacement of the other intrathoracic content to the left side. The presence of a small cyst (arrow) is suggestive of congenital cystic adenomatoid malformation, and the diagnosis was confirmed postnatally

Microcystic CCAMs are increasingly recognized as a cause of prenatal demise. Progressive enlargement can lead to compression of the esophagus, vena cava, and lungs, resulting in impaired swallowing, reduced venous return, and pulmonary hypoplasia. Affected fetuses develop polyhydramnios and hydrops fetalis. Prenatally detected CCAMs, especially microcystic CCAMs, should be closely followed. The development of polyhydramnios or hydrops is a harbinger of fetal demise; in a mature fetus, this is an indication for early delivery and in an immature fetus it is an indication for prenatal resection (Flake and Harrison 1995). Favorable results have been

1221

1222

10  MRI of the Fetal Body

reported with open fetal surgery for CCAM, with six survivors in a group of 11 fetuses treated at 21–27 weeks’ gestation (Quinn and Adzick 1997). The prenatal MRI appearances of CCAMs on T2-weighted images have been reported (Hubbard et al. 1999). CCAMs are seen as intrapulmonary masses of increased T2 signal intensity. Discrete cysts can be identified in macrocystic CCAMs, and are of fluid signal intensity. Microcystic CCAMs appear as solid masses that are not as bright as fluid, but are of higher T2 signal than is the adjacent lung. While CCAMs can be identified at prenatal MRI, it is unclear whether MRI has an incremental contribution to surgical planning. CCAMs can be diagnosed by ultrasound, and polyhydramnios and hydrops can be recognized on serial ultrasound surveillance. Potentially, MRI could be used to measure lung volume and monitor lung growth after fetal surgery. CCAM localization by MRI may also be helpful to surgeons, since large field-of-view images obtained in conventional anatomic planes may be helpful in planning the surgical approach.

amniocentesis, although MRI can be used to confirm equivocal ultrasound findings. A specific but poorly understood association with congenital cystic adenomatoid malformation and extralobar sequestration suggests a common underlying pathophysiology (Ryan et al. 1995). Cases of CDH associated with other anomalies have a mortality of 76% (Adzick et al. 1985a). The morbidity and mortality in isolated cases is due primarily to pulmonary hypoplasia, secondary to mechanical compression of the developing lungs. The prenatal diagnosis of CDH is usually established by detailed obstetric ultrasound. Accurate diagnosis is critical for parental counseling, especially with the development of in utero therapeutic tracheal occlusion (which is believed to promote lung growth by retention of bronchial secretions) (Harrison et al. 1998). Temporary tracheal occlusion, which can now be performed by endoscopic rather than open fetal surgery, has been shown to be beneficial in selected cases of CDH (Harrison et al. 1998). Currently, patients are eligible for surgical correction of CDH if the CDH is isolated and the diagnosis of CDH has been confidently established. Approximately 83% of CDH are left-sided, with ultra10.6.4  MRI for Fetal Lung Maturity sound or MRI demonstrating stomach and bowel loops Respiratory distress syndrome is a condition that affects in the left hemithorax, above the expected position of the 1 in 160 neonates, with 5% mortality. Prematurity is the left hemidiaphragm (Fig. 10.14). The stomach is seen as major risk factor because the disorder results in insuffi- a fluid filled gastric-shaped structure of low T1 and high cient surfactant, which begins to appear in the bronchial T2 signal intensity. Bowel loops appear as tubular serpigsecretions in later pregnancy. The current method of inous structures of either high or low T1 and T2 signal evaluating fetal lung maturity depends on amniocente- intensity, with the variable signal intensity of bowel presis, and examining the amniotic fluid for the presence sumably secondary to the presence or absence of mecoof surfactant. Of note, surfactant is composed primarily nium. Fetal lungs are of relatively high T2 signal intensity of phosphatidylcholine (lecithin), and at least in theory because they are filled with fluid and cardiomediastinal should be detectable by magnetic resonance spectroscopy. shift to the right and compression of both lungs may be Preliminary in vitro and in vivo studies at Georgetown best appreciated on axial images (Fig. 10.15). University have suggested that this can be done (Fenton In 57–86% of left CDHs, the herniated viscera include et al. 1998, 2000). Research is ongoing at the University a portion of liver (“liver-up”) (Guibaud et al. 1996; Metof California, San Francisco, and while our initial studies have also shown that choline can be detected both in vitro and in vivo, our experiences have shown that these studies are technically challenging and substantial additional work will be required before this modality may or may not become available as a non-invasive option for the evaluation of fetal lung maturity. 10.6.5  Congenital Diaphragmatic Hernia Congenital diaphragmatic hernia (CDH) is a developmental defect in the posterolateral diaphragm with herniation of abdominal viscera into the thorax. The incidence is 1 in 3,000–4,000 live births, with an overall mortality of 68%. Etiology is unknown, but a third of cases are associated with chromosomal or additional anatomic abnormalities. A wide spectrum of coexistent chromosomal or structural abnormalities may be seen and are primarily detected by detailed ultrasound, echocardiography, and

Fig. 10.14  Coronal T2-weighted single-shot RARE image in a fetus at 30 weeks’ gestation, with left CDH, demonstrating stomach (S) and bowel loops (arrow) in the left hemithorax, above the expected position of the left hemidiaphragm. The heart (H) is displaced to the right

10.6  MRI Findings in Pathologic Conditions of the Fetal Body

kus et al. 1996). Assessment of liver position is of major clinical importance because isolated liver-up and liverdown CDH have mortality of 57 and 7%, respectively (Metkus et al. 1996; Adzick et al. 1985b). In liver-up CDH, herniated liver appears in the left hemithorax as tissue that is continuous with and has the same signal characteristics as non-herniated liver (Fig. 10.16). In liver-down CDH, the liver remains inferior to the expected position of the left hemidiaphragm (Fig. 10.17). Fetal liver is of

relatively high T1 and low T2 signal intensity with coronal T1-weighted images particularly helpful for liver visualization. The position of the stomach in the chest, as seen on axial images, is an indirect indicator of liver position. As the liver herniates anteriorly, the stomach is displaced posteriorly (Fig. 10.18). Gastric distension is a recognized but unexplained finding in left CDH (R.A. Filly, personal communication). On MRI, the gastric outlet often appears stretched, and this may contribute to impair gastric

Fig. 10.15  Axial T2 single-shot RARE image of a fetus with a liver up left CDH. Herniated left hepatic lobe (L), stomach (S), and bowel (B) are seen at the left hemithorax with displacement of the heart (H) and right lung (R) to the right side of the chest. The left lung is not visible. Note that herniation of the left hepatic lobe upwards pushes the stomach more posteriorly into the midchest, which is a secondary sign that can assist in the distinction of liver up from liver down CDH

Fig. 10.16  a Spoiled GE T1 coronal image of a fetus with liverup left CDH. Herniation of the left hepatic lobe (white arrow) into the left hemithorax is visible. Note that the left hepatic lobe can be identified due to continuity and isointensity with the subdiaphragmatic liver, which is of high T1 signal intensity. In addition, the branching structure of the left portal vein (black arrow) can be seen within the herniated left hepatic lobe. b Coronal T2-weighted single-shot RARE image of the same fetus demonstrates the herniated left hepatic lobe to be of low T2 signal intensity (vertical arrow). The left portal vein (horizontal arrow) appears as a hyperintense T2 signal structure on this image

Fig. 10.17  a Coronal T1-weighted spoiled gradient-echo image of a fetus at 25 weeks’ gestation, with liver-down left CDH. The left hepatic lobe (arrow) is seen to lie below the expected location of the left hemidiaphragm. b Coronal T2-weighted singleshot RARE image again confirms the left hepatic lobe (arrow) to lie below the expected location of the left hemidiaphragm. The heart (H) is seen in the left hemithorax

Fig. 10.18  Axial T2-weighted image of a fetus with a liver-down left CDH herniated stomach (S) and bowel (B) are seen at the left hemithorax, with the displacement of the heart (H) and right lung (R) into the right hemithorax. The left lung is not visible. Note that the herniated stomach lies relatively anterior, because it has not been displaced posteriorly by herniation of the liver (compared to Fig. 10.15, showing an axial image in a liver-up left CDH)

1223

1224

10  MRI of the Fetal Body

emptying. Organoaxial volvulus of the herniated stomach has also been described (Beckmann and Nozicka 1999), and can be recognized when the greater curvature is superior to the lesser curvature (Fig. 10.19). Approximately 12% of CDHs are right-sided with liver herniation present in virtually all cases (Guibaud et al. 1996), so that the terms liver-up and liver-down are not appropriate. At MRI, right CDH is characterized by liver tissue above the expected position of the right hemidiaphragm (Fig. 10.20). Herniated bowel is less frequently seen, as liver frequently constitutes the entire hernia. Right CDH has a mortality of 80% (Adzick et al. 1985b). Liver herniation in right CDH may result in hepatic venous obstruction and ascites by a Budd-Chiari mechanism (Gilsanz et al. 1986), while localized edema of the head and neck may be due to obstruction of the superior vena cava (Giacocia 1994). Fetal ascites, hydrothorax, and integumentary edema can therefore be seen, without true hydrops. Fluid in the right hemithorax does not

Fig. 10.19  Coronal single-shot RARE T2-weighted image in a fetus with left CDH. Note the apparent volvulus of the stomach, such that the greater curvature (arrow) lies superior to the lesser curvature. This finding is a common observation, as is the gastric distention that is seen in this case

strictly constitute a pleural effusion, since the diaphragmatic defect allows ascites to track freely into the chest; in fact ascites may facilitate in the direct identification of the primary defect (Fig. 10.21), which otherwise is rarely seen directly. Approximately 5% of CDHs are bilateral (Fig. 10.2), and these are uniformly fatal (Adzick et al. 1985b). Ultrasound diagnosis can be difficult because little or no mediastinal shift is present, and because herniated liver can mimic lung tissue. MRI can readily identify herniated liver in both hemithoraces because of the characteristic differences in T1 and T2 signal intensity between lung and liver. Prenatal MRI can contribute to the management of CDH in several ways (Leung et al. 2000). First, MRI can be used to confirm the diagnosis of CDH in difficult or problematic cases; ultrasound may be limited by poor acoustic contrast between fetal lung and herniated abdominal viscera, small field of view, beam attenuation in maternal adiposity, operator dependency, and ultrasound mimics of CDH (Guibaud et al. 1996). Bowel, stomach, and liver can be easily identified and localized at MRI because of the excellent soft-tissue contrast. Second, MRI can be helpful to confirm the presence of additional structural anomalies that are contraindications to surgery (Fig. 10.22). Third, MRI can be used to evaluate liver position. Herniated liver can be identified and localized as above or below the expected position of the left hemidiaphragm. Fourth, MRI can be used to directly measure lung volume in CDH, rather than using the indirect measurement of LHR. Planimetric MR lung volumetry can be used to measure lung volumes in fetuses with normal lungs and fetuses with CDH (Fig. 10.23) (Coakley et al. 2000). Relative lung volume can be calculated by adjusting for fetal size, and relative lung volume is correlated with outcome (Paek et al. 2001). Fifth, MR volumetry can be used to follow response (Fig. 10.24), although there is currently insufficient data to establish volumetric criteria for an adequate response, or to correlate volumetric response with outcome.

Fig. 10.20  a Coronal single-shot RARE T2-weighted image in a fetus with a right CDH. Herniation of the right hepatic lobe (arrow) superiorly is evident, with displacement of the heart (H) to the left. b Sagittal T2-weighted single-shot RARE image showing herniated liver (L) and bowel (arrow) above the diaphragm

10.6  MRI Findings in Pathologic Conditions of the Fetal Body Fig. 10.21  Sagittal T2-weighted SS-RARE sequence in a fetus with a right CDH. The diaphragmatic remnant on the right side can be barely appreciated as a curvilinear structure (arrow) due to the outlining affect of free fluid above and below the diaphragm

Fig. 10.22  a Axial sonogram of the fetal head in a fetus with known left CDH. The cerebellar hemisphere closest to the transducer appears deficient, although it is unclear whether this represents near field artifact or a true morphologic abnormality. (Note that the finding of a second structural anomaly would preclude consideration for fetal surgical therapy of the congeni-

Fig. 23.  Coronal single-shot RARE image of the fetus with a left congenital diaphragmatic hernia. The right lung has been manually traced using an electronic PACS tool allowing calculation of the cross-sectional area of the lung on this image. Calculation of cross-sectional areas on successive images allows calculation of total lung volume, and ultimately of relative lung volume

tal diaphragmatic hernia at our institution.) b Axial T2-weighted single-shot RARE image of the fetal brain confirms unilateral cerebellar deficiency (arrow), and this finding contraindicated further consideration for fetal surgical therapy of congenital diaphragmatic hernia.

Fig. 10.24  a Axial single-shot RARE image of a fetus at 23 weeks’ gestations, with liver-up left CDH. The relative lung volume was calculated at 20%. b Axial T2-weighted single-shot RARE image at 30 weeks’ gestation after tracheal occlusion at 26 weeks’ gestation. The relative lung volume was calculated at 43%. This illustrates the utility of relative lung volume in therapeutic monitoring. While a relative lung volume of 40% or greater appears to carry a relatively good prognosis, the data remains provisional, as does the interpretation of serial changes in relative lung volume

1225

1226

10  MRI of the Fetal Body

tissue characterization, rather than in improved anatomic characterization (Fig. 10.25) (Hill et al. 2005). In Fetal abdominal diseases encompass a wide array of con- 8 of 422 fetuses (1.9%) that underwent MRI referred ditions that can arise from nearly every structure in the for evaluation of abdominal disease, we found MRI abdominal cavity. The prognosis and treatment of these was of supplemental value relative to ultrasound due diseases is equally variable. Management options include to improved tissue characterization in three of the six corrective fetal surgery (e.g., bladder outlet obstruction), cases with established diagnoses, specifically, in the fepostnatal resection, chemotherapy (e.g., neuroblastoma), tuses with congenital hemochromatosis, subdiaphragand surveillance (e.g., small ovarian cysts) (Walsh and matic sequestration, and cecal atresia with proximal Adzick 2000; Granata et al. 2000). Accordingly, accurate bowel dilatation (Fig. 10.26) (Hill et al. 2005). 10.6.6  Abdominal Masses

diagnosis is crucial for optimal treatment planning and parental counseling, and prenatal ultrasound is the primary modality for the detection and characterization of 1  0.6.6.1  Upper-Quadrant Masses these anomalies (Levine 2001). Fetal abdominal disease is a rare indication for MRI; Upper-quadrant masses seen in the fetus are usually due our research suggests the primary supplemental value to subdiaphragmatic extra lobar pulmonary sequestraof MRI relative to ultrasound lies in the improved tion or neuroblastoma. Subdiaphragmatic extralobar Fig. 10.25  a Coronal T2-weighted single-shot RARE image of the fetus with a solid intra-abdominal mass discovered on screening ultrasound. A large heterogeneous mass (M) is seen abutting the left hepatic lobe. b Axial T2-weighted single-shot RARE image through the mass demonstrates a somewhat irregular contour between the mass (M) and the liver (L). Postnatal resection confirmed that this mass was a large cavernous hemangioma arising from the liver. Note that a hepatic origin for the mass would be difficult to establish with confidence at prospective interpretation

Fig. 10.26  a Axial ultrasound image of fetus at 23 weeks’ gestation, with a nonspecific intra-abdominal hypoechoic mass (arrow). b Coronal T1-weighted spoiled gradient echo image shows the mass (vertical arrow) is of high T1 signal intensity, and lies adjacent to a tubular structure (horizontal arrow) of similar high T1 signal intensity. c Coronal single-shot RARE image shows the

mass (arrow) is of low T2 signal intensity. This combination of similar characteristics is suggestive of meconium, and therefore a gastrointestinal origin was postulated for this mass. At postnatal surgery, the mass was found to represent dilated small bowel proximal to ileocecal atresia

10.6  MRI Findings in Pathologic Conditions of the Fetal Body

pulmonary sequestration on ultrasound examination is typically echogenic, left-sided, and identified in the first trimester, with neuroblastoma cystic right-sided and recognized in the third trimester (Curtis et al. 1997). However, the distinction can be very difficult at ultrasound, whereas subdiaphragmatic extra lobar pulmonary sequestration is typically easily identified at MRI as a mass of uniformly very high T2 signal intensity (Fig. 10.12). 10.6.7  Gastrointestinal Tract Anomalies Gastrointestinal tract anomalies may also be encountered. Normal appearances of the fetal bowel have been described (Saguintaah et al. 2002). After 33 weeks, the jejunum typically contains ingested amniotic fluid exhibiting low T1 signal and high T2 signal. The signal intensity of distal small bowel is more variable and is dependent on meconium content and gestational age. Prior to 32 weeks, the bowel is T1 hyperintense in greater than half of cases, which reduces with gestation to 40% thereafter. Large bowel containing meconium is typically T1 hyperintense and T2 hypointense, with buildup of meconium after 20 weeks, facilitating identification of the rectum. The left colon is frequently identified after 24 weeks, with only 50% of transverse and ascending colon seen at 31 weeks. In duodenal atresia, stomach and duodenum contents are T2 hyperintense to the level of obstruction (Veyrac et al. 1004). In the largest study to date, nine cases of smallbowel atresia demonstrated dilated proximal small bowel (13–30 mm) at the level of the obstruction, with variable signal intensity within the dilated small bowel. In cases of meconium peritonitis, MRI allows for demonstration of associated small-bowel obstruction, with meconium pseudocysts having an intermediate signal on T1 and high signal on T2, with these signal characteristics allowing differentiation from other abdominal cystic collections including the bladder (Veyrac et al. 2004). Enteric duplication cysts are secondary to failure of the lumen to recanalize during embryogenesis. These cysts are of simi-

lar signal to the bladder on all sequences, lacking typical meconium signal on T1-weighted sequences. 10.6.8  Genitourinary Abnormalities Genitourinary abnormalities anomalies account for 14–40% of anomalies detected on ultrasound examination (Filly and Feldstein 2000). Although ultrasound is the primary imaging modality in detection of genitourinary abnormalities, oligohydramnios may be present in up to 50% of cases and can limit evaluation (Poutamo et al. 2000). In these cases, MRI can provide additional information. The ability of the radiologist to make a correct diagnosis can have an impact on major management decisions as urinary tract anomalies may vary from minor abnormalities to fatal conditions (Rapola 1991). Single-shot fast spin-echo sequences have been predominately used in the assessment of the genitourinary system (Poutamo et al. 2000; Caire et al. 2003), because this T2-weighted sequence delineates fluid-filled structures, such as amniotic fluid and urine. Ureteropelvic junction obstruction is the commonest cause of hydronephrosis detected prenatally (Guys et al. 1988). It appears that MRI and ultrasound are similar in the detection and characterization of hydronephrosis (Poutamo et al. 2000). MRI may however have a role to play in assessing the fetal pelvic anatomy in technically difficult cases where images are obscured secondary to the boney fetal pelvis. To date cases of duplicated collecting systems have been imaged with MRI, facilitating characterization of pelvic anatomy and identifying the insertion of duplicated ectopic ureters (Shinmoto et al. 2000). MRI has also correctly identified dilated posterior urethras in four cases of posterior urethral valves; none of these cases had hydroureter (Caire et al. 2003). In the same series, two fetuses with prune belly syndrome were correctly characterized by MRI, although in our experience complex congenital genitourinary anomalies can be difficult to characterize (Fig. 10.27).

Fig. 10.27  Sagittal T2-weighted single-shot RARE image in a fetus at 30 weeks’ gestation, with a complex cystic abdominal mass seen at ultrasound (same case as Fig. 10.4). The mass is seen to have two components, one anterior (vertical arrow) and one posterior (horizontal arrow). Prospectively, the anterior component was considered to represent a urachal diverticulum and the posterior component a dilated bladder, possibly secondary to posterior urethral valves. Postnatal surgery demonstrated that the anterior structure was the bladder and the posterior structure was a massively dilated seminal vesicle due to ectopic insertion of the ipsilateral ureter. This case illustrates the wide spectrum of pathology that may be encountered at fetal imaging, and the associated difficulty of maintaining high accuracy in prospective interpretation

1227

1228

10  MRI of the Fetal Body

In evaluating suspected genitourinary anomalies at prenatal MRI, it is important to assess both the urinary tract and the amniotic fluid volume. Axial T2-weighted imaging of the normal fetal abdomen typically demonstrates the kidneys at the same level as the gallbladder. The absence of bright signal on T2-weighted sequences from the renal pelvic urine at the level of the gallbladder would indirectly indicate renal agenesis. Caire et al. (2003) have defined normal amniotic fluid on MRI examination as three pockets greater than 2 cm in depth. Polycystic kidneys appear to be adequately imaged with ultrasound as demonstrated in a study of 24 fetuses with severe oligohydramnios or ultrasonographically suspected urinary tract abnormalities, where MRI did not provide additional information (Poutamo et al. 2000). In a series of seven cases of multicystic dysplastic kidneys imaged with MRI (Caire et al. 2003), the typical appearance of multiple cysts within the kidneys was well demonstrated. Amniotic fluid was assessed and in cases of anhydramnios, a lethal outcome was predicted secondary to pulmonary hypoplasia. In summary, it appears that MRI is equivalent to ultrasound in detecting renal anomalies, and that the role of MRI may be in detecting of associated extrarenal anomalies, or confirming ultrasound findings when diagnostic certainty may be lowered by confounding oligohydramnios. 10.7  Summary: Indications for MRI for the Fetal Body As previously noted, ultrasound is the primary modality for fetal evaluation and screening. Accordingly, MRI should be reserved for diagnostic confirmation of sonographic abnormality or for further evaluation of equivocal or inconclusive sonographic findings, in cases where the results will clearly influence parental counseling or pregnancy management. It is difficult to provide more detailed or more specific indications because of the emerging and investigational nature of the fetal MRI and the limited evidence basis in the literature. Other applications include fetuses with diagnoses or with differential diagnoses that cannot be made at ultrasound. Examples would include the diagnosis of congenital hemochromatosis and the distinction of subdiaphragmatic sequestration from adrenal neuroblastoma. Emerging indications include the performance of relative lung volumetry and evaluation of fetal lung maturity. Such latter applications should be regarded as highly investigational, and should probably only be performed at academic centers under appropriate research protocols. We have examined the management impact of fetal MRI stratified by the indication based on our initial experience of 44 cases (Tables 10.2, 10.3) (Coakley et al. 1999). Again, this data do not provide guidance on specific diagnoses, but do illustrate that studies performed with a very tailored or specific

Table 10.2  Impact of fetal MRI on clinical management in the first 54 fetal MRI studies performed at UCSF using SS-RARE technique (Guys et al. 1988). In general, studies performed with a focused and specific question were more likely to result in a change in management as compared to those that were less focused (“fishing expeditions”) Indication

Impact

Additional evaluation

2/34 (67%)

Conflicting results

0/4 (0%)

Pre-intervention confirmation

3/3 (100%)

Non-sonographic diagnosis

1/3 (33%)

Total

6/44 (14%)

Table 10.3  Accuracy of fetal MRI on clinical management in the first 54 fetal MRI studies performed at UCSF using SS-RARE technique (Guys et al. 1988), using pathological, clinical, or radiological follow-up as the standard of reference. In only one case was MRI inferior to ultrasound (a fetus with a small subcutaneous fluid collection in the scalp that mimicked a pocket of amniotic fluid at MRI, and was not recognized at prospective interpretation) Concordant

23

Supplementary information

12

Correct arbitration

3

Correct non-sonographic diagnosis

3

Discordant (outside US)—MRI correct

2

Discordant—ultrasound correct

1

clinical question are more likely to have a substantive management impact than those performed as a more unfocused study or “fishing expedition.” Fetal surgery is a risky procedure that is primarily reserved for surgically correctable anomalies that are life-threatening unless reversed prenatally. Isolated CDH with upward herniation of the liver is an excellent example of a condition in which prenatal correction (by tracheal occlusion) is appropriate and appears superior to standard postnatal treatment. Prenatal MRI complements ultrasound because of larger field of view, superior soft-tissue contrast, easier and more precise volumetric measurement, and greater accuracy in the demonstration of intracranial abnormalities. While ultrasound remains the primary modality for fetal imaging, these advantages of MRI make it a valuable adjunct to fetal surgery. Prenatal MRI can be used to assess the anomaly, exclude

10.7  Summary: Indications for MRI for the Fetal Body

other defects that might preclude surgery, and follow re- 14. Curtis MR, Mooney DP, Vacarro TJ et al (1997) Prenatal sponse and evaluate complications. Specific indications ultrasound characterization of the suprarenal mass: disand guidelines are likely to remain in flux, because of the tinction between neuroblastoma and subdiaphragmatic rapidly developing nature of both fetal surgery and preextralobar pulmonary sequestration. J Ultrasound Med natal MRI. 16:75–83 References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Adzick NS, Harrison MR, Crombleholme TM, Flake AW, Howell LJ (1998) Fetal lung lesions: management and outcome. Am J Obstet Gynecol 179:884–889 Adzick NS, Harrison MR, Glick PL, Nakayama DK, Manning FA, deLorimier AA (1985a) Diaphragmatic hernia in the fetus: prenatal diagnosis and outcome in 94 cases. J Pediatr Surg 20:357–361 Angtuaco T, Shah H, Mattison D, Quirk J (1992) MR imaging in high risk obstetric patients: a valuable complement to US. Radiographics 12:91–109 Baker PN, Johnson IR, Harvey PR, Gowland PA, Mansfield P (1994a) A three-year follow-up of children imaged in utero with echo-planar magnetic resonance. Am J Obstet Gynecol 170:32–33 Baker PN, Johnson IR, Gowland PA et al (1994b) Fetal weight estimation by echo-planar magnetic resonance imaging. Lancet 343:644–645 Baker PN, Johnson IR, Gowland PA, Freeman A, Adams V, Mansfield P (1994c) Estimation of fetal lung volume using echo-planar magnetic resonance imaging. Obstet Gynecol 83:951–954 Baker PN, Johnson IR, Gowland PA et al (1995) Measurement of fetal liver, brain, and placental volumes with echoplanar magnetic resonance imaging. Br J Obstet Gynecol 102:35–39 Beckmann KR, Nozicka CA (1999) Congenital diaphragmatic hernia with gastric volvulus presenting as an acute tension gastrothorax. Am J Emerg Med 17:35–37 Becmeur F, Horta-Geraud P, Donato L, Sauvage P (1998) Pulmonary sequestrations: prenatal ultrasound diagnosis, treatment, and outcome. J Pediatr Surg 33:492–496 Caire JT, Ramus RM, Magee KP, Fullington BK, Ewalt DH, Twickler DM (2003) MRI of fetal genitourinary anomalies. AJR 181:1381–1385 Chung HW, Chen CY, Zimmerman RA, Lee KW, Lee CC, Chin SC (2000) T2-weighted fast MR imaging with true FISP versus HASTE: comparative efficacy in the evaluation of normal fetal brain maturation. AJR Am J Roentgenol 175:1375–1380 Coakley FV, Hricak H, Filly RA, Barkovich AJ, Harrison MR (1999) Impact on management of breath-hold MR imaging of complex fetal disorders: preliminary clinical experience. Radiology 213:446 Coakley FV, Lopoo JB, Lu Y, Hricak H, Albanese CT, Harrision MR, Filly RA (2000) Volumetric assessment of normal and hypoplastic fetal lungs by prenatal single-shot RARE MR imaging. Radiology 216:107–111

15. Donaldson JS, Luck SR, Vogelzang R (1990) Preoperative CT and MR imaging of ischiopagus twins. JCAT 14:643–646 16. Fenton BW , Lin CS, Seydel F, Macedonia C (1998) Lecithin can be detected by volume-selected proton MR spectroscopy using a 1.5T whole body scanner: a potentially noninvasive method for the prenatal assessment of fetal lung maturity. Prenatal Diagnosis 12:1263–1266 17. Fenton BW, Lin CS, Ascher S, Macedonia C (2000) Magnetic resonance spectroscopy to detect lecithin in amniotic fluid and fetal lung. Obstet Gynecol 95:457–460 18. Filly RA, Feldstein VA (2000) Fetal genitourinary tract . In: Callen PW (ed) Ultrasonography in obstetrics and gynecology. Saunders, Philadelphia, pp 515–550 19. Flake AW, Harrison MR (1995) Fetal surgery. Ann Rev Med 46:67–78 20. Gandon Y, Guyeder D, Heautot JF et al (1994) Hemochromatosis: diagnosis and quantification of liver iron with gradient -echo MR imaging. Radiology 193:53–538 21. Giacoia GP (1994) Right-sided diaphragmatic hernia associated with superior vena cava syndrome. Am J Perinatol 11:129–131 22. Gilsanz V, Emons D, Hansmann M et al (1986) Hydrothorax, ascites, and right diaphragmatic hernia. Radiology 158:243–246 23. Goldstein R (2000) Ultrasound of the fetal thorax. In: Callen PW (ed) Ultrasonography in obstetrics and gynecology. Saunders, Philadelphia, pp 426–455 24. Gover P, Hykin J, Gowland P, Wright J, Johnson I, Mansfield P (1995) An assessment of the intrauterine sound intensity level during obstetric echo-planar magnetic resonance imaging. Br J Radiol 68:1090–1094 25. Granata C, Fagnani AM, Gambini C, Boglino C, Bagnulo S, Cecchetto G, Federici S, Inserra A, Michelazzi A, Riccipetitoni G, Rizzo A, Tamaro P, Jasonni V, De Bernardi B (2000) Features and outcome of neuroblastoma detected before birth. J Pedi Surg 35:88–91 26. Guibaud L, Filiatrault D, Garel L et al (1996) Fetal congenital diaphragmatic hernia: accuracy of sonography in the diagnosis and prediction of the outcome after birth. AJR 166:1195–1202 27. Guys JM, Borella F, Monfort G (1988) Ureteropelvic junction obstructions: prenatal diagnosis and neonatal surgery in 47 cases. J Pediatr Sur 23:156–158 28. Harrison MR, Mychaliska GB, Albanese CT et al (1998) Correction of congenital diaphragmatic hernia in utero IX: fetuses with poor prognosis (liver herniation and low lungto-head ratio) can be saved by temporary tracheal occlusion. J Pediatr Surg 33:1017–1023

1229

1230

10  MRI of the Fetal Body 29. Hill BJ, Joe BN, Qayyum A, Yeh BM, Goldstein R, Coakley FV (2005) Supplemental value of MRI in fetal abdominal disease detected on prenatal sonography: preliminary experience. AJR 184:993–998 30. Hubbard AM, Crombleholme TM, Adzick NS (1998) Prenatal MRI evaluation of giant neck masses in preparation for the fetal EXIT procedure. Am J Perinatol 15:253–257 31. Hubbard A, Adzick NS, Crombleholme TM et al (1999) Congenital chest lesions: diagnosis and characterization with prenatal MR imaging. Radiology 212:43–48 32. Leung JWT, Coakley FV, Hricak H, Harrison MR, Farmer DL, Albanese CT, Filly RA (2000) Prenatal MRI of congenital diaphragmatic hernia. AJR 174:1607–1612 33. Levine D (2001) Ultrasound versus magnetic resonance imaging in fetal evaluation. Top Magn Reson Imaging 12:25–38 34. Levine D, Barnes PD, Sherr S, Semelka RC, Li W, McArdle CR, Worawattanakul S, Edelman RR (1998) Fetal fast MR imaging: reproducibility, technical quality, and conspicuity of anatomy. Radiology 206:549–554 35. Lopoo JB, Goldstein RB, Lipshutz GS, Goldberg JD, Harrison MR, Albanese CT (1999) Fetal pulmonary sequestration: a favorable cystic lung lesion. Obstet Gynecol 94:567–571 36. Marcos HB, Semelka RC, Worawattanakul S (1997) Normal placenta: gadolinium-enhanced dynamic MR imaging. Radiology 205:493–496 37. Metkus AP, Filly RA, Stringer MD, Harrison MR, Adzick NS (1996) Sonographic predictors of survival in fetal diaphragmatic hernia. J Pediatr Surg 31:148–152 38. Mittermayer C, Blaicher W, Grassauer D, Horcher E, Deutinger J, Bernaschek G, Ulm B (2003) Fetal ovarian cysts: development and neonatal outcome. Ultraschall Med 24:21–26 39. Paek BW, Coakley FV, Lu Y, Filly RA, Lopoo JB, Qayyum A, Harrison MR, Albanese CT (2001) Congenital diaphragmatic hernia: prenatal evaluation with MR lung volumetry—preliminary experience. Radiology 220:63–67 40. Poutamo J, Vanninen R, Partanen K, Kirkinen P. Diagnosising fetal urinoary tract abnormalities: benefits of MRI compared to ultrasonography. Acta Obstet Gynecol Scand 2000:79:65–71 41. Quinn TM, Adzick NS (1997) Fetal surgery. Obstet Gynecol Clin North Am 24:143–157

42. Rapola J (1991) The kidneys and urinary tract. In: Wigglesworth JS, Singer DB (eds) Textbook of fetal and perinatal pathology, vol. 2. Blackwell Scientific, Boston, pp 1109–2243 43. Revel MP, Pons JC, Lelaidier C, Fournet P, Vial M, Musset D, Labrune M, Frydman R (1993) Magnetic resonance imaging of the fetus: a study of 20 cases performed without curarization. Prenat Diagn 13:775–799 44. Ryan CA, Finer NN, Etches PC, Tierney AJ, Peliowski A (1995) Congenital diaphragmatic hernia: associated malformations—cystic adenomatoid malformation, extralobar sequestration, and laryngotracheoesophageal cleft: two case reports. J Pediatr Surg 30:883–885 45. Saguintaah M, Couture A, Veyrac C et al (2002) MRI of the fetal gastrointestinal tract. Pediatr Radiol 32:395–404 46. Schwartz JL, Crooks LE (1982) NMR imaging produces no observable mutations or cytotoxicity in mammalian cells. AJR 139:583–585 47. Shinmoto H, Kashima K, Yuasa Y, Tanimoto A, Morikawa Y, Ishimoto H, Yoshimura Y, Hiramatsu K (2000) MR Imaging of Non CNS Fetal Abnormalities: A Pictorial Essay. Radiographics 20:1227–1243 48. Smith FW, Adam AH, Philips WDP (1983) NMR imaging in pregnancy. Lancet 1:61–62 49. Spencer JA, Tomlinson AJ, Weston MJ, Lloyd SN (2000) Early report: comparison of breath-hold MR excretory urography, Doppler ultrasound and isotope renography in evaluation of symptomatic hydronephrosis in pregnancy. Clin Radiol 55:446–453 50. Stocker JT (1986) Sequestrations of the lung. Semin Diagn Pathol 3:106–121 51. Turner RJ, Hankins GD, Weinreb JC et al (1986) Magnetic resonance imaging and ultrasonography in the antenatal evaluation of conjoined twins. Am J of Obstetr Gynecol 155:645–649 52. Veyrac C, Couture A, Saguintaah M, Baud C (2004) MRI of fetal GI tract abnormalities. Abdom Imaging 29:411–420 53. Walsh DS, Adzick NS (2000) Fetal surgical intervention. Am J Perinatol 17:277–283 54. Wolff S, Crooks LE, Brown P, Howard R, Painter R (1980) Test for DNA and chromosomal damage induced by nuclear magnetic resonance imaging. Radiology 136:707–710 55. Yip YP, Cappriotti C, Talagala SL, Yip JW (1994) Effects of MR exposure at 1.5 T on early embryonic development of the chick. J Magn Reson Imaging 4:742–748

Chapter 11

Whole-Body MRI

11.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1231 H.-P. Schlemmer

11.2 

Examination Techniques  .. . . . . . . . . . . .   1232

11.2.1 

Hardware  . . . . . . . . . . . . . . . . . . . . . . . . . . .   1232

11.2.2 

Sequence Protocols  .. . . . . . . . . . . . . . . . . .   1232

11.2.2.1  Cardiovascular System  . . . . . . . . . . . . . . .   1233 11.2.2.2  Musculoskeletal System  .. . . . . . . . . . . . . .   1233 11.2.2.3  Viscera  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1233 11.3 

Postprocessing/Reading  .. . . . . . . . . . . . .   1235

11.4 

Pathologic Findings  .. . . . . . . . . . . . . . . . .   1235

11.4.1 

Cardiovascular Diseases  . . . . . . . . . . . . . .   1235

11.4.2 

Oncology  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   1237

11.4.2.1  Metastases  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1239

11.1  Introduction H.-P. Schlemmer MRI is understood as a dedicated method for the evaluation of local disease. Despite its ability to image the whole body with high resolution and superior soft-tissue contrast, and consequently to detect systemic spread of diseases with high sensitivity, it is rarely applied for more than one body part within a single examination. The limited awareness of the clinical potential of whole-body MRI (WB-MRI) can be explained by the long examination times and patient discomfort associated with the test and, to a greater degree, by its relatively high cost, which is generally not adequately reimbursed by health insurance. Consequently, the cost-effectiveness of WB-MRI is considered questionable by most radiologists. Mainly because of practical considerations such as availability, time, and cost, CT is most often used as the standard whole-body imaging modality. The medical and economic potential of whole-body imaging is, however, of considerable importance, particularly because systemic illnesses such as cardiovascular diseases, cancer, and diabetes are responsible worldwide

11

11.4.2.2  Lung Cancer  .. . . . . . . . . . . . . . . . . . . . . . . .   1240 11.4.2.3  Prostate Cancer  .. . . . . . . . . . . . . . . . . . . . .   1240 11.4.2.4  Breast Cancer  .. . . . . . . . . . . . . . . . . . . . . . .   1240 11.4.2.5  Colorectal Cancer  .. . . . . . . . . . . . . . . . . . .   1240 11.4.2.6  Multiple Myeloma   .. . . . . . . . . . . . . . . . . .   1242 11.4.2.7  Less Common Tumors   .. . . . . . . . . . . . . .   1243 11.4.3 

Musculoskeletal   . . . . . . . . . . . . . . . . . . . . .   1243

11.4.4 

Adipose-Tissue Quantification  . . . . . . . .   1246

11.4.5 

Early Detection and Screening  . . . . . . . .   1247

11.4.5.1  Prevalence   . . . . . . . . . . . . . . . . . . . . . . . . . .   1250 11.4.5.2  Chance for Early Detection   .. . . . . . . . . .   1250 11.4.5.3  Valuable Examination Tools  .. . . . . . . . . .   1250 References   . . . . . . . . . . . . . . . . . . . . . . . . . .   1254

for substantial morbidity and mortality, not only in highincome but also in low- and middle-income countries (Epping-Jordan et al. 2005). Treatment and disability account for substantial healthcare costs, and may further increase with population aging, and changing social and environmental factors. Preconditions for assessing prognosis and planning individually optimized treatment strategies are early detection and accurate staging of disease. Treatment decisions in chronic and incurable diseases must consider individual disease characteristics and identify patients who are candidates for aggressive medical or surgical interventions, particularly with respect to quality of life and life expectancy. In this context, whole-body imaging provides important components of the decision-making process, including not only disease-specific information, but also accompanying findings (e.g., staging of lung cancer and detection of an unexpected paraneoplastic thrombosis of pelvic veins). Moreover, because treatment strategies are frequently dependent on the response to earlier treatments, tools for monitoring the success of systemic treatments are also urgently needed. Whole-body imaging may ensure that patients receive the most appropriate treatment.

1232

11  Whole-Body MRI

11.2  Examination Techniques 11.2.1  Hardware The field of view (FOV) of a single MRI examination is limited by the length of the magnet and the gradient coil system. As a consequence, state-of-the-art MRI of multiple body parts can be performed only with a step-bystep approach, and subsequent MRI examinations of different body parts are needed, with repeated repositioning of the patient and the surface coils. This approach is infeasible in clinical routine, because it is time-consuming and understandably cumbersome for patients as well as technicians. Furthermore, optimal intravenous contrast medium application is not possible because of the loss of time required for patient repositioning and repeated receiver and transmitter adjustments. Recently, however, technical advances in surface coil and integrated parallel imaging technology (iPAT) have enabled MR imaging of multiple body parts during a single examination. Three different WB-MRI techniques have been developed to subsequently extend the volume coverage from partial toward multiple anatomic regions, without the need for patient repositioning. As relatively long examinations times are major obstacles to routine application, MRI systems with 1.5 T and high-performance gradients have been used for to achieve fast data acquisition with sufficient signal-to-noise ratio (SNR). Patients are typically placed in the supine position with their arms lying beside their bodies. Whether the patient is positioned head or feet first depends on the particular technique used. 1 The radiofrequency (RF) signal is transmitted and received by the body coil. Whole-body coverage is generally achieved by acquiring coronal slabs of turbo STIR sequences at several (usually five) stations (Walker et al. 2000). The advantages of this method over the surface coil methods described below include the effortless performance of the examination and the greater comfort of the patient. The clinical impact for examining systemic diseases of the musculoskeletal system has been shown, e.g., for evaluating bone marrow metastases, for staging of multiple myeloma, or for examining polymyositis (Walker and Eustace 2001; Daldrup-Link et al. 2001; Goo et al. 2005). A drawback of this method as compared with the performance of a series of dedicated state-of-the-art MRI studies, however, is the comparatively moderate spatial resolution, which allows a limited diagnostic accuracy, particularly concerning the brain or moving organs such as the liver and lungs. 2 The RF signal is transmitted by the body coil and received anteriorly by one body phased-array surface coil and posteriorly by two spine coil elements embedded in the patient table (Ruehm et al. 2001). During the examination, the receiving coils remain stationary in the isocenter of the magnet while the patient

is gliding through the body phased-array surface coil on a rolling table platform, which is mounted on the original patient table on seven pairs of roller bearings (AngioSURF, MR-Innovation, Essen, Germany). The signal gain relative to the previously described technique enables an increase in the spatial and/or temporal resolution and contrast-enhanced whole-body–3D MR angiography (MRA) of the whole body is feasible within one single examination. Five contiguous 3D FLASH sequences from the level of the neck to the feet can be recorded within 72 s and repeated twice before and after intravenous contrast administration. One single 3D data set can be acquired within 12 s through partial k-space coverage, while an additional 3 s are needed for moving the table to the next imaging location. The method was initially described for evaluating atherosclerosis, but shortly after also successfully applied for tumor staging (Lauenstein et al. 2002). Promising results have been found for tumor staging in comparison to PET/CT or MSCT (Antoch et al. 2003). It has also been applied for screening purposes, whereby an additional MR sequence for virtual colonoscopy has been integrated for early detection of colon polyps (Goyen et al. 2003). 3 Recently, a novel WB-MRI technology was introduced, employing multiple phased-array surface coils in combination with multiple receiver channels and accelerated image acquisition by integrated parallel imaging technology (iPat) (Schlemmer et al. 2005). The1.5-T MR system enables the connection of up to 76 coil elements and signal acquisition, with up to 32 independent receiving channels for synchronic signal recording. For whole-body imaging a set of five to six phased-array surface coils with multiple individual coil elements are used independent of the body size: one head coil (12 coil elements), one neck coil (4 coil elements), two to three body phased-array coils for thorax, abdomen and pelvis (6 coil elements each), one peripheral angio coil for lower extremities (16 coil elements), and a set of spine coils with 24 elements imbedded in the patient table. The combination of an automatic moveable table, 500-mm FOV per imaging station, a high-performance gradient system (maximum amplitude: 45 mT/m), and parallel imaging technology in all three spatial directions enables highresolution WB-MRI in a state-of-the-art technique, with a maximal FOV length of 205 cm within a total examination time of approximately 1 h. 11.2.2  Sequence Protocols WB-MRI is not built on a unique examination protocol. The development of novel hardware technology and fast imaging techniques in combination with iPAT enables standard MRI sequence protocols to be used in whole-

11.2  Examination Techniques

body examination. The combination of different state-ofthe-art sequences applied in different body parts opens the window for a disease-specific approach by allowing comprehensive examination of those organ systems that are expected to be involved in a particular patho­physio­ logical process. In principle, three different types of whole-body MR protocols can be distinguished for consideration of the cardiovascular system, the musculoskeletal system, and the viscera. To answer an individual clinical question, different examination parts may of course be pieced together by focusing on the individual medical condition and accompanying risk-aggravating factors. 11.2.2.1  Cardiovascular System Three different multi-station techniques have been used for acquiring high-resolution WB-MRA. (1) Patients are positioned feet first and glide on a rolling table platform through one body phased-array surface coil, which is fixed in the isocenter of the magnet (Ruehm et al. 2001). (2) Patients are positioned head first for imaging of the head, neck, and thorax and then repositioned feet first for imaging of the abdomen, pelvis, and legs. A set of surface coils covering the respective body parts, iPAT, and an automatic moveable table is used for image acquisition (Kramer et al. 2005. (3) Patients are positioned head first, with multiple phased-array surface coils covering the entire body. Multiple receiver channels, iPAT, and an automatic moveable table allow high-resolution MRA of all body parts without repositioning (Kramer et al. 2005; Fenchel et al. 2006). The use of iPAT technology enables the extension of the WB-MRA examination protocol to include stateof-the-art cardiac and cerebral MRI for assessing prior myocardial and/or brain infarction probably indicating microvascular occlusive disease. In Table 5.4.16, Chapter 5.4.7 exemplary WB-MRA examination protocols are given for different WB-MRA techniques. The intracranial arterial system is assessed with 3D TOF sequences. Imaging of the arteries from the base of the scull to the foot is performed by contrast-enhanced MRA using 3D FLASH sequences in four to five subsequent stations. Using surface-coil technology the achieved spatial resolution of WB-MRA is comparable to that achieved in conventional MRA of one dedicated body region. Venous overlap, however, may be problematic, particularly in the lowerleg region. For achieving optimal arterial enhancement by avoiding concurrent venous overlap, different imaging strategies with sophisticated contrast media injection protocols are applied after test-bolus timing, whereas one single or two separate contrast media injections with generally biphasic injection protocols are applied depending on the particular WB-MRA technique used. Contrastenhanced images are as usual subtracted from baseline images to increase the vessel-to-background contrast.

11.2.2.2  Musculoskeletal System Coronal STIR sequences are particularly helpful for detecting anatomical abnormalities, pathologic signal in muscle, or bone marrow as well as joint effusion due to inflammation, degeneration, or trauma. The combination of STIR and T1-weighted images in the coronal plane is particularly useful for the detection of bone metastases. Osteoblastic lesions are better visualized on T1-weighted MR images and may be overlooked on STIR sequences. For the detection of possible rib metastases, breathhold T1-weighted GRE and T2-weighted-TSE sequences should be appended. The use of intravenous contrast and T1-weighted sequences with fat saturation is helpful for detecting inflammatory soft-tissue changes, e.g., synovitis, or mass lesions. Whole-spine MRI in the sagittal direction using STIR, T2-weighted or T1-weighted sequences is helpful for the assessment as well as therapy planning and follow-up of a variety of diseases, e.g., degenerative disc disease, osteochondrosis, spondylitis, spondylodiscitis, rheumatoid arthritis, seronegative spondyloarthropathies, or malignant disorders. Table 11.2.1 provides a list of suggested MRI sequences for whole-body and wholespine MRI. Additional axial imaging may be necessary in individual cases and should be determined accordingly during the examination. An individual WB-MRI sequence protocol may therefore be necessary to give consideration to an individual clinical problem. It is important to note, however, that small anatomic structures (e.g., the elbow, forearm, hand, and foot) cannot be sufficiently visualized within a WB-MRI examination. For the examination of those small anatomical structures, including cartilage or ligaments, additional dedicated MRI studies using specific surface coils may be necessary. 11.2.2.3  Viscera Due to the limited total measurement time, compromises have to be made regarding sophistication in support of standard MRI sequences. Screening for cerebral abnormalities should include multiplanar FLAIR, T2-weighted TSE, as well as pre- and post-contrast T1-weighted SE sequences. STIR, HASTE, or VIBE sequences have been proven useful for screening for lung nodules, infiltrates, pleural effusion, and mediastinal or hilar lymph node metastases, but how well their diagnostic accuracy compares to that of CT is not yet certain. For imaging of abdominal organs, axial free breathing, respiratorygated T2-weighted TSE, and breath-hold T1-weighted FLASH sequences with or without frequency selective fat saturation are useful. The pelvis should be examined by axial T2-weighted TSE and T1-weighted TSE or GRE sequences. Intravenous contrast media application can be accompanied by DCE studies using, for example, coronal VIBE sequences covering the abdomen and/or pelvis.

1233

1234

11  Whole-Body MRI Table 11.2.1  Exemplary MRI sequence combinations (modules) applicable within a whole-body MRI examination for assessing body parts Body part

Module

Sequence

iPat factor

Orientation

Spatial resolution (mm3)

Approximate measurement time (min:s)

Whole body

WB1

T2-weighted STIR TSE

2

Coronal

1.8 × 1.3 × 5

15:00

WB2

T1-weighted TSE

2

Coronal

1.7 × 1.3 × 5

10:00

WB3

CE T1-weighted TSE FS

2

Coronal

2 × 1.5 × 5

12:00

S1

T2-weighted STIR TSE

2

Sagittal

1.5 × 1 × 3

05:00

S2

T1-weighted TSE

2

Sagittal

1.4 × 1 × 3

03:20

S3

CE T1-weighted TSE FS

2

Sagittal

1.7 × 1.3 × 3

03:30

R1

T2-weighted STIR TSE, BH

2

Axial

1.8 × 1.2 × 6

00:45

R2

FLASH 2D, MBH

–

Axial

2.1 × 1.5 × 6

01:00

R3

CE FLASH 2D FS, MBH

–

Axial

2.1 × 1.5 × 6

01:30

B1

FLAIR

–

Axial

1.2 × 0.9 × 4

02:30

B2

T1-weighted SE

–

Axial

0.9 × 0.9 × 4

03:50

B3

CE T1-weighted SE

–

Axial

0.9 × 0.9 × 4

03:50

B4

CE T1-weighted SE

–

Coronal

0.9 × 0.9 × 4

04:20

N1

T2-weighted TSE

3

Axial

1.2 × 0.9 × 5

02:30

N2

T1-weighted FLASH 2D FS

2

Axial

1.1 × 0.8 × 5

00:50

N3

CE T1-weighted FLASH 2D FS

2

Axial

1.1 × 0.8 × 5

01:20

T1

T2-weighted STIR TSE, BH

2

Axial

1.8 × 1.2 × 6

00:45

T2

PD-weighted VIBE, MBH

–

Axial

2×2×2

00:18 (× 3)

T3

CE T1-weighted VIBE, MBH

–

Axial

2×2×2

00:18 (× 3)

A1

T2-weighted TSE FS, RT

2

Axial

1.6 × 1.2 × 6

02:00

A2

T1-weighted FLASH 2D FS, MBH

2

Axial

2.1 × 1.5 × 6

01:00

A3

CE T1-weighted FLASH 2D FS, MBH

2

Axial

2.1 × 1.5 × 6

01:00

A4

DCE T1-weighted VIBE, MBH

2

Axial

1.9 × 1.3 × 2

00:20 (× 4)

A5

DCE T1-weighted VIBE, MBH (MRC)

3

Coronal

2×2×2

00:25 (× 4)

P1

T2-weighted STIR TSE

2

Axial

1.5 × 1.2 × 4

4:30

P2

T1-weighted FLASH 2D FS

2

Axial

2.1 × 1.5 × 4

1:00

P3

CE T1-weighted FLASH 2D FS

2

Axial

2.1 × 1.5 × 4

1:00

P4

T2-weighted TSE

–

Axial

1.3 × 0.6 × 4

3:50

Whole spine

Ribs

Brain

Neck

Thorax

Abdomen

Pelvis

T2-weighted STIR TSE T2-weighted-weighted short-tau inversion recovery turbo spin echo, T1-weighted SE T1-weighted spin echo, FLAIR fluid-attenuated inversion recovery, PD-weighted and T1-weighted VIBE 3D BH proton density–weighted and T1-weighted volumetric interpolated, 3-dimensional breath-, T2-weighted TSE T2-weighted turbo spin echo, T1-weighted FLASH T1-weighted fast low-angle shot, FS frequency-selected fat saturation, MRC magnetic resonance colonography, BH breath hold, MBH multibreath hold, RT respiratory triggered, CE contrast enhanced, DCE dynamic contrast enhanced

11.4  Pathologic Findings

Finally, T1-weighted sequences with frequency-selective fat saturation can be performed from the head to the pelvic floor in at least the axial direction. Depending on the individual clinical question, optional sophisticated MR studies can be integrated; an example of such integration would be dark-lumen magnet resonance colonography with a VIBE-sequence, which, however, requires the administration of ca.1.5–2 l of warm tap water in hypotonia (intravenous scopolamine). The composed examination protocol should not require more than about 60 min. Table 11.2.1 gives a list of suggested partial-body MRI sequences that can be put together to compose an individual whole-body examination protocol. Novel sequence developments may further accelerate image acquisition. EPI sequences for whole-body application are limited in terms of spatial resolution, artifacts, and consequently diagnostic accuracy (Bader et al. 2002). 3D VIBE sequences have been successfully applied for rapid and dynamic whole-body imaging with respect to tumor staging (Lauenstein et al. 2004). To reduce artifacts from signal inhomogeneity at the edges of the scans, a partial overlapping of the fields of view is necessary, with consumption of measurement time. To overcome these problems and to further reduce image acquisition time, improvements in MRI sequence technology are currently being made. The sliding multislice (SMS) technique, which allows continuous data acquisition in the isocenter of the magnet while the table is moving with a constant velocity of about 4 mm/s, has recently been developed (Fautz et al. 2006) and resembles spiral CT scanner technology. A WB-MRI scan can be acquired in only about 7 min by continuously measuring data from a relatively small slab, enabling the acquisition of wholebody images even with short-bore magnets. Recently, it has been shown that even WB-MRA is achievable using this approach. However, clinical studies are needed to investigate the diagnostic accuracy of fast whole-body acquisition methods compared to standard (e.g., TSE or GRE) sequences. 11.3  Postprocessing/Reading WB-MRI significantly increases the number of acquired images per patient. A cardiovascular WB-MR examination comprises about 2,000 images, and the few relevant images have to be sorted out first, consuming a notable amount of time and concentration. A WB-MRI protocol to screen for metastases requires the interpretation of around 600 images, and every image has to be carefully reviewed for the presence of masses/lesions suggestive of tumor. The time required for reading, documentation, and discussion of the high number of images varies substantially, and 15–60 min are needed, particularly if additional images, e.g., from follow-up and/or multimodal diagnostic approaches with CT, PET, or PET/CT, have to be evaluated.

As referring clinicians become increasingly aware of a comprehensive whole-body approach that may cut down the total amount of time needed for imaging, radiologists will be faced with heavier workloads. Moreover, as therapeutic decisions are increasingly based on recommendations by multidisciplinary conferences, the essential imaging findings will have to be demonstrated in a fast and comprehensive manner. Logistical implications for work flow optimization need to be considered in order to minimize the time demanded not only for the patient examination but also for the reading and reporting process. Redesigning of the department’s workflow concepts is challenging but a reasonable prerequisite for utilizing the potential of whole-body imaging technology. Two basic aspects have to be considered: (1) the need for a cooperative team of radiologists, radiographers and further helpers along with standardized examination protocols to efficiently organize the patient examinations and achieve efficient patient throughput; and (2) the need for an effective way of handling data, which implies optimized image postprocessing, transferral to the PACS system, and computer-assisted reading and reporting. Thus, advances in picture archiving systems and image navigator tools are urgently needed. The rapid development of computer hardware and software technology opens new possibilities with respect to computer-assisted reading and reporting of whole-body imaging sets. The reading, reporting, and documentation workflow, structured into elementary reading steps according to anatomical regions and/or organ systems (e.g., bone marrow) should be assisted. Furthermore, the subsequent retrieval of previously reported findings should be facilitated for accelerated demonstration of imaging findings as well as evaluation of follow-up studies. A novel prototype software that assists the radiologist in imaging analysis, structured reporting, and the fast retrieval of previously described findings is currently being developed (through collaboration between the Dept. of Diagnostic Radiology, University Tübingen and Siemens Medical Solutions, Erlangen, and MeVis, Bremen, Germany) (Fig. 11.3.1) (Muller-Horvat et al. 2007). 11.4  Pathologic Findings 11.4.1  Cardiovascular Diseases Cardiovascular diseases and associated complications are the leading cause of mortality and morbidity in the industrialized countries. The diagnostic approach to study this systemic disease has remarkably changed in the past. Atherosclerosis is considered now as a chronic inflammatory process of the large arteries, which starts already at an early age and progresses without symptoms for many years. In the past, clinical studies predominantly focused on the diagnosis and treatment of clinical events, e.g., arterial occlusive disease, myocardial infarction, or stroke, and local imaging strategies were optimized accordingly.

1235

1236

11  Whole-Body MRI

Fig. 11.3.1  Desktop areas of prototype postprocessing software for whole-body MRI (developed through a collaborative research effort of the Department of Radiology, University Hospital Tuebingen and Siemens, Erlangen, and MeVis, Bremen, Germany). An anatomical navigator enables intuitive access to images, which are sorted according to anatomical regions and sequence types; for example, pre- and post-contrast series (above) or different orientations of the same sequence type (below, STIR) are

displayed next to each other automatically. A finding marked by the radiologist on one selected image is automatically assigned to all other images displaying the same anatomical region and stored in written form in a table. Conversely, by selecting previously marked findings from the table the respective images with marked findings can be retrieved directly, enabling accelerated demonstration of findings as well as follow-up evaluation

Although of tremendous clinical importance, however, these events mark only an endpoint of the progressive pathogenesis in systemic atherosclerosis including plaque formation, rupture, thrombosis, and subsequent tissue ischemia. Early detection of subclinical disease is considered now to be more important for preventing the development of final clinical events. As in oncology, an early diagnosis of cardiovascular diseases in an asymptomatic stage is of major importance for achieving the best treatment outcome. The role of diagnostic imaging complies now with the pathogenesis of atherosclerosis and shifts from purely detecting local vascular stenoses following an already occurred ischemic event toward revealing subclinical vascular abnormalities anywhere in

the body preceding those events and therefore indicating individual risk factors. WB-MRA may accordingly play an important role for detecting atherosclerotic abnormalities of the large arteries, monitoring their progression, and consequently supporting the identification of individuals who will benefit from prevention, medical treatment, or aggressive intervention. Patients are usually referred for dedicated local imaging according to their clinical symptoms (e.g., intermittent claudication). Due the systemic character of atherosclerosis, however, severe preclinical vascular abnormalities may be present additionally in any other arterial territory, e.g., the carotid arteries. The novel hardware developments and acquisition techniques allow now to visualize

11.4  Pathologic Findings

Fig. 11.4.1  63-year-old male patient with aortoiliac Y prosthesis and left femoropopliteal bypass graft. Whole-body contrast-enhanced 3D-FLASH MRA acquired with the multiple receiver channel system shows a patent aortoiliac Y prosthesis and left femoropopliteal bypass graft but reveals advanced systemic atherosclerotic disease with occlusion of the left superficial femoral artery and left distal part of the anterior tibial artery, as well as a transmural myocardial infarction of the posterior wall and a

cerebral infarction in the posterior territory of the right middle cerebral artery (cardiac late-enhancement with long-axis IR turbo FLASH and cerebral FLAIR and contrast-enhanced T1weighted-SE in axial direction). Cerebral TOF-MRA depicts an aberrant blood supply of the posterior from the middle cerebral artery on the right side. The examination protocol is given in Table 11.2.1, third row. (Courtesy of Achim Seeger, Tuebingen)

the entire arterial system from intracranial to the distal runoff vessels including furthermore state-of-the-art heart and brain studies within a single examination (Goyen et al. 2005; Kramer et al. 2005; Fenchel et al. 2006; see Chap. 5, Sect. 5.4). These developments enable to comprehensively examine either patients with vascular diseases or in asymptomatic clients to screen for suspected vascular disease, e.g., stenoses, aneurysms or infarcts (Fig. 11.4.1). A recent study comparing the clinical impact of conventional DSA and WB-MRA in patients with peripheral arterial occlusive disease showed that WB-MRA revealed in 75% of the patients clinically relevant findings leading to a subsequent change of therapy (Fenchel et al. 2006). It

is important to note, however, that the medical benefits as well as the cost–benefit ratio of preventive therapeutic interventions are still controversial. 11.4.2  Oncology A malignant tumor is intrinsically a potentially systemic disease. Early tumor detection, precise staging, and accurate therapy monitoring in individual patients are essential for assessing prognosis and achieving the best possible patient outcome in terms of survival and quality of life. Imaging is of fundamental importance for

1237

1238

11  Whole-Body MRI

initial and follow-up staging, as serum tumor markers cannot provide information about the localization of tumor tissue and secondary complications related to possibly damaged surrounding anatomic structures. Staging comprises the detection and precise anatomical localization of tumor tissue including its local infiltration into surrounding anatomical structures (T stage) as well as probable metastases caused by lymphatic (N stage) and hematogenous spread (M stage). Up until a few years ago, only nuclear medicine examination methods enabled whole-body imaging, for example, conventional skeletal scintigraphy with technetium-marked phosphate complexes and, more recently, positron emission tomography with the radioactive nuclide 18F-2-fluoro-2-deoxy-d-glucose (18F-PDG PET). FDG PET has rapidly gained a significant role in the clinical evaluation of cancer patients, because tumor tissue can be identified anywhere in the whole body due to its specifically increased metabolic activity. However, PET lacks anatomical landmarks for topographic orientation, and areas with suspicious metabolic activity may not be definitively assigned to specific anatomical structures. Moreover, the differentiation of normal and abnormal glucose metabolic activity can be particularly difficult in regions close to organs with physiological high FDG uptake. Spiral CT has gone through enormous technical advances over the last 10 years and has, therefore, made whole-body imaging within one single examination feasible. Whole-body spiral CT has become an indispensable tool in oncologic diagnosis, enabling high-resolution whole-body imaging within only a couple of seconds. However, the detection of tumor tissue by conventional CT basically relies on morphological criteria, which may be misleading and therefore inappropriate for discriminating between benign and malignant. The limited specificity of CT remains a fundamental problem in oncology. It is important to note, for example, the inadequate accuracy of CT for detecting small lymph node metastases or for discriminating recurrent tumor from treatment related scar tissue. Rapid advances in scanner and computer technology enabled the recent development of PET–CT hybrid systems. This technology uses the full capacity of both imaging methods, allowing the acquisition of metabolic and morphologic data within a single examination, with information superimposes on color-coded images. It should be noted, however, that the interpretation of PET information requires a thorough understanding of the normal physiological distribution of FDG in the body so that it will not be confused with the presence of malignant tissue. Furthermore, CT-based attenuation correction of PET images may generate some artifacts, which in turn may result in incorrect interpretation of PET images. Nevertheless, clinical and research applications are at present evolving and appear very promising (von Schulthess et al. 2006).

MRI provides anatomical detail with soft tissue contrast superior to that of CT and high isotropic spatial resolution. Moreover MRI is more sensitive than PET CT is for early detection of tumors in different organs, such as the brain, liver, or bone marrow because of due to the physiological high FDG uptake in these tissues. WB-MRI can therefore play an important part in evaluating metastatic disease and estimating the total tumor burden. Moreover, mass lesions can be characterized by assessing their characteristic signal behavior on different MR sequences, for example, STIR, diffusion-weighted (DWI), and dynamic contrast-enhanced (DCE) MR sequences. These methods enable identification of tumorinvolved tissue, using functional tissue parameters and visualization of its local relationship to surrounding normal tissues with high soft tissue contrast. For example, in patients with rectal carcinoma, MRI enables a more detailed depiction of rectal wall and mesorectal fascia and therefore a more detailed determination of local tumor extension than does CT. Furthermore, the lack of X-ray exposure is favorable, permitting dynamic imaging studies as well as closed follow-up examinations, even in young individuals and those anticipated to be healthy. However, because of technical factors and long measurement times, MRI was in the past used primarily to examine body parts or single organs, and this continues to be the case. In patients with rectal carcinoma, for example, either pelvic MRI is performed for evaluating the primary tumor, or abdominal MRI for detecting liver metastases, or cerebral MRI for detecting brain metastases, etc. WB-MRI needs to be considered as a novel diagnostic approach. The application of coronal turbo STIR sequences using the whole-body resonator for signal excitation/reception yielded sufficient spatial and contrast resolution along with reasonable short examination times (Walker and Eustace 2001). The potential clinical impact of this approach for detecting bone marrow involvement in malignant diseases, e.g., skeletal metastasis, multiple myeloma, or lymphoma, was quickly recognized and proven as an examination technique, which is even more sensitive than conventional skeletal scintigraphy (Ghanem et al. 2005). A further important advantage of WB-MRI is the feasibility of detecting tumor manifestations in viscera. Using only the body resonator, however, limitations relative to dedicated partial-body MRI examinations are expected because of reduced spatial resolution and artifacts from organ movement. For example, WB-MRI has been proven an important diagnostic tool particularly in pediatric radiology for detecting skeletal metastases, but accuracy for the detection of metastases outside of the skeletal system is considered limited (Goo et al. 2005). The performance of WB-MRI was also inferior to that of WB-CT in the detection of small lymph nodes (<1.2 cm) and pulmonary metastases in patients with lymphoma (Brennan et al. 2004).

11.4  Pathologic Findings

Reduced spatial and/or contrast resolution should be avoided in order to achieve comparable diagnostic accuracy comparable to that of a series of dedicated state-ofthe-art MRI examinations, particularly with respect to the detection of small metastases, e.g., in the brain or in moving organs such as the liver. The development of the AngioSURF technique enabled improved image quality and a reduction in the examination time (Lauenstein and Semelka 2005). An initial study comparing WB-MRI with the AngioSURF technique to PET-CT showed advantages of MRI in M staging, particularly regarding bone and liver metastases, although limitations in T as well as N staging were apparent (Antoch et al. 2003). It should be noted, however, that a quite heterogenic patient population, with a variety of tumors and a relatively high percentage of bronchogenic carcinoma (BC), was included, and that the image quality was limited compared to that of dedicated state-of-the-art MRI studies (e.g., in the brain). The use of surface coils and iPAT allows the application of various spin-echo and gradient-echo sequences for achieving equivalent high spatial and contrast resolution of each body part as compared with conventional MRI examinations of a single body part (Schlemmer et al. 2005). Consequently, relatively small metastases can be detected not only in the skeletal system, but also in the brain, neck, and body trunk, particularly even in moving organs by applying breath-hold or navigator techniques (Muller-Horvat et al. 2006). Exceptions are anatomical regions where dedicated coils have necessarily to be used, e.g., breast, prostate, and small joints of hand and foot. Although the clinical potential of WB-MRI in oncology and its feasibility in daily routine are becoming more and more apparent, there are no definitive recommendations about which sequences should be included in an examination protocol for most accurately establishing the individual diagnosis. The diagnostic accuracy of conventional local MRI is proven for different tumor entities, e.g., renal cell carcinoma or ovarian carcinoma. The feasibility of expanding the FOV with WB-MRI opens up the possibility of assessing both local and suspected systemic tumor spread within the same examination. Different solid tumors are characterized by predictable metastatic patterns dominated by specific sites of initial spread. An individual WB-MRI examination protocol should be based on the understanding of the specific biological pathophysiological behavior of the tumor with respect to both the local growth and the potential lymphatic and hematogenous spread. For example, prostate cancer often develops osteoblastic bone metastases, whereas colorectal carcinoma tends to spread into the liver and bronchogenic carcinoma into the brain. The employment of a disease-specific examination protocol can be based on a modular concept, enabling the combining of different organ specific examination parts according to the specific tumor entity and individual risk factors (Schaefer et al. 2006) (Table 11.4.1). For this, the

WB-MRI protocol can be developed from oncologic guidelines and extended according to the individual clinical situation. It is important to note, however, that the medical impact and economic benefits of WB-MRI for early detection and staging of various tumor entities have not yet been statistically evaluated. 11.4.2.1  Metastases Bone metastases are important sequelae of various solid malignancies and are often multifocal. In 80% of cases, breast, prostate, bronchogenic, thyroid, or renal cell carcinoma is the primary tumor. Recent studies have proven that the accuracy of coronal WB-MRI for the detection of skeletal metastases may even be superior to that of conventional skeletal scintigraphy (Ghanem et al. 2005). STIR images in coronal plane are most often applied for the detection of bone metastases. However, osteoblastic lesion may be overlooked on STIR sequences because of the low signal and are better visualized on T1-weighted sequences. For screening of rib metastases, breath-hold T1-weighted and T2-weighted sequences should additionally be performed to reduce moving artifacts and partial volume effects due to breathing. Regarding hematological malignancies, MRI plays an essential role for detection, assessment of spread and evaluation of therapy response (e.g., in multiple myeloma) (Baur-Melnyk et al. 2005). The specificity of MRI for assessing lymph node metastases is limited due to the known ambiguity of the applied morphologic criteria, particularly with respect to small lymph nodes. But the differentiation between benign and malignant may even be unachievable if conventional clear-cut morphologic criteria are visualized, e.g., a fatty hilus indicating benign tissue or a round shape with a fluid-like center suggesting malignant tissue. A substantial increase in the specificity of MRI for evaluating pelvic lymph node involvement in prostate cancer was recently achieved using magnetic nanoparticles as intravenous contrast material; the reported sensitivity and specificity were 90 and 100%, respectively (Harisinghani et al. 2003). Accordingly, it can be expected that the specificity of lymph node assessment by WB-MRI will improve similarly. Concerning organ metastases, only a few clinical studies have been published to date, but it is apparent that WB-MRI will be an important imaging tool for evaluating distant tumor spread in solid tumors and hematological malignancies. Initial disadvantages (i.e., relatively limited spatial resolution and movement artifacts) of WB-MRI as compared with designated partial-body examinations have been overcome through advances in hardware and software technology. WB-MRI with the BodySURF technology and contrast-enhanced 3D GRE sequences was successfully applied to also assess extraosseous tumor spread (Lauenstein et al. 2004). To achieve the highest di-

1239

1240

11  Whole-Body MRI

agnostic accuracy, however, a global evaluation of tumor spread to any part of the body, including, for example, the brain, requires an image quality comparable to that of subsequently performed dedicated state-of-the-art MRI studies. Using the multichannel MR technology, a clinical study of patients with advanced melanoma has proven the feasibility and clinical potential of high-resolution WB-MRI relative to WB-CT (Muller-Horvat et al. 2006). WB-MRI altered the therapeutic concept in approximately 25% of the included patients, mainly because of its higher sensitivity in detecting cerebral, hepatic, and skeletal metastases, most notably in cases of diffuse bone marrow involvement. It is particularly important to note that brain, liver, and bone metastases are important sequelae of many types of cancer that impact therapeutic decisions, and that MRI of these organs is more sensitive than CT or FDG PET (Fig. 11.4.2). 11.4.2.2  Lung Cancer Lung cancer is the leading cause of tumor death in the Western world. Smaller lesions may represent earlier stage disease, but unfortunately advanced stage disease is often detected. To date, however, MRI has had no definite routine role in early detection or staging of BC, but clinical studies comparing MRI and CT are currently being performed. Metastases to the lung are frequent sequelae of different tumors. At present, feasibility studies are available showing that the sensitivity of MRI to detect pulmonary lesions strongly depends on the applied MRI technique. HASTE or 3D GRE sequences, which can easily be integrated in a WB-MRI protocol, yielded sensitivity of about 90% for 4- to 10-mm nodules and approximately 100% for nodules larger than 10 mm (Schroeder et al. 2006). 11.4.2.3  Prostate Cancer Prostate cancer is the most common non-cutaneous cancer in men. The most favorable outcome that can be achieved is associated with complete tumor resection. For optimizing therapeutic outcome and avoiding unnecessary complications, precise information needs to be obtained regarding local tumor extension as well as distant tumor spread particularly to the most common sites (i.e., bone marrow and lymph nodes). Endorectal MRI including MRSI has been proven to be most accurate for predicting pathologic outcome of extracapsular extension and seminal vesicle invasion (Kurhanewicz et al. 2002). However, unexpected high failure rates are observed even after intended curative treatment and must be attributed to unknown systemic progression. A WBMRI protocol for either initial staging or tumor detection in the case of prostate-specific antigen (PSA) increase af-

ter treatment should include an examination of the pelvis for assessing the local tumor and pelvic lymph nodes as well as the skeletal system. The pelvis should be examined with multiplanar T2-weighted MRI sequences of the pelvis acquired at least with a phased-array surface coil; SNR and spatial resolution are better with endorectal MRI. On principle, endorectal MRI can be performed within a WB-MRI examination using multichannel technology. To assess the osteolytic and osteoblastic skeletal metastases, coronal T1-weighted SE and T2-weighted STIR sequences should be applied, and additional fast breathhold sequences of the thorax are essential for assessing possible rib metastases (Fig. 11.4.3). Metastases in other organs, for example in the liver or brain, are very unlikely, and screening should be subject to individual high-risk or symptomatic conditions. For initial and repeated staging, the combination of WB-MRI and endorectal MRI/ MR spectroscopy may consequently be advantageous for differentiating patients with localized disease from those with systemic progression. High-resolution MRI after intravenous application of magnetic nanoparticles will probably play an important role in detecting even small lymph node metastases with high sensitivity and specificity (Harisinghani et al. 2003). 11.4.2.4  Breast Cancer Breast cancer is the leading cause of cancer death in women. Standard T2-weighted and contrast-enhanced T1-weighted MRI are inadequate for assessing breast cancer. DCE MR mammography is infeasible within a WBMRI examination and furthermore reserved to selective clinical questions (Kneeshaw et al. 2003). Because of the prevalence of benign conditions such as fibroadenoma, a considerable number of ambiguous breast lesions will be detected during a WB-MRI examination. It is important to note that patients must be properly informed about the limited sensitivity and specificity of the method and that additional conventional breast examination, e.g., mammography and ultrasound, will be necessary. To screen for metastases, however, a WB-MRI protocol may be reasonable. 11.4.2.5  Colorectal Cancer Colorectal cancer is the third most common cancer in the world and second most common cause of tumor-related mortality. The aim of local staging is to classify the patients into different risk groups for local recurrence regarding preoperative and/or (neo-)adjuvant treatment planning. In a few studies, MR colonography (MRC) has achieved detection rates similar to those of CT colono­ graphy (CTC) for clinically relevant lesions ≥10 mm. For smaller lesions, however, the sensitivity appears lower

11.4  Pathologic Findings

Fig. 11.4.2  36-year-old male patient with cutaneous malignant melanoma American Joint Committee on Cancer stage IV. Whole-body MRI shows metastases to the brain, lung, liver, and bones. The examination is the fourth follow-up study monitoring response to treatment using the standardized whole-body

MR imaging protocol. The examination shows progression of the disease. Sequence protocol: “Metastases: bone marrow and viscera” (Table 11.4.1). Displayed MRI sequences: WB1, B1,3, T1,4, A1,3, P1,3 (Table 11.2.1)

1241

1242

11  Whole-Body MRI

Fig. 11.4.3  48-year-old male patient with prostate cancer. Whole-body MRI of the skeletal system demonstrates multiple bone metastases in the spine, ribs, pelvis, and left femur. Extramedullary tumor extension with spinal canal invasion is found in the fifth and tenth thoracic vertebrae. Note that even small

osteoblastic rib metastases can be detected on breath hold T1weighted FLASH 2D sequences. Sequence protocol: “Prostate cancer” (Table 11.4.1). Displayed MRI sequences: S1,2, WB1,2, R1,2 (Table 11.2.1)

(Ajaj et al. 2003). Pelvic MRI has been proven as a highly accurate tool for local staging of rectal cancer. In principle, MRC can be integrated in the WB-MRI examination protocol for revealing further colonic polyps. However, high spatial resolution, high CNR, and no motion artifacts in all parts of the colon have to be achieved for detecting even small polyps, and therefore the use of iPAT and large FOV is essential (Steidle et al. 2004). Coronal DCE VIBE sequences combined with dark-lumen MRC techniques may be combined to comprehensively examine the dynamic contrast uptake in the liver and colon polyps. At the time of diagnosis, however, 15–25% of patients have already developed liver metastases, and 35–45% will develop it during the course of the disease. Before resection with curative intention, metastases to distant lymph node (N2), lung, brain, and bone as well as peritoneal carcinomatosis have to be excluded. Initial staging should include breath-hold T1-weighted and breath-hold or respiratory-gated T2-weighted MRI of the abdomen and T1-weighted and T2-weighted MRI of the pelvis.

WB-MRI may play an important role in assessing initial diagnosis and therapy planning as well as in treatment monitoring to optimize efficacy by minimizing toxicity. 11.4.2.6  Multiple Myeloma The examination protocol in patients with multiple myeloma should contain T1-weighted SE or TSE as well as STIR sequences, whereby the intravenous administration of contrast medium can be advantageous in case of suspected diffuse bone marrow infiltration (Baur-Melnyk et al. 2005). WB-MRI in coronal orientation should be completed by sagittal imaging of the spine. Additional axial imaging may be necessary in individual patients depending on the findings, e.g., in case of pathological fractures or extramedullary tumor spread (Figs. 11.4.4, 11.4.5). WB-MRI can also be used to monitor response to therapy through the assessment of changes in the size, shape, signal intensity, and contrast uptake pattern of the lesions.

11.4  Pathologic Findings

Fig. 11.4.4  66-year-old male patient with multiple myeloma Salmon/Durie stage III. Whole-body MRI demonstrates multifocal bone marrow involvement in the spine, ribs, pelvis, and left femur. Extramedullary tumor extension as well as organ spread is detected in one left sided rib and the liver, respectively.

Sequence protocol: “Multiple myeloma” (Table 11.4.1) including A1, T1 according to the extramedullary tumor extension and organ involvement. Displayed MRI sequences: WB1,2, S1,2, A1, T1 (Table 11.2.1)

11.4.2.7  Less Common Tumors

mainly epithelial cancers like squamous cell carcinoma, adenocarcinomas, or neuroendocrine carcinomas without an established primary site and imaging is actually performed by using FDG PET in order to detect tumors with high metabolism anywhere in the body.

In principle, the local MR examination of any tumor can be included in a WB-MRI examination to screen for multifocal tumor growth or distant tumor spread (Figs. 11.4.6, 11.4.7). For renal cell carcinoma, for example, the detection rate and diagnostic accuracy of state-of-the-art MRI are equivalent to those CT, and local MRI of the kidneys can easily be included in a WB-MRI protocol for comprehensive staging, particularly in high-risk patients. At the present time, however, the potential of WB-MRI for detection and staging of different tumor entities has yet to be fully explored, and consequently, general algorithms cannot yet be determined. The medical and economic benefits of WB-MRI have to be evaluated in clinical studies focusing on various tumors types, particularly compared to PET/CT. It is particularly important to note that WB-MRI is so far not an appropriate imaging modality to evaluate patients with carcinoma of unknown primary site, which is one of the 10 most frequent cancers worldwide. Patients present with metastases from a hetero­genous group of

11.4.3  Musculoskeletal The intrinsic high soft-tissue contrast of WB-MRI is an essential advantage that enables to reveal individual patterns of multifocal or disseminated diseases. Major indications include the detection and therapy monitoring of tumor spreading into bones and/or soft tissue in case of solid tumors as well as hematological malignancies. Of particular importance is the evaluation of diseases, where conventional skeletal scintigraphy is insufficient and soft tissue involvement has to be assessed, e.g., in multiple myeloma or lymphoma. Further applications are the evaluation and follow-up of suspected multifocal diseases involving bones and soft tissue, e.g., in case

1243

1244

11  Whole-Body MRI

Fig. 11.4.5  61-year-old male patient with multiple myeloma Salmon/Durie stage III. Whole-body MRI demonstrates tumor involvement of the sixth thoracic with extramedullary tumor extension into the spinal cord with consecutive myelon compression. Additionally, metastasis to the spleen is clearly visu-

alized. Sequence protocol: “Multiple myeloma” (Table 11.4.1) including R3, A3 according to tumor invasion of the spinal canal and a spleen metastasis. Displayed MRI sequences: WB2, S1, R3, A3 (Table 11.2.1)

11.4  Pathologic Findings

Fig. 11.4.6  13-year-old male patient with chondroblastic osteosarcoma of the right tibia. Whole-body MRI is performed for staging and monitoring of chemotherapy. The primary tumor with soft tissue extension is located in the right proximal tibia. Multiple metastases are visualized in the right proximal distal femur, the spine (fifth lumbar vertebra, third sacral vertebra)

and the left humeral head. The sequence protocol was defined to screen for bone metastases/skip lesions: WB1,3, S1,3, R3, P3, followed by axial contrast-enhanced FLASH 2D fat saturation (FS) of the knee. Displayed MRI sequences: WB1, S1, R3, P3 (Table 11.2.1) and CE FLASH 2D FS of the knee

1245

1246

11  Whole-Body MRI

of multiple cartilaginous exostosis (Ollier’s disease) (Fig. 11.4.8), Paget’s disease, fibromatosis, or eosinophilic granuloma. Promising fields of application include the evaluation of systemic spread in case of inflammatory diseases involving muscle, bones and joints, e.g., polymyositis, osteomyelitis (either inflammatory or chronic recurrent multifocal osteomyelitis [CRMO]) (Fig. 11.4.9), spondylitis/spondylodiskitis, abscesses, rheumatoid arthritis, fasciitis (Fig. 11.4.10), or seronegative spondyloarthropathies. WB-MRI has been proven to be helpful in case of suspected polymyositis for detecting the optimal site for muscle biopsy as well as for therapy monitoring (Fig 11.4.11) (O’Connell et al. 2002). Whole-spine MRI has been proven as a sensitive method for detecting occult spine fractures in trauma patients (Green and Saifuddin 2004). One important application is accordingly the assessment of fractures in case of a suspected batteredchild syndrome, where MRI is more sensitive than skeletal scintigraphy, allows X-ray exposure to the child to be avoided and offers multiplanar imaging of detected fractures within the same examination. Whole-spine MRI in the sagittal plane is particular helpful for the assessment as well as therapy planning and follow-up of a variety of diseases involving diverse segments of the spine, e.g., degenerative disc disease, osteochondrosis, osteoporosis (Fig. 11.4.12), inflammation, or malignant disorders. Based on the whole-spine information, specific axial imaging can be planned during the course of the examination by focusing on involved segments. Employing 3D T2-weighted-TSE MR sequences are particularly helpful for the evaluation of scoliosis (Lichy et al. 2005). 11.4.4  Adipose-Tissue Quantification

11.4.7  32-year-old male patient with cutaneous T-cell lymphoma. Multifocal spreading of the disease in the subcutaneous fat tissue is obviously visualized by axial contrast-enhanced FLASH 2D MR imaging with frequency-selective FS obtained from thorax, abdomen, pelvis, and legs. The comprehensive whole-body information is used for monitoring therapy response

The total amount of fat and its regional distribution play an important role in the pathogenesis of diabetes and cardiovascular diseases. The quantitative analysis of adipose tissue and particularly the discrimination between visceral and subcutaneous fat is considered as important for assessing the individual risk to develop insulin resistance, hyperinsulinemia, dyslipidemia, and hypertension. Accordingly, standardized determination of the individual fat distribution will be important for monitoring lifestyle interventions. WB-MRI offers a unique strategy to quantify and follow-up the fat distribution in the entire body, with high accuracy and without radiation exposure. Different sequence protocols are proposed including T1weighted SE or GRE as well as chemical shift–selective (CHESS) sequences. Axial imaging with 10-mm slice thickness and a 10-mm gap is adequate and achievable within a total examination time of about 15 min (Machan et al. 2005). For standardized recording of individual adipose-tissue profiles, semi-automated approaches for segmentation of the subcutaneous and visceral fat compartments are essential.

11.4  Pathologic Findings

Fig. 11.4.8  34-year-old male patient with Ollier’s disease (multiple osteochondroma) with a high risk of developing chondrosarcoma. The left leg has already been exarticulated because of chondrosarcoma of the left femur. Repeated whole-body MRI examinations enable accurate follow-up of size, shape, signal in-

tensity, and contrast-uptake of different osteochondromas localized in the body trunk and proximal extremities, which bear the highest risk for malignant transformation. Sequence protocol: WB1,2, R1,3, A1,3, P1,3. Displayed MRI sequences: W1, R1, P1,3 (Table 11.2.1)

11.4.5  Early Detection and Screening

group. Three forms of prevention are distinguished: primary prevention (prophylaxis), secondary prevention (early detection), and tertiary prevention (hindering of disease progress). Secondary prevention intends the early detection of unknown disease in an asymptomatic individual, e.g., of a cerebral aneurysm or a renal cell carcinoma. Tertiary prevention means that individuals with an already known disease are examined in order to detect early any progression of the disease and associated complications, e.g., screening for a carotid stenosis in patients with known hypertension and atherosclerosis or screening for metastases in patients with a history of cancer. It is important to note, however, that by using whole-body imaging for screening the boundaries between secondary and tertiary prevention become obscured because a vari-

In the Western world, increases in health awareness and life expectancy, as well as recognition of the need to take responsibility for one’s health, have led to greater interest in early detection and screening. The advances in wholebody imaging technology now offer novel approaches for early disease detection. WB-CT was first offered commercially for secondary and tertiary prevention. However, the use of ionizing radiation may potentially be damaging to the health of patients otherwise assumed to be healthy. The progress in WB-MRI technology may offer a solution to this problem. The goal of prevention is to reduce disease-related mortality and thus extend life spans in the examined

1247

1248

11  Whole-Body MRI

Fig 11.4.9  12-year-old male patient with chronic recurrent multifocal osteomyelitis (CRMO). The whole-body examination was performed to assess the spread of the disease and its response to therapy. The STIR-sequences clearly show involvement of the

first 3 ribs on the left side, both distal femora, the left proximal and distal tibia, and the left talus. Involvement of the sacrum is shown on another coronal STIR slice on the left side. Sequence protocol and displayed MRI sequences: WB1, T1 (Table 11.2.1)

Fig. 11.4.10  65-year-old male patient with eosinophilic fasciitis. Clear visualization of thickened and contrast-enhancing fascia of different muscle groups allows differential diagnosis of necrotizing fasciitis, definition of a biopsy site and monitoring of therapy response. The sequence protocol was defined to depict the involved body parts for diagnosis as well as therapy follow-up: WB1; T1,3 (Table 11.2.1), and axial contrast-enhanced FLASH 2D MR imaging with frequency-selective fat saturation of the upper legs

11.4  Pathologic Findings

Fig. 11.4.11  70-year-old female patient with polymyositis. Whole-body MRI reveals involved muscles with edematous tissue changes, indicating inflammation. The examination particularly indicates the most favorable biopsy site for establishing the final diagnosis. Standardization of the examination protocol furthermore enables accurate therapy monitoring. The sequence protocol includes WB1, followed by axial STIR imaging of involved body parts, in this case P1, and axial STIR sequences of the lower leg (Table 11.2.1)

Fig. 11.4.12  79-year-old patient with osteoporosis showing vertebral fractures in thoracic vertebrae 7, 8, 11, and 12. Wholespine imaging allows visualizing and monitoring of pathologic vertebral fractures for an accurate assessment of the extent of the disease as well as its temporal changes during therapy. Sequence protocol and displayed MRI sequences: S1,2 (Table 11.2.1)

ety of diseases can be detected. Furthermore, those individuals who decide to undergo WB-MRI for prevention frequently present with a history of a disease or at least risk factors from family history or lifestyle. The central issue of the debate regarding the pros and cons of early detection is whether the health advantages outweigh the risks and costs. The cost–benefit relationship of WB-MRI screening depends on individual therapeutic consequences. The examination of selected risk groups with increased pretest probabilities of having a particular disease is most efficient. It is important to consider, however, that two completely different perspectives are left to oppose each other: On the one hand the perspective of health insurers and politicians being responsible for a restricted budget to serve the whole insured population and therefore paying only for examinations, which

reduce total costs in the long term, on the other hand the perspective of the individual client wishing to aim for optimal health and therefore most likely being insensible to statistical cost-effective considerations regarding an entire population. It is therefore important to differentiate between a more general screening program for assessing a specific disease in a larger population, e.g., the breast cancer screening program and an individual request for a WB-MRI examination to rule out probable diseases. A cost-efficient screening program is characterized by four requirements: (1) high prevalence of the disease, (2) chance for early detection, (3) relevant available examination tools, and (4) effective therapy (Gray 1997). WB-MRI can play an important role in this context for the secondary and tertiary prevention of particular cardiovascular and oncologic diseases.

1249

1250

11  Whole-Body MRI

11.4.5.1  Prevalence

11.4.5.3.1  Cardiovascular Screening

Cardiovascular and oncologic diseases have the highest prevalence in the industrialized countries. Although epidemiological data have shown that primary prevention of cardiovascular diseases is feasible by improving the lifestyle through weight reduction, sport activities, and healthy nutrition, secondary prevention of atherosclerosis remains essential to prevent heart attack or stroke. Carcinogenesis is complex, multifactorial, and inconsistent for different tumor types. Although some other risk factors are apparent, e.g., tobacco, alcohol, or high-fat nutrition, the most important risk factor is age, and early detection of cancer remains most important for achieving therapeutic success.

In a single examination, WB-MRA allows visualization of atherosclerotic-related abnormalities of all large arteries of the whole body except the coronaries, which are still better visualized by CTA. A cardiovascular examination protocol should include TOF MRA of intracerebral arteries, whole-body contrast-enhanced MRA, cardiac MRI including functional studies with cine trueFISP, and myocardial viability assessment with late-enhancement sequences as well as cranial MRI with FLAIR-, DW- and pre-/post-contrast T1-weighted SE sequences (Table 5.4.16, Chap. 5.4.7). A considerable number of subjects have been examined by Goede et al. (2005). By selecting patients with a high pre-test probability of vascular abnormalities, i.e., cardiovascular risk factors such as arterial hypertension, hyperlipidemia, or hyperglycemia, the efficiency of the examination can be increased. Cooperation with general practitioners or cardiologists should therefore be integrated into the selection of patients who undergo the prevention examination. But it must also be considered that efficiency considerations have no individual meaning, as the absence of risk factors does not automatically indicate the absence of the disease in the examined subject. For example, a cardiovascular examination of an asymptomatic 58-year-old client without any cardiovascular risk factors incidentally revealed a sub-acute cerebellar infarction. WB-MRA showed no abnormality, but cardiac MRI revealed a patent foramen ovale (Fig 11.4.13). The patient’s history revealed recurrent thrombophlebitis, most probably related to frequent flying. The final diagnosis of a paradoxical embolus had important therapeutical consequences, as there is a significantly increased risk for stroke. It is important to note that self-referring and asymptomatic patients take on the responsibility for their own health care themselves and might therefore be indifferent to efficiency considerations from an economical point of view.

11.4.5.2  Chance for Early Detection Vascular abnormalities causing major complications, like myocardial or cerebral infarction, can in principle be visualized by imaging. Since alterations of the vascular wall rather than the degree of narrowing are now considered critical risk factors regarding ischemic events, the role of cross-sectional has increased (Leiner et al. 2005). Early detection of cancer in the presymptomatic stage is considerably important. Achievement of this goal is in principle feasible for the most important solid tumors (i.e. bronchogenic, breast, prostate, colon, or renal cell carcinoma), the task is complex because of the large number of tumor types and their varying biological growth characteristics. Screening intervals for fast-growing, aggressive tumors must be closer compared to those for slower-growing, less aggressive tumors. Excessively long intervals between examinations increase the risk that fast tumors will become symptomatic without being detected by the screening program, whereas intervals that are too short are ineffective or even potentially harmful to patients. It has not yet been determined at what age and how often a general tumor screening examination by MRI should be performed. 11.4.5.3  Valuable Examination Tools The sensitivity and specificity of the examination method must be carefully balanced in defining an optimal intervention threshold. This addresses the question, how many false-positive or false-negative results have to be accepted and how many, in retrospect, unnecessary therapeutic interventions have to be taken into account. Exemplary examination protocols for cardiovascular and tumor screening are given in Tables 5.4.16 and 11.2.1, respectively. Both examination protocols require approximately 60–90 min, depending on the performance of MRC. The time subsequently required for the image reading, documentation and discussion with the patient vary substantially, but may require up to 30–60 min.

11.4.5.3.2  Tumor Screening Early tumor detection may have significant consequences for the examined client. A WB-MRI tumor protocol must be performed in state-of-the-art quality, including all different body parts, guaranteeing the highest sensitivity for detecting small mass lesions (Tables 11.2.1, 11.4.1). Among the most common malignant tumors are bronchogenic, breast, prostate and colon carcinoma, with mortality rates of around 10–25%. Less common malignant tumors include cancers of the ovaries, stomach, pancreas, kidneys and bladder, as well as leukemia and lymphoma, which have mortality rates of around 3–6%. WB-MRI is inadequate for assessing breast and prostate cancer, and the diagnostic accuracy for early detection of bronchogenic and colorectal carcinoma has not been definitively assessed. On the other hand, MRI offers high

11.4  Pathologic Findings

Fig. 11.4.13  58-year-old asymptomatic individual undergoing a whole-body cardiovascular MRI screening examination. Whole-body MRA including cerebral TOF-MRA is normal, but MRI of the brain showed an unexpected subacute right cerebellar infarction and cardiac cine trueFISP sequences revealed a patent foramen ovale. Anamnesis revealed recurrent throm-

bophlebitis related to frequent air travelling probably causing a paradoxical thromboembolic event. Preventive occlusion of the patent for­amen ovale has to be considered to eliminate the risk of stroke. The examination protocol is given in Table 5.4.16, third row (Chap. 5.4.7)

1251

1252

11  Whole-Body MRI

Metastases: bone marrow

WB1,2, S1,2, R1,2

Metastases: viscera

B1–4, N1–3, T1–3, A1–3/4,a P1–3

routine CT check-ups, not all the round lesions detected will be malignant, and radiological lung cancer screening has therefore not been recommended. MRI has at present no definite role for early detection of BC. Published studies concerning the diagnosis of round lesions with the aid of MRI have already shown that the sensitivity for a lesion size of 5 mm is 90% compared with CT, and the usefulness of intravenous contrast medium administration is a subject of research (Schaefer et al. 2003).

Metastases: bone and viscera

WB1,2, B1–4, N1–3, T1–3, A1–3/4 a, P1–3

11.4.5.3.4  Colorectal Carcinoma

Bronchial carcinomab

WB1,2, B1–4, T1–3, A1–3

Prostate carcinomac

WB1,2, S1,2, R1,2, P4

Breast carcinomad

WB1,2, B1–4, R2, T1–3, A1–4

Colorectal carcinoma

WB1,2, B1–4, R2, T1–3, A1–3, (A5), P1–3

Multiple myeloma

WB1,2, S1,2, (S3) e, R1,2

Tumor screening

WB1, B1–4, N1–3, T1–3, A1–3, (A5) f, P1–3

Table 11.4.1  Possible combination of examination modules for tumor-specific, whole-body MRI examination protocols. Examination protocols should be considered as a basic guideline for staging, and have therefore eventually to be modified according to the individual clinical situation

 If hypervascularized metastases are suspected (e.g., breast or renal cell carcinoma) A4 should be performed instead of A3 b  CT still remains standard for local staging c  Pelvic imaging (P4) primarily for assessing regional lymph nodes; local staging requires endorectal MRI; breath-hold sequences T1–3 are useful for assessing small rib metastases d  Eventually separate DCE MRI study with use of the dedicated breast coil necessary because local staging of breast carcinoma is not feasible within the whole-body MRI examination e  If diffuse bone marrow involvement is suspected but not visualized on T1-weighted and T2-weighted MR images f  If requested by the client after detailed information about the sensitivity and specificity of MRC compared with CTC and conventional optical colonoscopy a

accuracy for the detection of a variety of tumors, e.g., renal cell carcinoma, ovarian carcinoma, or brain tumors. An example from the author’s own experience is the detection of a large, but asymptomatic cystic meningioma, preventing complications and enabling early surgery with good prognosis. Dark-lumen MRC may be included after repositioning of the patient in the prone position using contrast-enhanced VIBE-sequences after the administration of approximately 1.5–2 l of warm tap water in hypotonia with intravenous contrast medium application. 11.4.5.3.3  Bronchogenic Carcinoma Although CT enables detection of small, early-stage tumors (Humphrey et al. 2004), it is has not been proven that the mortality rate can be reduced by CT-screening. In

Secondary prevention of colorectal carcinoma is reasonable, because cancer usually develops from adenoma within a relatively long latency period of about 10 years, and endoscopic polypectomy is relatively easy to carry out (Smith et al. 2004). Data from a clinical study involving 1,233 asymptomatic participants comparing virtual CT colonoscopy (CTC) with conventional optical colonoscopy have proven high sensitivity of 90–95% for CTC, depending on polyp size (Pickhardt et al. 2003). Regarding MRC only initial results are available indicating that polyps with a diameter of 10 mm can be detected at the same rate (Ajaj 2003). The examined patient should, however, be informed that the literature does not provide sufficient data to support the effectiveness of MR colonoscopy for early detection of colonic polyps. 11.4.5.3.5  Breast Carcinoma DCE MRI requires the use of a dedicated breast coil and therefore it is not feasible to include it in a WB-MRI examination. Furthermore, because of its low positive predictive value, it is not recommended for the primary diagnosis of breast cancer except in high-risk patients with a dense glandular parenchyma (Kneeshaw et al. 2003). It must be expected that especially in pre-menopausal women a variety of breast lesions may be visualized with WB-MRI that cannot be characterized definitively. Written informed consent must be obtained from the patient before the examination, acknowledging that the sensitivity and specificity of the method are insufficient for early detection of breast cancer. It is also important to emphasize that WB-MRI does not obviate the need for further routine X-ray and ultrasound examinations of the breast. 11.4.5.3.6  Prostate Carcinoma Although endorectal MRI is the most accurate method for local staging of biopsy proven prostate carcinoma, the specificity for characterizing signal abnormalities is limited. Benign disorders like prostatitis or benign prostatic hyperplasia may not confidently be differentiated

11.4  Pathologic Findings

from cancer, even in cases of increased PSA serum values. A WB-MRI examination may contain T2-weighted sequences of the pelvis, on which the prostate needs to be assessed. Written consent should therefore be obtained from the patient acknowledging that the method is inadequate for early detection of prostate cancer. It must be accentuated that a urologic examination is indispensable and that, if the results are suggestive of tumor, a dedicated endorectal MRI examination may be indicated afterwards for tumor localization and/or staging. Effective Therapy Therapeutic concepts of cardiovascular and oncologic diseases are considerably complex, interdisciplinary, and constantly changing. Despite considerations about therapeutic efficacies, it can be stated that, in general, early disease detection still represents one of the most important prognostic factors. However, slow rather than fast developing diseases tend to be detected by screening, because the latter are more likely to become symptomatic during the interval between repeated examinations. Accordingly, the prognosis of diseases detected by a screening examination may be improved only artificially (“length bias sampling”). 11.4.5.3.7  Patient Information Probable risks of the screening examination must be explained explicitly. It should also be stated that quality of life could eventually be lessened rather than improved by the examination. In addition to the obvious risk of morbidity from the examination itself (e.g., as a result of intravenous administration of contrast medium), there are the less obvious but graver risks that false-positive or false-negative results may lead to unnecessary further diagnostic testing or even treatment with potential morbidity or mortality. The patients need to be fully aware of the limitations of the examination, which are quite often underestimated due to the impressive imaging technology. The probability of false-positive or false-negative findings and the consequences of such findings should be explained in advance. Furthermore, it must be noted that clear specifications for intervals between screening repeated examinations are virtually impossible to define with respect to all kinds of diseases. All these factors lead to ethical problems and have to be considered in the case of a legal battle. The recording of detailed and comprehensive information about the patient and written informed consent are necessary before the beginning of the prevention examination: 1 A true-negative result is obviously desired by both the examined individual as well the Radiologist. However, it should be noted that even in this case, the development of a disease cannot be ruled out for the future. The momentary “feeling of safety” may lead to the situation that other routine examinations by general

practitioners, gynecologists, urologists, etc., may no longer be taken seriously. Furthermore, the motivation to make a positive turn in the lifestyle, e.g., by ceasing a harmful habit like smoking, may be reduced. Consequently, risk factors and the need to address them should still be discussed. The intervals at which WB-MRI examinations should be repeated cannot be determined at this stage. 2 A true-positive result is feared, but the possibility of it is in a strict sense the basic motivation for the examination, since it is hoped that the early diagnosis of a disease will allow a curative treatment and therefore improved life expectancy and/or quality of life. The detection of atherosclerotic vascular abnormalities may lead to positive lifestyle changes. Small tumors detected early can in principle be treated with a better success rate, although this is of course not true for all tumors, e.g., glioblastoma multiforme. An early diagnosis may only prolong the time period of knowing of the presence of the disease without extending the actual life expectancy (“lead-time bias”). On the other hand, in the case of less aggressive tumors in patients of advanced age, it cannot be known whether the disease would ever have become symptomatic within the life expectancy (“pseudo disease”). 3 A false-negative result can be either unavoidable (because of limitations of sensitivity and/or specificity) or avoidable (if due to an inadequate examination protocol or an overlooked result). Such a result may give the patient a false sense of safety, leading him or her to ignore symptoms and thus delaying the definitive diagnosis. 4 A false-positive result may lead to unnecessary pain and psychological suffering for the examined person and his/her relatives. In contrast to the screening examination, the extra expenses of (in retrospect, unnecessary) follow-up examinations and therapies could lead to loss of working hours and even eventual morbidity/mortality, all of which place a burden on the general public. The radiologist must consider that obtaining written informed consent, interpreting, and presenting the images and discussing the results with the patient may require a substantial amount of time. Self-referring patients in general expect that lifestyle changes or therapies that may be needed will be explicitly discussed. Those considerations may demand more time and specialized medical knowledge than the radiologist has to offer. In the case of pathologic findings, recommendations for supplementary diagnostic and/or therapeutic procedures should be available immediately. Moreover, the patient may be confronted with a wave of thoughts, expectations and anxieties regarding further examinations, the potential need for surgery, and so on; therefore, from an ethical perspective, it is important that the examination is embedded in a thoughtful and organized medical context.

1253

1254

11  Whole-Body MRI

Experience from breast carcinoma mammography screening has proven that the highest diagnostic accuracy can be achieved by specially trained and experienced radiologists. It has further been shown that the reliability of the diagnosis can be increased with double-reading by two independent radiologists. Medical and economical advantages of mammography screening have become apparent as specialized centers are now quality controlled. It may be reasonable to consider similar concepts for WB screening with MRI. References 1.

Ajaj W, Pelster G, Treichel U, Vogt F, Debatin J, Ruehm S, Lauenstein T (2003) Dark lumen magnetic resonance colonography: comparison with conventional colonoscopy for the detection of colorectal pathology. Gut 52:1738–1743 2. Antoch G, Vogt FM, Freudenberg LS, Nazaradeh F, Goehde SC, Barkhausen J, Dahmen G, Bockisch A, Debatin JF, Ruehm SG (2003) Whole-body dual-modality PET/ CT and whole-body MRI for tumor staging in oncology. JAMA 290:3199–3206 3. Bader TR, Semelka RC, Pedro MS, Armao DM, Brown MA, Molina PL (2002) Magnetic resonance imaging of pulmonary parenchymal disease using a modified breathhold 3D gradient-echo technique: initial observations. J Magn Reson Imaging 15:31–318 4. Baur-Melnyk A, Buhmann S, Durr HR, Reiser M (2005) Role of MRI for the diagnosis and prognosis of multiple myeloma. Eur J Radiol 55:56–63 5. Bongartz GM, Boos M, Winter K, Ott H, Scheffler K, Steinbrich W (1997) Clinical utility of contrast-enhanced MR angiography. Eur Radiol 7 Suppl 5:178–186 6. Brennan DD, Gleeson T, Coate LE, Cronin C, Carney D, Eustace SJ (2004) Comparison of whole-body MRI and CT for the staging of lymphoma. 185:711–716 7. Daldrup-Link HE, Franzius C, Link TM, Laukamp D, Sciuk J, Jurgens H, Schober O, Rummeny EJ (2001) Whole body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. Am J Roentgenol 177:229–236 8. Epping-Jordan JE, Galea G, Tukuitonga C, Beaglehole R (2005) Preventing chronic diseases: taking stepwise action. Lancet 366:1667–1671 9. Fautz HP, Kannengiesser SAR (2006) Sliding multi-slice (SMS): a new technique for minimum FOV usage in axial continuously moving table acquisitions. Magn Reson Med 55:363–370 10. Fenchel M, Scheule AM, Stauder NI, Kramer U, Tomaschko K, Nagele T, Bretschneider C, Schlemmer HP, Claussen CD, Miller S (2006) Atherosclerotic disease: whole-body cardiovascular imaging with MR system with 32 receiver channels and total-body surface coil technology—initial clinical results. Radiology 238:280–291

11. Ghanem N, Uhl M, Brink I, Schafer O, Kelly T, Moser E, Langer M (2005) Diagnostic value of MRI in comparison to scintigraphy, PET, MS-CT and PET/CT for the detection of metastases of bone. Eur J Radiol 55:41–55 12. Goehde SC, Hunold P, Vogt FM, Ajaj W, Goyen M, Herborn CU, Forsting M, Debatin JF, Ruehm SG (2005) Fullbody cardiovascular and tumor MRI for early detection of disease: feasibility and initial experience in 298 subjects. AJR Am J Roentgenol 184:598–611 13. Goo HW, Choi SH, Ghim T, Moon HN, Seo J J (2005) Whole-body MRI of paediatric malignant tumours: comparison with conventional oncological imaging methods. Pediatr Radiol 5:766–773 14. Goyen M et al (2003) Detection of atherosclerosis: systemic imaging for systemic disease with whole-body threedimensional MR angiography—initial experiences. Radiology 227:277–282 15. Gray JAM (1997) Evidence based health care: how to make health policy and management decisions. Churchill Livingstone, New York 16. Green RA, Saifuddin A (2004) Whole spine MRI in the assessment of acute vertebral body trauma. Skeletal Radiol 33:129–135 17. Harisinghani MG, Barentsz J, Hahn PF, Deserno WM, Tabatabaei S, van de Kaa CH, de la Rosette J, Weissleder R (2003) Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 348:2491–2499 18. Humphrey L, Teutsch S, Johnson M (2004) Lung cancer screening with sputum cytologic examination, chest radiography, and computed tomography: An update for the U.S. Preventive services task force. Ann Intern Med 140:740–753 19. Kneeshaw PJ, Turnbull LW, Drew PJ (2003) Current applications and future direction of MR mammography. Br J Cancer 88:4–10 20. Kramer H, Schönberg SO, Nikolaou K, Huber A, Struwe A, Winnik E, Wintersperger BJ, Dietrich O, Kiefer B, Reiser MF (2005) Cardiovascular screening with parallel imaging techniques and a whole-body MR imager. Radiology 236:300–310 21. Kurhanewicz J, Swanson MG, Nelson SJ, Vigneron DB (2002) Combined magnetic resonance imaging and spectroscopic imaging approach to molecular imaging of prostate cancer. J Magn Reson Imaging 16:451–463 22. Lauenstein TC, Semelka RC (2005) Whole-body magnetic resonance imaging. Top Magn Reson Imaging 16:15–20 23. Lauenstein TC, Goehde SC, Herborn CU, Goyen M, Oberhoff C, Debatin JF, Ruehm SG, Barkhausen J (2004) Whole-body MR imaging: evaluation of patients for metastases. Radiology 233:139–148 24. Lauenstein TC, Goehde SC, Herborn CU, Treder W, Ruehm SG, Debatin JF, Barkhausen J (2002) Three-dimensional volumetric interpolated breath-hold MR imaging for whole-body tumor staging in less than 15 min: a feasibility study. AJR Am J Roentgenol 179:445–449

11.4  Pathologic Findings 25. Leiner T, Gerretsen S, Botnar R, Lutgens E, Cappendijk V, Kooi E, van Engelshoven J (2005) Magnetic resonance imaging of atherosclerosis. Eur Radiol 15:1087–1099 26. Lichy MP, Wietek BM, Mugler JP III, Horger W, Menzel MI, Anastasiadis A, Siegmann K, Niemeyer T, Konigsrainer A, Kiefer B, Schick F, Claussen CD, Schlemmer HP (2005) Magnetic resonance imaging of the body trunk using a single-slab, 3-dimensional, T2-WEIGHTED-weighted turbospin-echo sequence with high sampling efficiency (SPACE) for high spatial resolution imaging: initial clinical experiences. Invest Radiol 40:754–760 27. Machan J, Thamer C, Schnoedt B, Haap M, Haring HU, Claussen CD, Stumvoll M, Fritsche A, Schick F (2005) Standardized assessment of whole-body adipose tissue topography by MRI. J Magn Reson Med 21:455–462 28. Michaely HJ, Herrmann KA, Kramer H, Laub G, Reiser MF, Schönberg SO (2004) The significance of MR angio­ graphy for the diagnosis of carotid stenoses. Radiologe 44:975–984 29. Muller-Horvat C, Radny P, Eigentler T K, Schafer J, Pfannenberg C, Horger M, Khorchidi S, Nagele T, Garbe C, Claussen CD, Schlemmer HP (2006) Prospective comparison of the impact on treatment decisions of whole-body magnetic resonance imaging and computed tomography in patients with metastatic malignant melanoma. Eur J Cancer 42:342–350 30. O’Connell M J, Powell T, Brennan D, Lynch T, McCarthy CJ, Eustace SJ (2002) Whole body MR imaging in the diagnosis of polymyositis. Am J Roentgenol 179:967–971 31. Pickhardt PJ, Choi JR, Hwang I, Butler JA, Puckett ML, Hildebrandt HA, Wong RK, Nugent PA, Mysliwiec PA, Schindler WR (2003) Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med. 2003 349:2191–2200 32. Ruehm SG, Goyen M, Barkhausen J, Kroger K, Bosk S, Ladd ME, Debatin JF (2001) Rapid magnetic resonance angiography for detection of atherosclerosis. Lancet 357:1086–1091

33. Schaefer JF, Vollmar J, Schick F, Vonthein R, Seemann MD, Aebert H, Dierkesmann R, Friedel G, Claussen CD (2004) Solitary pulmonary nodules: dynamic contrast-enhanced MR imaging—perfusion differences in malignant and benign lesions. Radiology 232:544–553 34. Schlemmer HP, Schafer J, Pfannenberg C, Radny P, Korchidi S, Muller-Horvat C, Nagele T, Tomaschko K, Fenchel M, Claussen CD (2005) Fast whole-body assessment of metastatic disease using a novel magnetic resonance imaging system: initial experiences. Invest Radiol 40:64–71 35. Schroeder T, Ruehm SG, Debatin JF, Ladd ME, Barkhausen J, Goehde SC (2005) Detection of pulmonary nodules using a 2D HASTE MR sequence: comparison with MDCT. AJR Am J Roentgenol 185:979–984 36. Schulthess GK von, Steinert HC, Hany TF (2006) Integrated PET/CT: current applications and future directions. Radiology 238:405–422 37. Smith R, Cokkinides V, Eyre H (2004) American cancer society guidelines for the early detection of cancer. CA Cancer J Clin 54:41–52 38. Steidle G, Schafer J, Schlemmer HP, Claussen CD, Schick F (2004) Two-dimensional parallel acquisition technique in 3D MR colonography RoFo 176:1100–1105 39. Walker RE, Eustace SJ (2001) Whole body magnetic resonance imaging: techniques, clinical indications, and future applications. Semin Musculoskelet Radiol 5:5–20 40. Walker R, Kessar P, Blanchard R, Dimasi M, Harper K, DeCarvalho V, Yucel EK, Patriquin L, Eustace S (2000) Turbo STIR magnetic resonance imaging as a whole-body screening tool for metastases in patients with breast carcinoma: preliminary clinical experience. J Magn Reson Imaging 11:343–350 41. Willinek WA, Born M, Simon B et al (2003) Time-of-flight MR angiography: comparison of 3.0-T imaging and 1.5-T imaging—initial experience. Radiology 229:913–920

1255

Chapter 12

Interventional MRI

12.1 

Technical Implementation  . . . . . . . . . . .   1258 M. Bock and F. Wacker

12.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . .   1258

12.1.2 

Interventional MR Systems   .. . . . . . . . . .   1258

12.1.2.1  C-Arm Magnets  . . . . . . . . . . . . . . . . . . . . .   1258 12.1.2.2  Closed-Bore Solenoid Magnets  .. . . . . . .   1259 12.1.2.3  Combined X-Ray and MRI Systems  . . .   1260 12.1.2.4  In-Room Image Displays  . . . . . . . . . . . . .   1260 12.1.2.5  Communication Systems  . . . . . . . . . . . . .   1261 12.1.3 

Instrument Tracking   .. . . . . . . . . . . . . . . .   1262

12.1.3.1  Contrast-Changing Markers  . . . . . . . . . .   1262 12.1.3.2  Direct Currents  .. . . . . . . . . . . . . . . . . . . . .   1263 12.1.3.3  Radiofrequency Coils   for Profiling and Tracking  .. . . . . . . . . . . .   1263 12.1.3.4  Inductively Coupled Coils  . . . . . . . . . . . .   1264 12.1.3.5  Gradient Field Measurement  .. . . . . . . . .   1264 12.1.4 

Instrument Navigation   .. . . . . . . . . . . . . .   1265

12.1.4.1  Manual Instrument Alignment  .. . . . . . .   1266 12.1.4.2  Robotic Assistance Systems  . . . . . . . . . . .   1266 12.1.4.3  MRI Instrument Steering   .. . . . . . . . . . . .   1267 12.1.5 

Real-Time Imaging  .. . . . . . . . . . . . . . . . . .   1267

12.1.5.1  Pulse Sequence Types  . . . . . . . . . . . . . . . .   1267 12.1.5.2  Sequence Control Interfaces  . . . . . . . . . .   1268 12.1.6 

Functional Imaging  . . . . . . . . . . . . . . . . . .   1269

12.1.6.1  Flow Measurements  .. . . . . . . . . . . . . . . . .   1269 12.1.6.2  Temperature Measurements  .. . . . . . . . . .   1271

12

12.1.7 

Safety Aspects  . . . . . . . . . . . . . . . . . . . . . . .   1272

12.1.7.1  Radiation Protection  . . . . . . . . . . . . . . . . .   1272 12.1.7.2  Static Magnetic Field  . . . . . . . . . . . . . . . . .   1273 12.1.7.3  Imaging Gradients  . . . . . . . . . . . . . . . . . . .   1273 12.1.7.4  Radiofrequency Fields  .. . . . . . . . . . . . . . .   1273 12.1.8 

Summary  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   1274 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1274

12.2 

Clinical Applications of Interventional and Intraoperative MRI  .. . . . . . . . . . . . .   1277 F.A. Jolesz and E. Samset

12.2.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . .   1277

12.2.1.1  Field Strength   and Clinical Applicability  .. . . . . . . . . . . .   1277 12.2.2 

Interventional MRI and Computer-Assisted Surgery  .. . . . . .   1278

12.2.3 

Interventional MRI and MR Imaging Techniques  .. . . . . . . . .   1280

12.2.4 

MRI-Guided Thermal Ablations   . . . . . .   1280

12.2.4.1  MR-Guided Focused   Ultrasound Surgery  . . . . . . . . . . . . . . . . . .   1282 12.2.5 

Current Clinical Applications of Interventional MRI  . . . . . . . . . . . . . . . .   1284

12.2.5.1  Intraoperative Guidance for Neurosurgery  .. . . . . . . . . . . . . . . . . . . .   1284 12.2.5.2  Vascular Interventions under MRI Guidance  .. . . . . . . . . . . . . . . .   1285 12.2.6 

Conclusion  .. . . . . . . . . . . . . . . . . . . . . . . . .   1285 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1286

1258

12  Interventional MRI

12.1  Technical Implementation M. Bock and F. Wacker 12.1.1  Introduction With the rapid progress of minimally invasive medicine in the recent decades, image guidance during inter­ ventional procedures plays an increasingly important role. Since the complexity of minimally invasive procedures is steadily increasing, the technical requirements grow for the imaging modalities that are used to monitor the procedure. Traditionally, interventions have been monitored with imaging modalities using either X-ray (fluo­roscopy, digital subtraction angiography [DSA], or computed tomography) or ultrasound. In particular, ultra­ sound is widely available and provides a cost-effective alternative to other imaging modalities. Unfortunately, both X-ray and ultrasound have distinct disadvantages: the unwanted ionizing radiation in X-ray techniques, the lack of soft tissue contrast, the fixed geometry of X-ray source and detector, and the sound reflections in ultrasound. Already in its early days MRI was considered as an alternative imaging modality for interventional pro­ce­ dures. Unfortunately, at that time the relatively slow image acquisition and the closed nature of the magnets pro­ hi­bited widespread use during interventional procedures. Nevertheless, MRI offers several advantages over other imaging modalities, which makes it an attractive imaging method for both diagnostic and therapeutic pro­cedures: MRI provides an excellent soft tissue contrast, imaging slices can be oriented without restrictions in all three dimensions, no ionizing radiation is used, and MRI yields morphologic as well as functional information such as the blood flow, tissue oxygenation, diffusion, perfusion, and temperature changes. This plethora of imaging tech-

niques is currently not available with any other imaging modality alone. Faster image acquisition and reconstruction have become possible with the development of new MR hardware. Real-time MRI is now implemented at nearly any existing MR system, and fast steady-state pulse sequences in combination with parallel imaging techniques allow acquiring images at frame rates of up to 10 images per second. In particular with the developments of new magnet designs, which provide an increased access to the patient, MR guided interventions have evolved over the last decade from a research tool to a preclinical method. An overview over the technical and clinical aspects of MR-guided interventions can be found in several textbooks (Lufkin 1999; Debatin and Adams 1999). In the following, the system design of interventional MR systems, the techniques for instrument tracking and navigation, image acquisition, and reconstruction strategies, as well as the safety aspects of interventional MRI will be discussed. 12.1.2  Interventional MR Systems The ideal interventional MR system combines optimal access to the patient with high image quality, fast image acquisition, and rapid image reconstruction and display. To this end several MR system designs have been proposed that are often used not only for interventional applications, but also for the imaging of obese or claustrophobic patients. 12.1.2.1  C-Arm Magnets The historically first interventional MR systems were designed as C-arm type magnets, which create a magnetic Fig. 12.1.1  Percutaneous procedure performed in a low-field open 0.2-T MR system

12.1  Technical Implementation

field between two magnet pole shoes in vertical directions (Grönemeyer and Lufkin 1999). These open-configuration MR systems offer—similar to conventional Carm fluoroscopy units—access to the patient from at least three directions and are thus especially suited for biopsies and thermal ablations (Fig. 12.1.1). Compared with the image detectors in X-ray fluoroscopy, the MRI pole shoes are significantly larger and therefore patient access is still limited. The C-arm-shaped MRI systems were often implemented with resistive or permanent magnets that operate at field strengths of B0 well below 1 T, and which were equipped with gradient systems that provided amplitudes Gmax < 25 mT/m. Compared with diagnostic MR systems with B0 ≥ 1.5 T and Gmax ≥ 45 mT/m, these C-arm systems are limited both in their achievable signal-to-noise ratio (SNR) as well as in the image acquisition speed. Recently, vertical field systems have been realized with superconducting magnets operating at field strengths up to 1 T (Fig. 2.5.6), so that a similar image quality will be achievable as with current high-field solenoid MR systems. Unfortunately, the increase in field strength also results in an increase in pole shoe diameter, so that patient access is further compromised. 12.1.2.2  Closed-Bore Solenoid Magnets For many MR-guided procedures such as intravascular interventions in the heart, both a high SNR and a short image acquisition time are very important. For this reason and due to the wide availability of diagnostic MR

systems, today an increasing number of interventions are performed in closed-bore super­conducting MR systems. Closed-bore solenoid magnets are especially suited for endovascular procedures since—contrary to per­cutaneous interventions—the target area of the intervention is separated by a distance of 40–80 cm from the entry point of the catheter into the vascular system. Thus, the interventionalist can be standing at the magnet opening while the target area is positioned at magnet isocenter. Nevertheless, short magnets are required to successfully perform an intervention in a solenoidal MR magnet. Fortunately, over the recent decade the length of the superconducting solenoid magnets has been significantly reduced from 2 m and more (for a 1.5-T magnet) to about to 160 cm. Recently, an ultrashort 125-cm-long 1.5-T magnet with a 70-cm-dia­meter opening (Siemens Magnetom Espree) has been introduced, which allows an even better access to the patient as conventional magnets with a bore of 60 cm diameter (Fig. 12.1.2). The larger dimensions of this short and relatively open magnet result in a higher field inhomogeneity so that only a limited field-of-view (FOV) of 35 cm is available (as compared with 50 cm in conventional MR systems). This limited FOV can be challenging in vascular applications where catheters need to be tracked over extended sections of the blood vasculature. Additionally, for stereotactic procedures such as biopsies a very careful calibration of the image distortion compensation is required to assure that the planned device trajectories coincide with the actual needle paths. Fig. 12.1.2  Short 1.5-T MR system with a 70-cm bore diameter solenoid magnet (Magnetom Espree, Siemens Medical Solutions, Germany). The radiologist is performing the intravascular intervention at the magnet opening and the current MR images are display via an MR-compatible in-room monitor on the other side of the patient table. (Image courtesy of Jens Heidenreich and Jeff L. Duerk, Case Western Reserve University, Cleveland, Ohio, USA)

1259

1260

12  Interventional MRI

Despite their availability another major advantage of the solenoid magnets are the gradient systems—in solenoid magnets, currently the strongest gradient systems with typical maximum gradient amplitudes of up to 45 mT/m are available. Combined with the high gradient slew rates that are made possible by the powerful gradient drivers, these systems allow fast image acquisition rates of 5–10 images per second, which are for example required for real-time imaging of the heart. 12.1.2.3  Combined X-Ray and MRI Systems At present, intravascular MR-guided interventions are rarely performed under MR guidance alone. For clinical applications, MRI is combined X-ray fluoroscopy. X-ray imaging is indispensable both as a backup modality and for patient safety, because the intervention needs in part to be performed under X-ray guidance due to the lack of MR-compatible instruments (e.g., guide wires). To facilitate the transport of the patient between the MR system and the X-ray, so-called XMR systems have been developed that combine a clinical MR system with an angiography unit (Fig. 12.1.3). In an experimental split magnet MR system (GE Signa SP, “double donut”) the X-ray tube was integrated in the gap between the two magnet halves, allowing a nearly simultaneous acquisition of the MR and X-ray images (Fahrig et al. 2001). The currently available commercial XMR systems, however, consist of an MRI system and an angiography unit that are connected by an optimized patient transport device for seamless transfer between the angio­graphy and MR patient table (Dick et al. 2005; Vogl et al 2002). The separation of the two imag-

ing modalities assures a high image quality both for the conventional angiography as well as MRI. Additionally, at times when a combined procedure is not scheduled, the imaging systems can be used inde­pendently. To make optimal use of the excellent soft tissue contrast of MRI and the rapid guide wire projection imaging provided by angiography, the MR images can be merged with the X-ray data after careful calibration for image distortion. These merged images are particularly helpful during cardiac applications (Rhode et al. 2005), where the X-ray contrast is used to steer the catheter, and the soft-tissue contrast of MR provides the anatomical information of the heart. 12.1.2.4  In-Room Image Displays An interventional MR system is not only defined by the magnet design and the gradient hardware. Rapid image acquisition is possible with nearly any MR system today; however, the images also need to be presented to the radiologist in the Faraday cage with a low latency of only a few hundred milliseconds. Additionally, the images should be seen by the whole personnel in the magnet room and need to be visible in an environment with high levels of ambient light. For this purpose, the manufacturers of the MR systems offer dedicated monitors that are radiofrequency shielded to avoid electromagnetic interference with the MRI (Fig. 12.1.2). The TFT monitors are either integrated in a trolley that can be placed near the magnet or are ceiling-mounted. Another cost-effective solution is the use of video projectors in combination with a backprojection

Fig. 12.1.3  XMR system with an angiography workplace in front and a commercial 1.5-T MR system in a separate room. The patient can be transferred from one system to the other via an optimized patient transfer solution. (Image courtesy of Prof. T.J. Vogel, University of Frankfurt, Frankfurt, Germany)

12.1  Technical Implementation

screen, which provides a much larger image than the currently available TFT monitors do. In a commercial operating room with integrated MR system (BrainSuite, BrainLab, Munich, Germany) the image display system has been realized with several back­pro­ jection systems in the cabin wall to simultaneously show the MR images with coregistered intraoperative images acquired with an operation microscope (Fig. 12.1.4). Lately, head-mounted displays have been introduced for image visualization (Wendt et al. 2003). Here, the MR images are displayed as a transparent overlay over the current anatomy when the patient is removed from the magnet bore. This image overlay provides the radiologist with the cross-sectional anatomical information that is required to safely insert an instrument (e.g., a needle) into the human body. In this implementation again careful registration of the MR images with the patient and the image display apparatus is needed. A major disadvantage

of this solution is the inability to correct for short-term patient motion caused, e.g., by breathing or the beating heart. 12.1.2.5  Communication Systems Excessive gradient noise during real-time imaging is a severe problem during MR-guided interventions. During rapid real-time imaging typi­cal­ly the highest gradient amplitudes are required, which lead to sound pressure levels of up to 120 dB. For this reason ear protection must be worn by both patients and personnel during those parts of the procedure when MR imaging is required (Kanal et al. 1990). Gradient noise makes a communication within the RF cabin and between RF cabin and the control room very difficult or even impossible. To this end, dedi­cated com-

Fig. 12.1.4  Image displays for interventional and intraoperative MRI. Top In a commercial operating room for MR-guided neurosurgery (BrainSuite, BrainLab, Germany) backprojection displays are integrated in the walls of the RF cabin. Bottom In a prototype setup a virtual reality system is used to display previously acquired MR images over the anatomy of the patient. To co-register the images a commercial optical tracking system is used. (Images courtesy of top BrainLab, bottom Frank Wacker)

1261

1262

12  Interventional MRI

munication systems have been developed that use optical microphones to avoid electromagnetic inter­ference with the image acquisition. Unfortunately, these headphone systems can become very uncomfortable to wear during longer procedures. For this reason, effective noise suppression within the gradient system is becoming increasingly important in the future (Edelstein et al. 2005). Noise cancellation has already been demonstrated using vacuum chambers around the gradients and external gradient mounting (e.g., Pianissimo gradients, Toshiba [Katsunuma et al. 2002]), which reduces the sound pressure levels by 20–30 dB. 12.1.3  Instrument Tracking During an interventional procedure instruments and materials such as catheters, guide wires, needles, embo­ li­za­tion material, and implants such as stents, coils, occluders, and inferior vena cava filters are introduced into the human body. An important prerequisite for any image-guided intervention is a rapid, precise, and reliable visualization and tracking of these instruments and materials. Although a large number of fast MR acquisition tech­ni­ques are available today, the visualization and tracking of interventional instruments is still technically challenging, as MR-guided interventions often require simultaneously a high spatial and temporal resolution. If for example a catheter needs to be inserted into the coronary arteries, image update rates of 5–10 images per second are required, and an image resolution of 1 mm or better should be used. For imaging and tracking of instruments in the MR environment, several techniques have been developed over the recent years, ranging from passive techniques that locally change the MR signal to methods where the imaging process itself is utilized for localization. 12.1.3.1  Contrast-Changing Markers The materials used to build instruments for interventions are not necessarily compatible with MRI: Either the instruments are nearly invisible on the MR image (as, e.g., with plastic catheters or guide wires made of Nitinol), or they cause significant artifacts (e.g., wire-stabilized catheters or ferromagnetic implants) that make it difficult to exactly define the instrument position. To detect interventional instruments on MR images, markers can be attached that either amplify or reduce the MR signal. A negative contrast or signal void can be created with small paramagnetic rings (e.g., made of dysprosium oxide), which are simple and cost-effective in their production (Bakker et al. 1996). A positive contrast has been achieved with paramagnetic T1-shortening contrast agents in the interior or at the surface (Frayne et al. 1999)

of the instrument (e.g., an intravascular catheter). If the instrument is used in a percutaneous intervention (e.g., a biopsy), then additional marker systems consisting of small reservoirs can be attached to the instrument, which are later visible in the MR image. For MR-guided prostate biopsies, such a marker has been realized as a cylinder filled with a gel (Beyersdorff et al. 2005), and in a robotic assistance system small spherical markers are attached to the instrument holder that are filled with a contrast agent solution. With this image-based tracking, the instruments (or their holders) are directly visualized on the MR image, and no ad­di­tional hardware is necessary—for this reason these localization techniques are often termed “passive.” If the marker is visible in the MR image, then the position can be extracted using dedicated image post-processing tech­niques (e.g., cross-correlation with a marker template image). The quality of the visualization is dependent on the material constants of the instruments as well as field strength, partial vo­lu­me effects, pulse sequence type, and imaging parameters. A disadvantage of passive tracking is the fact that instruments can only be detected when they are located in the current imaging slice. Thus, passive tracking is especially suited for applications where the instrument position is approximately known (e.g., during transrectal MR-guided biopsies), whereas it can be difficult to apply in situations where an extended instrument is to be localized (e.g., a catheter in a tortuous vessel). The temporal resolution is limited with passive localization techniques because a new image needs to be acquired for each update of the device coordinates. Even with fast MRI acquisition techniques such as parallel imaging and temporal interpolation, the required sampling rates of up to 10 images per second are not always achievable. In some applications, the instruments need to be identified automatically in the images for example to compare the actual with the planned position. In this case, the image acquisition time can amount to several seconds if 2D sampling techniques are used, since for a definite detection of the instrument several slices need to be acquired. If the instruments can be filled with a T1-shortening contrast agent (as is the case with intravascular catheters), heavily T1-weighted imaging techniques such as shortTR spoiled gradient-echo pulse sequences (FLASH) can be used to selectively visualize the instrument against a background this is nearly completely suppressed. For mar­ kers with negative contrast a separation is easily achieved between the markers’ signal void and the sur­rounding tissue. Unfortunately, in areas of the human body with low signal intensity as e.g., the lung, the in­strument’s susceptibility artifact cannot be distinguished from the tissue, and image-based automatic instrument tracking is not possible. Projection imaging has been proposed to overcome the temporal limitations of conventional imaging for de-

12.1  Technical Implementation

vice tracking. Therefore, thick 2D slices are acquired or non-selective RF excitations are used to detect MR signal from the whole sensitive volume of the MR system. With negative markers projection imaging can be combined with an incomplete refocusing of the transverse magnetization, so that only the signal in the strong static field gradients of markers is refocused. Thus, the negative contrast is converted into a positive signal from the vicinity of the markers (white-marker phenomenon [Seppenwoolde et al. 2003]); however, this technique requires that signal-generating material (e.g., tissue or blood) is present next to the marker. 12.1.3.2  Direct Currents To create a local signal change near the instrument a direct current can also be applied that is passed through an integrated wire—the current creates a local magnetic field distortion and thus a local signal void. Typically, in intravascular catheters, currents of several 10 mA are applied in coils of 1-mm diameter, and the size of the signal void is controlled by the amplitude of the current. Since the signal void can be adjusted and even be switched off, automatic instrument localization is possible: Projection images are acquired with and without signal void and a difference image is calculated that shows only signal from the vicinity of the coil (Glowinski et al. 1997). This difference technique increases the acquisition time by a factor of two, and requires that the switching of the direct current is synchronized with the pulse sequence.

the other two spatial directions, and the instrument position is determined within 10 to 20 ms. The instrument position can be reacquired with a frame rate of 50 updates per second and more, and the position can be visualized on a previously acquired image data set (roadmap) such as a 3D MRA. With a roadmap tech­ni­que only the position information and not the anatomical images are updated, which limits the use of this tech­ni­que to static anatomical situations as, e.g., intracranial interventions. For abdominal interventions where breathing and heart motion lead to a constantly changing anatomy the position, measurement is combined with a fast imaging pulse sequence (e.g., balanced SSFP, trueFISP). Here, the position of the instrument is detected in between the acquisition of two real-time images, and the information is used to automatically reposition the imaging slice (Leung

12.1.3.3  Radiofrequency Coils for Profiling and Tracking Another solution for automatic device tracking utilizes small RF coils that are attached to the in­stru­ments (Dumoulin et al. 1993). The coils are directly con­nec­ted to the receiver of the MR system via coaxial cables, so that the MR signal of the coils is read out separately from the imaging coils (Fig. 12.1.5). To localize the small coils pro­ jection data sets are acquired using a non-selective RF excitation (e.g., with the body coil) that creates transverse mag­netization in the whole imaging volume of the MR system. After the excitation, one spatial direction (x, y, or z) is encoded with a gradient echo that is only few milliseconds long. Due to the limited sensitivity profile of the small coil, in the one-dimensional projection only a single peak at the instrument position is observed, which can easily be detected. In practice, this signal is contaminated with background signal due to coil coupling; however, using an incomplete refocusing orthogonal to the projection direction (z-dephaser, twister gradients) the coil signal can be separated from the background (Unal et al. 1998). Projection data sets are then also acquired for

Fig. 12.1.5  Active tracking coils (top) at the tip of a 5-F catheter and (bottom) at the instrument holder of a robotic assistance system. The solenoid catheter coil has 10 windings and is connected to the MR receiver system via a coaxial cable of 300-µm diameter. The three tracking coils at the robotic assistance system are embedded in a container, which is filled with a contrast agent solution to provide the necessary MR signal. Additionally, a tuning and matching circuitry is connected to maximize the MR signal

1263

1264

12  Interventional MRI

et al. 1995; Bock et al. 2004). Thus the operator does not need to manually realign the image position (or with multiple coils, the image orientation), but controls the imaging process and even selected imaging parameters through the catheter motion (Wacker et al. 2004). Using fast imaging pulse sequences such as FLASH or trueFISP sampling rates of 3–6 images/s (at matrix sizes of 128–256) can be achieved, which can be increased by a factor of 2 or more with optimized parallel imaging methods (Bock et al. 2006a; Guttman et al. 2003; Müller et al. 2006). In intravascular interventions, the delineation of the catheter shaft can be difficult and looping of the catheter is hard to detect if only a single coil is available at the catheter tip. With an additional guide wire antenna in the catheter lumen, the total length of the catheter can be visualized (profiling). For guide wires, so-called extended inner-conductor antennas have been proposed where the guide wire consists of a coaxial cable on which the outer braiding is removed over a certain section near the tip (Serfaty et al. 2000). Alternatively, several small coils can be attached to the catheter shaft, which can be detected simultaneously if each coil is connected to a separate receiver of the MR system. In the past, catheters with up to 32 tracking coils have been realized; however, with an increasing number of coaxial cables in the catheter lumen the mechanical properties of the catheter are compromised. Tracking and profiling coils can also be combined as has been demonstrated in a coil design with a twisted pair coil (profiling) and a solenoid coil (tracking). Since tracking and profiling information is acquired at different times of the data acquisition, a small circuitry was added in this implementation that allowed switching between the two coils so that only a single connec­ting cable was required (Zuehlsdorff et al. 2004). Unfortunately, both profiling and tracking coils require long con­duc­ting elements, which are susceptible to severe RF heating and thus compromise patient safety. Fortunately, several modifications to the transmission lines have been proposed that reduce the risk of heating (cf. Sect. 12.1.7). The small coils at the devices are not only used for instrument tracking, but also for high-resolution imaging in endovascular interventions. Therefore, often the signals of two colinear solenoid coils with opposite circularity are com­bi­ned (opposed solenoid coils), which provide a coil system with an extended sensitivity profile (Hillenbrand et al. 2004). With this coil design not only tracking of the catheter is feasible, but also high-resolution imaging of the adjacent vessel wall. If the coils are read out individually, even parallel imaging in catheter direction becomes possible, which helps to reduce the total imaging time approximately by a factor of two. 12.1.3.4  Inductively Coupled Coils An alternative localization technique utilizes the local signal amplification in resonance circuits that are in­duc­

ti­vely coupled to the imaging coils (Burl et al. 1996). Similar to the tracking coils, a resonant radiofrequency coil is attached to the instrument; however, the coil is not directly connected to the receiver of the MRI system (Quick et al. 2005). Instead, the coil is interacting via magnetic induction with the transmit coil during RF excitation and with the receive coil during signal reception. Since inductively coupled coils are tuned to the resonance frequency of the MR system, they can be used at single field strength only. During RF excitation, the resonant coil amplifies the local B1 field by a factor of typically 10 and more so that in the vicinity of the coil a significantly larger flip angle is realized. Therefore, very low flip angles can be employed to selectively enhance the signal near the coil. This creates a measurable MR signal in the coil, whereas nearly no signal is created outside the coil. For instrument tracking, again projection data sets are acquired; however, in this case the signal of the imaging coils needs to be analyzed to detect the position of the coil (Flask et al. 2001). In addition to instrument tracking the concept of inductive coupling can also be utilized to perform highresolution imaging of the interior of implants. For this purpose implanted structures such as stents or vena cava filters have been realized in the form of inductively coupled coils (Kivelitz et al. 2003; Bartels and Bakker 2003). Therefore, capacitors were added to the metallic structures of the implants that form the inductors of the resonance circuit with the aim to detect re-stenosis or thrombus formation after implantation of the device. Inductively coupled coils do not utilize extended long conducting cable structures, and thus heating of the coils via E-field coupling is avoided if the wavelength of the electric RF field in tissue is longer than the coil dimension. At B0 = 1.5 T, the wavelength amounts to 23 cm, whereas at 7 T it is as short as 3.3 cm so that at high fields the safety of resonant structures needs further evaluation. Because inductively coupled coils and implants are not connected to the MR system, active detuning is difficult during imaging. Thus, flip angle amplification is often present during imaging, which leads to artifacts due to stimulated echoes. To enable detuning without compromising coil safety, the integration of a photoresistor in the resonant circuit has been proposed, which is illuminated with light via fiber optics and thus significantly lowers the quality factor of the inductively coupled coil (Wong et al. 2000). 12.1.3.5  Gradient Field Measurement So far, all presented techniques use the MR signal for position detection. Recently, alternative technologies have been proposed that directly measure the spatially varying gradient fields with small sensors. In one imple­ mentation, three orthogonal Hall probes were proposed that are connected to an external voltmeter via high resis-

12.1  Technical Implementation Fig. 12.1.6  Commercial tracking system based on the induced voltages in small wire loops (Robin Medical, Baltimore, Md., USA). a During ramp-up and ramp-down of the gradients in the three spatial direction (red, green, blue) a voltage is induced in the three pickup coils (sensor signals), from which the position of the coils can be calculated. These tracking systems can be integrated into nearly any existing interventional device. b Two orthogonal coil systems have been incorporated into an electrophysiology catheter close to the distal end (Images courtesy of Erez Nevo, Robin Medical)

tance cables (Scheffler and Korvink 2004). To determine the position the measured voltages are compared to the nominal gradient signal, which is supplied by the gradient control units of the MR system. Another approach uses three orthogonal coils, which measure the induced voltages during gradient ramping (Fig. 12.1.6). The small coils can be inte­gra­ted into a large variety of instruments ranging from endoscopes to needle holders, and even a miniature version of only 1 mm diameter has been real­ ized that has been integrated into an electrophysiology catheter (Robin Medical, Baltimore, Md., USA). Alternatively, a non-magnetic and non–electrically conducting optical sensor has been presented that measures the gradient fields using the Faraday effect (Bock et al. 2006b). Here, the sensor is constructed of an optically active crystal (e.g., Terbium-Gallium-Garnet), which rotates the polarization of a light beam in the presence of a colinear magnetic field. On modulation of the static field with the gradients, a corresponding modulation in the sensor output signal is seen, which can be detected with a lock-in amplifier. With this approach, not only the position, but also one orientation angle of the sensor can be detected.

A major advantage of all gradient field measurements over localization techniques that utilize the MR signal is the possibility to measure device position and MR image information simultaneously. During conventional MRI all three gradient axes are continuously required, and thus position information can be extracted at all times during the image acquisition. 12.1.4  Instrument Navigation To successfully guide the instrument to the target in the human body, additional technologies are required, if the intervention is to be performed in a closed-bore MR system. In particular during percutaneous interventions or during neurosurgery a precise alignment of the instrument with a planned trajectory is required, and the advancement of the instrument needs to be monitored continuously during the procedure. To facilitate these interventions manual and robotic assistance systems have been proposed. More recently, it has been investigated whether the magnetic forces of the MRI system can be utilized to steer the instrument to the target tissue.

1265

1266

12  Interventional MRI

12.1.4.1  Manual Instrument Alignment

12.1.4.2  Robotic Assistance Systems

The most simple and cost-effective method for a stable alignment of an interventional instrument such as a biopsy gun or an aspiration needle is realized by a mechanical instrument holder. The holder must be fully MR compatible, and it should contain MR-visible markers for position control. Additionally, an independent position measurement using, e.g., an integrated grid raster should be possible, if the device trajectory is planned prior to the intervention. For breast biopsies several commercial needle holder systems have been realized (Heywang-Köbrunner et al. 2000). The systems typically consist of an open-configuration breast coil in combination with a lateral needle access. At the beginning of the pro­cedure, MR images of the lesion and the surrounding breast tissue are acquired. After planning of the tra­jectory of the biopsy needle, the needle is inserted into a needle holder grid with the patient outside the magnet. The needle is then manually advanced to the target, and an MR image is acquired for position verification, prior to harvesting the biopsy sample. To perform transrectal prostate biopsies a needle holder system with a cylindrical marker has been proposed (Beyersdorff et al. 2005). The MR-compatible needle holder is connected to the patient table, and the distal end of needle holder is positioned in the rectum of the patient. Alignment of the needle with the target structure in the prostate is achieved by a manual motion and rotation of the joints of the holder system (Fig. 12.1.7). After each realign­ment, the position of the system is verified with fast pulse sequences. Recently, an automatic procedure for image re-alignment has been presented that utilizes the special appearance of the cylindrical markers in rapidly acquired localization images (Oliveira et al. 2007).

In closed-bore MRI systems, the space for a robotic assistance system is very limited. Therefore, different access pathways have been realized with the currently available robotic systems for percutaneous interventions (Cleary et al. 2006). To perform breast biopsies a robotic needle system (Robitom, Forschungszentrum Karlsruhe, Germany) has been integrated into the patient table (Kaiser et al. 2000). The needle is automatically advanced to the lesion in the breast with the patient in a supine position. Breast biopsy samples have been harvested automatically using the image information of the previously acquired MR images. Due to the construction of the robot no lateral access to the patient is possible, which can be problematic for certain locations of the lesions. Another robot for MR-guided interventions is the assistance system Innomotion (Innomedic, Herxheim, Ger­ many), which consists of a pneumatically driven arm that is mounted on an arc above the patient (Gutmann et al. 2002). The fully MR-com­patible system has 6 degrees of freedom, which allows positioning and orienting a device within the magnet (Fig. 12.1.8). At the end of the arm, the instrument is attached to an instrument holder, which is also equipped with passive markers for localization. Before the intervention anatomical images of the target point and the insertion point are acquired to plan the device trajectory. The assistance system then moves the instrument into position, and the insertion of the instrument is manually performed by the radiologist either in the magnet under real-time image control or outside the magnet. To perform neurosurgery in the MR system recently a robotic arm has been presented (neuroArm, University of Calgary, Calgary, Canada). This arm uses piezo motors

Fig. 12.1.7  Side view of a needle holder for a manually operated transrectal biopsy system (InVivo/Daum, Schwerin, Germany). The system can be rotated (four positions shown) and shifted to adjust the orientation of the passive marker at the tip. The MR-compatible biopsy gun (not shown) is inserted through the marker, and the sample is harvested

12.1  Technical Implementation Fig. 12.1.8  Pneumatically driven robotic assistance system for percutaneous interventions (Innomotion, Innomedic, Herxheim, Germany). The robotic arm is attached to the patient table. The instrument holder at the distal end is automatically positioned over the lesion after the needle trajectory is planned on the MR images. When the instrument is advanced into the human body, the current device position can be visualized with real-time MR imaging (Image courtesy of Yan De Andres, German Cancer Research Center, Heidelberg, Germany)

for motion, and can be controlled from outside the MR scanner room. When operated with a second arm of the same construction, it can be utilized for neurosurgical applications similar to existing robotic operating systems currently in use. 12.1.4.3  MRI Instrument Steering If a solenoid coil with a large number of turns is wound around a bendable instrument such as a catheter, a direct current in the coil creates a magnetic field that interacts with the static magnetic field of the magnet. Depending on the orientation of the coil and the current amplitude, a torque is acting on the coil that tends to rotate the coil and thus can be used to steer the catheter (Roberts et al. 2002). In prototypes this technique has been utilized to remotely steer an intravascular catheter in a vessel phantom. Unfortunately, relatively high currents of 100 mA and more are required for this technique, which result in Ohmic heating of the coil and which are potentially dangerous in case of a device failure. Another possibility to exert a force on an object in the MR system is offered by the gradient system: If a (preferably ferro-)magnetic piece of suitable size is integrated into the interventional device, then the gradient fields create a force, which is acting in gradient direction. Even though the gradient-induced force is typically small, it has been shown that it is strong enough to manipulate a small ferromagnetic sphere in the blood vessel of an animal (Martel et al. 2007). Whether this technique is applicable in a clinical device needs to be evaluated. In particular the risks associated with the ferromagnetic components must be assessed, which can lead to strong forces during the introduction of the device in the fringe field of the magnet.

12.1.5  Real-Time Imaging Rapid image acquisition with real-time image reconstruction and display is an essential prerequisite for many MRguided interventional procedures. In particular, endovascular procedures require real-time imaging because the moving anatomical structures of the heart and the arterial blood vessels need to be visualized at frame rates that are comparable to that of the organ motion, and the orientation and position of the imaging slice must be readjusted continuously to follow the motion of the instrument. 12.1.5.1  Pulse Sequence Types Over the recent decades, many pulse sequences have been proposed that can acquire an MR image in less than 1 s (Derakhshan and Duerk 2005). One of the earliest pulse sequences for real-time imaging is the spoiled gradientecho pulse se­quence (fast low-angle shot, FLASH). Here, low flip angles (α = 5–30°) are used for excitation so that only a fraction of the longitudinal magnetization is utilized. In between two RF excitations, the transverse magnetization is com­pletely spoiled using strong gradients (spoilers) or a random sequence of RF phases (RF spoiling), so that the image contrast is mainly determined by T1. This strong T1 dependency is making FLASH pulse sequences very attractive for interventional MRI, because they allow visualizing the distribution volume of therapeutic agents directly during the injection. For this purpose, the agent is mixed with diluted contrast agent that shortens T1, so that in heavily T1-weighted FLASH images the injected volume appears bright. Another application of T1-weighted FLASH pulse sequences is MR angiography (MRA). Again the transit of a contrast bolus is visualized; how­ever, with contrast-enhanced MRA often an intra-arterial injection

1267

1268

12  Interventional MRI

of the contrast agent is combined with a rapid serial acquisition of FLASH images of the vascular target area to assess the perfusion of the subsequent organ. Another important pulse sequence for real-time imaging is the balanced steady state free precession pulse se­quence (bSSFP, trueFISP, FIESTA) (Duerk et al. 1998). Conceptionally similar to the FLASH pulse sequence, here the trans­verse magnetization is not spoiled but completely refocused, so that the image contrast is dependent on both T1 and T2. If high flip angles (α = 70°) are used in combination with short repetition times, then the contrast is dependent on the ratio of T2/T1, which is high for liquids. In general, these pulse sequences provide images with a sig­nificantly higher SNR. Unfortunately, so-called banding artifacts are present in areas with strong offresonances, which can only partially be compensated by shimming, since the position and orientation of the realtime image slices might be changing continuously. Echo planar imaging (EPI) is often associated with fast imaging, since it most efficiently acquires data in k-space with only a single RF excitation per image. Unfortunately, single-shot EPI images suffer from a number of artifacts (off-resonance shifts, N/2-ghosts, distortion), which still limits the use of EPI to intracranial imaging. However, if not all k-space lines are acquired in a single shot but only a certain fraction (typically 3–7 echoes per excitation), the short acquisition times of EPI can be combined with the advantages of a segmented image acquisition. Seg­men­ tation helps reducing the off-resonance blurring, and the degree of image artifacts can be adjusted by chan­ging the number of echoes per RF excitation (McKinnon 1993). Spin-echo sequences are generally associated with

Real-time pulse sequences often differ from conventional diagnostic pulse sequences in that certain parameters can be changed interactively. If the real-time sequence is for example used to monitor the movement of the interventional device, then both slice position and orientation

Fig. 12.1.9  Left Planning image of a percutaneous intervention showing the abdominal anatomy of a pig. Around the needle trajectory only a small strip is imaged using a fast spin-echo technique (HASTE) in combination with a local-look preparation and a restore pulse (DEFT) to enhance the signal from tis-

sues with long T2. Right In the three images acquired with this sequence during needle insertion, the needle with stainless steel mandrel is visible as an artifact with a very limited diameter, which is only possible because a spin echo acquisition was used (Images courtesy Zimmermann et al. 2006)

long acquisition times, and are thus considered unsuitable for real-time imaging. However, with multiple refocusing of the spin echo using different phase-encoding steps (RARE) an MR image can be acquired in less than a one second. If combined with a half-Fourier data acquisition strategy, then this acquisition time can be further reduced. Spin-echo sequences are advantageous for interventional MRI since susceptibility artifacts from needles or other instruments are highly reduced. Nevertheless, real-time imaging with SE sequences remains challenging, because each 90° RF excitation saturates the longitudinal mag­ne­tization, so that long repetition times are required to reestablish the signal. For percutaneous interventions, often image update rates of less than one image per second are acceptable, so that SE MRI has been used effecti­ve­ly in this context. To accelerate the image acquisition, so-called inner volume imaging techniques can be used where only a certain area around the needle trajectory is imaged (Buecker et al. 1998). Furthermore, the remaining magnetization at the end of the echo train can be restored (DEFT), which allows shortening the repetition times for tissues with long T2 (Zimmermann et al. 2006) (Fig. 12.1.9). 12.1.5.2  Sequence Control Interfaces

12.1  Technical Implementation

permanently need to be readjusted (Quick et al. 2003). Unfortunately, it can become very challenging to update the slice position manually during an intervention. The radiologist needs to concentrate on the manipulation of the interventional instrument rather than interacting with a graphical interface for slice positioning. To facilitate the interaction with the pulse sequence several control interfaces have been presented (Fig. 12.1.10). In the sequence control interfaces for interventional procedures, typically only a subset of all MR parameters is displayed. This subset is selected depending on the application: During percutaneous interventions the readout bandwidth might be set up as a dynamically adjustable parameter to change the artifact sizes of metallic in­stru­ ments, whereas in intravascular procedures the FOV could be changed to zoom in on suspect lesions. Typi­ cally, all control interfaces have some sort of orientation control that allows changing the position and the orientation either graphically or through button clicks. To further minimize the interaction with the interfaces often automatic slice tracking is applied. Here, the po­si­tion the device is determined, and the device coordinates are then utilized to automatically position the imaging slices so that the radiologist is controlling the slice position by the interventional device itself. For localization active tracking coils, external markers, optical tracking systems or even control information of robotic assistance systems can be used as described above. In general it is advantageous if the position information is extracted from the MR signal or the measurement of the gradient field, because the non-linearities of the gradients cancel as they are present during both position measurement and slice orientation. Another method for automatic parameter control has been proposed that reduces the interaction of the radiologist with the control interfaces even further. In this method not only the position, but also the velocity of the inter­ventional instrument (here, an active catheter) is utilized for parameter control: If the catheter is moving with a high velocity during the initial placement of the instrument, a large imaging FOV is selected to gain an over­view over the vasculature. Later on during the insertion of the catheter into a vessel branching, the velocity is lowered, and the FOV is automatically reduced to visualize the local anatomy with a higher spatial resolution (Elgort et al. 2003). During the interventional procedure, different image contrasts are required at different stages of the intervention. When an intravascular catheter is maneuvered into the target vessel, balanced SSFP pulse sequences are used to visualize the blood in the vessels with positive contrast. Once the target position is reached, a contrast agent bolus can be injected, which is preferably imaged with a spoiled gradient-echo pulse sequence with high flip angles to suppress the background signal of the tissue. For this reason the image contrast can be changed via the user interface—in the example of the FLASH

and the trueFISP sequence the change in contrast can be achieved by the introduction of an additional re-winder gradient, so that the same gradient timing can be used for both contrasts. In other implementations, magnetization preparation pulses or saturation slabs can be dynamically inserted into the sequence to impose a T1 contrast on the images or to presaturate inflowing blood. 12.1.6  Functional Imaging An advantage of MRI over other imaging modalities for image-guided therapies is the availability of functional imaging techniques. With functional MRI, the therapyinduced change in organ function can be assessed during the intervention so that the procedure can be repeated in case of an incomplete result. If for example a flow measurement indicates that the renal artery flow is only partially restored after stent placement, the balloon catheter might be reinserted to fully expand the stent. A large variety of functional imaging techniques exist for example for the quantification perfusion, diffusion, brain activation (BOLD contrast), flow, and temperature. To apply these quantification techniques in an interventional procedure, the functional images need to be acquired in a short time, and the functional information should be reconstructed in nearly real time. For this reason, only selected functional techniques have been utilized during interventional procedures. 12.1.6.1  Flow Measurements Flow measurements are an indispensable tool to assess blood flow in vessels. With phase-contrast flow measu­re­ ments, the flow velocity can be measured even in small vessels with pulsatile flow in only 1–2 min. Espe­cially during an intravascular procedure the therapy-induced change in blood flow is an important parameter, which can be used, e.g., to define the end-point of an embolization procedure. In conventional phase-contrast flow measurements, velocity images are calculated from the phase difference be­tween flow-sensitized and flow-compensated data acquisition (cf. Sect. 2.7.7.1). In the images a region of interest is placed around the blood vessel, and the mean flow rate is calculated for each phase of the cardiac cycle. Even though real-time implementations exist for this procedure (van der Weide et al. 2000), it can still be timeconsuming and is often user depen­dent. To simplify and accelerate intravascular flow measurements an active catheter can be used in combination with a fast velocity-encoding pulse sequence, which acquires the flow-encoded signal without any further spatial enco­ding (Volz et al. 2004). Here, use is made of the fact that the small coil at the catheter tip is acquiring MR signal from its immediate vicinity, so that the catheter

1269

1270

12  Interventional MRI

Fig. 12.1.10a–c  User interfaces for real-time pulse sequence control. a In a simplified interface, only very few para­meters of the pulse sequence are accessible; however, this greatly reduces the complexity of the interaction for the radiologist. In the trueFISP image, the active catheter is seen as a white cross, and after manual change of the image contrast to FLASH imaging with a

high flip angle, the transit of a contrast agent bolus through the lift upper pole of the kidney is visualized. b After integration into the conventional user interface of the MR system (Siemens Magnetom Symphony), more parameters can be controlled in real time, and the current slice position is also visible on previously acquired localizer images

12.1  Technical Implementation

Fig. 12.1.10a–c  (continued) User interfaces for real-time pulse sequence control. c In a prototype of a dedicated interventional interface the slice positioning is achieved via graphical controls,

and again only the relevant parameters of the sequence can be changed (not shown) (Image © courtesy of Christine Lorenz, Siemens, Princeton, N.J., USA)

acts as a local velocity sensor. Since no spatial encoding is required, the velocity information can be measured in only two TR intervals, i.e., typically in less than 20 ms. With such a real-time flow measurement comparable results have been achieved as with conventional flow measurements that require several minutes of acquisition time.

of these thermal therapy concepts, all of them can be real­ ized under MR guidance. For thermal therapies, MR guidance is very helpful, since MRI is the only imaging modality that provides a quan­titative measure of the local temperature change (Quesson et al. 2000). MR temperature measurements are based on the tem­pe­ra­tu­re dependence of three different parameters: the longitudinal relaxation time T1, the diffusion coefficient D, and the proton resonance frequency. Above 60 °C the T1 relaxation time of tissue is increasing linearly with temperature according to T1(T+∆T) = T1(T) + m × ∆T (Graham et al. 1998). Unfortunately, the proportionality constant m is dependent on the tissue type, which requires cali­bra­tion and, in case of multiple tissue types, segmentation of the image information. Typically, the initial T1 value is determined prior to the procedure using for example a fast saturation-recovery pulse sequence with different saturation delays. During heating, a T1 measurement would be too time-consuming, and thus only the change of the signal intensity in a

12.1.6.2  Temperature Measurements Many therapy concepts utilize thermal effects for the destruction of tissue. Tissue is frozen with the help of needle-shaped cryoprobes (cryotherapy), it is coagulated using high-intensity laser pulses that are applied via optical fibers (laser-induced thermal therapy, or LITT), it is exposed to RF waves that are trans­mitted through dedicated needle systems, or it is heated using ultrasound waves that are focused using external transducers (highintensity focused ultrasound, or HIFU). Because MRcompatible applicators are commercially avai­lable for all

1271

1272

12  Interventional MRI

saturation recovery imaging pulse sequence is measured. To optimize the sensitivity of the experiment (i.e., to create a maximal signal change under heating) the saturation delay is set to the T1 of the target tissue. A major advantage of T1-based MR temperature measurements is that the technique is not sensitive to motion; therefore, it can also be applied in moving organs such as the liver. The second MR temperature measurement technique utilizes the temperature dependence of the diffusion co­ef­ficient D(T) ~ e Ea/kT (Le Bihan et al. 1989). Here, the activation energy Ea amounts to 0.18 eV, and k is the Boltzmann constant. For temperature mapping, the diffusion coefficient is continuously mapped before, during and after heating. The mea­su­rement of the diffusion coefficient is typically done with a Stejskal-Tanner pulse sequence that utilizes a pair of strong gradients to create a diffusion-dependent signal reduction. Diffusion mapping is often performed with single shot pulse sequences such as echo planar imaging to avoid artifacts from phase inconsistencies between the different k-space segments. The diffusion-sensitizing gradients make diffusion mapping very sensitive to motion, so that this technique is preferably applied in static organs. The proton resonance frequency (PRF) technique (DePoorter et al. 1994) is making use of the temperature dependency of the magnetic screening ∆σ(T) of the external magnetic field—with increasing temperature the screening increases, the nuclei ex­perience a lower local magnetic field, and the resonance frequency is shifted according to ∆ω(T) = γB ∆σ(T), with ∆σ(T) = α ∆T. An advantage of the PRF technique is the fact that the proportionality constant α = 0.008–0.009 ppm/°C is nearly tissue independent; only fat does not show a temperature dependence so that fat sup­pression techniques need to be employed to avoid systematic errors in fat-containing tissues (e.g., breast tissue). To measure the frequency shift ∆ω gradient echo techniques with echo times of typically 10–20 ms (at 1.5 T) are utilized. Temperature maps are calculated from the difference of the phase images acquired before and during treatment. Because the PRF method utilizes long echo times, it is also prone to systematic errors due to motion—for this reason, flow compensation should be included in the pulses sequences reduce the phase effects of linear motion. Since all temperature mapping techniques calculate difference images, motion compensation strategies are re­ qui­red in anatomical regions where for example breathing motion is present. One strategy acquires not only a single baseline image prior to treatment, but a series of reference images at different breathing levels. During treatment, the breathing motion is monitored (e.g., using a breathing belt), and the temperature map is calculated from the current data set and the baseline image acquired under the same breathing condition. Even though this strategy minimizes the motion-induced error, it cannot account for involuntary motion of the patient.

Thermal mapping is not only used to visualize the treated volume but also to minimize the treatment duration. Once a critical temperature is exceeded for a defined amount of time the thermal damage is sufficiently high so that a new target volume can be treated. This thermal dosimetry is especially important for HIFU treatments where the lesion is covered with many small ultrasound foci, which require up to several minutes of treatment time each. 12.1.7  Safety Aspects During interventional procedures not only the patient, but also the personnel is present in the magnet room. Thus, increased safety measures have to be taken to avoid accidents due to ferromagnetic objects, RF-induced device heating, and excessive noise (Shellock and Crues 2004). Fortunately, a typical safety concern associated with interventional procedures is not applicable to MR-guided interventions: the protection from ionizing radiation. 12.1.7.1  Radiation Protection In radiation protection one of the fundamental concepts is the “as-low-as-reasonably-achievable” (ALARA) prin­ciple: If achievable with reasonable technical measures, radiation doses should be kept at a minimum. The ALARA principle is especially relevant for children, as in these patients even low radiation doses lead to an increased pro­ba­bi­lity for the patients to experience the negative consequences of the radiation ex­posure during their lifetimes (Miller et al. 2003; Mooney et al. 2000). Thus, MR-guided interventions are particularly indicated in children and younger adults, since no radiation is applied with MRI. For example, in young patients suffering from congenital heart disease, arrhythmia, or vascular malformations MRI guided interventions could replace X-ray guided endovascular inter­ventions (Razavi et al. 2003). The personnel in the operation room or an angiography suite are often continuously exposed to ionizing radiation (William 1997); therefore, the ALARA principle should also strictly be applied to this group of people. With an increasing duration and complexity of the interventions (e.g., embolizations, percutaneous portosystemic shunt placements, invasive electrophysiology and catheter ablations) the applied radiation doses have also significantly increased, and in some of these procedures deterministic radiation effects such as erythema or loss of hair have been re­ported. Regardless of patient age, morbidity or life expectancy these damages are into­le­rable. For this reason it should always be considered whether a complex intervention cannot also be performed under MR guidance.

12.1  Technical Implementation

12.1.7.2  Static Magnetic Field Ferromagnetic objects can become lethal missiles in the vicinity of an MR magnet (Schenck 2000). The forces acting on the object increase with field strength, as well as with the static field gradient. Therefore, all devices must be tested for magnetic components in each MR system they will be used in. Also, special care is re­quired when an unshielded magnet is replaced by a shielded system of the same field strength as the static field gradients increase considerably. Around actively shielded MR systems higher forces might be present and instruments that could safely be used in unshielded magnets in the past might now be attracted to the magnet in an uncontrollable way. Despite the attractive force, also a torque can be present that tries to align the magnetic device with the static field. For saturated ferromagnetic devices, this torque is strongest at magnet isocenter. Both attractive forces and torque can pose a significant risk—if, for example, a magnetic stent experiences a torque while the patient is moved into the magnet soon after an implantation the vessel might rupture. For this reason many implantable devices have been tested, and the results of these tests are published (Shellock 2007) or can be accessed through a dedicated website (MRIsafety.com). To reduce the risks associated with magnetic devices all objects in the RF cabin need to be carefully screened for ferromagnetic components. The use of ferromagnetic instruments should be prohibited whenever possible. The personnel must be trained to keep magnetic instruments at a safe distance from the magnet (e.g., the 0.5-mT line) at any time during the intervention. Additionally, the labeling of MR-compatible instruments should be im­ple­ mented to reduce risks. Recently, a labeling standard has been proposed, which uses three different labels (American Society for Testing and Materials 2007): a green square (“MR safe”), a yellow triangle (“MR conditional”), and a red circle (“MR unsafe”). There is hope that this or similar labeling standards are increasingly adopted by device manufacturers. 12.1.7.3  Imaging Gradients During real-time MRI the imaging gradients are rapidly switched on and off. In large conductive loops—as they are formed by the human body—the changing magnetic fields induce currents that can lead to the stimulation of peripheral nerves (Schaefer 1998). With increasing gradient slew rate, the peripheral nerve stimulation is changing from a tickling sensation to painful nerve stimulation. The stimulation threshold varies individually; however, at present all manufacturers offer gradient systems that already reach these stimulation limits. For this reason, the slew rate of the imaging pulse sequences

are monitored, and real-time pulse sequences are sometimes limited in their per­formance due to slew rate restrictions. Gradient activity also leads to sound pressures of 120 dB and more with interventional pulse sequences (Ozturk et al. 2005), so that for both patients and interventional personnel the use of ear protection is mandatory. Disposable earplugs offer the most simple and cost-effective solution as they reduce noise by 10–30 dB, which in most cases provides sufficient ear protection. For the personnel an MR-compatible communication system is advisable to commu­nicate within the magnet room and with the MR control room. 12.1.7.4  Radiofrequency Fields In diagnostic imaging, regulatory limits have been established to restrict the amount of energy that the patient absorbs in an investigation. To this end the specific absorption rate (SAR) has been introduced, which is the absorbed power per kilogram tissue measured in W/kg. In general, higher limits for the SAR are acceptable if the patient is monitored during imaging (first level controlled operation mode), or if the RF energy is only locally applied. The whole body SAR limits for normal (first level) operation mode are currently 2 (4) W/kg averaged over 6 min’s scan time, and up to 10 W/kg are accepted for partial body exposure (International Electrotechnical Commission 2005). In interventional MRI conducting structures such as guide wires, needles, or metallic stents are used that interact with the electric and magnetic radiofrequency fields. In particular, the electric fields can cause displacement cur­rents in the tissue surrounding the instruments, which are heating the tissue through Ohmic losses (Liu et al. 2000; Nitz et al. 2001; Schaefers and Melzer 2006; Bassen et al. 2006). This hea­ting is particularly pronounced when the length of the electrically conducting structures approaches the B0-de­pendent resonance length. Temperature differences of up to 44 K have been ob­served at a field strength of 1.5 T. This effect can be reduced significantly if the conducting structures are separated into shorter segments using e.g., baluns (Ladd and Quick 2000) or transformers (Weiss et al. 2005; Vernickel et al. 2005; Kraft et al. 2006). RF-induced heating is especially problematic when commercially available guide wires are used. These guide­ wires are often constructed of the non-magnetic material Nitinol, which does not cause any artifacts in the MR images; however, with typical lengths of 1–2 m, these wires are suscep­tible to severe heating. The safety problems with metallic wires have led to the development of MRvisible guide wires made of fiberglass or PEEK (Mekle et al. 2006). These electrically non-conducting wires might replace metallic guide wires in MR-guided intravascular

1273

1274

12  Interventional MRI

inter­ven­tions if their mechanical properties such as torsion stiffness or fracture stability can be made compar­ able to that of current clinical guide wires. In the European Union (EU) recently, a directive was adopted that sets new electromagnetic field limits for wor­kers. The field limits are given for the static magnetic field, the time-varying gradient fields, and the RF fields. The directive applies to the personnel in interventional MRI and must be incorporated into do­mestic law by all EU member states by May, 2008. Unfortunately, the new exposure limits for workers will be very low so that it might become impossible to perform some interventional MRI procedures within the EU (Keevil et al. 2005). Even though safety issues always should be taken very seriously, this directive leads to the unwanted situation that, for legal reasons, within the EU more interventions will be performed under X-ray guidance with considerable levels of ionizing radiation than within an MRI system. 12.1.8  Summary MRI offers many advantages over other imaging modalities for monitoring of inter­ven­tional procedures. For many of the problems associated with the use of MRI in an interventional scenario, tech­ni­cal solutions have been proposed, so that there is hope that in the nearer future an increasing number of inter­ven­tions will be performed under MR guidance.

7.

8.

9.

10.

11.

12. 13.

14.

15.

References 1.

2.

3. 4.

5.

6.

American Society for Testing and Materials (ASTM) F2503–05, Standard Practice for Marking Medical Devices and Other Items for Safety in the MR Environment. Available at http://www.astm.org Bakker CJ, Hoogeveen RM, Weber J et al (1996) Visualization of dedicated catheters using fast scanning tech­niques with potential for MR-guided vascular interventions. Magn Reson Med 36:816–820 Bartels LW, Bakker CJ (2003) Endovascular interventional magnetic resonance imaging. Phys Med Biol 48:37–64 Bassen H, Kainz W, Mendoza G, Kellom T (2006) MRIinduced heating of selected thin wire metallic implants— laboratory and computational studies—findings and new questions raised. Minim Invasive Ther Allied Technol 15:76–84 Beyersdorff D, Winkel A, Hamm B, Lenk S, Loening SA, Taupitz M (2005) MR imaging-guided prostate biopsy with a closed MR unit at 1.5 T: initial results. Radiology. 234:576–581 Bock M, Volz S, Zuhlsdorff S et al (2004) MR-guided intravascular procedures: real-time parameter control and automated slice positioning with active tracking coils. J Magn Reson Imaging 19:580–589

16.

17.

18.

19.

20.

21.

22. 23.

Bock M, Muller S, Zuehlsdorff S et al (2006a) Active catheter tracking using parallel MRI and real-time image reconstruction. Magn Reson Med 55:1454–1459 Bock M, Umathum R, Sikora J et al A (2006b) Faraday effect position sensor for interventional magnetic resonance imaging. Phys Med Biol 51:999–1009 Buecker A, Adam G, Neuerburg JM, Glowinski A, van Vaals JJ, Guenther RW (1998) MR-guided biopsy using a T2-weighted single-shot zoom imaging sequence (Local Look technique). J Magn Reson Imaging 8:955–959 Burl M, Coutts GA, Young IR (1996) Tuned fiducial markers to identify body locations with minimal perturbation of tissue magnetization. Magn Reson Med 36:491–493 Cleary K, Melzer A, Watson V, Kronreif G, Stoianovici D (2006) Interventional robotic systems: Applications and technology state-of-the-art. Minim Invasive Ther Allied Technol 15:101–113 Debatin JF, Adam G, (eds) (1998) Interventional magnetic resonance imaging. Springer, Berlin Heidelberg New York DePoorter J, Wagter C, DeDeene Y, Thomsen C, Stahlberg F, Achten E (1994) The proton-resonance-frequency shift method compared with molecular diffusion for quantitative measurement of two-dimensonal time-dependent temperature distribution in a phantom. J Magn Reson 103:234–241 Derakhshan JJ, Duerk JL (2005) Update to pulse sequences for interventional MR imaging. Magn Reson Imaging Clin N Am 13:415–429 Dick A, Raman V, Raval A et al (2005) Invasive human magnetic resonance imaging during angioplasty: fea­si­bility in a combined X-ray/MRI suite. Catheterization and Cardiovascular Interventions 64:265–274 Duerk JL, Lewin JS, Wendt M, Petersilge C (1998) Remember true FISP? A high SNR, near 1-second imaging method for T2- like contrast in interventional MRI at .2 T. J Magn Reson Imaging 8:203–208 Dumoulin CL, Souza SP, Darrow RD (1993) Real-time position monitoring of invasive devices using magnetic resonance. Magn Reson Med 29:411–415 Edelstein WA, Kidane TK, Taracila V et al (2005) Active-passive gradient shielding for MRI acoustic noise re­ duction. Magn Reson Med 53:1013–1017 Elgort DR, Wong EY, Hillenbrand CM, Wacker FK, Lewin JS, Duerk JL (2003) Real-time catheter tracking and adaptive imaging. J Magn Reson Imaging 18:621–626 Fahrig R, Butts K, Rowlands JA et al (2001) A truly hybrid interventional MR/X-ray system: feasibility de­mon­stration. J Magn Reson Imaging 13:294–300 Flask C, Elgort D, Wong E et al (2001) A method for fast 3D tracking using tuned fiducial markers and a limited projection reconstruction FISP (LPR-FISP) sequence. J Magn Reson Imaging 14:617–627 Frayne R, Weigel C, Yanng Z et al (1999) MR evaluation of signal emitting coatings. In: ISMRM proceedings, p 580 Glowinski A, Adam G, Bucker A et al (1997) Catheter visualization using locally induced, actively controlled field inhomogeneities. Magn Reson Med 38:253–258

12.1  Technical Implementation 24. Graham SJ, Bronskill MJ, Henkelman RM (1998) Time and temperature dependence of MR parameters during thermal coagulation of ex vivo rabbit muscle. Magn Reson Med 39:198–203 25. Grönemeyer DHW, Lufkin RB (1999) Open-field magnetic resonance imaging. Equipment, diagnosis and inter­ ventional procedures. Springer, Berlin Heidelberg New York 26. Gutmann B, Gumb L, Goetz M, Voges U, Fischer H, Melzer A (2002) Principles of MR/CT compatible robotics for image guided procedures. Radiology (Suppl)677 27. Guttman MA, Kellman P, Dick AJ, Lederman RJ, McVeigh ER (2003) Real-time accelerated interactive MRI with adaptive TSENSE and UNFOLD. Magn Reson Med 50:315–321 28. Heywang-Köbrunner SH, Heinig A, Pickuth D, Alberich T, Spielmann RP (2000) Interventional MRI of the breast: lesion localisation and biopsy. Eur Radiol 10:36–45 29. Hillenbrand CM, Elgort DR, Wong EY et al (2004) Active device tracking and high resolution intravascular MRI using a novel catheter-based, opposed-solenoid phased array coil. Magn Reson Med 51:668–675 30. International Electrotechnical Commission (2005) IEC 60601-2-33 Medical electrical equipment—Part 2-33: Par­ ticular requirements for the safety of magnetic resonance equipment for medical diagnosis. Edition 2.1 consolidated with amendment 1:2005 31. Kaiser WA, Fischer H, Vagner J, Selig M (2000) Robotic system for biopsy and therapy of breast lesions in a highfield whole-body magnetic resonance tomography unit. Invest Radiol 35:513–519 32. Kanal E, Shellock FG, Talagala L (1990) Safety considerations in MR imaging. Radiology 176:593–606 33. Katsunuma A, Takamori H, Sakakura Y, Hamamura Y, Ogo Y, Katayama R (2002) Quiet MRI with novel acoustic noise reduction. Magma 13:139–144 34. Keevil SF, Gedroyc W, Gowland P et al (2005) Electromagnetic field exposure limitation and the future of MRI. Br J Radiol 78:973–975 35. Kivelitz D, Wagner S, Schnorr J et al (2003) A vascular stent as an active component for locally enhanced mag­ ne­tic resonance imaging: initial in vivo imaging results after catheter-guided placement in rabbits. Invest Ra­diol 38:147–152 36. Krafft A, Muller S, Umathum R, Semmler W, Bock M (2006) B1 field-insensitive transformers for RF-safe transmission lines. MAGMA 19:257–266 37. Ladd ME, Quick HH (2000) Reduction of resonant RF heating in intravascular catheters using coaxial chokes. Magn Reson Med 43:615–619 38. Le Bihan D, Delannoy J, Levin RL (1989) Temperature mapping with MR imaging of molecular diffusion: application to hyperthermia. Radiology 171:853–857 39. Leung DA, Debatin JF, Wildermuth S et al (1995) Intravascular MR tracking catheter: preliminary experimental evaluation. AJR Am J Roentgenol 164:1265–1270

40. Liu CY, Farahani K, Lu DS, Duckwiler G, Oppelt A (2000) Safety of MRI-guided endovascular guidewire applications. J Magn Reson Imaging 12:75–78 41. Lufkin RB (1999) Interventional MRI. Mosby, St. Louis 42. Martel S, Mathieu JB, Felfoul O et al (2007) Automatic navigation of an untethered device in the artery of a living animal using a conventional clinical magnetic resonance imaging system. Appl Phys Lett 90:114105 43. McKinnon GC (1993) Ultrafast interleaved gradient-echoplanar imaging on a standard scanner. Magn Reson Med 30:609–616 44. Mekle R, Hofmann E, Scheffler K, Bilecen D (2006) A polymer-based MR-compatible guidewire: a study to ex­ plore new prospects for interventional peripheral magnetic resonance angiography (ipMRA). J Magn Reson Imaging 23:145–155 45. Miller DL, Balter S, Cole PE et al (2003) Radiation doses in interventional radiology procedures: the RAD-IR study: part I: overall measures of dose. J Vasc Interv Radiol 14:711–727 46. Mooney RB, McKinstry CS, Kamel HA (2000) Absorbed dose and deterministic effects to patients from interventional neuroradiology. Br J Radiol 73:745–751 47. Müller S, Umathum R, Speier P, Zühlsdorff S, Ley S, Semmler W, Bock M (2006) Dynamic coil selection for realtime imaging in Interventional MRI. Magn Reson Med 56:1156–1162 48. MRIsafety.com. Available at http://mrisafety.com 49. Nitz WR, Oppelt A, Renz W et al (2001) On the heating of linear conductive structures as guide wires and catheters in interventional MRI. J Magn Reson Imaging 13:105–114 50. Oliveira A, Rauschenberg J, Beyersdorff D, Semmler W, Bock M (2007) Automatic passive tracking of a prostate biopsy device using phase-only cross correlation. Proc Intl Soc Magn Reson Med 15:3391 51. Ozturk C, Guttman M, McVeigh ER, Lederman RJ (2005) Magnetic resonance imaging-guided vascular inter­ ventions. Top Magn Reson Imaging 16:369–381 52. Quesson B, de Zwart JA, Moonen CT (2000) Magnetic resonance temperature imaging for guidance of thermo­ therapy. J Magn Reson Imaging 12:525–533 53. Quick HH, Kuehl H, Kaiser G et al (2003) Interventional MRA using actively visualized catheters, trueFISP, and real-time image fusion. Magn Reson Med 49:129–137 54. Quick HH, Zenge MO, Kuehl H et al (2005) Interventional magnetic resonance angiography with no strings attached: wireless active catheter visualization. Magn Reson Med 53:446–455 55. Razavi R, Hill DL, Keevil SF et al (2003) Cardiac catheterization guided by MRI in children and adults with congenital heart disease. Lancet 362:1877–1882 56. Rhode KS, Sermesant M, Brogan D et al (2005) A system for real-time XMR guided cardiovascular intervention. IEEE Trans Med Imaging 24:1428–1440 57. Roberts TP, Hassenzahl WV, Hetts SW, Arenson RL (2002) Remote control of catheter tip deflection: an opportunity for interventional MRI. Magn Reson Med. 48:1091–1095

1275

1276

12  Interventional MRI 58. Schaefer DJ (1998) Safety aspects of switched gradient fields. Magn Reson Imaging Clin N Am 6:731–748 59. Schaefers G, Melzer A (2006) Testing methods for MR safety and compatibility of medical devices. Minim Invasive Ther Allied Technol 15:71–75 60. Scheffler K, Korvink JG (2004) Navigation with Hall sensor device for interventional MRI. 11th ISMRM meeting and exhibition proceedings, p 950 61. Schenck JF (2000) Safety of strong, static magnetic fields. J Magn Reson Imaging 12:2–19 62. Seppenwoolde JH, Viergever MA, Bakker CJ (2003) Passive tracking exploiting local signal conservation: the white marker phenomenon. Magn Reson Med 50:784–90 63. Serfaty JM, Yang X, Aksit P, Quick HH, Solaiyappan M, Atalar E (2000) Toward MRI-guided coronary cathe­te­ rization: visualization of guiding catheters, guidewires, and anatomy in real time. J Magn Reson Imaging 12:590–594 64. Shellock FG (2007) Reference manual for magnetic resonance safety, implants and devices: 2007 edition. Biomedical Research, Los Angeles 65. Shellock FG, Crues JV (2004) MR procedures: biologic effects, safety, and patient care. Radiology 232:635–652 66. Unal O, Korosec F, Frayne R, Strother CM, Mistretta CA (1998) A rapid 2D time-resolved variable-rate k-space sampling MR technique for passive catheter tracking during endovascular procedures. Magn Reson Med 40:356–362 67. Vernickel P, Schulz V, Weiss S, Gleich B (2005) A safe transmission line for MRI. IEEE Trans BME 52:1094–1102 68. Vogl TJ, Balzer JO, Mack MG, Bett G, Oppelt A. Hybrid MR interventional imaging system: combined MR and angiography suites with single interactive table (2002) Feasibility study in vascular liver tumor procedures. Eur Radiol 12:1394–1400

69. Volz S, Zuehlsdorff S, Umathum R, Hallscheidt P, Fink C, Semmler W, Bock M (2004) Semiquantitative fast flow velocity measurements using catheter coils with a limited sensitivity profile. Magn Reson Med 52:575–581 70. Wacker FK, Elgort D, Hillenbrand CM, Duerk JL, Lewin JS (2004) The catheter-driven MRI scanner: a new approach to intravascular catheter tracking and imaging-parameter adjustment for interventional MRI. AJR 183:391–395 71. Weide R van der, Bakker CJG, Hoogeveen RM, Viergever MA (2000) On-line flow quantification by low-re­so­lu­tion phase-contrast MR imaging and model-based postprocessing. J Magn Reson Imaging 12:623–631 72. Weiss S, Vernickel P, Schaeffter T, Schulz V, Gleich B (2005) Transmission line for improved RF safety of interventional devices. Magn Reson Med 54:182–189 73. Wendt M, Sauer F, Khamene A, Bascle B, Vogt S, Wacker FK (2003) A head-mounted display system for aug­mented reality: initial evaluation for interventional MRI. RoFo 175:418–421 74. Williams JR (1997) The interdependence of staff and patient doses in interventional radiology. Br J Radiol 70:498–503 75. Wong EY, Zhang Q, Duerk JL, Lewin JS, Wendt M (2000) An optical system for wireless detuning of parallel resonant circuits. J Magn Reson Imaging 12:632–638. 76. Zimmermann H, Müller S, Gutmann B, Bardenheuer H, Melzer A, Umathum R, Nitz W, Semmler W, Bock M (2006) Targeted HASTE (TASTE) imaging with automated device tracking for MR-guided needle inter­ven­tions in closed-bore MR-systems. Magn Reson Med 56:481–488 77. Zuehlsdorff S, Umathum R, Volz S et al (2004) MR coil design for simultaneous tip tracking and curvature de­ lineation of a catheter. Magn Reson Med 52:214–218

12.2  Clinical Applications of Interventional and Intraoperative MRI

12.2  Clinical Applications of Interventional and Intraoperative MRI F.A. Jolesz and E. Samset 12.2.1  Introduction The main goal of interventional or intraoperative MRI is to provide dynamic, interactive image guidance for surgical and percutaneous interventional procedures (Jolesz 1998). The recent emergence of interventional and intraoperative MRI has largely been due to an integrated, coordinated, multidisciplinary translational research effort to develop several MRI-guided procedures tested in clinical trials (Alexander III et al. 1996: Black et al. 1997; Hinks et al. 1998; Jolesz and Blumenfeld 1994; Tronnier et al. 1997). Current efforts by several groups address critical areas of interventional and intraoperative MRI that may have a significant impact on the future of this emerging field (Kettenbach et al. 2000). Through the combined efforts of several research teams, intraoperative MRI has been successfully developed and implemented for multiple interventional and surgical procedures, including biopsies, craniotomies, resection or treatment of various tumors, drainage of intra-cranial cysts, and thermal ablation of malignant and benign tumors (Jolesz et al. 2002). Serial intraoperative imaging using frequent volumetric image updates offers physicians several advantages over other intraoperative guidance systems that use only a single database generated from preoperative images. Intraoperative imaging and navigation can be fully integrated with various MRI scan configurations, dynamic MR imaging methods, cardiovascular procedures, surgical tools, robotic devices, and thermal ablation systems such as MRI-guided laser or cryoablation and focused ultrasound (FUS). Image guidance, however, is a rather complex task that involves localization, targeting, navigation, monitoring, and therapy control. Current research efforts aim to improve these basic components of image guidance using advances in MR imaging technology and resolving issues related to limited access. It is anticipated that substantial improvements in MRI-guidance techniques will result in better clinical outcomes of MRI-guided surgeries and minimally invasive therapies. Within each clinical application, the successful development of interventional MRI demands innovative approaches, efficient use of MRI and computing technologies, and the integration of advanced therapy devices. This can only be accomplished with a multi-focus, crossdisciplinary effort aimed at developing and implementing these MR-guided interventions. It is important to combine the research and development with testing and evaluation of multiple technological advances related to the new, more advanced imaging platforms. We anticipate that further integration of intraoperative MRI guidance and computer-assisted surgery will greatly acceler-

ate the clinical utility of image-guided therapy in general and interventional MRI in particular. Based on the rapid advancement of technology, midand high-field-strength interventional magnets may become the standard in interventional medicine within the near future. In clinical practice, within the framework of multidisciplinary programs, the investigation of a wide range of interventional and surgical procedures is possible. However, the cost and technical support required for an intraoperative MRI system presently limits its use to only around a hundred sites worldwide. As new technology is developed, both researchers and clinicians must explore and refine intraoperative MRI, making it more cost-effective and widely accessible to a multidisciplinary community of users. 12.2.1.1  Field Strength and Clinical Applicability Interventional MRI was originally introduced in low- and mid-field systems. Most of these magnets had an open configuration allowing direct access to the patient during the procedures (Schenck et al. 1995). The introduction of higher field strength, closed bore systems represents a major paradigm shift from an open access application environment toward a high-field (up to 3 T) but limited-access setting. During the initial period of implementation of interventional MRI, two distinct types of paradigms were introduced. In the intraoperative paradigm and for percutaneous interventions when access to the patient was critical, some investigators applied a mid-field strength open magnet with a vertical or horizontal gap, allowing the patient to stay within the imaging field during the entire procedure. For non-invasive thermal ablations with focused ultrasound when direct access to the patient was not essential, only the table of a traditional close configuration 1.5-T magnet was modified. The field of intraoperative and interventional MRI has grown and changed significantly over the course of the last decade. It has become increasingly obvious that full access open MRI is not the only solution for guiding procedures. Indeed, numerous procedures require a higher field strength and more advanced image acquisition technology. For example, several research groups using intraoperative MRI for neurosurgery have moved to a higher field strength using closed-bore 1.5-T MRI systems (Hall et al. 2003; Nimsky et al. 2003; Sutherland et al. 2003). MRI-guided thermal ablations, interventional vascular applications, and prostate cancer brachytherapy have also been introduced at 1.5 T (Hirose et al. 2002; Susil et al. 2004). Proponents of the closed bore paradigm move most of the procedures (open surgeries, thermal ablations) to the advanced 1.5- to 3-T imaging platform to achieve faster and more flexible image acquisition, which in turn provides improved localization, targeting, monitoring,

1277

1278

12  Interventional MRI

and therapy control for those procedures adapted to this high-field application environment. There are both positive and negative consequences of this paradigm shift: The higher-field-strength system demands more advanced MRI technology, which in turn calls for a more flexible and cutting-edge imaging platform that can fulfill the complex needs of image-guided therapy. At the same time, relatively limited access to the patient requires innovative solutions for intraoperative guidance, and therapy methods need to be adapted for closed bore magnets. We believe that by developing novel imaging methods, guidance techniques and therapy devices, we can take advantage of the significantly improved MRI technology provided by an advanced higher field strength MRI system and that in parallel with this development, critical issues related to limited access will be resolved. In terms of selecting an existing design or developing a new magnet for interventions or intraoperative guidance, there are tradeoffs with respect to magnet configuration, field strength, and gradient strength. The higher signal-tonoise ratio (SNR) can be used to improve spatial and temporal resolution and can enable techniques such as temperature or flow-sensitive imaging, functional brain MRI, diffusion imaging, or MR spectroscopy for image guidance. Clearly, higher field systems are preferable for some intraoperative or interventional procedures. Although the cylindrical configuration of conventional high-field MR imaging systems precludes direct contact with the patient, such systems do provide high image quality and temporal resolution. The lack of direct access, however, prohibits real-time intraoperative imaging for open surgeries and certain minimally invasive procedures. Problems related to imaging artifact from devices used in the therapy are even more of an issue at higher field strengths. Nevertheless, infrequent image updates can be obtained during open procedures when the patient is moved in and out of the magnet. In addition, certain catheter-based applications and most thermal ablations can be monitored in real time within the closed bore of the high-field magnet. The current trend in intraoperative MRI is definitely toward more widespread use of higher-field magnets and more advanced imaging techniques presently not feasible for less expensive low and mid-field systems. 12.2.2  Interventional MRI and Computer-Assisted Surgery An essential component of MRI-guided therapy is computer technology for image processing, visualization, and navigation. It has been demonstrated that these tools can be integrated to provide more effective minimally invasive therapies. Surgical planning and intraoperative navigation systems have been used to deploy these technologies in clinical practice.

In current surgical practice, the localization of a lesion and the surrounding anatomy rely exclusively upon preoperative image data. Advanced computer technology has already provided neurosurgeons with the means to overcome the inherent inconvenience of frame-based systems by using neuronavigation systems (Gildenberg 1990). Although preoperative image data can be used for surgical planning and intraoperative navigational guidance, the utility of this information is limited because of the unavoidable deformation of the anatomy during surgery (Dorward et al. 1998, 1999; Drzymala and Mutic 1999; Hill et al. 1998; Yoo et al. 2004). Fortunately, intraoperative MRI can resolve this problem if appropriate navigational tools are implemented (Gering et al. 2001; Kansy et al. 1999; Nabavi et al. 2002; Samset and Hirschberg 1999; Samset et al. 2002, 2005; Wirtz et al. 1999a, b). Tracking is the process by which interactive localization is achieved within the patient’s coordinate system. Tracking systems that are used inside, or in the proximity of the MRI imaging volume are active or passive optical tracking, ultrasonic tracking, and electromagnetic tracking and MR tracking. Active optical trackers use multiple video cameras to triangulate the 3D location of flashing light-emitting diodes (LEDs) that can be mounted on any surgical instrument. Passive optical tracking systems use a video camera (or multiple video cameras) that emit infrared light and localize reflective markers that have been placed on surgical instruments. These systems do not use a cable attached to the handheld localizer. Unfortunately, both LED and passive vision localization systems require at least a partial line of sight between the landmarks or emitters and imaging sensor at all times when an object is tracked. Ultrasonic trackers work by transmitting pressure wave impulses that are being picked up by an ultrasonic microphone. Although they avoid line-of-sight issues, they do not provide simultaneous extra- and intra-corporal tracking. Two tracking systems have emerged that exploit the magnet and gradient hardware of the MRI scanner. MR tracking is a method in which small coils are being used in conjunction with a non-selective RF excitation (Bock et al. 2004; Dumouling et al. 1993; Wacker et al. 2004). During acquisition, readout gradients are applied and a peek signal can be measured corresponding to the position of the coil. This simple pulse sequence can be used with an arbitrary number of coils, and up to 32-channel tracking has been demonstrated. Electromagnetic tracking can be performed by measuring the current that is being induced in tracking coils as the image gradients are being activated. This signal is compared to the gradient driver signal, and with the help of a pre-acquired gradient map, converted to the position and orientation of the tracking sensor. This type of tracking does not require a special pulse sequence; it can be used with any imaging sequence that activates all gradients.

12.2  Clinical Applications of Interventional and Intraoperative MRI

Image-based tracking is also possible with exploitation of device artifacts or positive signal from devices (Omary et al. 2000). Out-of-plane tracking can be performed by interleaving imaging with projection or thick-slab scans in multiple angles. Surgical navigation based on intraoperatively updated image datasets is a powerful tool, and complements the ability to do real-time imaging during the treatment (Samset and Hirschberg 1999). However, there is still a need to integrate and navigate with pre-operatively acquired image information, such as information from CT, fMRI, PET, diffusion tensor MRI, or MRI spectroscopy (Alexander et al. 1995) (Fig. 12.2.1). Advanced MRI techniques can be difficult or impossible to employ intra-operatively because of limitations of the scanner or because they cannot be performed with an anaesthetized patient. In such cases, the solution may be to co-register pre-operative images with intraoperative images Gering et al. 2001). Several different algorithms have been developed for solving rigid-body registration problems (Chan et al. 2003; Cohen et al. 1995; Ferrant et al. 2002). The multiple data sets are aligned using a multimodal registration method based on the maximization of the inherent mutual information contained by the images originating

from the same patient. Image registration techniques have improved steadily in recent years, although most methods are confined to rigid structures. Clinical experience with image-guided therapy in deep brain structures and with large resections has revealed the limitations of existing rigid registration and visualization approaches. For example, the deformations of brain anatomy during surgery obviously require the application of non-rigid registration algorithms and the updating of anatomic changes using intraoperative imaging (Fig. 12.2.2). The work on non-rigid registration has produced a family of finite element-based algorithms using linear and non-linear elasticity behavior (Bharata et al. 2001; Ferrant et al. 2001). One example of the practical clinical application of this technology is the ability to provide real-time feedback regarding tissue deformation that occurs during surgery. This has resulted in the development of a non-rigid registration algorithm using a biomechanical model of the brain to support navigation during image-guided surgery. An ideal intraoperative MRI combines MRI with interactive localization of the surgical instruments, intraoperative displays, and computer workstations. However, software tools for visualizing the segmented models and the MRI scans in concert with the tracked instruments

Fig. 12.2.1  shows a result from analysis of diffusion tensor MRI (DT MRI) data. A sagittal image of fractional anisotropy (FA) is shown together with the result from DT MRI tractography. The trace lines from tractography have automatically been grouped and labeled according to major white matter fiber tracts, using spectral clustering and a high-dimensional atlas description of white matter (O’Donnell and Westin 2005). FA is high when the tensors are very anisotropic as in the white matter where the diffusion is restricted to only one direction, the direction of the axons (axial, sagittal, and coronal views bottom)

1279

1280

12  Interventional MRI

Fig. 12.2.2  Intrasubject alignment of preoperative structural MRI, functional MRI, and diffusion tensor MRI. This type of visualization enables ready appreciation of the 3D relationship between anatomical structures, the tumor, and critical fiber pathways

are needed to provide direct feedback to the surgeon. The 3D model of the patient must correlate directly to the actual images. The tracked probe enables us to depict the position of the probe relative to the segmented structures and to the original scan. The surgeon is thus equipped with an enhanced view of the surgical field relative to the entire anatomical model of the patient. This view can be further augmented with accurately registered preoperative functional and anatomical images using advanced deformable registration methods. 12.2.3  Interventional MRI and MR Imaging Techniques The development of MR imaging technology for interventional MRI has been closely related to the progress in fast imaging, interactive imaging, and dynamic imaging in the general field of MRI. Dynamic and interactive MRI research has been directed specifically at needs posed by the unique technological aspects of MR-guided therapy (Hoge et al. 2001). Not only has interventional MRI taken advantage of available methods from other MR imaging areas, but also imaging methods developed specifically for interventional MRI have become available for diagnostic purposes. The functionality of interventional MRI is unique, but the techniques developed for it can be used for other diagnostic applications. Because of the number of high-technology tools readily available during procedures, a premium is placed on the ability to combine the capabilities of imaging tools while preventing the technical aspects of use from becoming overwhelming to

the interventionalist. Technologically, the interventional setting calls for information integration and ease of use without information overload. The development of a dynamic imaging methodology known as “dynamically adaptive imaging” was driven by the unique demands of MRI-guided therapy. This research has progressed into novel image RF-based encoding techniques such as multi-resolution wavelet encoded MRI and SVD-encoded MRI (Panych et al. 1998; Shimizu et al. 1999). In parallel with the development of dedicated imaging platforms, real-time and adaptive imaging has also matured significantly. Several methods such as nonFourier imaging and reduced field-of-view imaging have significant potential for application in MRI-guided therapies (Kyriakos et al. 2000; Mitsouras et al. 2001, 2004). Concerted efforts are required to develop real-time platforms for visualization and control of data acquisition with dynamic imaging and interventional MRI (Panych et al. 1999, 2001; Zientara et al. 1998, 1999). These imaging platforms for real-time interactive and adaptive MRI facilitate research and development in all applications of MRI-guided therapy. It is obvious that the use of faster, more complex imaging techniques will improve image guidance. However, this technical challenge requires the development of a new imaging platform and, likewise, enhanced parallel MRI data acquisition methods. Since higher field strengths allow better localization, improved tissue characterization, and more exact targeting, an integrated, image-based system should be developed that allows more focal, MR-guided targeted interventions. Intraoperative MRI has a number of applications with specific requirements for dynamic MRI, particularly MR fluoroscopy, which can operate on an “open” MR scanner without extraordinary gradient coil or RF coil hardware parameters in place. The typical interventional openmagnet configuration disallows current dynamic MRI methods. For example, it cannot be outfitted with specialized gradient coils with the high slew rate necessary for echo planar–based fast MRI. Thus, the need exists for innovative, dynamic MRI approaches. The most important applications that have specific requirements for dynamic MR, particularly MR fluoroscopy, include (1) monitoring thermal therapies, (2) catheter tip tracking, and (3) monitoring the progress of surgical resections. With 2D fluoroscopic imaging capability, the surgeon can visually, or with computer-assistance, guide and monitor therapy or surgery for greater effectiveness and safety (Samset et al. 2001b). 12.2.4  MRI-Guided Thermal Ablations Thermotherapeutic tools have been in use for centuries. This type of treatment is compelling due to its minimally invasive nature, and it can also be a resort when surgery

12.2  Clinical Applications of Interventional and Intraoperative MRI

is not an option. Several systems exist to conduct thermal therapy, with different means of delivering or removing energy. Current heating modalities include RF ablation, microwave ablation, focused ultrasound, and laser ablation (Jolesz et al. 1988). Heating beyond 60°C produces permanent tissue damage due to denaturation of protein. Depending on tissue type, heating to lower temperatures may also be lethal. Tissue destruction may also be obtained by systems that reduce tissue temperature below the freezing point, enabling the creation of a cryolesion using Joule-Thomson engines or circulating cryogens. In cryosurgery, the tissue temperature is lowered so that cell membranes are ruptured or the cells are dehydrated. This will in turn lead to necrosis and inflammatory response. With the introduction of MRI as a monitoring method for thermal therapies, a novel mechanism for controlling energy deposition was developed (Jolesz et al. 2004; McDannold and Jolesz 2000; Samset 2006). Many MRI parameters are sensitive to temperature changes, making MRI suitable for monitoring thermal ablations non-invasively (Kuroda et al. 2000). Furthermore, the physician can take advantage of diffusion MRI, which detects changes in water mobility and compartmentalization and identifies reversible as well as irreversible thermally induced tissue changes. However, MRI monitoring of thermal ablations is only feasible if the imaging and therapy delivering systems are integrated. The role of MRI during thermal ablations is to monitor temperature levels, to restrict thermal coagulation to the targeted tissue volume, and to avoid ablation of normal

tissue. MRI can also detect irreversible tissue necrosis and demonstrate permanent changes within the treated tissue. Physiologic effects such as perfusion or metabolic response to elevated temperature can also be used for monitoring the ablation. Both flow and tissue perfusion can affect the rate and extent of energy delivery and the size of the treated tissue volumes. Monitoring can optimize treatment protocols (Fig. 12.2.3). Since the original description of MRI monitoring and control of laser–tissue interactions, MRI-guided interstitial laser therapy (ILT) and other MRI-guided thermal ablation methods have been clinically tested and accepted as minimally invasive treatment options (Jolesz et al. 1988; Kettenbach et al. 1998). ILT is a relatively simple, straightforward method, which can be well adapted to the interventional MRI environment. Overall, early results suggest that ILT is a safe therapy method (Kahn et al. 1994). Although no definitive conclusion can be drawn based on the currently available data, it appears that ILT can be of benefit in patients with low-grade gliomas (Hata et al. 1998). In malignant gliomas, thermal therapy has been essentially unsuccessful, a predictable outcome, since such tumors extend far beyond the area of MRI contrast enhancement. Cryosurgery garnered renewed interest in the 1990s partly because of the development of intraoperative ultrasound, and its use to monitor the freezing process. Advanced cryosurgical equipment was developed, involving a small probe driven by super cooled liquid nitrogen or high-pressure argon gas. A limitation of ultrasound imaging in this application is that the images of the cryole-

Fig. 12.2.3  Interstitial laser therapy of a brain tumor. a Temperature map computed by phase subtraction, b laser fiber inserted into the contrast-enhanced tumor

1281

1282

12  Interventional MRI

Fig. 12.2.4  Cryoablation of a liver metastasis using several cryoprobes (Silverman et al. 2005)

sion are distorted by artifacts. The most dominant artifact is the shadow distal to the frozen region (Gilbert et al. 1986; Tacke et al. 1999). MR imaging gives an excellent visualization of the frozen region. Due to the ultra-short T2* relaxation time of frozen tissue a sharp border can be seen between the non-frozen tissue and the signal void of the frozen tissue (Pease et al. 1995) (Fig. 12.2.4). Neither ultrasound nor MR directly gives information on sub-zero temperatures. The critical temperature assumed necessary for hepatic tumor ablation is –40°C (Gage and Baust 1998; Rubinsky et al. 1990). Above this limit, cell destruction cannot be guaranteed. The interesting isotherm surface for such procedures will thus be lower than the 0°C isotherm, which MR can give directly. This problem can be solved by estimating 3D temperature maps based on thermal models (Gilbert et al. 1997; Hong et al. 1994; Samset et al. 2001b, 2005b) or using customized pulse sequences to measure signal from super cooled water in the tissue (Daniel et al. 1999). The major applications of cryosurgery to date have included the treatment of prostatic cancer, kidney cancer, and liver metastasis (Silverman et al. 2000). 12.2.4.1  MR-Guided Focused Ultrasound Surgery One of the most promising methods of treatment with MRI-guided thermal ablation is centered on the use of non-invasive focused ultrasound surgery (FUS), which has been developed over the last decade (Jolesz and Hynynen 2002). Unlike the above-described probe-delivered thermal energy deposition methods, FUS uses extracorporeal acoustic energy for heating tissue only within the focal volume where most of the energy is absorbed. The probe-delivered thermal ablation methods deposit

energy at a large tissue volume within which there is a wide temperature gradient, with high temperature at the probe and lower temperature at the periphery. Due to this large gradient, the biological effects are also variable within the treated volume, and the boundary within which the tissue is irreversibly destroyed is, unfortunately, ill defined. In FUS, the treatment volume is small but the thermal gradient is very narrow, thus ensuring a more complete coagulation of the targeted tissue volume. Correct targeting is achieved by identifying the temperature elevations at the targeted tissue volume below the level of thermal coagulation (under 56°C) (McDannold et al. 2000, 2003) and, if the alignment is correct, repeating the sonication at a higher temperature to assure the desired tissue kill effect. MRI-based targeting and temperaturesensitive imaging provides a closed-loop control of this thermal ablation method (Cline et al. 1994; Mahoney et al. 2001; McDannold et al. 1998, 2001). The feasibility of MRI-guided FUS was originally demonstrated in a series of animal experiments (McDannold et al. 1999, 2002) and then in the treatment of benign fibroadenoma of the breast (Hynynen et al. 2001b). Clinical trials using the commercial system (Exablate 200 Insightec, Haifa, Israel) have been completed for the treatment of uterine fibroids (McDannold et al. 2003b, 2004a; Stewart et al. 2003; Tempany et al. 2003) (Fig. 12.2.5). Based on initial experience treating more than 1,000 women worldwide, it has been shown that the completely non-invasive technique of MRI-guided thermal ablation has significant potential for replacing invasive tumor surgery and radiosurgery. The feasibility of using this MRI-guided technique to effectively treat prostate cancer has also been demonstrated (Smith et al. 2001). The potential therapeutic use of ultrasound energy for intracranial pathology has long been acknowledged. There is no more convincing example of the benefits of FUS than in the brain, where deep lesions can be induced without any associated damage along the path of the acoustic beam. In the brain, where most injuries have detectable functional consequences, it is extremely important to limit tissue damage to the targeted area. This necessitates the use of an imaging technique for localization, targeting, real-time intraoperative monitoring, and control of the spatial extent (Vykkhodtseva et al. 2000). Indeed, by combining FUS with MRI-based guidance and control, we might well achieve complete tumor ablation without any associated structural injury or functional deficit. Since the skull scatters and attenuates the propagation of the ultrasound beam, most clinical trials have been performed following craniotomy in order to provide an ultrasound window. However, the transcranial application of FUS, although challenging, is not impossible (Hynynen and Jolesz 1998; McDannold et al. 2003). Although bone scatters and absorbs most of the acoustic energy, a small fraction can penetrate through the skull. Recent simulation and experimental studies have dem-

12.2  Clinical Applications of Interventional and Intraoperative MRI Fig. 12.2.5  Treatment planning and monitoring during FUS fibroid ablation. a,b Sonication sizes and location (green lines) were determined and thermal dose (yellow area) was displayed on top of planning images. c,d Estimation of thermal dose for each sonication (Tempany et al. 2003)

Fig. 12.2.6  FUS ablation through the intact skull. The development of large-scale, high-power phased arrays and the amplifiers to drive them has made it possible to correct the beam distortions induced by the skull (Connor et al. 2002)

onstrated the feasibility of accurately focusing ultrasound through the intact skull by using an array of multiple ultrasound transducers arranged over a large surface area (Clement and Hynynen 2000; Hynynen et al. 2004) (Fig. 12.2.6). To correct for beam distortion, the driving sig-

nal for the transducer elements of the array is individually adjustable, based on measurements obtained with an invasive hydrophone probe, or better, based on detailed MRI. Unfortunately, the skull thickness is uneven, causing variable delays of the acoustic waves originating from

1283

1284

12  Interventional MRI

individual phased array elements. Phase incoherence can be corrected, however, if the skull thickness is known from preoperative X-ray computed tomography scans. Because of the large surface area, the ultrasound energy is distributed in such a manner as to avoid heating and consequent damage of skin, bone, meninges, or surrounding normal brain parenchyma while at the same time coagulating the tissue at the focus. The experimental data are extremely promising and a clinical trial is in progress at Harvard Medical School. Based on these preliminary results, the thermal coagulation of brain tumors through the intact skull under MRI thermometry control using MR-compatible arrays appears feasible. Beyond thermal coagulation of tissue, FUS has various other effects that can be therapeutically exploited, and thus may open the way for potentially innovative vascular and functional neurosurgery applications as well as targeted drug delivery to the central nervous system. Among the most important is focused ultrasound’s ability to occlude vessels, which could make FUS a viable therapeutic tool for the treatment of vascular malformations (Hynynen et al. 1996). Groundbreaking studies also show that FUS can open the blood–brain barrier selectively without damaging the surrounding brain parenchyma (Hynynen et al. 2001a, 2003; McDannold 2004b). To achieve this effect, preformed gas bubbles must be introduced into the vasculature, as is routinely done with ultrasound contrast agents. The gas bubbles implode and release cavitation-related energy, which transiently inactivates the tight junctions. Consequently, large molecules can pass through the artificially created “window” in the blood–brain barrier. These large molecules can be chemotherapeutic or neuropharmacological agents. FUS-based, targeted selective drug delivery to the brain could result in novel therapeutic interventions for movement and psychiatric disorders. Such MRI-guided focal opening of the blood–brain barrier, combined with ultrasound technology that permits sonication through the intact skull, will open the way for new, non-invasive, targeted therapies. Specifically, it would provide targeted access for chemotherapeutic and gene therapy agents, as well as monoclonal antibodies, and could even provide a vascular route for performing neurotransplantations. 12.2.5  Current Clinical Applications of Interventional MRI There is no published number, but it can be estimated from the literature that various investigators have performed more than 25,000 MRI-guided procedures throughout the world. These have included body and brain procedures. MRI guidance has been provided for open surgeries as well as endoscopies (Hsu et al. 1998). Among them, the most predominant clinical applications are MRI-guided brain surgery, MRI-guided sinus endos-

copy, and MRI-targeted brachytherapy (D’Amico et al. 1998, 2000, 2003; Tempany et al. 1999). Several groups have performed various MRI-guided ablations using percutaneous procedures (laser, radiofrequency, microwave, and cryoablation) as well as non-invasive focused ultrasound methods for both malignant and benign tumors. This latter method appears to be a major advance and a so-called “disruptive technology” that may change the face of the entire surgical field. From a clinical standpoint, several teams have acquired substantial experience in intraoperatively guided prostate interventions (Hata et al. 2001a, b). In the development of each of these applications and with multiple clinical trials, it has been proven that MRI is a superior therapeutic tool for localization, targeting, monitoring, and therapy control. 12.2.5.1  Intraoperative Guidance for Neurosurgery The fundamental principle of image-guided neurosurgery is to target, access, and remove intracranial lesions, without injuring normal and functioning brain tissue or intact blood vessels. The overall concern is the preservation of neurological function, which requires precise delineation of functional anatomy and correct definition of tumor margins. During surgery, the visual appearance of the infiltrating malignant tumors is sometimes indistinguishable from that of adjacent normal brain tissue. Because of the difficulty in recognizing exact tumor margins, complete resection is, in most cases, problematic. For the neurosurgeon, the goal is to achieve complete tumor resection while maximally preserving normal brain tissue and function. Ideally, the physician should be able to precisely localize the lesion, choose the optimal trajectory of approach, and accurately determine the margins of the tumor and its separation from the normal brain. Using advanced computing technologies, these surgical steps are now undertaken with a remarkable degree of confidence and accuracy (Samset and Hirschberg 2002). Specifically, surgical planning uses multimodality and multiparametric lesion characterization, and it includes the full depiction of the relevant anatomical structures and their related functions. The 3D multimodality image fusions represent an enrichment of the information provided by the MRI slices alone. They do not change the diagnosis, but can contribute substantially to surgical planning by providing additional information regarding (1) the optimal craniotomy and corticotomy sites, (2) surgical excision margins, and (3) access trajectories to the targeted tumors. In addition to image-guidance, the most novel aspect of neurosurgical planning is the intraoperative use of the 3D model for interactive surgical simulation such as trajectory optimization, access route selection, and in the case of thermal therapy, 3D thermal dosimetry.

12.2  Clinical Applications of Interventional and Intraoperative MRI

Current advances in MRI, specifically functional MRI (fMRI) and diffusion tensor imaging (DTI) significantly improve localization and targeting within the cortical gray matter (functional anatomy) and along deep white matter structures (connectivity) (Jolesz et al. 2002; Maier et al. 2003; Park et al. 2003, 2004a, b; Ruiz-Alzola et al. 2002). In addition, contrast-enhanced dynamic MRI, magnetic resonance spectroscopy (MRS), positron-emission tomography (PET), and single-photon-emission computed tomography (SPECT) provide complementary physiologic and/or metabolic data, allowing further differentiation of brain tissue and improved characterization of brain tumors. Processing, registration, and intraoperative visualization of multi-modal imaging (anatomic, functional, diffusion tensor MRI, CT, ultrasound) and elec­tro­physio­log­ical monitoring results (intraoperative cortical stimulation, and electrocorticography) play a key role in identifying the relationship between tumor and functionally critical cortical and white matter structures (Talos et al. 2003b). Availability of accurate, multi-modal morphologic and functional information throughout the surgical procedure is a prerequisite for achieving the goal of maximally safe brain tumor removal while preserving the integrity of functionally critical cortical and white matter areas. Results to date include the successful use of preoperative functional MRI and DTI, integrated with intraoperative MR image updates and electrophysiological brain mapping for surgical guidance in brain tumor resections and biopsies (Talos et al. 2003a). To enable the use of preoperatively acquired multimodal images for surgical guidance, a mechanism cap­ able of compensating for brain shift must be put in place (Nimsky et al. 2000). Development and implementation of non-rigid registration algorithms for this purpose has continued to be a central research focus. The maximal preservation of normal tissue may contribute to decreased surgical morbidity. Specifically, intraoperative MRI can decrease surgical complications by identifying normal structures, such as blood vessels, white matter fiber tracts, and cortical regions with functional significance (Lunsford et al. 1996; Moriarty et al. 1996). However, intraoperative complications, such as hemorrhage, ischemia, or edema are possible and can directly affect the outcome. 12.2.5.2  Vascular Interventions under MRI Guidance Currently, cardiac and peripheral vascular interventions are typically guided by X-ray fluoroscopy. To depict the vessels, this imaging modality uses ionizing radiation and nephrotoxic iodine-based contrast, both of which may be harmful to the patient. Occupational health regulations limit the radiation exposure time of clinicians in a year,

and the heavy lead shielding worn by clinicians is apt to cause musculoskeletal problems. Furthermore, fluoro­ scopy yields only limited anatomic information. Use of MRI to guide such procedures eliminates the concerns of both exposure and contrast. In addition, MRI can depict the normal anatomy, target tissue, and the effect of the treatment via multi-slice and multi-planar capabilities. A recent clinical application that is currently being developed by several groups is cardiac electrophysiology ablation. For this treatment, areas of myocardial scar that contain the arrhythmia substrate can be directly identified on MRI (Soejima et al. 2002). Radiofrequency ablation lesions can be imaged with MRI in real time (Lardo et al. 2000), making it possible to determine the creation of lesions at target sites and the creation of continuous lines of ablation lesions. A crucial point is the visualization of the catheter. It is important to clearly depict the location of the ablation catheter relative to the abnormal regions and cardiac structures that often serve as borders for arrhythmia reentry circuits, such as the mitral or aortic valve annulus, pulmonary veins, and muscular ridges. Several groups have worked on image-based tracking of the catheter, which either assumes that the catheter is in the MRI scan plan or uses multiple projection images, thus reducing temporal resolution of the anatomical imaging (Kozerke et al. 2004; Omary et al. 2000; Serfaty et al. 2000). It has also been demonstrated that miniature MRI coils can be used to detect the position of these coils along the catheter. This technique requires specialized pulse sequences and occupies receive channels of the MRI system (Doumouling et al. 1993). Similarly, small coils can be used to track the position and orientation of a sensor mounted inside a catheter by measuring the current being induced in these coils from the imaging gradients. The deployment of MRI as an intraoperative imaging modality in vascular applications is currently limited by the availability of customized MRI compatible catheters. 12.2.6  Conclusion Since its introduction as a diagnostic tool in the mid1980s, MRI has evolved into the premier imaging modality. With the use of higher field magnets, we are now able to achieve spatial resolution of such superb quality that anatomy can be visualized in exquisite detail. The implementation of intraoperative MRI has had a major impact on conventional surgical approaches (Gould and Darzi 1997). Not only can we monitor brain shifts and deformations, but we can also achieve intraoperative navigation using intraoperative image updates (Jolesz 2003). Special features unique to MRI both enhance the information available during a procedure and guide therapy. These include flow sensitivity, evaluation of perfusion, administration of contrast agents to detect a breakdown of the

1285

1286

12  Interventional MRI

blood–brain barrier, or communication among CSF compartments. In addition, the ability of MRI to detect temperature changes can be exploited to monitor and control thermal ablations. Images acquired intraoperatively can also be combined and correlated with preoperatively acquired studies such as SPECT or PET scans, prior CT studies, diffusion tensor imaging, and functional MRI studies. The ultimate goal of intraoperative image guidance is to combine preoperative and intraoperative image data into a comprehensive “package” of information that will be indispensable to accurate surgical decision making. Indeed, this data package offers several benefits: (1) with intraoperative MRI, images can be obtained at each stage of a given procedure without moving the patient and without significantly extending the surgery; (2) a lesion can be accurately and, even more importantly, directly localized; (3) changes in the anatomy due to anatomical shifting can be immediately recognized; (4) the correlation between the surgeon’s field of view and the image allows confirmation of the exact location of pathologic tissue; and (5) serial images allow evaluation of the extent of excision and aid complete removal when possible (Nabavi et al. 2001). One of the most promising new therapeutic approaches under MRI guidance is FUS, allowing non-invasive thermal ablation of tumors. With the clinical introduction of such advances, intraoperative MRI is rapidly changing the face of surgery and interventional radiology today. In modern image-guided interventions, there is a growing need to image both targeted tissue and the effects of therapy. To achieve this, imaging must be comprehensively integrated with therapy and the various components of the operating environment must be completely reassessed. Successful integration entails the introduction of interactive dynamic imaging, high-performance computing, and real-time image processing in the operating room. Novel intraoperative imaging techniques are being aggressively developed and tested for their diagnostic and clinical utility. These techniques will be applied to several revolutionary image-guided therapy methods that are currently being explored and that will likely be incorporated into standard clinical practice eventually. The broad medical community has accepted the role of imaging in both diagnosis and therapy. Increasingly, minimally invasive procedures are viewed favorably, and there is a strong demand for their widespread implementation across numerous surgical disciplines. Nowhere is this demand more evident than in neurosurgery, where advances in intraoperative MRI and computing technology have marshaled in a new and exciting era in the treatment of brain tumors. We are especially encouraged by the ability of MRI-guided FUS to penetrate the blood– brain barrier selectively. This breakthrough technology holds great promise for a host of interventions, from vascular occlusion, to targeted drug delivery for cancer, Alzheimer’s disease, and epilepsy, to gene therapy.

Although radiology has combined imaging with various novel therapeutic methods, the full utilization of advanced imaging technology has not yet been accomplished. The current trend is to focus on the creation of integrated therapy delivery systems in which advanced imaging modalities are closely linked with high-performance computing. Obviously, the operating room of the future will accommodate various instruments, tools, and devices that are attached to the imaging systems and controlled by image-based feedback. We are confident that these innovative technologies, when applied in an integrated, multi-modality imaging environment, will produce a range of minimally invasive and non-invasive therapies for the brain as well as other organs and systems. References 1.

2.

3.

4.

5.

6.

7.

8.

9.

Alexander E III, Kooy HM, van Herk M, Schwartz M, Barnes PD, Tarbell N, Mulkern RV, Holupka EJ, Loeffler JS (1995) Magnetic resonance image-directed stereotactic neurosurgery: use of image fusion with computerized tomography to enhance spatial accuracy. J Neurosurg 83:271–276 Alexander E III, Moriarty TM, Kikinis R, Jolesz FA (1996) Innovations in minimalism: intraoperative MRI. Clini Neurosurg 43:338–352 Bharata A, Hirose M, Hata N, Warfield SK, Ferrant M, Zou KH et al (2001) Evaluation of three-dimensional finite element-based deformable registration of pre- and intra-operative prostate imaging. Med Phys 28:2560 Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ, Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997) Development and implementation of intraoperative magnetic resonance imaging and its neurosurgical applications. Neurosurgery 41:831–842; discussion 842–845 Bock M, Volz S, Zuhlsodorff S, Umathum R, Fink C, Hallscheidt P, Semmler W (2004) MR-guided intravascular procedures: real-time parameter control and automated slice positioning with active tracking coils. J Magn Res Imaging 19:580–589 Chan HM, Chung A, Yu S, Norbash A, Wells W (2003) Multi-modal image registration by minimizing KullbackLeibler distance between expected and observed joint class histograms. IEEE Conference on Computer Vision and Pattern Recognition, 2003 Clement GT, WJ, Hynynen K (2000) Investigation of a large-area phased array for focused ultrasound surgery through the skull. Phys Med Biol 45:1071–1083 Cline HE, Hynynen K, Hardy CJ, Watkins RD, Schenck JF, Jolesz FA (1994) MR temperature mapping of focused ultrasound surgery. Magn Reson Med. 31:628–636 Cohen DS, Lustgarten JH, Miller E, Khandji AG, Goodman RR (1995) Effects of coregistration of MR to CT images on MR stereotactic accuracy. J Neurosurg 82:772–779

12.2  Clinical Applications of Interventional and Intraoperative MRI 10. Connor CW, Clement GT, Hynynen K (200) A unified model for the speed of sound in cranial bone based on genetic algorithm optimization. Phys Med Biol 47:3925–3944 11. D’Amico AV, Cormack R, Tempany CM, Kumar S, Topulos G, Kooy HM, Coleman CN (1998) Real-time magnetic resonance image-guided interstitial brachytherapy in the treatment of select patients with clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 42:507–515 12. D’Amico A, Cormack R, Kumar S, Tempany CM (2001) Real-time magnetic resonance imaging-guided brachytherapy in the treatment of selected patients with clinically localized prostate cancer. J Endourol 14:367–370 13. D’Amico AV, Tempany CM, Schultz D, Cormack RA, Hurwitz M, Beard C, Albert M, Kooy H, Jolesz F, Richie JP (2003) Comparing PSA outcome after radical prostatectomy or magnetic resonance imaging-guided partial prostatic irradiation in select patients with clinically localized adenocarcinoma of the prostate. Urology 62:1063–1067 14. Daniel BL, Butts K, Block WF (1999) Magnetic resonance imaging of frozen tissues: temperature-dependent MR signal characteristics and relevance for MR monitoring of cryosurgery. Magn Reson Med.41:627–630 15. Dorward NL, Alberti O, Velani B, Gerritsen FA, Harkness WF, Kitchen ND, Thomas DG (1999) Postimaging brain distortion: magnitude, correlates, and impact on neuronavigation. J Neurosurg 88:656–662 16. Dorward NL, Alberti O, Palmer JD, Kitchen ND, Thomas DG (1990) Accuracy of true frameless stereotaxy: in vivo measurement and laboratory phantom studies. Technical note. J Neurosurg 90:160–168 17. Drzymala RE, Mutic S (1999) Stereotactic imaging quality assurance using an anthropomorphic phantom. Comput Aided Surg 4:248–255 18. Dumouling CL, Souze SP, Darrow RD (1993) Real-time position monitoring of invasive devices using magnetic resonance. Magn Reson Med 29:411–415 19. Ferrant M, Nabavi A, Macq B, Jolesz FA, Kikinis R, Warfield SK (2001) Registration of 3-D intraoperative MR images of the brain using a finite-element biomechanical model. IEEE Trans Med Imaging 20:1384–1397 20. Ferrant M, Nabavi A, Macq B, Black PM, Jolesz FA, Kikinis R, Warfield SK (2002) Serial registration of intraoperative MR images of the brain. Med Image Anal 26:337–359 21. Gage AA, Baust J (1998) Mechanisms of tissue injury in cryosurgery. Cryobiology 37:171–186 22. Gering DT, Nabavi A, Kikinis R, Hata N, O’Donnell LJ, Grimson WE, Jolesz FA, Black PM, Wells M III (2001) An integrated visualization system for surgical planning and guidance using image fusion and an open MR. J Magn Reson Imaging 13:967–75 23. Gilbert JC, Onik GM, Hoddick WK, Rubinsky B, Ferrell LD (1986) Ultrasound monitored hepatic cryosurgery: longevity study on an animal model. Cryobiology 23:277–285

24. Gilbert JC, Rubinsky B, Wong ST, Brennan KM, Pease GR, Leung PP (1997) Temperature determination in the frozen region during cryosurgery of rabbit liver using MR image analysis. Magn Reson Imaging 15:657–667 25. Gildenberg PL (1990) The history of stereotactic neurosurgery. Neurosurg Clin N Am1:765–780 26. Gould SW, Darzi A (1997) The interventional magnetic resonance unit–the minimal access operating theatre of the future? Br J Radiol 70(Spec no.):S89–S97 27. Hall WA, Liu H, Maxwell RE, Truwit CL (2003) Influence of 1.5-Tesla intraoperative MR imaging on surgical decision making. Acta Neurochir Suppl 85:29–37 28. Hata, N, Morrison, PR, Kettenbach, J, Kikinis, R, Jolesz, FA (1998) Computer-assisted intra-operative MRI monitoring of interstitial laser therapy in the brain: a case report. J Biomed Optics 3:304–311 29. Hata N, Jinzaki M, Kacher D, Cormak R, Gering D, Nabavi A, Silverman SG, D‘Amico AV, Kikinis R, Jolesz FA, Tempany CM (2001a) MR imaging-guided prostate biopsy with surgical navigation software: device validation and feasibility. Radiology 220:263–268 30. Hata N, Jinzaki M, Kacher D, Cormak R, Gering D, Nabavi A, Silverman SG, D’Amico AV, Kikinis R, Jolesz FA, Tempany CM (2001b) MR imaging-guided prostate biopsy with surgical navigation software: device validation and feasibility. Radiology 220:263–268 31. Hill DL, Maurer CR Jr, Maciunas RJ, Barwise JA, Fitzpatrick JM, Wang MY 1999 Measurement of intraoperative brain surface deformation under a craniotomy. Neurosurgery 43:514–26; discussion 527–528 32. Hinks RS, Bronskill MJ, Kucharczyk W, Bernstein M, Collick BD, Henkelman RM (1998) MR systems for imageguided therapy. J Magn Reson Imaging 8:19–25 33. Hirose M, Bharatha A, Hata N, Zou KH, Warfield SK, Cormack R, D’Amico A, Kikinis R, Jolesz FA, Tempany CMC (2002) Quantitative MR imaging assessment of prostate gland deformation before and during MR imaging-guided brachytherapy. Academic Radiology 9:906–912 34. Hoge WM, Miller EL, Lev-Ari H, Brooks DH, Karl WC, Panych LP (2001) An efficient region of interest acquisition method for dynamic magnetic resonance imaging. IEEE T Imag Proc 10:1118–1128 35. Hong JS, Wong S, Pease G, Rubinsky B (1994) MR imaging assisted temperature calculations during cryosurgery. Magn Reson Imaging 12:1021–1031 36. Hsu L, Fried MP, Jolesz FA (1998) MR-guided endoscopic sinus surgery. AJNR Am J Neuroradiol 19:1235–1240 37. Hynynen K, Jolesz F (1998) Demonstration of potential noninvasive ultrasound brain therapy through intact skull. Ultrasound Med Biol 24:275–283 38. Hynynen K, CV, Chung A, Jolesz FA (1996) Noninvasive artery occlusion using MRI guided focused ultrasound. Ultrasound Med Biol 22:1071–1077 39. Hynynen K, MN, Vykhodtseva N, Jolesz FA (2001a) Noninvasive MR imaging-guided focal opening of the bloodbrain barrier in rabbits. Radiology 220:640

1287

1288

12  Interventional MRI 40. Hynynen K, Pomeroy O, Smith DN, Huber PE, McDannold NJ, Kettenbach J, Baum J, Singer S, Jolesz FA (2001b) MR imaging-guided focused ultrasound surgery of fibroadenomas in the breast: a feasibility study. Radiology 219:176–185 41. Hynynen K, McDannold N, Vykhodtseva N, Jolesz FA (2003) Non-invasive opening of BBB by focused ultrasound. Acta Neurochir Suppl 86:555–558 42. Hynynen K, Clement GT, McDannold N, Vykhodtseva N, King R, White PJ, Vitek S, Jolesz FA (2004) 500-element ultrasound phased array system for noninvasive focal surgery of the brain: a preliminary rabbit study with ex vivo human skulls. Magn Reson Med. 200452:100–107 43. Jolesz, FA (1998) Interventional and intraoperative MRI: a general overview of the field. J Magn Reson Imaging 8:3–7 44. Jolesz FA (2003) Future perspectives in intraoperative imaging. Acta Neurochir Suppl 85:7–13 45. Jolesz FA Blumenfeld SM (1994) Interventional use of magnetic resonance imaging. Magn Reson Q 10:85–96 46. Jolesz FA, Talos IF (2004) MRI-guided thermal therapy for brain tumors. In: Proctor MR, Black PM (eds), Minimally invasive neurosurgery. Humana, Totowa, N.J. 47. Jolesz FA, Bleier AR, Jakab P, Ruenzel PW, Huttl K, Jako GJ (1988) MR imaging of laser-tissue interactions. Radiology 168:249–53 48. Jolesz FA, Hynynen K (2002) Magnetic resonance imageguided focused ultrasound surgery. Cancer J (Suppl 1): S100–S112 49. Jolesz, FA, Talos, IF, Schwartz, RB et al (2002) Intraoperative magnetic resonance imaging and magnetic resonance imaging-guided therapy for brain tumors. Neuroimaging Clin N Am 12:665–683 50. Kahn T, Bettag M, Ulrich F et al (1994) MRI-guided laserinduced interstitial thermotherapy of cerebral neoplasms. J Comput Assist Tomogr 18:519–532 51. Kansy K, Wisskirchen P, Behrens U, Berlage T, Grunst G, Jahnke M, Ratering R, Schwarzmaier HJ, Ulrich F (1999) LOCALITE—a frameless neuronavigation system for interventional magnetic resonance imaging systems. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 832–841 52. Kettenbach J, Kacher DF, Koskinen SK, Silverman SG, Nabavi A, Gering D, Tempany CM, Schwartz RB, Kikinis R, Black PM, Jolesz FA (2000) Interventional and intraoperative magnetic resonance imaging. Annu Rev Biomed Eng 2:661–690 53. Kettenbach J, Silverman SG, Hata N et al (1998) Monitoring and visualization techniques for MR-guided laser ablations in an open MR system. J Magn Reson Imaging 8:933–943 54. Kozerke S, Hegde S, Schaeffter T, Lamerichs R, Razavi R, Hill DL (2004) Catheter tracking and visualization using 19F nuclear magnetic resonance. Magn Reson Med 52:693–697 55. Kuroda K, Mulkern RV, Oshio K, Panych LP, Nakai T, Moriya T, Okuda S, Hynynen K, Jolesz FA (2000) Temperature mapping using water proton chemical shift: self-referenced method with echo planar spectroscopic imaging. Magn Reson Med 43:220–225

56. Kyriakos WE, Panych LP, Kacher DF, Westin C-F, Bao SM, Mulkern RV, Jolesz FA (2000) Sensitivity profiles from an array of coil for encoding and reconstruction in parallel (SPACE RIP). Magn Reson Med 44:301–308 57. Lardo AC, McVeigh ER, Jumrussirikul P et al (2000) Visualization and temporal/spatial characterization of cardiac radiofrequency ablation lesions using magnetic resonance imaging. Circulation 102:698–705 58. Lunsford LD, Kondziolka D, Bissonette DJ (1996) Intraoperative imaging of the brain. Stereotact Funct Neurosurg 66:58–64 59. Mahoney K, Fjield T, McDannold N, Clement G, Hynynen K (2001) Comparison of modeled and observed in vivo temperature elevations induced by focused ultrasound: implications for treatment planning. Phys Med Biol 46:1785–1798 60. Maier SE, Mamata H, Mulkern RV (2003) Characterization of normal brain and brain tumor pathology by chi-squares parameter maps of diffusion-weighted image data. Eur J Radiol 45:199–207 61. McDannold NJ, Jolesz FA (2000) Magnetic resonance image-guided thermal ablations. Top Magn Reson Imaging 11:191–202 62. McDannold N, Hynynen K, Wolf D, Wolf G, Jolesz F (1998) MRI evaluation of thermal ablation of tumors with focused ultrasound. J Magn Reson Imaging 8:91–100 63. McDannold NJ, Jolesz FA, Hynynen KH (1999) Determination of the optimal delay between sonications during focused ultrasound surgery in rabbits by using MR imaging to monitor thermal buildup in vivo. Radiology 211:419–426 64. McDannold NJ, King RL, Jolesz FA, Hynynen KH (2000) Usefulness of MR imaging-derived thermometry and dosimetry in determining the threshold for tissue damage induced by thermal surgery in rabbits. Radiology 216:517–523 65. McDannold NJ, Hynynen K, Jolesz FA (2001) MRI monitoring of the thermal ablation of tissue: effects of long exposure times. J Magn Reson Imaging 13:421–427 66. McDannold N, King RL, Jolesz FA, Hynynen K (2002) The use of quantitative temperature images to predict the optimal power for focused ultrasound surgery: in vivo verification in rabbit muscle and brain. Med Phys 29:356–365 67. McDannold N, Moss M, Killiany R, Rosene DL, King RL, Jolesz FA, Hynynen K (2003a) MRI-guided focused ultrasound surgery in the brain: Tests in a primate model. Magn Reson Med 49:1188–1191 68. McDannold N, Tempany CM, Stewart EA, Jolesz F, Hynynen K (2003b) Thermal dosimetry during MRI-guided focused ultrasound surgery in uterine fibroids. Eleventh Meeting of the International Society for Magnetic Resonance in Medicine Toronto, ON 2003b; 1207 Med 2004b; 51:913–923 69. McDannold N, Vykhodtseva N, Martin H, Jolesz F, Hynynen K (2004b) Investigation of the threshold for tissue damage in the rabbit brain using MRI-derived temperature information. Eleventh Meeting of the International Society for Magnetic Resonance in Medicine, Toronto, p 1208

12.2  Clinical Applications of Interventional and Intraoperative MRI 70. McDannold N, Tempany C, Stewart E, Jolesz FA, Hynynen K (2003b) MRI-based thermometry and thermal dosimetry during focused ultrasound thermal ablation of uterine leiomyomas. Ultrasonics Symposium 71. McDannold N, Vykhodtseva N, Jolesz FA, Hynynen K (2004a) MRI investigation of the threshold for thermally induced blood-brain barrier disruption and brain tissue damage in the rabbit brain. Magn Reson Med 51:913–923 72. Mitsouras D (2001) Near real-time 2D dynamic adaptive MRI using near-optimal spatial encoding. In Proc ISMRM Workshop on Minimum Data Acquisition Methods, Marco Island Fla, October 2001, pp 71–75 73. Mitsouras D, Hoge WS, Rybicki FJ, Edelman A, Zientara GP (2004) Non-Fourier-encoded parallel MRI using multiple receiver coils. Magn Res Med 52:321–328 74. Moriarty TM, Kikinis R, Jolesz FA, Black PM, Alexander E III. (1996) Magnetic resonance imaging therapy. Intraoperative MR imaging. Neurosurg Clin N Am 7:323–331 75. Nabavi A, Black PM, Gering DT, Westin CF, Mehta V, Pergolizzi RS Jr, Ferrant M, Warfield SK, Hata N, Schwartz RB, Wells WM III, Kikinis R, Jolesz FA (2001) Serial intraoperative magnetic resonance imaging of brain shift. Neurosurgery 48:787–797; discussion 797–798 76. Nabavi A, Gering DT, Kacher DF, Talos IF, Wells WM, Kikinis R, Black PM, Jolesz FA (2003) Surgical navigation in the open MRI. Acta Neurochir Suppl 85:121–125 77. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G, Fahlbusch R (2000) Quantification of, visualization of, and compensation for brain shift using intraoperative magnetic resonance imaging. Neurosurgery 47:1070–1079 78. Nimsky C, Ganslandt O, von Keller B, Fahlbusch R (2003) Preliminary experience in glioma surgery with intraoperative high-field MRI. Acta Neurochir Suppl. 88:21–29 79. O’Donnell L, Westin CF. White matter tract clustering and correspondence in populations (2005) In Eighth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI’05), Lecture Notes in Computer Science 3749 Palm Springs, Calif., pp 140–147 80. Omary RA, Unal O, Koscielsk DS, Frayne R, Korosec FR, Mistretta CA, Strother CM, Grist TM (2000) Real-time MR imaging-guided passive catheter tracking with use of gadolinium-filled catheters. J Vasc Interv Radiol 11:1079–1085 81. Panych LP, Zientara GP, Saiviroonporn P, Yoo SS, Jolesz FA (1998) Digital wavelet-encoded MRI: a new wavelet-encoding methodology. J Magn Reson Imaging 8:1135–1144 82. Panych LP, Zientara GP, Jolesz FA (1999) MR image encoding by spatially selective excitation: An analysis using linear systems models. Int J Imag Syst Technol 10:143–150 83. Panych LP, Zhao L, Jolesz FA, Mulkern FA (2001) Dynamic imaging with multiple resolutions along phase-encode and slice-select dimensions. Magn Reson Med 45:940–947 84. Park HJ, Kubicki M et al (2003) Spatial normalization of diffusion tensor MRI using multiple channels. Neuroimage 20:1995–2009

85. Park HJ, Mamata H, Talos IF, Yoo SS (2004a) Integration of diffusion tensor and functional MRI: a preliminary study on motor circuits (abstract). Human Brain Mapping Conference, Budapest, 2004 86. Park HJ, McCarley RW, Westin CF, Kubicki M, Talos IF, Brunn A, Pieper S, Kikinis R, Jolesz FA, Shenton ME (2004b) Method for combining information from white matter fiber tracking and gray matter parcellation. AJNR Am J Neuroradiol 25:1318–1324 87. Pease GR, Wong ST, Roos MS, Rubinsky B (1995) MR image-guided control of cryosurgery. J Magn Reson Imaging 5:753–760 88. Rubinsky B, Lee CY, Bastacky J, Onik G (1990) The process of freezing and the mechanism of damage during hepatic cryosurgery. Cryobiology 27:85–97 89. Ruiz-Alzola J, Westin CF, Warfield SK, Alberola C, Maier S, Kikinis R (2002) Nonrigid registration of 3D tensor medical data. Med Image Anal 6:143–161 90. Samset E (2006) Temperature mapping of thermal ablation using MRI. Minim Invasive Ther Allied Technol 15:36–51 91. Samset E, Hirschberg H (1999) Neuronavigation in intraoperative MRI. Comput Aided Surg 4:200–207 92. Samset E, Hirschberg H (2002) Stereotactic target localization accuracy in interventional magnetic resonance imaging. Stereotact Funct Neurosurg. 79:191–201 93. Samset E, Mala T, Edwin B, Gladhaug I, Soreide O, Fosse E (2001) Validation of estimated 3D temperature maps during hepatic cryo surgery. Magn Reson Imaging 19:715–721 94. Samset E, Talsma A, Kintel M, Elle OJ, Aurdal L, Hirschberg H, Fosse E (2002) A virtual environment for surgical image guidance in intraoperative MRI. Comput Aided Surg 7:187–196 95. Samset E, Høgetveit JO, ten Cate G, Hirschberg H (2005a) Integrated neuronavigation system with intra-operative image updating. Minim Invasive Neurosurg 48:73–76 96. Samset E, Mala T, Aurdal L, Balasingham I (2005b) Intraoperative visualisation of 3D temperature maps and 3d navigation during tissue cryo ablation. Comput Med Imaging Graph 29:499–505 97. Schenck JF, Jolesz FA, Roemer PB et al (1995) Superconducting open-configuration MR imaging system for image-guided therapy. Interv Radiol 195:805–814 98. Serfaty JM, Yang X, Aksit P, Quick HH, Solaiyappan M, Atalar E (2000) Toward MRI-guided coronary catheterization: visualization of guiding catheters, guidewires, and anatomy in real time. J Magn Reson Imaging 12:590–594 99. Shimizu K, Panych LP, Mulkern RV, Yoo S-S, Schwartz RB, Kikinis R, Jolesz FA (1999) Partial wavelet encoding: A new approach for accelerating temporal resolution in contrastenhanced MR imaging. J Magn Reson Imaging 9:717–724 100. Silverman SG, Tuncali K, Adams DF, vanSonnenberg E, Zou KH, Kacher DF, Morrison PR, Jolesz FA (2000) MR imaging-guided percutaneous cryotherapy of liver tumors: initial experience. Radiology 217:657–64 101. Silverman SG, Tuncali L, Morrison PR (2005) MR imaging-guided percutaneous tumor ablation. Acad Radiol 12:1100–1109

1289

1290

12  Interventional MRI 102. Smith NB, Merrilees NK, Dahleh M, Hynynen K (2001) Control system for an MRI compatible intracavitary ultrasound array for thermal treatment of prostate disease. Int J Hyperthermia 17:271–282 103. Soejima K, Stevenson WG, Maisel WH, Sapp JL, Epstein LM (2002) Electrically unexcitable scar mapping based on pacing threshold for identification of reentry circuit isthmus. Circulation 106:1678–1683 104. Stewart EA, Gedroyc WM, Tempany CM, Quade BJ, Inbar Y, Ehrenstein T, Shushan A, Hindley JT, Goldin RD, David M, Sklair M, Rabinovici J (2003) Focused ultrasound treatment of uterine fibroid tumors: safety and feasibility of a noninvasive thermoablative technique. Am J Obstet Gynecol 189:48–54 105. Susil RC, Camphausen K, Choyke P, McVeigh ER, Gustafson GS, Ning H; Miller RW, Atalar E, Coleman CN, Menard C (2004) System for prostate brachytherapy and biopsy in a standard 1.5 T MRI scanner. Magn Reson Med 52:683–687 106. Sutherland GR, Kaibara T, Louw DF (2003) Intraoperative MR at 1.5 Tesla—experience and future directions. Acta Neurochir Suppl. 85:21–28 107. Tacke J, Speetzen R, Heschel I, Hunter DW, Rau G, Gunther RW (1999) Imaging of interstitial cryotherapy–an in vitro comparison of ultrasound, computed tomography, and magnetic resonance imaging. Cryobiology 38:250–259 108. Talos IF and Black PM (2003a) MRI-guided stereotactic biopsy in the CNS. In: D’Amico A, Loeffler J (eds) Image-guided diagnosis and treatment of cancer. Humana, Totowa, N.J., pp 111–116 109. Talos IF, O’Donnell L, Westin CF et al (2003b) Functional and diffusion tensor MRI fusion with anatomic MRI for image-guided neurosurgery, lecture notes in computer science. Springer, Berlin Heidelberg New York, pp 407–415 110. Tempany C, D’Amico A, Cormack R, Kumar S, Silverman S, Jolesz F (1999) MR-guided prostate brachytherapy: a new approach to seed implantation. New England Section of the American Urological Society, Beverly, Mass.

111. Tempany CM, Stewart EA, McDannold N, Quade BJ, Jolesz FA, Hynynen K (2003) MR imaging-guided focused ultrasound surgery of uterine leiomyomas: a feasibility study. Radiology 226:897–905 112. Tronnier VM, Wirtz CR, Knauth M et al (1997) Intraoperative diagnostic and interventional magnetic resonance imaging in neurosurgery Neurosurgery 40:891–900; discussion 900 113. Vykhodtseva N, Sorrentino V, Jolesz FA, Bronson RT, Hynynen K (2000) MRI detection of the thermal effects of focused ultrasound on the brain. Ultrasound Med Biol 26:871–80 114. Wacker FK, Elgort D, Hillenbrand CM, Duerrk JL, Lewin JS (2004) The catheter-driven MRI scanner: a new approach to intravascular catheter tracking and image-parameter adjustment for interventional MRI. AJR 183:391–395 115. Wirtz CR, Bonsanto MM, Knauth M et al (1997a) Intraoperative magnetic resonance imaging to update interactive navigation in neurosurgery: method and preliminary experience. Computer Aided Surg 2:172–179 116. Wirtz CR, Tronnier VM, Bonsanto MM et al (1997b) Image-guided neurosurgery with intraoperative MRI: update of frameless stereotaxy and radicality control. Stereotact Funct Neurosurg 68:39–43 117. Yoo SS, Talos IF, Golby A, Black PM, Panych LP (2004) Evaluating requirements for spatial resolution of fMRI for neurosurgical planning. Human Brain Map 21:34–43 118. Zientara GP, Panych LP, Jolesz FA (1998) Applicability and efficiency of near-optimal spatial encoding for dynamically adaptive MRI. Magn Reson Med 39:204–13 119. Zientara GP, Panych LP, Jolesz FA (1999) Near-optimal spatial encoding for dynamically adaptive MRI: mathematical principles and computational methods. Intl J Imaging Syst Technol 10:151–165

Chapter 13

Functional MRI

13

13.1 

Basics  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1292 L.R. Schad

13.1.1 

Introduction  .. . . . . . . . . . . . . . . . . . . . . . . .   1292

13.1.6.3  Quantification   of the Stimulation Effect  . . . . . . . . . . . . . .   1301

13.1.2 

Physiology  . . . . . . . . . . . . . . . . . . . . . . . . . .   1292

13.1.6.4  Artifacts and Motion Correction  . . . . . .   1302

13.1.3 

Contrast  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1292

13.1.7 

13.1.6.2  Statistical Methods  .. . . . . . . . . . . . . . . . . .   1299

13.1.3.1  BOLD Contrast  .. . . . . . . . . . . . . . . . . . . . .   1292

Comparison of Methods  .. . . . . . . . . . . . .   1303 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1306

13.1.3.2  Field Strength Dependency of the Contrast  .. . . . . . . . . . . . . . . . . . . . . .   1294

13.2 

Clinical Applications  . . . . . . . . . . . . . . . .   1308 K.K. Peck and A.I. Holodny

13.1.3.3  Alternative Contrast  .. . . . . . . . . . . . . . . . .   1294

13.2.1 

Introduction   . . . . . . . . . . . . . . . . . . . . . . . .   1308

Measuring Techniques  .. . . . . . . . . . . . . . .   1294

13.2.2 

Normal Functional Anatomy  .. . . . . . . . .   1308

13.1.4 

13.1.4.1  GE Technique (FLASH)  . . . . . . . . . . . . . .   1294

13.2.2.1  Sensory–Motor  . . . . . . . . . . . . . . . . . . . . . .   1308

13.1.4.2  Echo Planar Imaging  . . . . . . . . . . . . . . . . .   1295

13.2.2.2  Language  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   1308

13.1.4.3  Turbo Spin Echo   . . . . . . . . . . . . . . . . . . . .   1295

13.2.2.3  Memory  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1309

13.1.4.4  Blood-Bolus Tagging   (STAR Technique)   .. . . . . . . . . . . . . . . . . .   1295

13.2.3 

13.1.5 

Sequence Optimization  .. . . . . . . . . . . . . .   1295

13.1.5.1  FLASH  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1296 13.1.5.2  EPI  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1298 13.1.5.3  TSE  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1298 13.1.5.4  STAR  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1298 13.1.6 

Data Analysis  .. . . . . . . . . . . . . . . . . . . . . . .   1298

13.1.6.1  Subtraction Method  .. . . . . . . . . . . . . . . . .   1298

Functional MRI in Disease   . . . . . . . . . . .   1309

13.2.3.1  Paradigm Selection   . . . . . . . . . . . . . . . . . .   1309 13.2.3.2  Patient Preparation  .. . . . . . . . . . . . . . . . . .   1312 13.2.3.3  fMRI Consideration in Patients  . . . . . . .   1312 13.2.4 

Specific Pathology  .. . . . . . . . . . . . . . . . . . .   1314

13.2.4.1  Tumor  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1314 13.2.4.2  Stroke  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1316 13.2.4.3  Epilepsy  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1318 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1318

1292

13  Functional MRI

13.1  Basics L.R. Schad 13.1.1  Introduction Functional MRI (fMRI) allows non-invasive indirect measurement of neuronal activity and imaging of activated cortical areas. Measurements are based on the fact that brain stimulation is correlated with an increased local brain metabolism. This metabolic activity causes local changes of the magnetic properties of blood, which can be imaged by fMRI due to a hemodynamic effect (changes in blood flow and blood volume). The technique of visualizing activated areas of the visual cortex was introduced in 1991 (Belliveau et al. 1991), when changes of blood perfusion were detected by firstpass techniques using intravenous contrast agents during optical stimulation. Since then, fMRI methods where blood acts as an intrinsic contrast agent have been developed with great interest. Early animal studies performed in 1990 showed that changes in blood oxygenation are correlated with changes in image contrast (Ogawa et al. 1990a, b). These changes are related to the magnetic properties of blood, where local blood oxygenation is the most important parameter influencing signal intensity. The first in vivo experiments based on this method were performed with specially equipped EPI tomographs or using special high-field scanners (Kwong et al. 1992; Blamire et al. 1992; Bandettini et al. 1992; Ogawa et al. 1992; Turner et al. 1993). Shortly after these experiments, it was shown that this effect of image contrast could also be reached by conventional gradient-echo techniques (Frahm et al. 1992). Since then the method of fMRI has been further developed and optimized on standard clinical scanners operating at 1.5 T (Schad et al. 1993; Constable et al. 1993; Connelly et al. 1993). Different paradigms of cerebral activation can be detected by fMRI effects in different cortical areas. Now fMRI is being used in basic research but is beginning to come into clinical use with different applications. Promising applications of fMRI are pre-therapeutical localization of critical cortical areas and therapy monitoring after ischemic infarction, craniocerebral injury, or drug therapy. While in the beginning of fMRI, methodological developments were focused on the visual and sensorimotor cortex, complex paradigms have since been developed and successfully tested to activate not only primary and secondary areas of the brain, but also subcortical areas and the cerebellum. In this chapter, the physiological and physical basics of functional magnetic resonance imaging will be described followed by exemplary clinical applications.

13.1.2  Physiology Stimulation of a brain area, i.e., motor cortex, by finger tapping or other stimuli results in a local increase of metabolic activity. The energy consumption of neurons is balanced by a chemical reaction where adenosine triphosphate (ATP) is converted into adenosine diphosphate (ADP). ADP can be transformed back into ATP under oxygen consumption, and free ADP molecules act as vasodilators for the capillaries surrounding a neuron. Consequently, an increase of the relative cerebral blood volume (rCBV) and the relative blood flow (rCBF) can be detected. With the help of PET studies, this leads to a local increase of rCBV and rCBF of more than 30% (Fox and Raichle 1986). The oxygen consumption for restoring of the ATP molecule is balanced by deoxygenation of blood’s hemoglobin, where diamagnetic oxyhemoglobin is transformed into paramagnetic deoxyhemoglobin. This process of increased local deoxygenation in the brain yields only about 5%, i.e., the main entrée of oxygen for neurons in activated cortical areas is dominated by increasing cerebral blood flow. Typical time scales for changing rCBV and rCBF are in the order of seconds. 13.1.3  Contrast 13.1.3.1  BOLD Contrast Functional magnetic resonance imaging uses the hemodynamic effect of increasing rCBF and rCBV with decreasing deoxyhemoglobin concentration for imaging activated cortical areas. In the majority of clinical studies to date, the techniques used have been based on the change in deoxyhemoglobin concentration, i.e., dependent on the oxygen saturation (BOLD: blood oxygen–level dependent). The so-called BOLD contrast was detected in animal studies in 1990 for the first time using high-field scanners operating at 7 T. The BOLD contrast also changes the local spin-spin relaxation time T2* because of the different magnetic properties of oxy- and deoxyhemoglobin. Two antidromic processes are responsible for the change in deoxyhemoglobin concentration during cortical activation: On the one hand the absolute amount of deoxyhemoglobin increases in the capillaries, and on the other hand an overcompensation of the oxygen extraction takes place because of an increase in CBF and rCBV, resulting in an effective reduction of the deoxyhemoglobin concentration in the blood. Because deoxyhemoglobin has paramagnetic properties, additional magnetic field gradients are created, resulting in larger local differences of the magnetic field in contrast to fully oxygenated oxyhemoglobin with diamagnetic properties. These differences in the local magnetic field are the dominating factor in fMRI and appear between the inner and outer

13.1  Basics

space of the red blood cells and also between the blood vessels and the brain parenchyma. The diffusion of water molecules through these magnetic field gradients leads to a change of the resonance frequency of water protons, resulting in a phase shift. This results in an additional dephasing of spins and shortens the effective spin–spin relaxation time T2*. Thereby the decrease in T2* depends on the strength of local field gradients and the time interval where the water protons “see” this additional magnetic field. This means that the reduction of deoxyhemoglobin concentration during brain stimulation creates a more homogeneous local magnetic field, which can be detected by a prolongation of the T2* time. Therefore measurement sequences with high sensitivity to susceptibility variations, i.e., gradient echo techniques like FLASH or EPI, are an ideal tool for imaging activated cortical areas since they are very sensitive to T2* variations. Using these measuring techniques a signal increase can be detected in activated brain regions in susceptibility-weighted MRimages under stimulation. Since draining venous vessels are also loaded by decreased deoxyhemoglobin concentration, they appear likewise brighter on susceptibilityweighted fMRI during stimulation. The influence of all these parameters (rCBV, rCBF, hematocrit value, and oxygen partial pressure in arterial and venous vessels) is very complex, but only very simple theoretical models exist, with different assumptions describing the effect on the BOLD contrast (Boxerman et al. 1995; Yablonskiy and Haacke1994; Kennan et al. 1994; Ogawa et al. 1993). In the simplest case, blood vessels are described by cylinders with a length of infinity where the susceptibility of these cylinders is different from that of the surrounding tissue (Fig. 13.1.1). Under this assumption the change in the frequencies of protons in and out of the cylinder can be calculated by, in the cylinder

Example. The typical susceptibility difference between fully oxygenated and fully deoxygenated blood in a vessel perpendicular to the main magnetic field is about ∆χ = 0.08 ppm (Thulborn et al. 1982), assuming an oxygen saturation Y of 0.6 at a main magnetic field of 1.5 T, with a frequency shift at the vessel’s surface (r = a) of about 13 Hz (~0.2 ppm). Note. In Eq. 13.1.2 the term (1–Y) ∆χ describes the susceptibility difference of blood as a function of oxygen saturation Y, which is dependent on the oxygen partial pressure pO2 given by

with p50 the oxygen partial pressure where half of the oxygenbinding positions of the hemoglobin molecules are occupied.

Numeric simulations using the cylinder model with diffusion effects (random walk) result in different dependencies of the transversal relaxation rate 1/T2* with respect to field strength. Different factors are responsible for this dependency: vessel diameter, strength of gradients and walking distance of the protons in the time interval between HF pulse and readout of SI. In case of small vessels (capillary radius < 8 µm) theoretical models using fast diffusion (“fast exchange” [Gillis and Koenig 1987]) can be used. Under special assumptions with respect to the structure of the field inhomogeneities and the walk of the protons (protons “see” the whole range of the field inhomogeneities), a quadratic dependency of the relaxation rate from the resonance frequency can be calculated. For large vessels where the walk of the protons is typically small with respect to the vessel diameter (“slow exchange” [Majumdar and Gore 1988]) field inhomogeneities can be described basically as static gradient fields. In this case, a linear dependency of change of the relaxation rate from resonance frequency can be calculated.

(13.1.1)

out of the cylinder (13.1.2) with

ωB being the change of resonance frequency, ω0 the resonance frequency of the main magnetic field, ∆χ the maximum of susceptibility difference between vessel and surrounding tissue, Y the oxygen saturation of blood (0 = completely deoxygenated, 1 = completely oxygenated), a the vessel radius, r distance between vector position and cylinder axis, φ the angle between position r and normal component of the main magnetic field to the cylinder axis, and θ the angle between cylinder axis and main magnetic field.

Fig. 13.1.1  Schematic illustration of the geometric assumptions of Eq. 13.1.1. a vessel radius, r distance between vector position and cylinder axis, φ angle between position r and normal component of the main magnetic field to the cylinder axis, θ angle between cylinder axis and main magnetic field

1293

1294

13  Functional MRI

Estimating the dependency of the resonance frequency by using the change of frequency at the surface of the blood vessel: ν = γ · B0 · Δχ · (1 – Y) ,

(13.1.3)

a simple model containing volume elements with tissue and capillaries/venules can be used for calculating the dependencies of the effective transversal relaxation rate: 1 / T2* ∝ ν ∙ bL for large vessels,

(13.1.4)

1 / T2* ∝ ν2 ∙ (bS)γ ∙ p for capillaries, with γ ν bL bS p

being the model specific constant, the frequency shift at the surface of the blood vessel, the fraction of large venous vessels in the volume of interest, the fraction of capillaries in the volume of interest, and the fraction of “active” capillaries of all capillaries in the volume of interest.

13.1.3.2  Field Strength Dependency of the Contrast With respect to Eq. 13.1.3, the susceptibility effect causes a shift of the resonance frequency at the surface of the blood vessel, which is proportional to the main magnetic field B0 so the change of the effective transversal relaxation rate can be calculated using Eq. 13.1.4: 1 / T2* ∝ B0 for large vessels, and

(13.1.5)

1 / T2* ∝ B02 for capillaries. Example. A field strength of 4 T results in a 7.1-fold change of the relaxation rate for capillaries compared to a only 2.7-fold effect at 1.5 T.

Under this assumption, a field strength dependency of the stimulation effect is estimated with linear to quadratic correlation with the main magnetic field. At high field strength, a separation of the stimulation effect is possible because of the quadratic increasing component of the capillaries and a differentiation of activated brain parenchyma and macroscopic draining veins can be performed. This is confirmed by experimental studies where field strength dependencies have been measured at 1.5 T and 4 T, resulting in exponents between 1.6 and 2 (Turner et al. 1993). 13.1.3.3  Alternative Contrast The first fMRI measurements showing functional SI changes were performed with administration of contrast

agents (Gd-DTPA). Nowadays this technique has no relevance in fMRI because of the possibility of measuring non-invasively by using the BOLD effect and the advantage of instant repetition. By using different measuring parameters, distinct properties of the stimulation effect can be weighted into fMRI, i.e., the inflow effect can be pronounced on flowsensitive sequences because an increase in blood flow is correlated with higher perfusion. This technique has been called FOLD (flow level dependent contrast) in the literature and can be designed either by a special choice of measuring parameters in conventional sequences or by HF preparation of inflowing spins into the slice of interest (STAR-technique [Edelman and Siewert 1994]). 13.1.4  Measuring Techniques The BOLD-contrast measuring techniques are based on repetitive measurements where the MR SI differences are caused by the susceptibility difference between oxygenated and deoxygenated blood. Different requirements need to be met by scanners (long-term stability, homogeneity of the main magnetic field by shimming) and by the measuring technique (sequence optimization for high sensitivity of small susceptibility changes). Below, these requirements are discussed in detail for different measuring sequences and a further description of the sequences’ structures can be found in Chap. 2. 13.1.4.1  GE Technique (FLASH) fMRI performed on clinical scanners is often based on gradient echo techniques like FLASH (Haase et al. 1986). The advantages of FLASH sequences are availability, simplicity, high signal-to-noise ratio, and high spatial resolution. On the other hand, a measuring time of seconds for this technique is the limiting factor for the temporal resolution of the BOLD effect. This results in a separation of activated and non-activated phases in the design of a stimulation experiment and does not allow measurement of the transition between these two states at higher resolution. In practice, because of the relatively long measurement time, a compromise has to be found between the number of measured slices and the number of stimulation cycles (increasing the statistical stability). Also, because of the long acquisition time, small periodical signal intensity variations are detectable using FLASH techniques originating from breathing and the cardiac cycle (liquor pulsation). These signal variations are averaged during the long acquisition of the FLASH technique but can also contribute to the measured signal intensities. Correction strategies are being developed; the published methods use reference echoes for correction of periodical signal modulations (Hu and Kim 1994). On the other

13.1  Basics

hand, one of the most important advantages of FLASH techniques is their insensitivity to image distortions, i.e., functional parameter images measured by FLASH techniques correlate well with morphological images and can be directly superimposed, resulting in a clear depiction of activated areas. 13.1.4.2  Echo Planar Imaging Echo planar imaging (EPI) methods allow acquisition of one image within subseconds, since the total image data are acquired after a single HF excitation. Because the BOLD effect is measured in seconds, additional slices can be acquired sequentially in between consecutive HF excitations of the same slice. Keeping total measuring time constant the use of EPI techniques allows a higher temporal resolution for each slice and the acquisition of additional slices, which increase the statistical stability in data evaluation. Thereby, as with FLASH sequences, slice thicknesses are generally larger because of lower signal-to-noise ratio, and neighboring slices should not overlap due to slice profiling. EPI images are insensitive to periodical physiological activities like breathing or the cardiac cycle because of the very fast data acquisition. On the other hand, inhomogeneities of the main magnetic field can cause image distortions on fMRI since all the data for an image are acquired after a single HF excitation. Therefore, the correlation of activated cortical areas with the underlying anatomy is difficult, and direct superimposition on corresponding morphological images is not possible without correction of image distortions. Methods for correction of image distortions are experimental and under development for routine use (Jezzard and Balaban 1995). 13.1.4.3  Turbo Spin Echo The use of turbo spin-echo (TSE) techniques was proposed in the beginning of fMRI (Constable et al. 1994) but is still experimental. TSE sequences start with an initial HF pulse followed by a series of spin echoes, which are used to measure a set of lines of the Fourier space (RARE [Hennig et al. 1986]). Based on the diffusion effect, the spin-echo technique allows discrimination of SI differences between capillaries and larger vessels. Thus the magnetic field of large vessels can be described by an extended linear field gradient (see Sect. 13.1.3.1), where the phase dispersion of the spins can be rephased if the timing between successive echoes is short enough. This results in fMRI data where only capillaries contribute to the SI changes. In typical experiments these changes are very small (~2%), resulting in high stability of the scanner and very good patient fixation since small head movements can delete the effect. Because of these problems, the TSEfMRI technique is rarely used in clinical practice.

13.1.4.4  Blood-Bolus Tagging (STAR Technique) The acronym STAR (signal targeting with alternating radiofrequency) was introduced in 1994 (Edelman and Siewert 1994). The idea of this measuring technique is tagging of a blood bolus by a 180° inversion pulse followed by its sequential detection in a neighboring slice of interest. The time delay between tagging and readout of the blood bolus can be varied in the measurements, with values of 800 to 1200 ms for a typical fMRI experiment. A further reduction of signal contributions of the stationary tissue in the slice of interest can be achieved by subtracting two identical measurements with and without inversion pulse. Data readout is based either on a segmented FLASH sequence with flow compensation or on an EPI technique. The inversion of the magnetization is performed with flip angles of varying algebraic signs leading to a nearly doubled SI of the tagged blood bolus created by its transversal magnetization. Furthermore, by complex adding of SI of both measurements an excellent suppression of stationary tissue is obtained. Using this technique inflow effects in fMRI experiments can be measured (FOLD contrast). In addition fMRI at low field strength (≤ 1 T) seems to be possible but still needs to be investigated. 13.1.5  Sequence Optimization Techniques based on the BOLD contrast are detecting the influence of small susceptibility differences on the MR signal caused by oxygenated and deoxygenated blood. Gradient-echo techniques (FLASH or EPI) are the method of choice because of their high sensitivity to susceptibility effects. In contrast to conventional imaging where low sensitivity to susceptibility differences, short echo, and repetition times resulting in short measuring times are required, measuring techniques in fMRI have to be optimized to maximal sensitivity to small susceptibility changes caused by the BOLD effect. In consequence, sequence optimization should try all possibilities for maximizing the effect of susceptibility-related dephasing of the spins. Possible sequence parameters are echo time, slice thickness, field-of-view (FOV), and matrix size. The size of voxel volume is dependent on the choice of resolution of anatomic details for the measurement. Thereby the measurable stimulation effect depends on the ratio between stimulated and non-stimulated tissue in the voxel. This partial volume effect reduces the measurable stimulation effect at large voxel size (Schad et al. 1994). With respect to sequence parameters, the only possibility for increasing the sensitivity to spin dephasing is a prolongation of the echo time. Thereby, the measurable MR signal at readout time drops down exponentially according to the relaxation mechanism, which is the limiting factor of the echo timing.

1295

1296

13  Functional MRI

For sequence optimization, an important aspect is the optimal echo time used in the experiments. The aim of sequence optimization should focus on a maximum signal difference in fMRI images between stimulated and non-stimulated periods. With respect to echo formation, FLASH and EPI techniques belong to the same family of gradient-echo sequences. Optimization of echo time will be described in the following by using the FLASH sequence. An analogous estimation can be performed for the EPI fMRI technique. 13.1.5.1  FLASH Estimation of the optimal echo time is performed by calculating the SI of the transversal magnetization of the measured slice under steady state condition: S = α exp (–TE / T2*),

(13.1.6)

where all parameters that are independent of T2* were collected in the constant a according to

(13.1.7)



(13.1.9)

If T2r * and T2s * are in the same order of magnitude, it follows that TEopt ≅ T2r* ,

(13.1.10)

i.e., the optimal echo time is comparable to the effective transversal relaxation time. Example. Evaluation of the optimal echo time has been performed in a volunteer at 1.5 T, where the effective transversal relaxation time of the gray matter in the motor cortex area has been determined under stimulated and non-stimulated conditions (finger tapping). Echo times TE ranging from 15 to 95 ms in increments of 10 ms have been selected for the stimulation experiments. The results of the measurement are demonstrated in Figure 13.1.2, where the error bars in the diagram represent the standard deviation of the mean value of the signal. Errors of the evaluation result from errors of the least-squares fit. Differences of the measured effective transversal relaxation time between stimulation and non-stimulation are not very pronounced with only 10 ms. Optimal echo time results in about 70 ms.

with

ρ(x) being the proton density at x, a the flip angle, TR the repetition time, T1 the long relaxation time.

Signal differences of voxel elements in FLASH images measured during stimulated and non-stimulated conditions are: ΔS = as exp (–TE / T2s *) – ar exp (–TE / T2r *),

(13.1.8)

with as the constant for images during stimulation, ar the constant for images during non stimulation, T2s * the effective transversal relaxation time during stimulation, and T2r * the effective transversal relaxation time during non-stimulation. Thereby, the quantities ar and as describe the influence of parameters that are not dependent on T2* during stimulated and non-stimulated periods. Under the assumption that the BOLD contrast is influenced only by susceptibility effects, i.e., that SI changes related to inflow effects are negligible, the optimal echo time where the signal differences of Eq. 13.1.8 between stimulated and non-stimulated images measured in fMRI are maximized is given by:

Fig. 13.1.2  Determination of the effective transversal relaxation time T2* of gray matter of the motor cortex. Echo times of a FLASH sequence ranging from 15 to 95 ms (in increments) have been acquired during stimulation and non-stimulation. The effective transversal relaxation time T2* under stimulated and non-stimulated conditions has been calculated by mean values of signal intensities of a region-of-interest using a leastsquare-fit according to S(TE) = a*exp(–TE/T2*). The dashed line represents the curve fitting of the measurement under nonstimulated conditions (squares), while the dotted line represents the measurement under stimulation (triangles). Measurements have been performed using a FLASH sequence (flip angle = 40°, TR = 150 ms, TH = 3 mm, FOV = 200 mm)

13.1  Basics

The amplitude of the stimulation effect measured by MRI is strongly dependent on the chosen echo time. Using shorter echo times will improve the signal-to-noise ratio since higher signal amplitude is available at readout time, while the measurable stimulation effect will be reduced since the echo time is not optimal. Further increase of the echo time to more than TEopt will reduce the measurable stimulation effect and the MR signal due to the exponential signal decay. Example. The relative amount of the stimulation effect has been evaluated for an echo time of TE ≠ TEopt using data from Fig. 13.1.3. Using an echo time of 60 ms results in 98% of the maximal possible stimulation effect reached at the optimal echo time of about 70 ms, while the measured signal in a single slice at readout time is increased compared to TE = TEopt. In the experiment described above, signal improvement of about 20% was possible.

Therefore, a reduced echo time increases the absolute MR signal intensity at the expense of the measurable signal difference between stimulated and non-stimulated periods. On the other hand, a shorter repetition time is possible and will result in improvement of temporal resolution.

Fig. 13.1.3  Simulation of the dependency of the observed BOLD stimulation effect as a function of echo time TE. A reduction of the measurable effect is predicted for echo times differing from the optimum TE = T2* (Eq. 13.1.9). Echo times longer than the optimum will further reduce the measured signals due to signal decay with T2* while echo times shorter than the optimum will lead to measurements with higher signal intensities. Using an echo time of only 60 ms results in 98% of the possible maximal stimulation effect, which can be reached at an optimal echo time of about 70 ms (Fig. 13.1.2). On the other hand, the measured signal in a single image is increased at this readout time compared to TE = TEopt  . In this experiment a 20% signal increase was observed

Example. The dependency of the stimulation effect on the echo time has been evaluated experimentally using 40 and 60 ms (Fig. 13.1.4). Sixty images have been acquired for each echo time, alternating 10 images without and 10 images with finger tapping of the right hand. This results in a total measuring time of about 30 min for both measurements, with 14 s per image. The mean signal change under stimulation was about 5% for TE = 40 ms and about 10% for TE = 60 ms.

An additional signal increase is generated by fully relaxed inflowing spins into the slice of interest during data acquisition while stationary spins reach the steady state after a few HF excitations. The sensitivity of the sequence to inflow effects can be minimized by relatively low flip angles (about 15°) or maximized by large flip angles (about 40°). In addition, optimization of the signal-to-noise ratio using FLASH sequences can be performed by reducing the bandwidth of the readout frequency. This is based on the fact that the acquired MR signal is low-pass filtered after acquisition. Lower readout bandwidth corresponds to a lower cut-off frequency of the low-pass filter with better suppression of the randomly distributed noise in the frequency domain. Reduction of the readout bandwidth requires reduction of the readout gradient, resulting in higher sensitivity to inhomogeneities of the main magnetic field of the sequence. On the other hand, the chemical shift of protons in water and fat increases with reduction of the readout bandwidth and, for example, an overlapping of signals from protons of fatty tissue of the scalp with signals from protons of brain parenchyma can

Fig. 13.1.4  Measurement of signal intensity of the stimulation effect in the motor cortex as a function of echo time. Stimulation experiments have been performed using a FLASH sequence with echo times of TE = 40 and 60 ms. A total of 60 images has been acquired including alternating series of 10 images under non-stimulated and 10 images under finger tapping conditions. Measurement with reduced echo time shows a signal increase of about 20% and a reduced stimulation effect of about 5%

1297

1298

13  Functional MRI

occur. At the same time higher sensibility of the sequence to flow artifacts (acceleration terms) can be observed when prolongation of the readout time and repetition time increases. In practice, a lower limit down to 16 Hz/ pixel of the readout bandwidth has been established.

ging of the whole amount of inflowing blood during data acquisition. The distance of the readout slice (thickness 4–6 mm) parallel to, and above the tagging slab should be in the range of 2–4 mm to avoid overlapping of the tagging slab and the readout slice due to slice profiling.

Note. Signal of protons of fatty tissue can be effectively reduced by using special fat suppression techniques at low readout bandwidth.

13.1.6  Data Analysis

13.1.5.2  EPI In principal the same argument for optimal echo time can be made for EPI measuring techniques but with careful attention to the T2* sensitivity resulting from the special chronological order of the measured Fourier lines. Thereby, the Fourier lines that dominate the image contrast are acquired at the time TE. While FLASH techniques build up a steady state condition for the magnetization, EPI methods create a developing steady state of the magnetization during readout of SI, and the successive lines of the k-space have different signal weighting because of T2* relaxation. Standard EPI techniques use a flip angle of 90° for excitation. The resulting transversal component of the magnetization is readout by alternating gradients (FID EPI). Using modern whole body scanners with high-end gradient systems, measuring times of about 200–300 ms/image are possible with matrix size of 128 × 128 (spatial resolution of about 1.2 mm, slice thickness of about 5 mm). 13.1.5.3  TSE Since stimulation procedures are also associated with T2 changes, the estimation with respect to echo time can be transferred from T2* to T2. The best choice of echo time is therefore in the range of the transversal relaxation time of gray matter. Typical experiments (Constable et al. 1994) use 16 echo trains in the range of 100–200 ms, and echo spacing of about 20–30 ms, with repetition times of about 1,500–2,500 ms. 13.1.5.4  STAR In STAR experiments, SI is dependent on the time difference between tagging and readout of the inflowing spins and therefore inversion time TI has to be selected properly. The range of TI is from 800 to a maximum of 1,200 ms because the T1 relaxation time of blood is about 1,200 ms at 1.5 T. STAR measurements are performed with transversal plane imaging in most cases. The thickness of the inversion slab is in the range of 6–10 cm to guarantee tag-

The aim of data analysis of fMRI images is the detection and quantification of activated areas caused by brain stimulation. Figure 13.1.5 shows a general schema for fMRI data acquisition and data analysis. Statistical methods (i.e., subtraction methods) are used for calculation of so-called statistical parametric maps (Friston et al. 1991, 1994), based on alternating series of images measured under stimulated and non-stimulated conditions where regions with cortical activity are represented by higher parameter values. Highly sophisticated methods are necessary for detection of the BOLD effect with concomitant suppression of noise because of the small difference between stimulated and non-stimulated fMRI series seen with the BOLD effect (<10%) and the different noise characteristics of the individual fMRI images of the image series. Artifacts in the fMRI data, such as those from head movements, can destroy the statistical analysis of an fMRI data series. All methods of analysis are based on the assumption that incoherent noise is statistically averaged while accumulation of coherent signal takes place continuously in activated areas. Different methods of data analysis will create significant differences in the absolute value and distribution of activation of the stimulated areas. This is the reason why sometimes a clear determination of activated areas can be difficult. For better localization, these areas are transferred and overlaid with anatomical MR images. In the following section, a brief overview of the most commonly used data analysis strategies is presented. Each of the presented techniques has to be executed for each slice separately in multi-slice experiments. 13.1.6.1  Subtraction Method The simplest way to analyze data from fMRI series involves subtracting acquired stimulated and non-stimulated MR images and thus defining two groups of data sets corresponding to acquired functional images under stimulated and non-stimulated conditions. Using SI mean values according to and

µ

s



(13.1.11)

13.1  Basics

Fig. 13.1.5  Schematic illustration of fMRI data acquisition and data evaluation. Functional MRI data acquisition starts after anatomic localization of the stimulated area. Alternating series of non-stimulation and stimulation are measured. Data evaluation of the measured series results in parametric images where

activated cortical areas are highlighted. Parametric maps can be superimposed to corresponding morphological images. Signal intensities versus time courses of defined regions reflect the temporal evolution of non-stimulated and stimulated phases (quality control)

with

Thus, areas with the largest SI changes under stimulation correspond to the highest intensities in the subtraction image. A significant drawback of this method is the lack of statistical significance of the activated region.

µr (i,j) mean value of signal intensity of pixel (i,j) acquired in non-stimulated images, and µs (i,j) mean value of signal intensity of pixel (i,j) acquired in stimulated images, nr the number of non-stimulated images, ns the number of stimulated images, Rk(i,j) the signal intensity of pixel (i,j) of the kth non-stimulated image, and Sk(i,j) the signal intensity of pixel (i,j) of the kth stimulated image.

The intensity of the subtraction image can be calculated: D(i,j) = μs (i,j) – μr (i,j).

(13.1.12)

In the majority of cases, the absolute value of Eq. 13.1.12 is used for easier representation of the data instead of the simple difference in SI mean values: DABS(i,j) = |μs (i,j) – μr (i,j)|.

(13.1.13)

13.1.6.2  Statistical Methods 13.1.6.2.1  Student’s t-Test and Z-Score Analysis Statistical tests allow an assignment of significance values to the measured SI changes. Under the assumption of Gaussian distribution of SI changes, both data sets (stimulated and non-stimulated) can be tested to show statistically significant variations of the mean values by using the Student’s t test. This statistical test tries to reject the null hypothesis of equality of mean values at a defined significance level. For pixel (i,j), the corresponding test value of the Student’s t is given by (Press et al. 1988):

(13.1.14)

1299

1300

13  Functional MRI

where the standard deviation of the data can be expressed by:

weighting of the differences of the mean values with the variance of the data pointing to the corresponding mean value. Compared with the subtraction method, these approaches leads to a clear improvement in the stability of the statistical analysis, seen as the suppression of flow effects in the sagittal sinus (Constable et al. 1993; Baudendistel et al. 1950. Therefore, these methods are used    in practice instead of the subtraction method. However, (13.1.15) all statistical methods for data evaluation benefit from an increasing number of data sets, and the quality of the Using the empirical standard deviation the standard er- statistical parametric maps can be improved significantly. ror can be calculated by: Fig. 13.1.6 shows in a direct comparison this dependency on number of data sets for the subtraction and t-test statistical parametric maps. 13.1.6.2.2  Correlation Analysis

     with

(13.1.16)

SDr being the empirical standard deviation of pixel (i,j) of fMRI images under non-stimulation, and SDs the empirical standard deviation of pixel (i,j) of fMRI images under stimulation.

The quantity Student’s t describes the probability p extracted from the probability distribution of getting a t value t1 with t ≥ |t1| under validation of the null hypothe­ sis. All significant pixels, i.e., pixels with a probability p(t) < p, are defined as activated pixels under stimulation. An equivalent parameter description is expressed by the Z-score statistic:

(13.1.17)

which differs from the Student’s t test in principal only by a (normalization) factor which is dependent on the number of data sets. Similar to the subtraction method, absolute values of Student’s i or Z-score parameters are used for better pictorial illustration. By evaluating in each pixel the corresponding statistical parameters of Student’s t or Z score, a so-called statistical parametric map can be calculated (Friston et al. 1991, 1994). Since the statistical parametric map depends on the results of the statistical tests of each single pixel— each single test is afflicted with a discrete probability of error—an appropriate correction with respect to multiple tests should be performed to minimize pixels with wrongly assigned significant activation. Now, correction algorithms for fMRI data are not suitable for routine use since standard correction methods used in statistics (i.e., Bon­ferroni correction) lead to overcorrection because of the large amount of data. Both statistical tests lead to a parameter image that, in contrast to a subtraction parameter image, contains a

Activated areas can also be detected using a pixel-bypixel correlation with a reference function, which can be defined by assuming a temporal signal response curve of the data series of length N. Using this technique, the degree to which the sequence of measured data points and reference function coincide is measured. The quantity of conformity called the linear correlation coefficient can be defined as (Bandettini et al. 1993):

(13.1.18)

with

fk (i,j) being the signal intensity of pixel (i,j) of kth image of fMRI series, rk the value of reference function of kth image of fMRI series, N the number of images of fMRI series, µf (i,j) the mean signal intensity of pixel (i,j) of fMRI series, and µr mean signal intensity of the reference function,

where the statistical significance is given by:

(13.1.19)

Note. In case of correlation using a boxcar function (0 = nonstimulated, 1 = stimulated) with the same number of images for stimulation and non-stimulation all methods (t-test, Z-score, and correlation analysis) result in identical statistical parametric maps that differ only by multiplicative constants.

The choice of the reference function is the main problem in the use of the correlation method, since the observed signal response is characterized by individual fluctua-

13.1  Basics

Fig. 13.1.6  Comparison of statistical analyses of motoric stimulation using the subtraction method and the Student’s t-test as a function of the total number of evaluated images. Upper series Student’s t-test parametric maps, lower series subtraction maps containing absolute parametric values. The stimulation experiment was performed using a total of 60 images including alternating series of 10 images under non-stimulated and 10 images under stimulated conditions Left to right parametric maps us-

ing 20, 40, and 60 images for data evaluation. Image quality of parametric maps is improved by increasing number of evaluated images for both statistical methods. A comparison of parametric maps calculated by the subtraction method and the Student’s t-test show a clear reduction of activated values in sagittal sinus in the Student’s t parametric maps due to statistical weighting of data scatter

tions, which can depend on the duration and mode of the stimulus. This also causes problems when using a reference function directly from the data set for statistical data analysis of different studies. Fig. 13.1.7 shows two different statistical parametric maps as a result of reference functions of the activated motoric cortex and the sagittal sinus. For better comparison, the same data set used in Fig. 13.1.6 has been used. In practice, correlation analysis with boxcar function is generally used and statistical data evaluation is equivalent to statistical tests using Student’s t or Z-score.

13.1.6.3  Quantification of the Stimulation Effect Beside the localization of the activated areas, a quantification of the stimulation effect is necessary for interpretation of the results of a stimulation experiment. At the same time, controlling of the data analysis can be performed by using the observed stimulated and non-stimulated periods. Thereby, SI differences of activated areas are evaluated based on parametric images and used as a measure of the strength of activation. This can be represented by, for example, the percentage of signal increase

1301

1302

13  Functional MRI

Fig. 13.1.7  Data evaluation using the correlation method. Reference function and resulting correlation parametric maps are illustrated. Data series of Fig. 13.1.6 have been evaluated using two different reference functions. The reference function has

been defined by signal-time courses of the activated right motor cortex area (top) and the sagittal sinus (bottom). The choice of reference function influences the intensity distribution and size of the detectable activated cortical areas of the parametric map

detected in activated areas, the size of activated areas, or the mean value evaluated in the parametric image. Problems appear in activated areas with ambiguous contours, i.e. the mean signal value of an area (region of interest, or ROI) depends strongly on the number of activated pixels in the ROI. With respect to this fact, methods are established that are nearly independent of the individual observer and allow a self-sufficient comparison of the absolute stimulation effects of different individuals. The use of statistical tests allows a comparison of the activation of single examinations and different patient studies by using the corresponding parametric values. A simple procedure for quantification of the stimulation effect is based on calculating Student’s t-test parametric images with a superimposed reference grid. Only pixels with values higher than a given significance threshold are counted as activated under stimulation. Regions of interest can be defined by combination of different grid segments containing activated areas. Absolute values of activation (defined by the mean Student’s t value of activated

pixels) and the sizes of activated areas can be evaluated in the so-defined regions (Wenz et al. 1994; Baudendistel et al. 1995). Note. The issue of absolute quantification in fMRI and the comparison of different scanners, measuring techniques, and data analyzing strategies are the focus of ongoing international studies.

13.1.6.4  Artifacts and Motion Correction Analysis of fMRI studies is based on data evaluation pixel-by-pixel with relatively small SI changes of measured images and is therefore very sensitive to image artifacts. Image artifacts can be created by technical insufficiencies (inhomogeneities of the main magnetic field, instabilities of the scanner) and also by physiological effects (signal variations due to periodical physiological processes, i.e., cardiac and pulmonary activities). Significant artifacts are produced in parametric images by small

13.1  Basics

instabilities of head fixation due to SI differences between vessels and surrounding tissue, leading to questionable application of the method (Hajnal et al. 1994). This effect of a periodic motion that is related to the stimulation phases, and its influence on the observed SI changes is a subject of controversy (MOLD: motion-level-dependent contrast). Because of this problem, an effective fixation of the head or applicable algorithms for retrospective detection and correction of head motion during examination are very important (“image registration”) (Hajnal et al. 1994; Woods et al. 1992; Friston et al. 1996). In the simplest case, methods for image registration are based on the minimization of differences of signal intensities between different fMRI images. Two different strategies with different criteria for optimal alignment of fMRI images are under investigation using, for example, maximization of the correlation between predefined regions of different fMRI images, or minimization of the absolute differences of signal intensities, or optimization of the pixel-wise ratio of single images, as a criteria for the quality of image alignment. In addition, methods working in the Fourier space have also been proposed (Chen et al. 1994). For single-slice techniques, motions perpendicular to the image section are the main problem, since parts of the imaged slice are not excited and it is not possible to reconstruct the missing information. On the other hand, using multi-slice techniques motions parallel to the slice section direction can be corrected by data interpolation in the measured volume (Woods et al. 1992; Friston et al. 1996). Uncorrected and corrected Student’s t parametric images are shown in Fig. 13.1.8 for comparison. The applied motion correction algorithm uses maximization of the correlation of a region covering the central corrugation of the brain. The corrected parametric image shows a clear improvement in the detection of the activated areas because of maximized correlation and homogeneous background (Baudendistel et al. 1995). Correction of artifacts caused by head motion occurring between single slices of an fMRI series can be difficult, since the acquired images show large noise and relatively poor contrast of fine brain structures, which helps to improve image registration due to the long echo time. Furthermore, derivation of different algorithms requires special conditions with respect to the acquired images, i.e., image intensities are a continuous function in space of discrete pixels. Since individual noise can suggest artificial structures in single slices, low image noise with sufficient brain structures, i.e., contrast differences, are a prerequisite for motion correction, which compares these structures in all images. In practice, fMRI images are filtered before motion correction to reduce image noise but at the cost of losing image details seen on the evaluated parametric images. A serious problem of motion correction algorithms is the necessary accuracy of image alignment. SI differences between sulci and gray

Fig. 13.1.8  A typical example of motion correction: comparison of uncorrected (right) and corrected (left) Student’s t parametric maps. The applied motion correction algorithm uses maximization of the correlation of a region covering the central corrugation of the brain. The corrected parametric image shows a clear improvement in detection of the activated areas because of maximized correlation and homogeneous background

matter are relatively large and small inaccuracies of the motion correction algorithm (in the range of one pixel) can lead to signal mixing and apparent activation. To improve accuracy images can be interpolated to higher image resolution where the resulting image quality is different for distinct interpolation techniques. In consequence, the motion corrected images are dependent on the filter and interpolation algorithm used. In addition, depending on the location of the region of interest, tissue–air interfaces can produce spurious artifacts due to signal mixing, making it impossible to evaluate data from the fMRI examination. In this special case, a critical visual inspection of the corrected image data is necessary. Note. Algorithms used for motion correction, filtering, and interpolation are subjects of ongoing intensive discussions. Therefore recommending an optimal motion correction algorithm is difficult. On the other hand, a practicable motion correction algorithm should be used for retrospective data evaluation to detect head motion during the fMRI examination and to evaluate the accuracy of the calculated results.

13.1.7  Comparison of Methods A critical evaluation of the presented measuring sequences has to be performed with respect to clinical applicability. Advantages and disadvantages of the different measuring techniques have to be balanced with regard to the experimental setup. The following section presents short and clear guide to measuring techniques, with a comprehensive conclusion given in Table 13.1.1.

1303

1304

13  Functional MRI Table 13.1.1  Comparison of different fMRI measuring techniques Technique

Measuring parameters

Advantages

Disadvantages

TR

TE

α (°)

TI

FLASH

60–100 ms

40–60 ms

10–40

–

High spatial resolution, good availability, superimposition on morphological images

Low temporal resolution (5–15 s), mostly singleslice measurements, large total acquisition time

EPI

2–4 s

>40 ms

90

–

High temporal resolution, multi-slice technique (about 20 slices)

Sensitive to image artifacts, problems of image distortions: EPI images can only be superimposed with morphological images after correction. Dead time of the scanner due to data evaluation of large data sets

TSE

1.5–2.5 s

100–200 ms (16-echo trains, echo spacing of 20–30 ms)

90

–

Suppression of noncapillary vessels

Small signal change, relatively substantial requirements with respect to the scanner (stability) and head fixation (motion artifact), large total acquisition time

0.8–1.2 s

Robust technique, possible application at low magnetic field strength, dynamic examinations of the fMRI effect by variation of TI

Single-slice technique, reduced temporal resolution due to waiting period between blood-bolus labeling and readout of SI, limitation of possible waiting period due to T1 relaxation of blood

STAR

Measuring parameters are defined by FLASH or EPI readout

FLASH • Measuring parameters: echo time ~ T2* (40–60 ms); flip angle between 10 and 40°; TR minimal (60–100 ms); mostly single slice measurement • Advantages: high spatial resolution; good availability; high signal-to-noise ratio of a single image; fMRI images can be superimposed with morphological images • Disadvantages: relatively low temporal resolution (5–15 s); insufficient volume size (number of slices/ time); large total measuring time (about 15 min for 60 fMRI images) EPI • Measuring parameters: echo time ~ T2* (> 40 ms); flip angle 90°; multi-slice technique possible (about 20 slices); TR in the order of seconds (2–4 s) • Advantages: high temporal resolution; multi-slice technique possible (large volume size)

• Disadvantages: sensitive to artifacts; problems of image distortions. EPI images can only be superimposed with morphological images after correction; dead time of the scanner due to data evaluation of large data sets TSE • Measuring parameters: conventional 16-echo train with echo times of 100–200 ms and echo spacing of 20–30 ms; TR in the range of 1,500–2,500 ms; other parameters used as conventional • Advantages: suppression of contribution from noncapillary vessels is possible • Disadvantages: relatively substantial requirements with respect to the scanner (stability) and head fixation (motion artifacts) because of the small amount of observed effects and long measuring time per single image (20–40 s)

13.1  Basics Fig. 13.1.9  Comparison of functional imaging using EPI BOLD and EPI STAR performed in a healthy volunteer with finger tapping of the left hand. Illustrated are the conventional T1-weighted spin echo image, the results of EPI BOLD measurements (white areas) superimposed on the morphological image, and the superimposed results of EPI STAR measurements at TI = 800 and 1,000 ms. A clear change in activated areas at different inversion times TI can be detected due to the increasing flow sensitivity of the EPI STAR technique (FOLD contrast)

STAR • Measuring parameters: waiting period between inversion and readout of about 800–1,200 ms; readout of slice of interest can be performed either using segmented FLASH or EPI technique • Advantages: robust technique; possible application at low magnetic field strength; dynamic examinations of the fMRI effect are possible by variation of the waiting period between blood bolus labeling and readout of SI (FOLD contrast) • Disadvantages: only single slice measurements are possible; reduced temporal resolution due to waiting period between blood bolus labeling and readout of SI; limitation of possible waiting period due to SI decay with respect to T1 relaxation time of blood (about 1.2 s at 1.5 T)

Note. Application of TSE and STAR techniques are of experimental character and the clinical applicability has to be demonstrated. Because of their insensitivity to image artifacts and wide availability, FLASH techniques have been used in the past for fMRI trials performed on clinical scanners with reduced bandwidths (down to about 16 Hz/pixel), flip angles of 10–40°, and echo times of 40–60 ms. Slice thicknesses of 3–10 mm and FOV of 180–250 mm have been used in most cases. The reachable temporal resolution for a 128 × 128 matrix is about 10–15 s/image. Today, mostly EPI sequences are used in fMRI because of widely used modern scanners with high-performance gradient systems, allowing matrix size of 64 × 64 to 128 × 128 with about 20 slices at a total temporal resolution of about 2–4 s with slice thickness of 5–10 mm. Using series of 100 fMRI images/slice (128 × 128 matrix, 2 byte/pixel, 20 slices) data sets of about 64 MB are acquired. Figure 13.1.9 represents the results of an experimental fMRI study for comparison of EPI BOLD and EPI STAR at 1.5 T.

1305

1306

13  Functional MRI

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Bandettini PA, Wong EC, Hinks RS, Tikofsky RS, Hyde JS (1992) Time course EPI of human brain function during task activation. Magn Reson Med 25:390–397 Bandettini PA, Jesmanovicz A, Wong EC, Hyde JS (1993) Processing strategies for time-course data sets in functional MRI of the human brain. Magn Reson Med 30:161–173 Baudendistel K, Schad LR, Friedlinger M, Wenz F, Schröder J, Lorenz WJ (1995) Postprocessing of functional MRI data of motor cortex stimulation measured with a standard 1.5T imager. Magn Reson Imaging 13(5):701–707 Belliveau JW, Kennedy DN, McKinstry RC, Buchbinder BR, Weiskoff RM, Cohen MS, Vevea JM, Brady TJ, Rosen BR (1991) Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716–719 Blamire AM, Ogawa S, Ugurbil K, Rothman D, McCarthy G, Ellermann JM, Hyder F, Rattner Z, Shulman RS (1992) Dynamic mapping of the human visual cortex by highspeed magnetic resonance imaging. Proc Natl Acad Sci USA 89:11069–11073 Boxerman JL, Bandettini PA, Kwong KK, Baker JR, Davis TL, Rosen BR, Weisskoff RM (1995) The intravascular contribution to fMRI signal change: Monte Carlo modeling and diffusion-weighted studies in vivo. Magn Reson Med 34:4–10 Chen QS, Defrise M, Deconinck (1994) Symmetric phaseonly matched filtering of Fourier-Mellin transforms for image registration and recognition. IEEE-PAMI 16(12):1156–1168 Conelly A, Jachson GD, Frackowiak RS, Belliveau JW, Vargha-Khadem F, Gadian DG (1993) Functional mapping of activated human primary cortex with a clinical MR imaging system. Radiology 188:125–130 Constable RT, McCarthy G, Allison T, Anderson AW, Gore JC (1993) Functional brain imaging at 1.5T using conventional gradient echo MR imaging techniques. Magn Reson Imaging 11:451–459 Constable RT, Kennan RP, Puce A, McCarthy G, Gore JC (1994) Functional NMR using fast spin echo at 1.5 T. Magn Reson Med 31:686–690 Edelman RR, Siewert B (1994) Signal targeting with alternating radiofrequency (STAR) se­quen­ces. Magn Reson Med 31:233 Fox PT, Raichle ME (1986) Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc Natl Acad Sci USA 83:1140–1144 Frahm J, Bruhn H, Merboldt KD, Hänicke W (1992) Dynamic MR imaging of the human brain oxygenation during rest and photic stimulation. J Magn Reson Imaging 2:501–505 Friston KJ, Frith CD, Liddle PF, Frackowiak RSJ. (1991) Comparing functional (PET) images: the assessment of significant change. J Cereb Blood Flow Metab 11:690–699 Friston KF, Jezzard P, Turner R (1994) The analysis of functional MRI time series. Human Brain Mapp 1:153–171

16. Friston KJ, Williams S, Howard R, Frackowiak RSJ, Turner R (1996) Movement-related effects in fMRI time-series. Magn Reson Med 35:346–355 17. Gillis P, Koenig SH (1987) Transverse relaxation of solvent protons induced by magnetized spheres: application to ferritin, erythrocytes, and magnetite. Magn Reson Med 5, 323–345 18. Haase A, Frahm J, Matthaei D, Hänicke W, Merboldt KD (1986) FLASH imaging. Rapid NMR imaging using low flip-angle pulses. J Magn Reson 67:258–266 19. Hajnal JV, Myers R, Oatridge A, Schwieso JE, Young IR, Bydder GM (1994) Artifacts due to stimulus correlated motion in functional imaging of the brain. Magn Reson Med 31:283–291 20. Hennig J, Naureth A, Friedburg H (1986) RARE Imaging: a fast imaging method for clinical MR. Magn Reson Med 3, 823–833 21. Hu X, Kim SG (1994) Reduction of signal fluctuation in functional MRI using navigator echoes. Magn Reson Med 3:495–503 22. Jezzard P, Balaban RS (1995) Correction for geometric distortions in echo planar images from B0 field variations. Magn Reson Med 34:65–73 23. Kennan RP, Zhong J, Gore JC (1994) Intravascular susceptibility contrast mechanisms in tissues. Magn Reson Med 31:9–21 24. Kwong KK, Belliveau JW, Chesler DA, Goldberg IE, Weisskoff RM, Poncelet BP, Kennedy DN, Hoppel BE, Cohen MS, Turner R, Cheng HM, Brady TJ, Rosen BR (1992) Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc Natl Acad Sci USA 89:5675–5679 25. Majumdar S, Gore JC (1988) Studies of diffusion in random fields produced by variations in susceptibility. J Magn Reson 78:41–55 26. Ogawa S, Lee TM, Kay AR, Tank DW (1990a) Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci USA 87:9868–9872 27. Ogawa S, Lee TS, Nayak AS, Glynn P (1990b) Oxygenation-sensitive contrast in magnetic resonance imaging of rodent brain at high magnetic fields. Magn Reson Med 26:68–78 28. Ogawa S, Tank DW, Menon R, Ellermann JM, Kim SG, Merkle H, Ugurbil K (1992) Intrinsic signal changes accompanying sensory stimulation: functional brain mapping using MRI. Proc Natl Acad Sci USA 89:5951–5955 29. Ogawa S, Menon RS, Tank DW, Kim SG, Merkle H, Ellermann JM, Ugurbil K (1993) Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. Biophys J 64:803–812 30. Press WH, Flannery BP, Teukolsky SA, Vetterling WT (1988) Numerical recipes: the art of scientific computing. Cambridge University Press, Cambridge, pp 465–469

References 31. Schad LR, Trost U, Knopp MV, Müller E, Lorenz WJ (1993) Motor cortex stimulation measured by magnetic resonance imaging on a standard 1.5T clinical scanner. Magn Reson Imaging 11:461–464 32. Schad LR, Wenz F, Knopp MV, Baudendistel K, Müller E, Lorenz WJ (1994) Functional 2D and 3D magnetic resonance imaging of motor cortex stimulation at high spatial resolution using standard 1.5T imager. Magn Reson Imaging 12:9–15 33. Thulborn KR, Waterton JC, Mathews PM, Radda G (1982) Oxygenation dependence of the transverse relaxation time of water in whole blood at high field. Biochem Biophys Acta 714:265–270

34. Turner R, Jezzard P, Wen H, Kwong KK, Le Bihan D, Zeffiro T, Balaban RS (1993). Functional mapping of the human visual cortex at 4 and 1.5 Tesla using deoxygenation contrast EPI. Magn Reson Med 29:277–279 35. Wenz F, Schad LR, Knopp MV, Baudendistel KT, Flömer F, Schröder J, van Kaick G (1994) Functional magnetic resonance imaging at 1.5T: activation pattern in schizophrenic patients receiving neuroleptic medication. Magn Reson Imaging 12:975–982 36. Woods RP, Cherry SR, Mazziotta JC (1992) Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992:620–633 37. Yablonskiy DA, Haacke EM (1994) Theory of NMR signal behavior in magnetically inhomogeneous tissues: the static dephasing regime. Magn Reson Med 32: 749–763

1307

1308

13  Functional MRI

13.2  Clinical Applications K.K. Peck and A.I. Holodny 13.2.1  Introduction Since the introduction of the blood oxygenation–leveldependent (BOLD) contrast in MRI in 1990, BOLD functional MRI (fMRI) has been increasingly utilized as a tool to investigate human brain function. With this advanced neuroimaging technique, a rapidly increasing number of studies have been published that attempt to map the brain function obtained during specific functional tasks including sensory, motor, and cognitive tasks. More recently, fMRI has expanded from the domain of the MRI laboratory into the clinical setting. As a result, efforts to understand the diseased brain and its dynamics have greatly accelerated. Currently, this technique has been adopted almost universally by disciplines that endeavor to understand how the brain works, including clinicians, such as radiologist, psychologists, psychiatrists, neurologists and neurosurgeons, as well as the basic scientists such as cognitive neuroscientists and physicists. In this section, clinical application of fMRI is described. Issues including paradigm selection, patient preparation, and fMRI in neurologic disease are discussed. 13.2.2  Normal Functional Anatomy 13.2.2.1  Sensory–Motor The primary sensory (S1) and motor cortex (M1) is commonly studied. The initial applications of fMRI involved

Fig. 13.2.1  Motor task using bilateral finger tapping at 1.5 T demonstrates activation of gray matter regions of the primary motor cortex in both hemispheres and the supplementary motor area in a healthy control (Reprinted with permission from Topics in MRI (2004) 15:325–335, Lippincott, Williams & Wilkins)

the sensory–motor system because the stimulation parameters of such tasks can be easily standardized. The localization of the primary motor cortex using fMRI has been studied precisely and reliably (Rao et al. 1995; Yousry et al. 1997) and has been verified through intraoperative direct cortical stimulation (Yousry et al. 1995). Furthermore, fMRI can also demonstrate the activation of cortical networks in secondary motor areas during complex motor tasks, including the supplementary motor area (SMA) and the premotor cortex (Moriyama et al. 1998; Van Oostende et al. 1997). Figure 13.2.1 is an example showing activation in the primary motor cortex and supplementary motor area that are superimposed on high-resolution T1-weighted 3D anatomic images. 13.2.2.2  Language Language is closely related to cognitive functions, such as memory, attention, and perception. fMRI paradigms including word generation, object naming, reading and semantic decision are commonly used for their ability to activate both receptive and expressive language areas as well as their ability to activate large cerebral regions. Chee et al. (1999) previously showed that semantic tasks with both visual and auditory cues produce similar activation patterns. The classic model of language-related functional neuroanatomy places the motor output of language in the left inferior frontal gyrus (Broca’s area) and the reception and perception of language in the left posterior–superior temporal gyrus (Wernicke’s area). These two main language areas (Fig. 13.2.2) have been successfully mapped using fMRI in normal subjects, where it was confirmed that language has a predominantly unihemi-

Fig. 13.2.2  An example showing activation in the left inferior frontal gyrus (Broca’s area) and the left posterior temporal lobe (Wernicke’s area) while generating verbs to visually presented nouns in healthy control subject

13.2  Clinical Applications

spheric representation, (usually in the left hemisphere) in right handed people (Springer et al. 1999). However, the precise location of Broca’s and Wernicke’s areas is variable (Amunts et al. 1999; Bookheimer et al. 2000).

ies on patients with neurological conditions needs special consideration.

13.2.2.3  Memory

The paradigms that are administered must be selected based on the location of the lesion and/or behavioral assessment of the patient. It is important to review the patient’s previous brain scan to determine the location of the lesion and the parts of the brain involved. This will indicate the type of paradigms that will be administered during the fMRI study. Below are examples of some of the most commonly requested fMRI exams in clinic.

Memory paradigms can be difficult to perform and interpret. In part, this is due to MR signal loss from susceptibility artifacts in the parts of the brain associated with memory tasks including, but not limited to the hippocampus and the medial temporal lobe. Golby et al. (2001) separates verbal memory from spatial memory, as these functions may be lateralized differently. Verbal memory often lateralizes to the dominant hemisphere and spatial memory to the non-dominant hemisphere, though this organization is not steadfast. In this study, memory paradigms most often consist of novel versus repeated faces, patterns, and words. For example, the patient is presented with pictures of male and female faces, some repeated and some novel. The response to the repeated faces is contrasted statistically. A similar design can be employed using indoor and outdoor scenes and word pairs (Fig. 13.2.3a,b). 13.2.3  Functional MRI in Disease fMRI in the clinical setting is markedly different from that in healthy control subjects. Performing fMRI stud-

13.2.3.1  Paradigm Selection

13.2.3.1.1  Sensory–Motor Sensory–motor paradigm is one of the most common clinical fMRI tasks performed. The activation patterns do not vary significantly for different hand movement tasks for clinical purposes (Kober et al. 2001; Thickbroom et al. 2001; Beisteiner et al. 2000), measuring cerebral reorganization following stroke (Ward et al. 2003, Zemke et al. 2003; Cao et al. 1998), and demonstrating the location of a tumor or vascular malformation relative to eloquent cortex (Holodny et al. 2000; Krings et al. 2001; Baciu et al. 2003). Motor and sensory paradigms are fairly straightforward and highly lateralized to the contralateral hemisphere. Paradigms should be tailored to the motor hoFig. 13.2.3  Activation maps for patterns and word encoding in the hippocampal and parahippocampal ROIs. a Patterns encoding (contrast novel images vs. repeated images, p = 0.05); note lateralization to the left to the left anteriorly and to the right posteriorly. b Words encoding (contrast novel images vs. repeated images, p = 0.01); note overall lateralization to the right (Courtesy of Dr Daniel Branco, Golby Lab, Brigham and Women’s Hospital, Boston, Mass., USA)

1309

1310

13  Functional MRI

munculus. For example, lesions closer to the midline should involve leg and foot stimulation, lesions closer to the reverse omega portion of the central sulcus should include finger tapping, and lesions more inferior should include tongue motion. The most common sensory–motor paradigms include movements of the hand, e.g., finger tapping (Ulmer et al. 2003), hand closing and opening or squeezing (Kober et al. 2001), flexion–extension of hand and fingers (Johansen-Berg et al. 2002) as well as tongue movements (Preibisch et al. 2003), toe movements (Marquart et al. 2000), and near-isometric wrist extension movements (Peck et al. 2001). Pure sensory functions have been captured by scrubbing or massaging a part of the body (Hoeller et al. 2002). It is often useful to localize the motor strip prior to neurosurgical intervention in order to save the surgeon time during direct cortical stimulation and or supplant the invasive mapping all-together (Fig. 13.2.4). Figure 13.2.5 demonstrates an example of a patient with a high-grade tumor. Significant activation is seen in both hemispheres during a bilateral finger-tapping task. The relationship of the tumor to the motor cortex is clear where it was uncertain before fMRI mapping.

13.2.3.1.2  Language Unlike the sensory–motor system, the language system is more complex in its fMRI measurement and interpretation because it involves a widely distributed network of structures both supportive and essential to the task. To obtain a good fMRI result, it requires more involved cognitive participation on the part of the patient. Language fMRI paradigms should be designed to target productive or frontal areas, as well as receptive or temporoparietal areas (Fig. 13.2.6). Numerous variations of fMRI paradigms are described. Some of the paradigms used in clinics are (1) word generation, such as generating a verb associated with a noun (Burton et al. 2004; Cao et al. 1999); (2) phonemic fluency, generating various words from a given letter (Staudt et al. 2002; Adcock et al. 2003) or from a given category (semantic fluency) (Peck et al. 2004; Sabbah et al. 2003); (3) object naming, where you identify images that are displayed as line drawings (Cao et al. 1999); and (4) reading a single noun or a complete sentence (Gaillard et al. 2002). A lesion in the frontal lobe (near Broca’s area) necessitates a productive speech paradigm like phonemic fluency. A lesion in the temporoparietal area (near Fig. 13.2.4  a High-resolution 3D rendered images, b three-dimensional image with coregistered fMRI data (in red) from a bilateral finger-tapping paradigm in a patient with a large frontal tumor. Pathologic analysis revealed a glioblastoma multiforme (Reprinted with permission from Topics in MRI (2004) 15:325–335, Lippincott, Williams & Wilkins)

Fig. 13.2.5  Activation from a finger-tapping paradigm coregistered to an axial T1-weighted image in a patient. Pathologic analysis revealed a left parietal glioblastoma multiforme. There is a robust activation in the supplementary motor area and right sensory–motor cortex but little detectable activity in the left motor cortex

13.2  Clinical Applications

Fig. 13.2.6  Patient with a high-grade tumor in right hemisphere (thick arrow) was experiencing stuttering. Functional MRI was used to verify that the patient’s language area (thin arrows) was left dominant (Reprinted with permission from Topics in MRI (2004) 15:325–335, Lippincott, Williams & Wilkins)

Wernicke’s area) calls for receptive tasks and/or tasks that assess reading or the ability to perform confrontation naming (Fig. 13.2.7). Covert (silent speech) as opposed to overt (vocalized speech) language paradigms are often used in fMRI to minimize head motion artifacts that often occur during vocalization. The problem with silent speech paradigms, however, is that they may not accurately represent the whole of the patient’s speech network. In addition, overt responses allow the experimenter to monitor the patient’s responses, which is important for subsequent analysis and interpretation (Huang et al. 2001). Overt responses may also be more appropriate for intraoperative guidance in that they are better predictor of the location of speech arrest. Importantly, however, it should be noted that despite careful consideration of the tasks in the context of the lesion location, many tasks activate both Broca’s and Wernicke’s areas nearly equally.

Fig. 13.2.7  Activation is observed in Wernicke’s area during a phonemic fluency paradigm in a patient with a tumor in that vicinity (Reprinted with permission from Topics in MRI (2004) 15:325–335, Lippincott Williams & Wilkins)

13.2.3.1.3  Memory Advances in both experimental design and acquisition have improved the ability to detect hippocampal activity associated with memory encoding in patients with epilepsy (Fig. 13.2.8). Several groups have used complex visual scene encoding as an activation paradigm for fMRI of memory encoding (Golby et al. 2001; Machulda et al. 2003). While a majority of studies have demonstrated activation of medial temporal lobe structures during memory encoding, activation in the hippocampus proper has been difficult to demonstrate reliably. Possible explanation for this include challenges in effectively modulating hippocampal activity and increased susceptibility effects in this region that reduce sensitivity to the BOLD effect (Stark and Squire 2001).

Fig. 13.2.8  A patient with a left mesial temporal sclerosis in whom a memory Wada test lateralized memory function to the right. fMRI activity confirmed lateralization to the right mesial temporal lobe (p = 0.05) (Courtesy of Dr Daniel Branco, Golby Lab, Brigham and Women’s Hospital, Boston, Mass., USA)

1311

1312

13  Functional MRI

13.2.3.2  Patient Preparation It is important to assess the clinical status of the patient to determine if the patient will be able to perform the planned fMRI paradigms. In the case of an impaired patient, the paradigm should be modified to reflect the patient’s capabilities. For example, one could plan a finger-tapping task for a patient with a tumor in the precentral gyrus in the expected location of the homunculus of the hand. However, if the patient’s hand is severely weak or paralyzed, one must replace the planned paradigm with an externally controlled motor paradigm such as manipulation of that hand by a technician (Fig. 13.2.9). In another example, the clinician may want to know the location of Broca’s area in a patient with a large fronto-insular glioma. However, if the patient is severely aphasic, he/she will not be able to perform the required language tasks overtly. In such a case, a silent generation (covert) language task would be better suited. If a patient cannot adequately perform the task, the fMRI map will be difficult to interpret. In this way, pre-fMRI screenings can save an enormous amount of frustration and lost time on the scanner. The paradigms to be administered must be rehearsed by the patient prior to the fMRI scan. During this preparatory time, all of the possible things that could go wrong should be addressed. This includes obvious but crucial factors such as hearing difficulty, problems with visual activity, as well as problems with following commands and language barriers. These situations could cause patients to get frustrated, miss stimuli, and stop the scan. Adequate patient preparation and clear guidelines as to

Fig. 13.2.9  The activation map of a patient with an oligodendroglioma. The patient was not able to perform the finger-tapping task using affected right hand. A passive hand sensory stimulation paradigm, demonstrated robust post central gyrus activity (arrow identifies the precentral gyrus)

when to begin performing the task and when to stop are essential. Such efforts will increase the chance that the patient will perform the paradigm correctly in the magnet and provide a good functional MRI output. 13.2.3.3  fMRI Consideration in Patients 13.2.3.3.1  Susceptibility Effects Including Prior Surgery As discussed in detail in an earlier chapter, the change of MR signal intensity is due to the sensitivity of MR signal to the paramagnetic deoxygenated hemoglobin (Hb). This is the basis of the contrast in fMRI. In fact, a fast gradient-echo sequence such as echo planar imaging (EPI) is commonly used to emphasize the susceptibility difference between deoxygenated and oxygenated Hb, reflecting that the source of BOLD fMRI contrast utilizes the susceptibility effect. However, this strategy also means that measurement is also susceptible artifacts, which can degrade the study (Kim et al. 2005). The loss in signal and geometric distortions due to the susceptibility artifacts can mask the BOLD signal (Fig. 13.2.10). Susceptibility artifacts, for example, are often shown at junctions between air and tissue, such as at the temporal lobes or orbitofrontal cortex (from the sinuses). The neuro­radiol­ ogist may make a mistake in determining the location of the functional cortex due to the artifact if the fMRI data is interpreted based on the volume of activation. As a result, the T2* image should always be inspected prior to the interpretation of the results. Susceptibility artifacts are common in patients with prior neurosurgery. The presence of titanium plates to secure skull flaps, metallic staples to close surgical incisions, hemorrhage, and residual metal from the skull drill could cause an increase in the susceptibility artifact, which, in turn, can lead to a decrease in the accuracy of BOLD fMRI (Fig. 13.2.11). The main support for the idea that the decrease in the volume of fMRI activation is due to susceptibility artifacts from prior surgery is from visual inspection of the T2* images. In our experience, in each case, visual inspection of the T2*-weighted images demonstrated that the decrease in fMRI signal was due to the presence of prominent signal dropout from the susceptibility artifact. Therefore, it is imperative to examine the T2* images and not just the coregistered images of the analyzed fMRI data onto the T1-weighted images. 13.2.3.3.2  Loss of Autoregulation It has been shown that there may be a loss of autoregulation of the vasculature in malignant gliomas (Holodny et al. 2000) and a subsequent degradation of the BOLD

13.2  Clinical Applications Fig. 13.2.10  A bilateral fingertapping paradigm in a patient with prior surgery for a brain tumor. There is robust activation in the right motor cortex but little detectable activity in the left motor cortex (a). The signal drop off is seen on the T2*-weighted image (b) (arrow) and corresponds to a titanium plate placed during prior operation. The susceptibility artifact is not visible in the T1-weighted image (a)  (Reprinted with permission from AJNR (2005) 26:1980–1985)

spond to increased neural activity by a corresponding increase in blood flow. Consequently, the area of interest may not show significant activity. In a patient with a tumor in the primary motor cortex, for example, the vasculature in the motor cortex could be affected by the tumor, causing a loss of autoregulation. This may prevent an increase in blood flow in the expected area of activation that normally occurs in addition to motor activity. A reduced blood flow to the expected area would significantly limit the ability of BOLD fMRI to detect activation. 13.2.3.3.3  Artifacts Due to Motion

Fig. 13.2.11  Bilateral finger tapping in a patient with prior surgery for a brain tumor. The location of the susceptibility artifact is further away from the motor cortex than in Fig 13.2.10. As a result, the effect on the volume of motor fMRI activity is less (Reprinted with permission from AJNR (2005) 26:1980–1985)

response. In angiographic and MR studies, the vasculature of gliomas revealed an abnormal response to various physiological and pharmacological challenges (Pronin et al. 1997). BOLD fMRI is based on increased neural activity inherently linked, both temporally and spatially, to an increase in blood flow and the resultant changes in deoxyhemoglobin concentration. If, however, the brain’s ability to autoregulate the blood flow is lost in brain tissue that is still functioning, then this area may not re-

Head motion during the fMRI experiment is a major problem, causing both false-positive and false-negative artifacts and uninterpretable results. Artifacts due to head movement are present when voxels of higher signal intensity are shifted into positions occupied by voxels with lower signal intensity. This issue may be worse in clinical studies, as patients often exhibit more involuntary movements than healthy controls. Motion in the axial plane typically has a ring-like spatial appearance at the edges of the brain and high-contrast regions of the brain after statistical analysis. These artifacts are false positives and cannot be differentiated from real activation since they artificially lead to higher signal changes. Task-correlated motion may be more problematic than non–task-correlated motion. Task correlated head motion could occur, for example, if a subject moves the head while performing a finger tapping, or a toe movement task during a stimulus period in a block paradigm. In such a case, head motion could mask real (true positive) activation by altering the signal time course in such a way that the task

1313

1314

13  Functional MRI

correlated real signal becomes difficult to separate from motion induced false positives. However, not all is lost in the case of head motion. The BOLD hemodynamic response to a functional task typically has a delay of 3–6 s before onset, and the peak is reached around 4–6 s. The time course due to motion related artifacts does not follow this pattern and shows increased signal abruptly reaching its maximum without a delay. Percent signal changes during head motion are typically higher than 5%. Field et al.’s (2000) study suggested that when correlation between stimulus designed waveform and motion exceeds 0.52, false activations are increasingly likely despite the application of motion correction procedures. There is a high degree of correlation between motion artifacts and the degree of paresis (Krings et al. 2001). In patients with paresis, different para­ digms could be used to minimize artifacts while keeping functional information intact such as passive movement or sensory stimulation. This also has lead to the development of devices for immobilizing the head (Edward et al. 2000), as well as motion correction algorithms for preprocessing prior to statistical analysis (Cox 1996; Friston et al. 1995; Jiang et al.1995). 13.2.3.3.4  Draining Veins Another major consideration for fMRI is to distinguish between sites of cortical activation and small parenchymal venules that are in close proximity to these sites (maximally 1.5 mm apart), and large draining veins distant from the active parenchyma (Frahm et al. 1994). This problem becomes especially important as fMRI is performed for pre-surgical mapping of functional cortical areas. The most common approach used to differentiate between draining veins and functional cortical sites has been visual inspection in correlation with high-resolution anatomic images. However, difficulties may arise when spatially smoothed data is used. Differences in the onset of signal change between large draining veins and cortical microvasculature have also been studied (Lee et al. 1995). It has been shown that increases in signal around large draining veins during stimulation are time delayed, compared to those from the parenchymal vasculature corresponding to the sites of neuronal activity. The longer delay was consistent with the time required for blood to pass smaller veins and reach larger vessels. The differences between macrovasculature and microvasculature in the rise time were 1–3 s. With a temporal resolution of 4 s for example (TR = 4,000 ms), one cannot expect to detect robust differences between both types of activation. Larger draining veins show a high percent signal change compared with parenchymal venules, so one should be cautious if the data shows a low signal-to-noise ratio.

13.2.4  Specific Pathology 13.2.4.1  Tumor 13.2.4.1.1

Neurosurgical Planning and Intraoperative Guidance

The primary goal in brain tumor surgery is to maximize tumor resection while preserving important eloquent cortices (Maldjian et al. 1997). It has been shown that the length and quality of survival of brain tumor patients are improved with maximized tumor resection in many tumor types (Ammirati et al. 1987). To preserve important neurologic function controlled by the cortex directly adjacent to or invaded by the tumor, it is important for the neurosurgeon to be able to determine the anatomic location of this eloquent cortex intraoperatively. Morphologic identification of the eloquent cortices during surgery can be complicated by distortion of the anatomy by mass lesions and by the fact that the functional cortex does not always correspond to its expected anatomic location. Traditionally, the eloquent cortices, such as the motor cortex, have been identified intraoperatively by electrophysiological methods such as intraoperative cortical mapping (Schulder et al. 1998). This method remains the gold standard for defining the location of eloquent cortices. However, this requires a craniotomy and only limited brain areas can be mapped. The most direct clinical application of fMRI has been presurgical mapping for patients with brain tumors near essential functional cortical areas. fMRI data can be superimposed on high-resolution structure images. With contrast enhancing regions, usually, gadoliniumenhanced T1-weighted images serve as the high-resolution ‘‘background’’ images on which to superimpose the functional images. However, it should be stressed that the fMRI data can be coregistered to any sequence and that one should choose the most optimal sequence based on the location of the pathology and the imaging characteristics. For example, if one is evaluating a low-grade glioma that is barely perceptible on a T1-weighted image and does not enhance, one may choose to coregister the fMRI data to the FLAIR sequence where such a lesion is typically much better seen. Similarly, if the lesion is in the temporal lobe, one may choose to coregister the fMRI data to a coronally acquired sequence, which is sometimes optimal for visualizing structures in the temporal lobe. fMRI can identify eloquent cortices prior to the surgery, thus guiding the intraoperative cortical stimulation (Fig. 13.2.12). Performing preoperative fMRI in concert with intraoperative cortical mapping has a number of distinct advantages. First, preoperative fMRI allows the neurosurgeon to make the decision of whether to attempt a resection, a stereotactic biopsy or not to operate at all.

13.2  Clinical Applications Fig. 13.2.12  Concordance between intraoperative cortical mapping and functional MRI finding. fMRI showed activity in the left temporal lobe (arrow) in a patient with a low grade glioma while performing a language (phonemic fluency) task. The fMRI results were confirmed by intraoperative cortical mapping, causing a speech arrest in that same area. The cross hairs show the location of speech arrest during intraoperative cortical stimulation

Second, a preoperative fMRI may influence the trajectory that a surgeon will take in approaching the tumor. For example, if the tumor were shown to be posterior to the motor cortex by fMRI, the surgeon could take their approach accordingly. Third, having a preoperative fMRI also decreases the chances of intraoperative surprises. Fourth, preoperative fMRI mapping could provide information about which patients should have an awake craniotomy. Information such as this can change the character of the entire operation including extending the craniotomy or having to worry about anatomic structures such as large arteries or veins that one thought could be avoided. One can avoid many of these issues with accurate preoperative functional MRI mapping of the eloquent cortices adjacent to the tumor. Figure 13.2.13 shows a schema for presurgical planning fMRI. 13.2.4.1.2  Cortical Reorganization Cortical reorganization can be defined as follows: as a disease process damages part of the brain, rendering it in­cap­ able of performing its function, another part of the brain may then take over that function. There are several important questions currently being asked regarding cortical reorganization. Is there reorganization of cortical function due to the growth of a brain tumor or other brain lesions in adults? Is there reorganization of cortical function due to rehabilitation after stroke? When and under what conditions does it occur? Can this be measured by fMRI? These concepts are important in clinic. Neuronal connections and cortical maps are continuously remodeled by our experience and may be beneficial for compensatory recovery after injury. Neuroimaging studies demonstrate altered cortical activation patterns that suggest cortical reorganization in areas that deal with motor function in

Fig. 13.2.13  A presurgical planning fMRI schema

1315

1316

13  Functional MRI

patients suffering from tumor (Reings et al. 2005), stroke (Calautti and Baron 2003), arteriovenous malformations (AVM) (Alkadhi et al. 2000), and other lesions. Cortical reorganization of motor function could be shown by a decrease in activity in the lesioned hemisphere, and increase in activity in the intact hemisphere or some combination of these effects. Reorganization has been shown to occur across hemispheres, where the ipsilateral primary motor area as well as non-primary motor areas, are recruited. Intrahemispheric reorganization is also thought to occur where increased functional activity is observed in the contralateral primary motor area or adjacent cortices (Calautti and Baron 2003; Alkadhi et al. 2000). The magnitude and anatomic location of activation depends on many factors, such as lesion size and location, as well as extent of therapy after the incident. There are two tumor cases described where one of the dominant language areas was translocated to the contralateral side. In this setting, translocation is defined as the switch of most cortical function to the homologous region in the contralateral hemisphere. In the first case (Holodny et al. 2002) the growth of the brain tumor in the left inferior frontal lobe led to an inability of this area to function properly. This led to the transfer of the functional Broca’s area to the contralateral side (Fig. 13.2.14). The other case (Petrovich et al. 2004) showed the translocation of Wernicke’s area in an adult to the contralateral side due to the presence of a long standing low-grade glioma. Both cases were documented by fMRI and the later surgically confirmed with direct cortical stimulation. Further, in the pre- and postoperative neuropsychological testing showed no new language deficits.

Translocation of one of the dominant language areas to the contralateral side has also been reported after strokes (Thulborn et al. 1999) and in arteriovenous malformations (Lazar et al. 2000). In these cases, as well as in the tumor presented above, the pathologic process developed in adulthood and involved only one of the two language areas. The mechanism by which such transfer occurs is unknown. Because disease can alter the map of language function in an unexpected fashion, fMRI has the potential to reveal cortical reorganization that could affect the neurosurgeon’s decision to offer surgery. These cases can serve to emphasize another important issue regarding the preoperative assessment of eloquent cortices adjacent to brain tumors. Assumptions regarding language dominance that are based solely on handedness may be misleading. This may result in an unnecessarily conservative treatment approach for certain patients with brain tumors in whom surgery is, in fact, safe and clinically desirable. These occurrences show that fMRI should routinely be done preoperatively in patients with lesions in the language cortex, particularly when brain tumors are deemed inoperable because of their proximity to essential language centers. 13.2.4.1.3  Radiation Treatment Planning fMRI-guided treatment planning provides a new approach to preserving eloquent cortex near brain tumors from unnecessary radiation exposure. fMRI may provide physicians with pre-therapeutic treatment planning along with direct 3D identification of eloquent cortices from anatomical and functional brain images (Liu et al. 2000; Witt et al. 1996). This new technique allows one to decrease the dose to adjacent eloquent cortices without decreasing the dose to the tumor. Studies have reported the incorporation of fMRI activation maps into the stereo­ tactic radiosurgery treatment planning to identify eloquent cortices and decrease the radiation dose to adjacent normal functional tissues (Liu et al. 2000). 13.2.4.2  Stroke

Fig. 13.2.14  fMRI data shows translocation of Broca’s area to the opposite side from its expected location, now occupied by the large tumor. Wernicke’s area is still on the expected side (Reprinted with permission from Topics in MRI (2004) 15:325–335, Lippincott, Williams & Wilkins)

Two methods of functional imaging that have contributed most to our understanding of cortical reorganization after stroke in humans are fMRI and PET. Using fMRI technique, a non-invasive and repeatable imaging method, it is possible to evaluate and visualize recovery after stroke in a longitudinal as well as cross sectional fashion. fMRI has been used for spatial mapping after functional recovery in motor and language performance (Rijntjes and Weller 2002). Recovery of lost function after a stroke has been associated with brain plasticity and has been one of most important issues in stroke therapy. The

13.2  Clinical Applications

Fig. 13.2.15  Activation volume in the stroke hemisphere during movement of a stroke-affected limb is related to degree of behavioral recovery. Patients with chronic stroke underwent fMRI scanning while alternating rest with tapping the index finger on the right, affected hand (Zemke et al. 2003). Six patients had full recovery, and five had excellent but only partial motor recovery.

Note that activation volume in the supplementary motor area was independent of recovery. However, the patients with partial recovery showed activation volume in the left primary sensory motor-premotor cortex that was 37% of the volume activated in patients with full recovery (Reprinted with permission from Stroke (2004) 35:2695–2698, Lippincott, Williams & Wilkins)

fMRI activation changes obtained from the longitudinal study can also be correlated with behavioral recovery. There has been stirring evidence that certain patterns of change in brain activity in motor or aphasia stroke are correlated with recovery of function, and these changes can be influenced by intensive therapy programs. In some models of stroke recovery, the contralateral hemisphere is employed leading to a reduction in the laterality of brain activity. Early reports emphasized a less lateralized pattern of activation after stroke than normal, i.e., the effect of stroke is to increase the extent to which both hemispheres are recruited rather than just the contralateral hemisphere (Cramer 2004). However, the idea that language compensation occurs interhemispherically has been challenged. In stroke patients, it has been shown that the degree of perilesional integrity on PET positively correlates with the patient’s functional prognosis. This model implies perilesional compensation ipsilateral to the lesion. Therefore, it may be that the mechanism of reorganization is a function of the type of lesion inducing the deficit.

ing motor networks, including the disconnected motor cortex in subcortical stroke and the infarct region after stroke (Fig. 13.2.15). fMRI studies have reported a shift in the cortical activation site in association with stroke recovery, showing an inferior shift (Cao et al. 1998) or a posterior shift (Rossini et al. 1998) in the activation site within sensorimotor cortex of the stroke hemisphere during recovery of motor function. Involvement of secondary and contralateral motor areas has been reported. There is an emerging view that the greater the participation of the ipsilateral motor network, the better the recovery. This hypothesis is supported by the increased BOLD fMRI activation of the ipsilateral primary motor cortex induced by motor rehabilitation (Calautti and Baron 2003).

13.2.4.2.1  Motor Stroke There are a number of fMRI investigations performed in cross sectional (Cramer 2000; Pineiro et al. 2001) and longitudinal studies (Marshall et al. 2000; Feydy et al. 2002). Both cross-sectional and longitudinal studies have demonstrated that lesioned brain is able to compensate for motor deficit. Rather than a complete replacement of function, the main mechanism underlying recovery of motor function may deal with increased activity in exist-

13.2.4.2.2  Aphasia Stroke fMRI studies offer the potential to better understand the neuronal mechanisms that aid in language recovery after stroke. Various fMRI language paradigms are used on stroke patients. Passive listening and silent reading were adapted mostly using visual or auditory stimulation delivery systems. Overt speech generation, overt repetition of heard speech or overt picture naming is rarely used in fMRI paradigms, even though it allows monitoring of patient’s task performance. This is due to the difficulty agrammatic patients experience in keeping their head still while speaking overtly. There are a number of longitudinal fMRI studies performed on aphasia stroke patients (Price and Crinion 2005; Crosson et al. 2005) to investigate how the functional recovery stage changes with treatment intervention (Fig. 13.2.16).

1317

1318

13  Functional MRI

Fig. 13.2.16  Pre- and post-intention treatment fMRI images through the frontal lobes. Participants generated a single word for a category presented acoustically during fMRI sessions immediately before and immediately after the intention treatment. Left is on the left side of the images. Pre-treatment images are shown for A008 (a) and for X030 (b). Post-treatment images are shown for A008 (c) and for X030 (d). Before the intention treatment, A008 showed bilateral lateral frontal activity, but preSMA activity was lateralized to the left hemisphere (a). After the intention treatment, A008’s lateral frontal activity was lateralized to the right hemisphere, and pre-SMA activity was no longer lateralized (c). In other words, A008 showed a shift of

both lateral frontal and pre-SMA activity toward the right hemisphere. Before the intention treatment, 100% of X030’s lateral frontal activity was in the right hemisphere, and pre-SMA activity (not shown) also was lateralized to the right hemisphere (b). After treatment, lateral frontal activity for X030 was reduced but still 100% lateralized to the right hemisphere. Pre-SMA activity was no longer significantly lateralized. Basal ganglia activity increased dramatically over pre-treatment activity and was strongly lateralized to the left hemisphere (d). y-Axis coordinates are in Talairach space (Reprinted with permission from Journal of Cognitive Neuroscience (2005) 17:392–406, MIT Press)

13.2.4.3  Epilepsy

in regards to hemispheric dominance for language and increasingly memory (Gaillard et al. 2002, Springer et al. 1999). Therefore, it is hoped that the much less invasive fMRI will replace the Wada test for hemispheric dominance.

The intracarotid sodium amobarbital procedure (IAP) also known as the Wada test has previously been the gold standard for establishing hemispheric dominance for language (Wada 1949) and memory in epilepsy patients. Typically, an anesthetic agent is injected into one hemisphere of the brain at a time while a neuropsychologist monitors the patient’s performance in a task. The precipitous inability to perform the task indicates that the anesthetized hemisphere is responsible for the function (Woermann et al. 2003; Rihs et al. 1999). However, due to its invasive nature, the need to replace the Wada test with non-invasive techniques has been recognized (Abou-Khalil et al. 2002). fMRI has shown excellent concordance with both Wada and intraoperative mapping

References 1. 2.

Abou-Khalil B, Schlaggar B (2002) Is it time to replace the Wada test? Neurology 59:160–161 Adcock J, Wise R, Oxbury J, Oxbury S and Mattews P (2003) Quantitative fMRI assessment of the differences in lateralization of language-related brain activation in patients with temporal lobe epilepsy. Neuroimage 16:954–967

13.2  Clinical Applications 3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15. 16.

17.

18.

Alkadhi H, Kollias SS, Crelier G et al (2000) Plasticity of the human motor cortex in patients with arteriovenous malformations: a functional MR imaging study. AJNR Am J Neuroradiol 21:1423–1433 Ammirati M, Vick N, Liao Y et al (1987) Effect of the extent of surgical resection on survival and quality of life in patients with supratentorial glioblastomas and anaplastic astrocytomas. Neurosurgery 21:201–206 Amunits K, Schleicher A, Burgel U, Mohlberg H, Uylings H, Zilles K (1999) Broca’s region revisited: cytoarchitectonic and intersubject variability. J Comp Neurol 412:319–341 Baciu M, Le Bas J, Segebarth C, Benabid A (2003) Presurgical fMRI evaluation of cerebral reorganization and motor deficit in patients with tumors and vascular malformations. Eur J Radiol 46:139–146 Beisteiner R, Lanzenberger R, Novak K, Edward V, Windischberger C, Erdler M et al (2000) Improvement of presurgical patient evaluation by generation of functional magnetic resonance risk maps. Neurosci Lett 290:13–16 Bookheimer S, Zeffiro T, Blaxton T, Gaillard P, Theodore W (2000) Activation of language cortex with automatic speech tasks. Neurology 55:1151–1157 Burton H, Snyder A and Raichle M (2004) Default brain functionality in blind people. Proc Natl Acad Sci USA 101:15500–15505 Calautti C, Baron J (2003) Functional neuroimaging studies of motor recovery after stroke in adults: a review. Stroke 34:1553–1566 Cao Y, Vikingstad E, George K, Johnson A, Welch K (1999) Cortical language activation in stroke patients recovering from aphasia with functional MRI Stroke 30:2331–2340 Cao Y, D’Olhaberriague L, Vikingstad E, Levine S, Welch K (1998) Pilot study of functional MRI to assess cerbral activation of motor function after poststroke hemiparesis. Stroke 29:112–122 Chee M, O’Craven K, Bergida R, Rosen B, Savoy R (1999) Auditory and visual word processing studied with fMRI. Human Brain Mapp 7:15–28 Cox RW (1996) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res 29(3):162–173 Cramer SC (2004) Functional imaging in stroke recovery. Stroke 35:2695–2698 Cramer S, Moore C, Finklestein S, Rosen B (2000) A pilot study of somatotopic mapping after cortical infarct. Stroke 31:668–671 Crosson B, Moore A, gopinath K, White K, Wierenga C, Gaiefsky, Fabrizio K, Peck K, Soltysik D, Milsted C, Briggs R, Conway T, Rothi L (2005) Role of the right and left hemispheres in recovery of function during treatment of intention in aphasia. J Cognitive Neuroscience 17:391–406 Edward V, Windischberger C, Cunnington R, Erdler M, Mayer D, Endl W, Beisteiner R (2000) Quantification of fMRI artifacte reduction by a novel plaster cast head holder. Human Brain Mapp 11:207–13

19. Feydy A, Carlier R, Roby-Brami A, Bussel B, Cazalis F, Pierot L, Burnod Y, Maier M (2002) Longitudinal study of motor recovery after stroke: recruitment and focusing of brain activation. Stroke 33:1610–1617 20. Field A, Yen , Burdette J, Elster A (2000) False cerebral activation on BOLD functional MR images: Study of low-, amplitude motion weakly correlated to stimulus. AJNR Am J Neuroradiol 21:1388–1396 21. Frahm J, Merboldt K, Hanicke W et al (1994) Brain or vein: oxygenation or flow? On signal physiology in functional MRI of human brain activation. NMR Biomed 7:445–53 22. Friston K, Ashburner J, Frith C, Poline J, Heather J, Frackowiak (1995) Spatial registration and normalization of images. Human Brain Mapping 3:165–189 23. Gaillard W, Balsam L, Xu B, Grandin C, Braniecki S, Papero P et al (2002) language dominance in partial epilepsy patients identified with an fMRI reading task. Neurology 59:256–265 24. Golby A, Poldrack R, Brewer J, Spencer D, Desmond J, Aron A, Gabrieli J (2001) Material-specific lateralization in the medial temporal lobe and prefrontal cortex during memory encoding. Brain 124:1841–1854 25. Hoeller M, Krings T, Reinges M, Hans F, Gilsbach J, Thron A (2002) Movement artifacts and MR BOLD signal increase during different paradigms for mapping the sensorimotor cortex. Acta Neurochir 144:279–284 26. Holodny A, Schulder M, Liu W, Wolko J, Maldjian J, Kalnin A (2000) The effect of brain tumors on BOLD functional MR imaging activation in the adjacent motor cortex: Implications for image-guide neurosurgery. AJNR Am J Neuroradiol 21:1415–1422 27. Holodny A, Schulder M, Ybasco A et al (2002) Translocation of Broca’s area to the contralateral hemisphere due to a growth of a left inferior frontal glioma. J Comput Assist Tomogr 26:941–943 28. Huang J, Carr T, Cao Y (2001) Comparing cortical activations for silent and overt speech using event-related fMRI. Human Brain Mapp 15:39–53 29. Jiang A, Kennedy D, Baker J et al (1995) Motion detection and correction in functional MR imaging. Human Brain Mapp 3:224–235 30. Johansen-Berg H, Dawes H, Guy C, Smith S, Wade D, Mattews P (2002) Correlation between motor improvements and altered fMRI activity after rehabilitative therapy. Brain 125:2731–2742 31. Kim M, Holodny A, Hou B, Peck K, Chaya M, Gutin P (2005) The effect of prior surgery on BOLD fMRI in the pre-operative assessment of brain tumors. AJNR Am J Neuroradiol 26:1980–1985 32. Kober H, Nimsky C, Möller M, Hastreiter P, Fahlbusch R, Ganslandt O (2001) Correlation of sensorimotor activation with functional magnetic resonance imaging and magnetoencephalography in presurgical functional imaging: a spatial analysis. Neuroimage 14:1214–1228

1319

1320

13  Functional MRI 33. Krings T, Reinges M, Erberich S, Kemeny S, Rohde V, Spetzger U, Korinth M, Willmes K, Gilsbach J, Thron A (2001) Functional MRI for presurgical planning: problems, artifacts, and solution strategies. J Neurol Neurosurg Psychiatr 70:749–760 34. Lazar R, Marshall R, Pile-Spellman J et al (2000) Interhemispheric transfer of language in patients with left frontal cerebral arteriovenous malformation. Neuropsychologia 38:1325–1332 35. Lee A, Glover G, Meyer C (1995) Discrimination of large venous vessels in time course spiral blood oxygen level dependent magnetic resonance functional neuro. Magn Reson Med 33:745–754 36. Liu W, Schulder M, Narra V, Kalnin A, Cathcart C, Jacobs A, Lange G, Holodny A (2000) Functional magnetic resonance imaging aided radiation treatment planning. Med Phys 27:1563–1572 37. Machulda M, Ward H, Borowski B et al (2003) Comparison of memory fMRI response among normal, MCI and Alzheimer’s patients. Neurology 61:500–506 38. Maldjian J, Schulder M, Liu W et al (1997) Intraoperative functional MRI using a real-time neurosurgical navigation system. J Comp Assist Tomogr 21:910–912 39. Marquart M, Birn R, Haughton V (2000) Multiple-event paradigms for identification of motor cortex activation. AJNR Am J Neuroradiol 21:94–98 40. Marshall R, Perera G, Lazar R, Krakauer J, Constantine R, DeLaPaz R (2000) Evolution of cortical activation during recovery from corticospinal tract infraction. Stroke 31:656–661 41. Moriyama T, Yamanouchi N, Kodama K (1998) Activation of non-primary motor areas during a complex finger movement task revealed by functional magnetic resonance imaging. Psychiatr Clin Neurosci 52:339–343 42. Peck K, Sunderland A, Peters A, Butterworth S, Clark P, Gowland PA (2001) Cerebral activation during a simple force production task: changes in the time course of the haemodynamic response. Neuroreport 12:2813–2816 43. Peck K, Moore A, Crosson B. Gaiefsky M, Gopinath K, White K, Briggs R (2004) Pre and post fMRI of an aphasia therapy: shifts in hemodynamic time to peak during overt language task. Stroke 35:554–559 44. Petrovich N, HolodnyA, Brennan C et al (2004) Isolated translocation of Wernicke’s area to the right hemisphere in a 62 year man with a temporo-parietal glioma. AJNR Am J Neuroradiol 25:130–133 45. Pineiro R, Pendlebury S, Johansen-Berg H, Mattews P (2001) Functional MRI detects posterior shifts in primary sensorimotor cortex activation after stroke: evidence of local adaptive reorganization?. Stroke 32:1134–1139 46. Preibisch C, Pilatus U, Lanfermann H (2003) Functional MRI using sensitivity-encoded echo planar imaging (SENSE_EPI). Neuroimage 19:412–421 47. Price C, Crinion J (2005) The latest on functional imaging studies of aphasic stroke. Curr Opin Neurol 18:429–434

48. Pronin I, Holodny A, Kornienko V et al (1997) The use of hyperventilation in contrast-enhanced MR of brain tumors. AJNR Am J Neuroradiol 18:1705–1708 49. Rao SM, Binder JR, Hammeke TA (1995) Somatotopic mapping of the human primary motor cortex with functional magnetic resonance imaging. Neurology 45:919–924 50. Reings M, Krings T, Rohde V, Hans F, Willmes K, Thron A, Gilsbach J (2005) Prospective demonstration of short motor plasticity following acquired central pareses. Neuroimage 15:1248–1255 51. Rihs F, Sturzenegger M, Gutbrod K, Schroth G, Mattle H, Determination of language dominance: Wada test confirms functional transcranial Doppler sonography, Neurology, 1999, 52:1591–1596 52. Rijntjes M, Weiller C (2002) Recovery of motor and language abilities after stroke: the contribution of functional imaging. Prog Neurobiol 66:109–122 53. Rossini P, Caltagirone C, Castriota-Scandlerbeg A, Cicinelli P, DelGratta C, Demartin M, Pizzella V, Traversa R, Romani G (1998) Hand motor cortical reorganization in stroke: a study with fMRI, MEG and TCS maps. Neuroreport 9:2141–2146 54. Sabbah P, Chassoux F, Leveque C, Landre E, Baudoin-Chial S, Devaux B et al (2003) Functional MR imaging in assessment of language dominance in epileptic patients. Neuroimage 18:460–467 55. Schulder M, Maldjian J, Liu W et al (1998) Functional image guided surgery of intracranial tumors located in or near the sensorimotor cortex. J Neurosurgery 89:412–418 56. Springer J, Binder J, Thomas H, Swanson S, Frost J, Bellgowan P, Brewer C, Perry H, Morris G,, Mueller W (1999) Language dominance in neurologically normal and epilepsy subjects: a functional MRI study. Brain 122:2033–2045 57. Stark C, Squire L (2001) When zero is not zero: the problem of ambiguus baseline conditions in fMRI. Proc Natl Acad Sci USA 98:12760–12766 58. Staudt M, Lidzba K, Grodd W, Wildgruber D, Erb Mabd Kragelohmann I (2002) Right-hemispheric organization of language following early leftsided brain lesion functional MRI topography. Neuroimage 16:954–967 59. Thickbroom G, Byrnes M, Archer S, Nagarajan L, Mastaglia F (2001) Differences in sensory and motor cortical organization following brain injury early in life. Ann Neurol 49:320–327 60. Thulborn K, Carpenter P, Just M (1999) Plasticity of language-related brain function during recovery from stroke. Stroke 30:749–754 61. Ulmer J, Krouwer H, Mueller W, Ugurel M, Kocak M, Mark L (2003) Pseudo-reorganization of language cortical function at fMR imaging: a consequence of tumor-induced neovascular uncoupling. AJNR Am J Neuroradiol 24:213–217 62. Van Oostende S, Van Hecke P, Sunaert S, Nuttin B, Marchal G (1997) fMRI studies of the supplementary motor area and the premotor cortex. Neuroimage 6:181–190

References 63. Wada J (1949) A new method for determination of the side of cerebral speech dominance: a preliminary report on the intracarotid injection of sodium Amytal in man. Igaku to Seibutsugaki 14:221–222 64. Ward NS, Brown MM, Thompson AJ, Frackowiak RSJ (2003) Neural correlates of outcome after stroke: a crosssectional fMRI study. Brain 126:1430–1448 65. Witt T, Kondziolka D, Baumann S, Noll D, Small S, Lunsford D (1996) Preoperative cortical localization with functional MRI for use in stereotactic setup. Magn Reson Imaging 14:1007–1012

66. Woermann F, Jokeit H, Luerding R, Freitag H, Schulz R, Guertler S, Okujava M, Wolf P, Tuxhorn I, Ebner A (2003) Language lateralization by Wada test and fMRI in 100 patients with epilepsy. Neurology 61:699–701 67. Yousry TA, Schmid UD, Jassoy AG (1995) Topography of the cortical motor hand area: prospective study with functional MR imaging and direct motor mapping at surgery. Radiology 195:23–29 68. Yousry TA, Schmid UD, Alkadhi H (1997) Localization of the motor hand area to a knob on the precentral gyrus. A new landmark. Brain 120:141–157 69. Zemke A, Heagerty P, Lee C, Cramer S (2003) Motor cortex organization after stroke is related to side of stroke and level of recovery. Stroke 34:e23–e28

1321

Chapter 14

Computer Aided 3D Radiation Planning Using MRI

14.1 

Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain  .. . . . .   1323 L.R. Schad

14.1  Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain L.R. Schad Accurate planning of stereotactic neurosurgery (Schlegel et al. 1982), interstitial radiosurgery, or radiotherapy (Schlegel et al. 1984) re­quires precise spatial information about the location and size of the lesion. MRI, because of its good soft-tissue contrast and multipla­nar tomographic format, is a logical choice for displaying the pertinent anatomy and pathology, and it seems to be ideally suited for defining the target volume and delineating critical structures at risk in radiotherapy treatment planning of the brain and also the body stem. However, the use of MRI alone as the basis for treatment planning is often not advisable because pixel intensities are unrelated to electron densities and an exact dose calculation is not possible; therefore, CT data are still needed. With correlation of images from different modalities, the excellent soft tissue contrast of MRI can be combined with electron densities evaluated by CT as well as physiological information measured by PET. However, an indispensable prerequisite for such image correlation is the measurement and correction of spatial image distortions on MRI. Depending on the individual MR system, inhomogeneities and nonlinearities induced by eddy currents dur­ing the pulse sequence can distort the images and pro­duce spurious displacements of the stereotactic coordinates in both the x–y-plane and the z-axis. If neces­ sary, these errors in po­sition can be assessed by means of two phantoms placed within the stereo­tactic guid­ance system—a “two-dimensional (2D) phantom” displaying

14

14.1.1 

2D-Phantom Measurement  . . . . . . . . . . .   1323

14.1.2 

3D-Phantom Measurement  . . . . . . . . . . .   1325 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1329

“pincushion” distortion in the image, and a “three-dimensional (3D) phantom” displaying dis­placement, warp, and tilt of the image plane itself. The pincushion distortion can be “corrected” by calculations based on modeling the distortion as a fourth-order 2D polyno­mial. Accurate spatial representation demands uniform main magnetic fields and linear orthogonal field gradients. Inhomogeneity of the main magnetic field and nonlinearity of the gradients produce image distortion (O’Donnell and Edelstein 1985). A significant source of these geometric artifacts is eddy currents produced during the imaging sequence. Correc­tion of these distortions is usually unneces­sary for clinical diagnosis but is important for stereotaxy and in planning radioisotopic or radiation therapy. This chapter deals with the as­sessment of and correction for geometric distortions of MR images using phantom meas­urements. 14.1.1  2D-Phantom Measurement The 2D phantom is used to measure the geometrical distor­tions within the imaging plane (Fig. 14.1.1a). It consists of a water-filled cylinder 17 cm in radius and 10 cm in depth, containing a rectangular grid of plastic rods spaced 2 cm apart and ori­ented in the z-direc­tion (i.e., perpendicular to the imaging plane). Since the exact positions (x, y) of these rods are known a priori (plus symbols in Fig. 14.1.1b), their positions (u, v) in the 2D phantom (multiplication symbols in Fig. 14.1.1b) reflect the geometric distortion in the ima­ging plane and may be used to calculate the coordinate trans­formation that mathematically descri­bes the distortion process. The measuring sequence used was a 3D-GRE technique (3D

1324

14  Computer Aided 3D Radiation Planning Using MRI

FISP see Chap, 2, Sect. 2.4.3; TR/TE/α 40/7/15°, image ma- tion is provided by the distortion vector, which measures trix 256 × 256, 64-partition slab in the z-direction, slab the discrepancy between the true (x,y)k position of a pin thick­ness 64 mm, FOV = 260 mm). In or­der to image the in the 2D phantom and its apparent position in the imreference points of the stereotactic localization system, age (u,v)k:dk = |(x,y)k – (u,v)k| (i.e., the vector distance FOV was restricted to 260 mm, resulting in an effective between the true and apparent positions). Reduction in voxel size of about 1 mm3. dis­tortion is demonstrated in Fig. 14.1.1c—a comparison The x–y-plane distortion of axial MR images of the of the distribution of dk in the uncorrected and corrected 2D phantom can be correc­ted (reducing dis­placements images. The distribution of dk in the cor­rected 2D-phanto the size of a pixel) by cal­culations based on modeling the dis­tortion as a fourth order two-dimensional polyno­ mial:

(14.1.1)

(14.1.2)

where (x,y) are the true pin positions of the 2D phantom (Fig. 14.1.1b; plus symbols), (u,v) are the dis­torted positions measured on the image (Fig. 14.1.1b; multiplication symbols), N (= 4) is the order of the polynomial, and M (= 249) is the total number of pin positions of the 2D phantom. Note that the origin of the u–v-coordinate system corresponds to the origin of the x–y-coordinate system and lies in the center of the image, the area free of distortion. Equations 14.1.1 and 14.1.2 applied to all M pin positions of the 2D phantom form a system of simulta­neous linear equations from which the coefficients Uij and Vij can be calculated and the dis­tortion correc­ted. Thereby, the selection of N = 4 is a compromise between computational burden and reduction in distortion. A more quantitative, direct measurement of distor-

Fig. 14.1.1   a The “two-dimensional (2D)” phantom for measuring the geometrical distor­tions within the imaging plane. It consists of a water-filled cylinder 17 cm in radius and 10 cm in depth, contai­ning a rectangular grid of plastic rods spaced 2 cm apart and oriented in the z direc­tion (i.e., per­pendicular to the imaging plane). b Typical example of a 2D-phantom measurement using a velo­city-compensated FISP sequence. The a priori known regular grid of the plastic rods (+ calculated points) is deformed to a pincush­ion-like pattern (× measured points) from which the 2D-distor­tion polynomial can be derived. Note that the origin of the coordinate sys­tem of both the calculated and measured points lies in the center of the im­age, the area free of distortion. c Distributions of lengths of distortion vec­tor (magnitude of positional errors) of the 2D-phantom pins in the un­cor­rected (solid line) and in corrected image (dotted line) calculated with N = 4 (expansion order of the 2D polynomial). Magnitude of positional errors in uncorrected image is about 2–3 mm at outer range of phantom (i.e., at the position of the reference points of the stereo­tactic guidance system). These errors are reduced to about 1 mm in corrected image, which corre­ sponds to the pixel resolution of the image (from Schad 1995)

14.1  Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain

tom image shows that the positional errors are reduced stereotactic zero plane) was determined by the distance from about 2–3 mm (Fig. 14.1.1c, solid line) to about between two reference points pro­duced by the obliquely 1 mm across the entirety of the 2D phantom (Fig. 14.1.1c, oriented plastic tubes of the guidance system (Schad et dotted line). No further attempts were made to re­duce al. 1987a) (Fig. 14.1.3b). The stereo­tactic x- and y-coordithis re­sidual error since this approximates the dimen- nates were measured with re­spect to the stereotactic zero sions of the image pixels (256 × 256 ma­trix). point (midpoint of the guidance sy­stem) (Fig. 14.1.3c). The manually meas­ured exact stereotactic coordi­nates of the plastic sphere from CT are within about 1 mm of the MR-based coordinates, whereas dis­crepancies of about 14.1.2  3D-Phantom Measurement 2–3 mm appeared mainly in the z-coordinate, without In contrast to CT, where the imaging plane is defined correction of geomet­ric distortion. The magnitude of the mechanically by the rotating ring containing the detector error in the z-coordinate increases with the z-co­ordinate. system and X-ray source, the imaging plane in MRI is This is related in part to the pincushion-like distortion defined by the gradient system and the local resonance pattern of the axial images, since z-coordi­nates are calfrequency of the protons (see Chap. 2). After correcting culated from and therefore ultimately dependent on the for distortion in the imaging plane, the three-dimen- fidelity of the x–y-com­ponents. In both the uncorrected sional position of the im­aging plane has to be assessed and corrected images, the calculated z-compo­nent cor­ with the 3D phantom (Fig. 14.1.2a). This is nec­essary be- responds to the distance between the x–y-components of cause inhomogeneities of the gradient fields that define two reference points. Ide­ally, and in the cor­rected images, the imaging plane can pro­duce deformation or til­ting this distance increases linearly with the true z-compoof the plane in space not apparent in the 2D-phantom nent. However, in the case of the uncorrected images the meas­urements. For the measurements, the 3D phan- distance, due to the pincushion effect, which is not negtom was fixed in the stereotactic guidance system, and ligible at the po­sition of the refe­rence points, increases axial-orientated MR images were obtained and initially nonlinearly—the z-component measured lying on an arc corrected using the 2D polynomial. These corrected im- rather than on a proper straight line. Hence, the dispar­ity ages display a regular pattern of points emanating from between these two measure­ments increases as the z-coor­ the water-filled boreholes (Fig. 14.1.2c). Each reference dinate increases. A further check of the correction method can be perpoint produ­ced by a rectangular borehole is encircled by several measurement points coming from neighboring formed by comparison of anatomical structures seen in oblique boreholes. The distance between the reference CT and corrected MR images (i.e., correction of anatomiand measurement points is a direct measure of the z-coor­ cal images by phantom measurements). Thereby the podinate of the imaging plane at the reference point (Fig. sition of an artery of the brain (middle cerebral artery) 14.1.2b). In this manner, the z-coor­dinates of the imag- has been assessed in the MR image corrected by phaning plane were measured at every reference point and the tom measurements described above and compared to the imaging plane was reconstructed and its position, shape, corresponding CT image using the same stereotactic sysand orientation assessed. Figure 14.1.2d is a typical ex- tem (Schad et al. 1987a, b). The comparison proves that ample of the initial reconstruction of the imaging plane after correction the stereotactic coordinates correspond showing a nearly horizontal imaging plane with some to the same anatomic focus. After correction, the accudiscrepancies in z-components to the order of 1 mm, racy of geometric information from MR images is only which approximates again the dimensions of the image limited by the pixel resolution of the image (= 1 mm). pixels (1-mm slice thickness). Mathematical correction Thus, this correction method allows the accurate transfer of these discrepancies in the z-direc­tion was not pursued of anatomical/pathological information and target point since properly adjusted shims avoid de­formation or tilt- coordinates into the isocenter of the linear accelerator so ing of the imaging plane. A tilting of the 3D phantom of essential for the stereotactic radiation technique or for about 3° from the trans­verse to the sagittal and coronal stereotactic-guided operation planning. planes can be clearly detected. Patient positioning is performed using a special steThe precision of MR stereotaxy and the importance reotactic head frame with different patient fixation either of geometric corrections may be tested by as­sessing the using a special mask (Fig. 14.1.4a) or by means of skull stereotactic coordinates of a plastic sphere (1-mm inner fixation using carbon-fiber pins. This allows a reproducdiameter) in a watermelon, with and without correction ible evaluation of coordinates of lesions (i.e., tumors or and comparing these with those measured by CT. For arteriovenous malformations) with respect to the referthis purpose the watermelon was fixed in the stereotac- ence system defined by the stereotactic head frame at tic guidance system and the stereotactic coordinates of different reproducibility depending on the kind of fixathe plastic sphere were measured (3D FISP se­quence) tion and organ of interest. The accuracy of the mask fixafrom axial MR images (Fig. 14.1.3a). The stereotactic z- tion of the head is within about 1–2 mm, whereas that coordinate (i.e., the distance of the image plane from the of skull fixation using carbon-fiber pins is within about

1325

1326

14  Computer Aided 3D Radiation Planning Using MRI

14.1  Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain 9 Fig. 14.1.2  a The “three-dimensional (3D)” phantom mounted in the stereotactic guid­ance system for measuring 3D position of MR imaging plane. This is com­prised of a regular grid of water-filled rectangular boreholes, with oblique water-filled boreholes in between. This produces a pattern of reference points surrounded by measurement points in axial image from which the 3D position of the imaging plane can be reconstructed. b Schematic illus­tration of 3D-phantom measurement. The distance d between reference point (rectangular water-filled borehole) and measurement point (oblique water-filled borehole) is a direct measure of the z-coordinate of the imaging plane at the re­ference point, since the angle between oblique water-filled borehole is α = 2 arctan(0.5). Systematic discrepancy in the distance measurements of d1 > d2 > d3 > d4 > d5 > d6 would be detected if the imaging plane were tilted. c Axial image of the 3D phantom. The 3D position of the imaging plane can be reconstructed from the measured points. d Typical example of the 3D position of the imaging plane with properly adjusted shims. Deviations in z-direction are reduced to about 1 mm, which ap­proxi­mates again the dimensions of the image pixels (1-mm slice thick­ness). Mathematical correc­tion of these discrepancies in the z-direc­tion was not pursued since properly adjusted shims avoid deformation or tilting of the imaging plane. e Tilting of the 3D phantom of about 3° from transversal to sagittal and coronal direction can be clearly detected (from Schad 2005)

7 Fig. 14.1.3  a The “watermelon phantom” mounted in the stereotac­tic guidance system for measuring the precision of MR stereo­taxy. A plastic sphere (1-mm inner diameter) was implanted in the watermelon and the stereotac­tic coordi­nates of the plastic sphere have been evaluated in transaxial images at CT and MR. b Evaluation of the stereo­tactic z-coordinate. Plexiglas squares embed­ded with a steel wires for CT or b plastic tubes filled with Gd-DTPA solution for MR are attached to the stereotactic head frame (arrow) and give refer­ence points in tran­saxial images. The z-coordinate of an image is deter­mined by the distance between two reference points. The angle between the reference tu­bes is α = 2arctan(0.5). The cros­sing point of two tubes lies in the zero-plane of the ste­reotactic head frame. c Schema­tic drawing of a transaxial image showing the tube re­ference points of the stereotactic gui­dance system. The z-coordinate directly measu­res the di­stance of the im­age from the stereo­tactic zero plane. The stereotactic x- and y-coor­dinates are measured with respect to the midpoint of the guidance system (= stereotactic zero point). These reference points serve as landmarks in all imaging modalities for the accurate superimposition of CT, MRI, and PET seen in Fig. 14.1.4b (from Schad et al. 1987a)

1327

1328

14  Computer Aided 3D Radiation Planning Using MRI Fig. 14.1.4  a Patient positioning by mask fixation of the head in the stereotactic guidance system in the head coil of the MR scanner. b Definition of the target vol­ume on a set of CT, PET (18FDG), and MRI (T1- and T2-weighted) images in a pa­tient with astrocytoma grade II. The target volume can only be defined by the therapist interac­tively with a digitizer. Thereby the computer automatically calculates the stereotactic z-coordinate (= slice position) and the stereotac­tic coordi­nate system in all axial slices with the help of reference points seen as landmarks in the images. After correction of geometrical distortions in MRI data, an accurate transfer of features such as target volume, calcu­lated dose distribution or organs at risk can be done from one set of imag­ing data to another. Complicating factors such as pixel size, slice position, or slice thickness differences are taken into account by the stereotactic co­ordinate system (from Schad 2001)

1 mm. Reproducibility is significantly reduced for fixa- in the images, from which the stereotactic coordinate systion of the body stem since fixation of the patient is much tem can be derived. Since the geometry of these reference more difficult and less accurate than fixation of the head points is known a priori, the information on the position and breathing motion of organs and lesions is difficult to of the individual lesion can be evaluated with respect to compensate for or suppress. The relation of the stereotac- the stereotactic coordinate system of the patient. Theretic frame, i.e. patient reference system, to the coordinate fore, information on the size, shape, and localization of system of the MRI or CT scanner is defined by special the target volume can be transferred precisely from the Plexiglas squares embedded with steel wires for CT, plas- imaging modalities to the stereotactic coordinate system tic tubes filled with Gd-DTPA solution for MRI, or ra- defined by the stereotactic frame. dioactive solution for PET. These Plexiglas squares are Metabolic data provided by PET imaging offer imattached to the stereotactic frame and serve as landmarks portant information for the delineation of tumors that is

14.1  Image Correlation and Coregistration of MRI, CT, and PET in Stereotactic Treatment Planning of the Brain

complementary to CT and MR imaging. With the help of the stereotactic reference points all imaging modalities are represented in the same anatomic focus, providing the therapist with an improved definition of target volume on a set of CT, MRI, and PET images in the treatment planning process (Fig. 14.1.4b).

4.

5.

References 1.

2.

3.

O’Donnell M, Edelstein WA (1985) NMR imaging in the presence of ma­gnetic field inhomoge­neities and gradient field non­linearities. Med Phys 12:20–26 Schad L, Lott S, Schmitt F, Sturm V, Lorenz WJ (1987a) Correction of spatial distortion in MR ima­ging: a prerequisite for accurate stereotaxy. J Com­put Assist Tomogr 11:499–505 Schad LR, Boesecke R, Schlegel W, Hartmann G, Sturm V, Strauss L, Lorenz WJ (1987b) Three di­mensional image correlation of CT, MR, and PET studies in radiotherapy treatment plan­ning of brain tumours. J Comput Assist Tomogr 11: 948–954

6.

7.

8.

Schad LR (1995) Correction of spatial distortion in magnetic resonance imaging for stereotactic op­eration/treatment planning in the brain. In: Hartmann GH (ed) Quality assurance program on stereotactic radiosurgery. Springer, Berlin Heidelberg New York, pp 80–89 Schad LR (2001) Improved target volume characterization in stereotactic treatment planning of brain lesions by using high-resolution BOLD MR-venography. NMR Biomed 14:478–483 Schad LR (200) Magnetic resonance imaging for radiotherapy planning. In: Schlegel W, Bortfeld T, Groscu AL (eds) New technologies in radiation oncology. Springer, Berlin Heidel­berg New York, pp 99–111 Schlegel W, Scharfenberg H, Doll J, Pastyr O, Sturm V, Netze­band G, Lorenz WJ (1982) CT-Images as the basis of opera­tion planning in stereotactical neurosurgery. In: Procee­d­ings of 1st Internatio­nal Symposium on Medical Imaging and Image In­terpretation. Silver Spring: IEEE Computer Society, pp 172–177 Schlegel W, Scharfenberg H, Doll J, Hartmann G, Sturm V, Lorenz WJ (1984) Three dimen­sional dose planning using tomo­graphic data. In: Proceedings of 8th International Con­ference on the Use of Com­puters in Radiation Therapy. Sil­ver Spring: IEEE Com­puter So­ciety, pp 191–196

1329

Chapter 15

Clinical Spectroscopy

15.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1331 W. Semmler; H.-P. Schlemmer

15.2 

Basics  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1332

15.2.1 

Nuclei for MRS in vivo  . . . . . . . . . . . . . . .   1332

15.2.1.1  1H MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1332 15.2.1.2  31P MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1332 15.2.1.3  13C MRS  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1332 15.2.1.4  19F MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1332 15.2.1.5  Other Nuclei (23Na, 9Li, 39K)  . . . . . . . . . .   1332 15.2.2 

Technical Requirements  . . . . . . . . . . . . . .   1332

15.2.2.1  Static Magnetic Field  . . . . . . . . . . . . . . . . .   1332 15.2.2.2  Radiofrequency System  .. . . . . . . . . . . . . .   1333 15.2.3 

Spectra Analysis  . . . . . . . . . . . . . . . . . . . . .   1333

15.2.4 

Localization Techniques  . . . . . . . . . . . . . .   1334

15.2.4.1  Localization by Means of Surface Coils    1334 15.2.4.2  Single-Volume Techniques  .. . . . . . . . . . .   1334 15.2.4.3  Spectroscopic Imaging (Chemical-Shift Imaging)  .. . . . . . . . . . . .   1336 15.2.5 

Spin–Spin Coupling and Double-Resonance Techniques  .. . .   1336

15.2.5.1  Spin–Spin Coupling  .. . . . . . . . . . . . . . . . .   1336

15.1  Introduction W. Semmler; H.-P. Schlemmer Intensive nuclear resonance signals can be gained from freely moving protons of cells and fatty acids that are used for MR imaging and can be used to receive morphologic and functional information. Metabolic processes in living tissue can be observed if the spectral resolution is improved for protons and if the frequency range is broadened also for other nuclei for example phosphorus-31 (31P), carbon-13 (13C), fluorine-19 (19F), potassium-39

15

15.2.5.2  Double-Resonance Techniques  .. . . . . . .   1338 15.2.6 

Limitations of MRS in vivo  .. . . . . . . . . . .   1339

15.2.6.1  Detection Sensitivity for Cellular Metabolites  .. . . . . . . . . . . . . .   1339 15.2.6.2  Absolute Quantification of Metabolite Concentrations  . . . . . . . . .   1342 15.3 

Clinical MRS  .. . . . . . . . . . . . . . . . . . . . . . .   1342

15.3.1 

1H MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1342

15.3.2 

31P MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1345

15.3.3 

13C MRS  . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1348

15.3.4 

19F MRS  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1348

15.4 

Clinical Applications in Selected Organs and Tumors  .. . . . . .   1350

15.4.1 

Brain  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1350

15.4.2 

Liver  .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1354

15.4.3 

Heart  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1356

15.4.4 

Skeletal Muscles  .. . . . . . . . . . . . . . . . . . . . .   1358

15.4.5 

Urogenital Tract  . . . . . . . . . . . . . . . . . . . . .   1360

15.4.6 

Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1360 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1369 Suggested Reading  . . . . . . . . . . . . . . . . . . .   1379

(39K), and others. This experimental method, using highfrequency-resolution nuclear resonance signals, is called magnetic resonance spectroscopy (MRS). Application of this method in living tissue (in vivo) makes possible to observe non-invasive biomolecules and concentrations of metabolites as well as intracellular pH and pharmacokinetic parameters. In this chapter only some basics are explained necessary to understand the today’s application of MRS in vivo in humans. For more advanced description of basics in NMR physics see the list “Suggested reading” at the end of the chapter.

1332

15  Clinical Spectroscopy

15.2  Basics

15.2.1.4  19F MRS

15.2.1  Nuclei for MRS in vivo

The gyromagnetic ratio of the 19F nucleus is comparable to that of protons (cf. Table 2.2.1), and therefore the resonance frequency of 19F is close to protons, minimizing technical challenges, e.g., transmitter, receiver, and coils may have similar layouts and constructions. This predestines 19F nuclei for MRS in vivo. However, the physio­logic concentration of MR visible fluorine in the human body is extremely low and below the detection limits. Consequently, 19F MRS is only possible after exogenous administration, resulting in a background-free resonance signal.

For in vivo MRS in humans, spin 1/2 nuclei are preferred. These are 1H (gyromagnetic ratio of protons: γ1H = 2.675 · 10–8 rad s–1T–1), 13C (γ13C = 0.252 · γ1H), 19F (γ19F = 0.941 · γ1H) and 31P (γ31P = 0.405 · γ1H) used. The relevant parameters for MR experiments are given in Chap. 2, Table 2.2.1. 15.2.1.1  1H MRS The proton has highest sensitivity of all stable atomic nuclei and very high concentration in tissue. On the other hand, the chemical shift is small (approximately 10 ppm) and therefore 1H spectroscopy demands high spectral resolution. This high spectral resolution demands high magnetic fields, high magnetic field homogeneity, and sufficient eddy current compensation to minimize signal distortions. The MR signal originates mainly from protons of tissue water as well as from triglycerides of fatty tissues. The proton signals of other metabolites with their low concentrations and small chemical shifts make meas­ uring more complex. 1H MRS in vivo necessitates pulse sequences that the dominate proton resonances of tissue water selectively suppresses to make possible to detect metabolites of interest. 15.2.1.2  31P MRS Second to protons spectroscopy, 31P MRS can be relatively easily performed in humans in vivo. This is due to the relatively high MR sensitivity of 31P, the high chemical shift (approximately 200 ppm) and to the observed metabolites, which are relevant metabolites of the high-energy metabolites and intermediates of the phospholipid metabolism. The spectra can be easily evaluated and interpreted. 15.2.1.3  13C MRS 13C is one of the most interesting MRS nuclei because carbon is present in almost all biochemical compounds. However, its MR sensitivity is low due to the low gyromagnetic ratio (cf. Table 2.2.1) and the low abundance of the isotope (1.1 %), resulting in general low signal-tonoise (S/N) spectra. Exogenous administration of selectively 13C-labeled compounds (e.g., 13C-labeled glucose) can downsize the detection limitations (Beckmann et al. 1991). The line splitting caused by strong scalar spin– spin coupling between 13C and 1H nuclei results in more complex spectra and makes difficult identification and assignments of lines (cf. Chap. 15.2.5).

15.2.1.5  Other Nuclei (23Na, 9Li, 39K) 23Na has second-highest MR signal in living tissue; however, only a single line of the free Na+ ions can be observed in MRS with little information. To date, no clinical useful MR spectroscopic examination using 23Na has been published. On the other hand, the strong 23Na resonance allows for fast MR imaging. Different relaxation times for extra- and intracellular Na+ are observed. Current clinical research is directed to make use of these differences of the 23Na relaxation times to receive information about the Na+/K+ pump. Only few in vivo patient studies were performed with spin 3/2 nuclei 7Li and 39K, both nuclei possessing electric quadrupolar moments, which inter alia introduce additional relaxation effects and make detection of the resonances more complicated (Komoroski 2005). 15.2.2  Technical Requirements 15.2.2.1  Static Magnetic Field In contrast to MRI, which can be performed at magnetic fields less B0 < 1.5 T, high-resolution MRS requires higher fields. The S/N ratio and spectral resolution increases with the magnitude of the magnetic field. The overwhelming number of examination in humans were performed at a field B0 = 1.5 T and in the recent past, at B0 = 3 T. At some centers, where whole-body systems up to 10.4 T are available, first MRS experiments were carried out at ultra-high magnetic fields (>3 T) (Mangia et al. 2006; Du et al. 2007). An important prerequisite for high MR spectral quality is the homogeneity of the external magnetic field, which has to be much better than that for MR imaging. Inhomogeneous resonance line spread spoils the spectral resolution significantly. The line width is determined by the full width at half maximum (FWHM): ∆ν1/2 = 1/(π T2*), with T2* being the effective spin–spin relaxation time. For this reason, the local field homogeneity of the external magnetic field B0 is optimized be-

15.2  Basics

fore MRS examinations, i.e., by means of auxiliary coils producing additional magnetic fields, the line widths of the water protons ∆ν1/2(1H) can be minimized. This is called the shim procedure. 15.2.2.2  Radiofrequency System For proton imaging the radiofrequency (RF) system has to operate at only one frequency; this is ν1H = 63.866 MHz at 1.5 T (transmit and receive). For multinuclear MRS an RF system is necessary with a wide frequency range to excite the nuclear spins of different isotopes/elements, which have all their own resonance frequency (B1 field) (cf. Table 2.2.1; e.g., at 1.5 T: ν31P = 25.881 MHz, ν13C = 16.062 MHz, etc.). The sensitivity is strongly dependent on the probe head, consisting of the RF antenna (coil) and the preamplifier. Double-resonant coils make possible sequential acquisition of proton images (1H imaging) and MRS with other nuclei than 1H. This is useful to control the localization of the volume of interest (VOI). Frequently used coils are surface coils for regions near the body surface and volume resonators (e.g., saddle, bird cage, Helmholtz coils, etc.) for deeper-lying structures. For application of X–(1H) MR double-resonance spectroscopy, a second RF channel is inevitable (B2 field). Both RF resonant circuits of the double-resonant coils have to be decoupled by RF filters. 15.2.3  Spectra Analysis The number of clinical MRS studies increased in the last years; however, approved standard procedures do not exist. Numerous approaches for evaluation of in vivo MR spectra were developed, which are based on different preposition and boundary conditions and that use also prior knowledge about the resonances of the metabolites (Bartha et al. 1999; Mierisová et al. 1998; Provencer 1993, 2001; Slotboom et al. 1998; Soher et al. 1998; Stanley et al. 1995; Zandt et al. 2001). In principle the evaluation is possible in the frequency (Mierisová et al. 2001) and the time domain (Vanhamme 2001) or a combination of both methods (Slotboom et al. 1998; Pels et al. 2006). To evaluate sets of spectra acquired by spectroscopic imaging, pattern recognition methods as well as neural networks are in use (cf. el-Deredy 1997; Opstad et al. 2007; Hagberg 1998; Kelm et al. 2007). The evaluation of the MR spectra in vivo in the frequency domain is the older method, having the advantage to review visually the resonance lines of the metabolites after Fourier transformation of the free-induction decay (FID). Quantification of such spectra simply by integrating the resonance line (Meyer et al. 1988) (summation of the signal in a particular frequency range) is in

most instances not adequate and even incorrect, because baseline distortions, resonance line overlap, and low S/N ratio influence the result of the integration. This is true especially in phosphorous spectroscopy in vivo. The determination and quantification of the spectral parameter can be optimized by fitting the data to model functions (in ’t Zandt et al. 2001).The evaluation in the frequency domain often requires several interactive steps (Bovée et al. 1992): • Improvement of the digital resolution by zero filling (extending of the FID by filling data points with zeros) • Improvement of the S/N ratio by digital filtering (multiplication of the FID with a damping function of Lorentzian or Gaussian profile) • Phasing to achieve a pure absorption signal of the resonance (0 or first order) • Elimination of artifacts and signal contributions of immobile molecules by base line correction (mostly spline or polynomial function) • Fitting of the data to a model function yielding the parameters: resonance position, resonance width, and resonance area (= resonance integral) The phasing and the base line correction are especially prone to user-dependent systematic errors, which restrict the objectivity and reproducibility of this kind of data evaluation. The fit function and the experimental data of frequency domain evaluation can be displayed simultaneously. The result of the frequency-domain fitting can be easily visually controlled, and results can be used in the next iterative step of the evaluation. However, for quantification of the parameter, the evaluation algorithm for in vivo spectroscopic data should work automatically and robustly—also in case of low S/N ratio and overlap of resonance lines. Evaluation in the time domain circumvents some of the above-mentioned drawbacks of the frequency-domain fitting. Advantages are first, that no start parameters have to be defined and second, no interactive steps are necessary. This reduces the user-dependent systematic errors that can arise from phase and baseline corrections in the frequency domain. Furthermore, signal contributions of fast relaxing components responsible for the broad background signal can be eliminated. However, a great disadvantage of the linear prediction and singular value decomposition (LPSVD) method is the high error probability at low S/N ratio. The extracted concentration values for metabolites deviating strongly from those determined in the frequency domain (Kreis et al. 1993a). The integration of prior knowledge (e.g., resonance positions) results in a more robust fitting, especially for MR spectra acquired with short echo times (Mierisová et al. 1998). The automatic evaluation of spectroscopic imaging data has special requirements (Soher et al. 1998). These data have often low S/N ratio and broad background

1333

1334

15  Clinical Spectroscopy

signals. The automatic differentiation of the relative signal contributions of the metabolite resonances from the broad, unknown background signal may be difficult because the S/N ratio and the background signal may be distributed differently, e.g., signals from the surface can contain high signals of the subcutaneous fatty tissue; signals from nearby the bone–air interfaces may be influenced by susceptibility artifacts, etc. Furthermore, the selective water-suppression pulse may have spatially different and unforeseeable influence on the signal. Parameterized model functions to simulate background signal are not available because the reason for and the strength of the background signal cannot be predicted (see Fig. 15.3.1c). Chemical-shift imaging (spectroscopic imaging) is increasingly applied. MRS imaging requires the evaluation of hundreds of MR spectra per patient, a formidable task in clinical routine. An alternative evaluation strategy for MR spectra that reduces user interaction is based on pattern recognition methods. Supervised pattern recognition methods require a training data set, i.e., a representative collection of example MR spectra that have been assigned to one of several predefined tissue classes by an expert. The machine classifier then learns a mapping from spectral patterns to classes that imitates the expert’s decision on previously unseen MR spectra. In this way, quantification can be avoided (Kelm et al. 2007). Once trained, the machine classifier can predict the probability of a volume element belonging to a certain tissue class, such as tumorous or healthy tissue. The corresponding probability maps can be computed automatically from MRS images and can be overlaid transparently on conventional MR imagery, such as T2-weighted MR images (cf. Fig. 15.2.1). Recent algorithmic advances permit the

determination of tumor probability maps, with much robustness with respect to artifacts and patient-to-patient variability. The evaluation and fitting of MR spectra in vivo are discussed intensively by Vanhamme et al. (2001) and new approaches are described by Opstad et al. (2007). 15.2.4  Localization Techniques To receive MR signals from a defined region of the body, special localization techniques are necessary. Out of the many MR localization techniques for spectroscopy in vivo proposed and developed, only a few have been frequently used for volunteer and patient examinations: among these are surface coils, single-voxel techniques (e.g., ISIS, STEAM, double spin echo) and chemical-shift imaging (CSI) (= spectroscopic imaging [SI]). 15.2.4.1  Localization by Means of Surface Coils The most simple localization technique is the application of surface coils. This method is applied in MRS examinations of skeletal muscles, liver, and other tissues. The penetration depth of such antenna systems is roughly given by the diameter of the coil. With this technique a defined tissue volume (VOI) can be delineated. The drawback is the decay of the sensitivity as function of depth (inhomogeneity of the magnetic components of the RF field), resulting in high signal contribution of surface near tissue and low contribution of deeper-lying tissue (cf. Fig. 15.2.2). An improved delineation of the VOI can be achieved using selective saturation pulses as proposed by Sauter et al. (1987) or selective excitation of a defined slice (Bottomley et al. 1984). 15.2.4.2  Single-Volume Techniques Single-voxel localization techniques render possible the measurement of MR spectra out of cuboids or cubes. For 1H MRS, two sequences are commonly used for localization (STEAM, PRESS). Localization of 31P MRS spectra is rarely done using single-voxel localization techniques ISIS; nowadays mainly spectroscopic imaging is used for localization (see below). 15.2.4.2.1  STEAM

Fig. 15.2.1  Tumor probability map obtained with pattern recognition from 1H MRS images in the prostate. Red areas indicate volume elements with typical tumor spectra, whereas green areas contain spectra as usually found in healthy prostate tissue (Kelm et al. 2007)

The most commonly used single-voxel localization technique for 1H MRS is the stimulated echo acquisition mode (STEAM) (Fig. 15.2.3a; Frahm et al. 1987, 1989). This technique uses a train of three slice-selective 90° pulses, which produce a stimulated echo. This acquisition

15.2  Basics 7 Fig. 15.2.2  Localized MR spectroscopy with surface coils. The surface coil (SFC) is positioned above the region of interest. After excitation of the spins with a RF pulse, the MR signal (S) is detected (1-pulse sequence); magnetic field gradients (G) are not applied. The coil’s sensitivity decreases with the distance to the coil’s winding (B1 inhomogeneity)

technique allows for a localized shim (= minimization of the inhomogeneity of the static magnetic field B0 in the VOI), resulting in resonance line width of the tissue water less than 0.1 ppm (FWHM 6 Hz at B0 = 1.5 T) in the human brain. The suppression of the intensive resonance of the water protons is realized by low bandwidth, socalled chemical-shift–selective (CHESS) pulses, which are applied prior to the three slice-selective 90° pulses. Only a high homogeneity of B0 field, a careful compensation of eddy currents induced by the gradient pulses, and adequate use of spoiler gradients for suppression of the unwanted echo signals allow for the acquisition of high resolution localized in vivo 1H MR spectra of excellent quality.

Fig. 15.2.3  Localized MR spectroscopy with single voxel techniques. a By means of a series of three-slice selective 90° pulses (stimulated echo-pulse sequence [STEAM]: 90°–TE/2–90°–TM– 90°–TE/2-STE) the signal will be read out of a cuboidal or cubical volume of interest (VOI) (gray slice-selecting gradient, R rephasing gradient, D dephasing gradient). The resonance of the

water protons will be selectively suppressed by a narrow-band Gaussian pulse (CHESS). b The second Hahn spin echo of a series of slice selective 90° and 180° pulses (double spin-echo [DSE]: 90°–TE1/2–180°–TE1/2–SE1–TE2/2–180°–TE2/2–SE2) is the signal of the spins, which are localized in the area of overlap of the three orthogonal slice planes

1335

1336

15  Clinical Spectroscopy

15.2.4.2.2  PRESS The PRESS technique allows higher 1H MRS signal intensity in comparison to the STEAM sequence. The localized Hahn spin echo is produced by the slice-selective pulse train 90°–T1–180°–T2–180° (Fig. 15.2.3b). 15.2.4.3  Spectroscopic Imaging (Chemical-Shift Imaging) Spectroscopic imaging (Brown et al. 1982; Maudsley et al. 1983) uses field gradients for spatial encoding of the MR signals almost in the same manner as in MR imaging; however, during readout, no gradient is applied, and so the spectral information is preserved. The 3D spectroscopic imaging needs three phase-encoding gradients. The gradient strength is increased stepwise and for each combination of the gradient values, a MR signal is acquired. The resulting phase keeps the spatial information. For a complete set of a 3D spectroscopic image and n gradients values (phase-encoding steps) n3 measurements have to be performed. The measuring time will be reduced significantly when a 2D data set is acquired during the initial slice excitation.

than in Fig. 15.2.4a, so the spin–spin coupling becomes visible. The methylene resonance is split into a quartet with the intensity ratio 1:3:3:1, and the methyl resonance is split into a triplet with the relative intensities 1:2:1. As the protons of the hydroxyl (-OH) group are exchanged rather quickly between the different molecules, the averaged spin–spin interaction is zero, and the -OH resonance line is not split.

In order to explain the principle of the spin–spin coupling, we will consider a molecule consisting of two nuclei, A and X, which are coupled via an electron bound (e.g., hy-

15.2.5  Spin–Spin Coupling and Double-Resonance Techniques As mentioned already, MRS in vivo is characterized by a low S/N ratio of the spectra and the limited spectral resolution. Reasons for this are local susceptibility differences in tissues and an overlap of the manifold metabolite signals, as well as line broadening and line splitting due to spin–spin interaction. 15.2.5.1  Spin–Spin Coupling The scalar spin–spin interaction (J coupling) causes line broadening (long-range, weak coupling) and splitting of resonance lines in multiplets (short-range, strong coupling). To understand this splitting of the lines, some basics concerning spin–spin coupling should be mentioned here. In many MR spectra, one can observe, apart from the chemical shift, a split of resonance lines in multiplets with typical distribution of intensity. This multiplet split is caused by the so-called spin–spin coupling. This term represents a weak interaction between the nuclei of a molecule, which is mediated indirectly (namely via the electrons of the chemical bound). Example: Fig. 15.2.4b shows a proton spectrum of ethanol (CH3–CH2–OH), with a decisively higher spectral resolution

Fig. 15.2.4  Proton-spectrum of ethanol (CH3–CH2–OH). a These three 1H resonances can be linked to the protons of the hydroxyl, methylene, and methyl groups. The ratio of the area under the resonance lines is 1:2:3 and thereby corresponds to the number of protons in the three atom groups. b If the spectral resolution is improved, one can observe (apart from the chemical shift) a splitting of the resonance lines into multiplets (spin– spin coupling). The methylene group is split into a quartet with the intensity ratio 1:3:3:1, and the the methyl group into a triplet with the intensity ratio 1:2:1

15.2  Basics

drofluoric acid, HF). The magnetic moment μ X of the nucleus X is linked to a weak magnetic field, which disturbs the electron bound. Due to this change of the electron orbital, an additional field at the nucleus A is generated, which is proportional to the magnetic moment μX; the resonance frequency thus depends on the magnetic spin quantum number mx of the nucleus X. For a given value Ix, one can observe the splitting of the resonance line of nucleus A into a multiplet of 2 Ix + 1 lines, with the intensity of the resonance lines being quasi-identical, as the 2 Ix + 1 spin states of nucleus X occur practically of equal probability (see Chap. 2). Analogously one can consider the splitting of the resonance line of nucleus X. The gaps between neighbored lines are equidistant and identical in both spectra; it is called the (scalar) coupling constant JAX and is given in Hertz (Hz). As the spin–spin interaction is determined solely by the magnetic moments μ A and μ X,

the magnitude of the coupling constant JAX remains—as opposed to the chemical shift— independent of strength of the static magnetic field.

Fig. 15.2.5  Spin–spin coupling in a two-spin system AX(IA = 1/2, IX = 1). Without spin–spin coupling, one only observes single resonance lines in the A and the X spectra at a resonance frequency of νA or νX (ν = ω/2π). However, if the two nuclei are coupled, the A resonance will split into a triplet and the X res-

onance into a doublet. The splitting scheme is determined by the number of spin states of the adjacent nucleus. The distance JAX between the adjacent lines is equidistant and equal in both spectra

Example: Fig. 15.2.5 shows the splitting of an A–X spin system with the spin quantum numbers IA = 1/2 and IX = 1. The A resonance is split into a triplet, the X resonance into a doublet.

These considerations can also be transferred to more complicated spin systems. As an example, we will consider the coupling of a spin 1/2 nucleus A with three equivalent spin 1/2 nuclei X (AX3 spin system), for which there are all together eight possibilities of spin orientations. If the two permitted spin states of a single X nucleus are denoted ↑ (μXZ = + 1/2) and ↓ (μXZ = – 1/2), respectively, then these states are ↑↑↑, ↑↑↓ , ↑↓↑, ↓↑↑, ↑↓↓, ↑↓↑, ↓↓↑, and ↓↓↓. As the spin states ↑↑↓, ↑↓↑, ↓↑↑, and ↑↓↓, ↓↑↓ ,

1337

1338

15  Clinical Spectroscopy

↓↓↑ are energetically equivalent to one another, one can observe a splitting of the A resonance into a quartet with the relative intensity distribution 1:3:3:1 (Fig. 15.2.6). Example: To render the splitting pattern of ethanol in Fig. 15.2.4 intelligible: The two protons of the methanol couple with the three CH3 protons of the methylene group and vice versa, so the methylene resonance will be split into a quartet and a triplet. In opposition to the H–H coupling, the H–C and H–O coupling is generally not detectable with MRS, because the isotopes C and O with a natural abundance of 98.89 and 99.76% do not possess a magnetic moment (see Chap. 2, Table 2.2.1).

of a spectrum of higher order can therefore be rather difficult. For this reason, special double-resonance techniques for spin–spin decoupling were developed. If, for instance, during the detection of an FID of the A nucleus in a coupled A–X spin system, one irradiates an additional RF field B2(t) at the resonance frequency of the nucleus X, then the multiplet splitting of the A resonances will disappear. 15.2.5.2  Double-Resonance Techniques

However, the aforementioned laws only apply for weakly coupled spin systems, with the coupling constants J being small compared with the differences in the resonance frequencies due to chemical shift (spectra of first order). If this condition is not fulfilled (spectra of higher order), then the relative signal intensities within the multiplets cannot be described by the explained laws. The analysis

Double-resonance techniques, which are widely applied in high-resolution NMR, can improve S/N ratio and spectral resolution. By means of this technique not only the resonance spectrum of the nuclei of interest (e.g., 13C, 19F, or 31P) is excited, but also are the protons; however, the protons are not observed. Two methods of heteronuclear X–(1H) double-resonance techniques are in use: spin decoupling (broadband 1H decoupling) and the dynamic nuclear spin polarization (nuclear Overhauser effect [NOE]), both of which are significantly increasing the spectral quality (Bachert and Bellemann 1992; Bachert et al. 1992; Bachert-Baumann et al. 1990; Gonen et al. 1997; Krems et al. 1995; Luyten et al. 1989; Bachert 1997; Heerschap et al. 1989).

Fig. 15.2.6  Splitting scheme of the A resonance for a coupled four-spin system AX3 (IA = IX = 1/2). The nucleus A is surrounded by three equivalent adjacent nuclei X, for which there are eight different possibilities of spin orientations. They are marked by arrows, with an arrow characterizing the two per-

mitted spin orientations of a single X-nucleus (↑ :(μXZ  = + 1/2), (↑ :(μXZ  = – 1/2)). The spin-combinations, ↑↑↓, ↑↓↑, ↓↑↑ and ↑↓↓, ↓↑↓, ↓↓↑ are energetically equal to one another, so that one can observe in the A spectrum a quartet with the intensity distribution of 1:3:3:1

Remark: If a nucleus A couples to several magnetically nonequivalent nuclei (e.g., to an A–M–X spin system with coupling constants JAM and JAX), the A resonance splits into a multiplet of multiplets. The splitting pattern can be constructed successively in this case.

15.2  Basics

These effects are especially pronounced in 13C MRS because the direct coupled protons in the carbon compounds, e.g., (–CH2–) groups of the fatty acids, are strongly coupled to the 13C nuclei. The proton spin system is irradiated during detection of the signal of the nuclei of interest (e.g., 13C) to decouple both spin systems. The multiplets collapse if the second RF field (B2) has high-enough intensity. The total signal, which is distributed of all components of the multiplet, condenses into a single, narrow resonance (Figs. 15.2.7, 15.2.8). PME and PDE resonances of in vivo 31P MRS can be resolved by means of 1H decoupling (Fig. 15.2.9). Selective irradiation with the resonance frequency of protons induces an enhancement of the signal of the nuclei of interest (13C, 19F 31P) if nuclear spins can relax due to heteronuclear dipole–dipole interaction (Bachert 1997). Signal intensity changes observed in liquids in double-resonance experiments of dipolar coupled nuclei are NOEs. In contrast to spin decoupling experiments, the selective saturation of the resonance transitions of the proton spin system can be performed before the detection of the nuclei of interest. The resulting signal intensity enhancement depends on the delay time between excitation of the two spin systems (Bachert and Bellemann 1992; Bachert 1997). (Remark: If short repetition times are used, also in spin decoupling experiments NOE can be observed.) In the same manner, the NOE can be used to improve S/N ratio in 13C and 19F MRS (Ende and Bachert 1993; Gonen et al. 1997; Krems et al. 1995).

15.2.6  Limitations of MRS in vivo MRS has physical limitations, originating from its very low physical sensitivity. Besides others, these limitations concern the detection limits for metabolites, measurement times, spatial and temporal resolution (Bottomley 1989). The sensitivity of nuclear magnetic resonance is caused by the very low energy of the quanta, several orders of magnitude smaller than those of optical or γspectroscopy, e.g., the energy of the MR detectable photons are about 1012 times smaller than those annihilation photons used for positron emission tomography (PET). The S/N ratio can be improved applying longer meas­ urement times; however, for clinical purposes extended measurement times are not practicable—they have to be as short as possible. Therefore other methods to improve the S/N ratio have been developed and applied. They were described in Sect. 15.2.5. 15.2.6.1  Detection Sensitivity for Cellular Metabolites The sensitivity of MR is determined by the gyromagnetic ratio (γ), the natural abundance of the isotope, and the concentration of the spins in tissue. The concentration of water in tissue is about 40–50 mol/kg, whereas the concentration of phosphocreatine (PCr) in below 40 mmol/kg. The resonance signal of PCr

Fig. 15.2.7  31P–(1H) double resonance. 31P MR spectra of an aqueous solution methylene-diphosphonic acid (MDPA, chemical formula CH2[P(O)(OH)2]2) acquired in a whole body-MR imager at a magnetic field of B0 = 1.5 T (a) without and (b) with 1H decoupling. The coupling constants of the scalar 31P-1H spin–spin interaction of this molecule is JPH = 20.1 Hz. The splitting of the triplet (a) is cancelled by the excitation of the protons during detection of the signals of the phosphorus spins using WALTZ 8-pulse trains. Additional a signal amplification is observed [31P–(1H)–NOE] (Bachert-Baumann et al. 1990)

1339

1340

15  Clinical Spectroscopy

Fig. 15.2.8  13C-MR spectroscopy. a 13C MR spectrum in vivo of the calf of a healthy volunteer, acquired using proton decoupling. All signals originate from carbon nuclei of free fatty acids in the tissue. The splitting of the resonances into multiplets (due to scalar spin–spin coupling of 13C spins with protons in fatty acid molecules) and the superposition of signals make difficult the assignment and quantification of the resonances. b 13C–(1H)

double–resonance, proton-decoupled 13C MR spectrum of a healthy volunteer’s calf. Exciting protons with the WALTZ 4-pulse sequence significantly results in a decoupling and simplified signals of 13C spins. All resonances can be assigned to 13C nuclei in molecular groups of fatty acids (Ende and Bachert 1993). c In comparison, proton-decoupled 13C–(1H) MR spectrum of vegetable oil

15.2  Basics Fig. 15.2.9  31P–(1H) double resonance. 31P MR spectrum in vivo of the calf of a healthy volunteer acquired without (a) and with (b) 1H decoupling (B0 = 1.5 T). The excitation of the protons during the detection of the free-induction decay of the phosphor spins results in a reduction of the line width in the range of the phosphodiester (PDE) and a signal amplification of all resonances. [phosphomonosters (PME)] c Lorentzian line fit of a 1H-decoupled 31P MR spectrum (section) of the calf muscle. Within the spectral range of the PDE the signals of glycerophosphorylcholine (GPC) and glycerophosphorylethanolamine (GPE) are resolved

1341

1342

15  Clinical Spectroscopy

is smaller compared with the resonance signal of protons, not only because of the smaller gyromagnetic ratio γ of 31P compared with that of 1H , but also because of the lower physiological concentration, which is about four orders of magnitude smaller than that of protons. To achieve about the same S/N ratio as with protons in 31P MRS in vivo, the VOI has to be increased significantly. The most abundant carbon isotope is 12C, which possesses a nuclear spin of zero and in turns does not contribute to the MR signal. In comparison to the protons (1H), the sensitivity of the rare 13C isotope is about four orders of magnitude smaller. Besides the above-mentioned facts, the mobility of the molecules is important, described by the correlation time τc of the motion to which the nuclei of interest are bound. The motion of the molecules is restricted in membranes and solid structures (bone), this means τc is long. The resulting reduction of spin–spin relaxation time T2* results in a strong line broadening, so that the individual resonances cannot be resolved. In consequence, metabolites with high concentrations in tissue may not be detected if the motion is restricted. 15.2.6.2  Absolute Quantification of Metabolite Concentrations A serious problem of the MRS in vivo is the absolute quantification of metabolite concentrations of MRS-detectable compounds. The area under the resonance line is a measure of the metabolite concentration. This is true only if the flip angle (proportional to B1 field strength) is constant in the entire sensitive volume of the coil including the area, where the reference sample is located and furthermore the longitudinal relaxation of the magnetization is complete (repetition time TR > 5 · T1). In practice, local differences of the B1 field strength are present, and in turn the calibration with a reference sample located inside the coil but outside of the organ of interest causes errors. In other words, from the intensity ratio of the reference sample and the metabolite, signal of the organ cannot be directly concluded regarding the metabolite concentration of the tissue. Theoretical calculations of the coil’s B1 field distribution are possible using the Biot-Savart law (interrelation of magnetic field and current in the coil wire); however, the B1 field distribution in tissue can hardly be estimated or measured, and additional effects (e.g., dielectric effects) modify the field distribution and are not precisely known. Not only the spatial inhomogeneity of the B1 field, but also the long spin–lattice relaxation times (e.g., T1 ≅ 6 s for PCr) hinder the precise determination of the absolute phosphorous metabolite concentrations in tissue by means of 31P MRS. MRS examinations in vivo with repetition times of 30 s result in excessively long mea-

surement times, which cannot be used in patients. Examinations with shorter TR shorten measure times. Additionally, however, the influence on signal intensities must be corrected for partial saturation introducing further uncertainties and systematic errors (cf. Bottomley et al. 1996 for quantification of phosphorous metabolite quantification in vivo). Localization by means of slice-selective excitation (e.g., STEAM or 2D CSI) introduces further problems when using nuclei with strong chemical shifts (e.g., 13C, 19F). Depending on the chemical shift, the excited VOI for different metabolites are spatially shifted, making quantification even more complicated and prone to systematic errors. For 1H MRS, where the latter problem is not significant, robust methods for the absolute quantification of tissue metabolites are being developed, taking into account the above mentioned difficulties (Ernst et al. 1993; Husted et al. 1994; Kreis et al. 1993a; Michaelis et al. 1993). For 1H MRS in the brain, the reproducibility of the determination of brain metabolites was demonstrated (Marshall et al. 1996; Geurts et al. 2004). The concentration of fluorine-containing compounds in patients during and after chemotherapy with 5-fluoruracil was estimated from 19F MRS data (Li et al. 1996b; Schlemmer et al. 1994). 15.3  Clinical MRS At present, 1H MRS in vivo is most common clinically applied in brain (Ross and Michaelis 1994; cf. Sect. 15.4.1; see also Chap. 3, Sect. 3.2.1) and in prostate (Swanson et al. 2006; cf. Sect. 15.4.5; see also Chap. 7, Sect. 7.2). 15.3.1  1H MRS In brain high local magnetic field homogeneities can be achieved due to the relatively homogeneous organ and the fact that the region of interest can be located in the isocenter of the magnet. Furthermore, only suppression of the water resonance is necessary, because under normal conditions no mobile lipids are present. In Fig. 15.3.1b, c, localized 1H MR spectra are displayed, acquired using short and long repetition time, respectively. The resonances of N-acetyl-l-aspartate (NAA) at 2.02 and 2.70 ppm, creatine and phosphor creatine ((P)Cr) at 3.03 and 3.93 ppm, choline (Cho)-containing substances at 3.22 ppm, and myo-inositol (mI) at 3.56 ppm can be recognized and assigned. Resonance of glutamine and glutamate (Glx) are located in the region of 2.1–2.5 ppm and 3.6–3.9 ppm (Michaelis et al. 1991; Ross and Michaelis 1994a). The resonances of the freely rotating methyl groups of Cho, Cr, NAA, lactate (Lac), and lipids (Lip) do have at 1.5 T T1 relaxation times of 1 s and T2 relaxation times

15.3  Clinical MRS

Fig. 15.3.1  1H MR spectroscopy in vivo. a Transversal T2weighted MR image (SE 2,800/90) of the brain of a 26-year-old volunteer. The 1H MRS localized tissue regions are marked (VOI = 2 × 2 × 2 cm3). b Localized water-suppressed 1H MR spectrum in vivo of the normal brain in a. The residual resonances of the water protons (chemical shift δ = 4.70 ppm) are outside of the displayed frequency range. The 1H MRS-detectable molecules groups of the metabolites are given. Resonances of free fatty acids (chemical-shift range δ = 0 – 2.4 ppm) are generally not observed in 1H MR spectra in vivo of the brain. Compared with normal findings, tumor spectra show increased signal intensities of choline (δ = 3.22 ppm) and reduced of NAA (δ = 2.01 ppm).

The β-methyl-proton doublet of lactate at δ = 1.33 ppm is resolved (head coil, stimulated echo-pulse sequence (STEAM), VOI = 3 × 3 × 3 cm3, repetition time TR = 1,500 ms, echo time TE = 45 ms, middle interval TM = 30 ms, number of acquisitions AC = 256, B0 = 1.5 T). c 1H MR spectrum, however, with longer echo time TE = 270 ms. The resonances of Cho, [P]Cr, and NAA are dominating. The other resonances are suppressed because of the short T2 relaxation times (Bruhn et al. 1989). d 1H MR spectrum of the gastrocnemius muscle of a healthy volunteer. The resonances of the tissue water protons and the CH2 and CH3 protons of the lipids (free fatty acids, triacylglycerides) superimpose the weak signals of other metabolites

1343

1344

15  Clinical Spectroscopy

of more than 200 ms, whereas the other metabolites have significantly shorter echo times. For this reason the resonances of Cho, Cr, NAA, Lac, and Lip can be observed using long echo times (TE = 135 ms or 270 ms) (cf. Figs. 15.3.1c and 15.4.15b, c, respectively). Application of long echo times has the additional advantage of: • Disturbing components with short T2 relaxation times are suppressed because those components are already fully relaxed. • Water suppression becomes more effective. • Overlap of resonance lines is less pronounced. • The baseline distortions are significantly reduced. The presence of large amounts of NAA in brain tissue was discovered by Tallan et al. in 1956 (Tallen et al. 1956). NAA was mainly identified in neurons (Tallen 1957; Birken and Oldendorf 1989). Therefore it is assumed that 1H MR spectroscopic NAA signals observed in brain are originating from neurons and axons, even though hints are present that allow to assign NAA also to progenitor cells of the oligodendrocytes (Cheng et al. 2001; Ross and Bluml 2001). To date the function of NAA is not known free of doubt (Birken and Oldendorf 1989). Recent observations suggest that NAA and its metabolite N-acetylaspartylglutamate (NAAG) may play an important role in the neuronal-glial information exchange (Baslow 2000). Besides possible function of NAA as part of protein and neurotransmitter biosynthesis, it was also discussed that NAAG may be a storage form of glutamate in the neurons (Birken and Oldendorf 1989; Choi 1988). A loss of NAA signal intensity was observed in all neurological diseases that are accompanied by a loss of neurons and/or axons, e.g., ischemia (Federico et al. 1998; Gillard et al. 1996; Graham et al. 1995; Saunders 2000), dementia (Ross et al. 1997, 1998), tumors (McKnight et al. 2001; Nelson et al. 1999; Preul et al. 1998), multiple sclerosis (Arnold et al. 2000; Fu et al. 1998; Gonen et al. 2000), or temporal lobe epilepsies (Ende et al. 1997; Capizzano 2001). For this reason, NAA is commonly considered to be a so-called neuronal marker. Of all observed 1H MR spectroscopic metabolites, NAA has the most significant diagnostic value, because it yields information about the neuronal integrity. For instance, the intensity of the NAA signal is prognostic for traumatic, inflammatory, and ischemic CNS injuries of newborns and children (Holshouser et al. 1997). Resonances of (P)Cr have identical chemical shift and appear at two positions in 1H MR spectra, at 3.03 and at 3.93 ppm. For quantification, only the resonance at 3.03 ppm is of importance, because the other resonance is close to the water-suppression pulse and may be influenced by that. Adenosine triphosphate (ATP) is the most important source of chemical energy, PCr, however, is another energy-rich phosphate, which is a readily available energy reserve for the cell, when fast and high energy demand takes place. ATP can be provided by trans-

formation of Cr and vice versa, regenerated using the enzyme creatine kinase. PCr is the reserve for the high energy phosphates in brain and in muscles mainly and buffers the ATP–adenosine diphosphate (ADP) equilibrium, because in these organs the fastest changes in energy consumption occur. (P)Cr resonance is often used as an internal reference. However, in ischemia and brain tumors significant changes of the (P)Cr resonance integral were observed (Kinoshita and Yokota 1997; Lowry et al. 1977; Saunders et al. 2000; Tien et al. 1996). Furthermore, the creatine concentration is depending on the complex biosynthesis of creatine in liver and kidneys as well as on the transport processes to brain and muscles (Ross and Michaelis 1994a). In normal circumstances, the almost entire choline is bound in membrane lipids (phosphatidylcholine). These molecules are mainly bound in cell membranes, resulting in a restricted mobility and hence short T2 relaxation times, and therefore cannot be detected in vivo. The resonance of the Cho-containing compounds include signal contributions from water-soluble cholinic compound as phosphor-Cho (PC), glycerophosphorylcholine (GPC), free Cho, CDP-choline, acetylcholine, and Cho plasmalogens. However, the results of patient studies (Michaelis et al. 1993; Ross and Michaelis 1994a) and animal experiments (Dross and Kewitz 1972; Tunggal et al. 1990; Barker et al. 1994; Jimenez et al. 1995) demonstrating that the predominant signal contribution is due to PC and GPC. Separate observation of PC and GPC is not possible by means of 1H MRS, but by using proton decoupled 31P MRS (Fig. 15.2.9). The sugar alcohol myo-Inositol (mI) is known as an astrocytal marker and may have the function of an intracellular osmolyte (Häussinger et al. 1994). mI is the center of complex metabolic changes and therefore the interpretation of the changes in the mI concentration appears extremely difficult; furthermore, mI resonance contains contributions of mI monophosphate, scylloinositol, and glycine, making interpretation even more complex (Laubenberger et al. 1998; Ross and Michaelis 1994a; Ross and Bluml 2001). Resonances of the methyl protons of glutamate and glutamine are detectable at 2.1–2.5 ppm (β- and γ- methyl groups) and 3.6–3.8 ppm (α-methyl groups) in 1H MR spectra acquired with short echo times (Bovée 1991; Ross 1991). Differences in the concentration of these metabolites are often observed in patients with hepatic encephalopathy and dementia (Chamuleau et al. 1991; Kanamori et al. 1996; Moats et al. 1994; Ross et al. 1996, 1998). Under pathological conditions, free mobile fatty acids (Lip) are observed in the range of 0.9–1.3 ppm and the resonance of lactate (Lac) at 1.3 ppm as well. Usually these membrane lipids are part of the biomembranes and therefore because of their structure, restricted in their mobility and hence, the T2 relaxation times are too short to be detected in vivo. However, resonances

15.3  Clinical MRS

of mobile lipids are observed in necrotic, inflammatory, and neoplastic processes (Graves et al. 2001; Gotsis et al. 1996; Jüngling et al. 1993; Koopmans et al. 1993; Sijens et al. 1996; cf. Fig. 15.4.15b). Lactate is the end product of the anaerobic glycolysis and can be observed in pathological conditions by means of 1H MRS in vivo (Veech 1991). Detection of lactate is an indication of insufficient supply of oxygen (e.g., ischemia, tumors) (Graham et al. 1995; Schlemmer et al. 2001; Sijens et al. 1996) or a disturbance of the respiratory chain (De Stefano et al. 1995). The resonances of the lactate consists of the characteristic β-methyl proton doublet with a line splitting of J ≅ 7 Hz, which shows phase inversion at echo times of TE = 135 ms (Fig. 15.4.15c). In cerebral abscesses and brain tumors, in addition to lipids and lactate the resonances of the amino acids alanine, valine, leucine, isoleucine, as well as the resonances of acetate and succinate are detected meonsistently (Preul et al. 1998; Grand et al. 1999: Kim et al. 1997). 1H MRS spectra of muscles demonstrate methyl protons in subcutaneous fat and in muscular fatty septa (Fig. 15.3.1d). 15.3.2  31P MRS MRS examination in humans can be performed with ease not only with protons (1H), but also with 31P nuclei. This is based on one hand at the relatively high MR sensitivity of phosphorus (cf. Table 2.2.1), and on the other hand on the fact that the 31P MR spectra can be doubtlessly interpreted. 31P MRS allows the in vivo study of cellular energy metabolism, phospholipid metabolism, as well as the intermediate steps thereof. 31P MR spectra usually display the resonances of α-, β- and γ-nucleoside-5′-triphosphate (NTP), phosphocreatine (PCr) (not in liver spectra), and inorganic phosphate (Pi). The resonance bands of the intense signals of phosphomonoesters (PME) and phosphodiesters (PDE) are composed of several components (cf. Fig. 15.3.2). The assignments are deduced from tissue extracts measured at high field strength B0. In these experiments phosphoethanolamine (PE) and phosphocholine (PC)—precursors of the membrane phospholipids sn-3-phosphatidylethanolamine and sn-3-phosphatidylcholine, respectively—are identified as signal components of the PME resonance of tumor spectra. The degradation products of theses phospholipids —glycerophosphorylethanolamine (GPE) and glycerophosphorylcholine (GPC)—are assigned to the PDE resonance band (Daly et al. 1987; Luyten et al. 1989). In general, the resonances of nucleoside-5′-diphosphates (NDP), adenosine-5′-monophosphates (AMP), inosine-5′-monophosphates (IMP), nicotine-amide-adenine-dinucleotides (NAD+, NADH, NADP+, NADPH), and glucose-6′-phosphates (G-6-P) cannot be resolved in 31P MR spectra in vivo at field strength of 1.5 T.

NTP resonance mainly contains the signal of ATP. ATP is the most important cellular carrier of chemical energy and is the counterpart of phosphocreatine, a temporary storage for high-energy phosphate groups (Fig. 15.3.3). ATP is produced via the aerobic glycolysis within the cell (cytoplasm) and oxidative phosphorylation within the mitochondria if the oxygen supply is sufficient. The concentration level of ATP is kept almost constant. If energy demand is high, ATP is quickly synthesized via a creatine kinase reaction: PCr2– + MgADP– + H+ ↔ ATP + Cr + MgATP2–. Simplified: ATP ↔ ADP + Pi PCr + ADP + H+ ↔ ATP + Cr In summary: PCr ↔ Cr + Pi The increase of the hydrolyzed product ATP triggers the glycolysis, and—if oxygen supply is not sufficient (e.g., ischemic conditions)—ATP is synthesized via the anaerobic metabolic pathway with its end product lactate (Lac). The resulting lactate acidosis of the tissue can be detected non-invasively by means of 31P MRS because the resonance frequency of the phosphorous nuclei of the inorganic phosphate is pH dependent. On the other hand, lactate can be directly detected using 1H MRS. The relative chemical shift of Pi in respect to PCr—the latter resonance position is pH independent—can be used to determine the intracellular pH value in vivo and noninvasively: The orthophosphorus acid (H3PO4) dissociates in water and gives H2PO4–, with pKa ≅ 2; this compound dissociate further and gives HPO42– with pKa ≅ 7. The pKa value is related to the pH value (measure for hydrogen ion concentration) via the Henderson-Hasselbalch equation:

pK a = −log = −log

     

     = −log

[A −] − [H+ ] [AH] [A −] − log[H + ] [AH]

[A− ] + pH . [AH]

This relation may be modified for the application in 31P MRS. The proton exchange between both dissociation products is very fast, so only one resonance of Pi in the MR spectrum is observed. The chemical shift of the Pi resonance relatively to the PCr resonance (δobs) mirrors the dissociation equilibrium of the pH value. The Henderson-Hasselbalch equation can be rewritten and δAH and δA being the protonated and dissociated form of the molecule, respectively (Moon and Richards 1973):

1345

1346

15  Clinical Spectroscopy

Fig. 15.3.2  Signal components (schematically) of a 31P MR spectrum of muscle tissue (a) and tumor (b). For some of the metabolites, the molecule structure is given. The sum spectra are given at the bottom of the figure, and the superposition of all

31P MR-detectable phosphometabolites is presented. The area under the resonance line represents the concentration of the respective metabolite if ideal conditions exist (like full relaxation [TR ≥ 5 · T1], B1 homogeneity during signal acquisition, etc.)

15.3  Clinical MRS

Fig. 15.3.3  Energy metabolism of the cell (schematically). Adenosine-5′-triphosphate (ATP), the primary and universal carrier of the free energy in all organisms is formed during glycolysis and—with higher effect—during oxidative phosphorylation. The enzymes catalyzing these processes are localized within the

. The titration curve for living tissue cannot be determined with sufficient precision and consequently, the absolute determination of the pH value in vivo may be prone to systematic errors. However, relative changes of the pH value can be measured with fairly high precision. A major drawback of 31P MRS is the low spatial resolution, and therefore the pH value can be determined only for larger tissue volumes. Thus the pH value is an average value of several tissue compartments. In the literature slightly different expressions for the function of pH(δobs) in the physiological pH range are given:

cytoplasm and the mitochondria, respectively. ATP is used for biosynthesis, signal amplification, active transport, and muscle contraction as well as other cell motions. Phosphocholine act as a temporal storage for energy rich phosphate groups

, with δobs being the observed difference of the chemical shift (in parts per million) of Pi and PCr (Ng et al. 1982). This equation yields an intracellular pH value of 7.11 ± 0.05 for the gastrocnemius muscle of humans, measured by means of 31P MRS. It is worth mentioning that complex bindings of high-energy phosphates to metal ions causes chemical shifts in 31P MR spectra. The frequency difference of α- and β-NTP signals depends on the complexation of ATP with Mg2+ ions. Therefore, if the β-NTP resonance is used as an internal chemical shift reference, additional errors may occur.

1347

1348

15  Clinical Spectroscopy

Fig. 15.3.4  Schematics of the metabolic pathways of 5-fluora-uracil (5-FU). 5-FU is catabolized via the intermediates 5-fluoro-5,6-dihydrouracil (DHFU) and α-fluoro-β-ureidopropanoic acid (FUPA) and excreted as α-fluoro-β-alanine (FBAL) via the kidney. The cytotoxic effect is twofold; first via inhibition of the thymidylate synthetase, cutting off the deoxythymidine5′-monophosphate (dTMP) synthesis and, conse­quently, the

DNA replication and furthermore by integration of 5-fluorouridine-5′-triphosphate (FUTP) into rRNA. The resonance of the intermediates of the catabolic pathways cannot be resolved in vivo and assigned as 5-FUranuc (5-fluoro-uracil-nucleosides and nucleotides) (cf. Fig. 15.3.5; for further abbreviation see abbreviation list in that figure)

15.3.3  13C MRS

15.3.4  19F MRS

Because of the limitations mentioned above (cf. Sect. 15.2.1.3), exogenous administration of selectively 13Clabeled compounds (e.g., 13C-labeled glucose) might be necessary. In 13C MR spectra in vivo prominent and characteristic are the resonances of free fatty acids (triacylglycerides), e.g., methylene signals with chemical shifts δ between 23  and 34 ppm (relative to the chemical shift reference tetramethylsilane at δ = 0). 13C MR spectra can be simplified using special pulse techniques of the highresolution nuclear magnetic resonance, e.g., broadband proton decoupling (Ende and Bachert 1993; Fig. 15.2.8). At present, clinical 13C MR spectroscopy has only minor influence on patient examinations, diagnosis, and treatment follow-up. 13C MRS is used mainly in experimental set ups and some examinations in humours (Ross et al. 2003).

The first 19F examinations in vivo were performed in patients with liver metastases of colorectal cancer after administration of the cytotoxic agent 5-fluorouracil (5FU) (Glaholm et al. 1990; Schlemmer et al. 1994; Semmler et al. 1990; Wolf et al. 1987; Bachert 1998). These studies demonstrated that the anabolic pathway for 5-FU synthesis in vivo of the cytotoxic 5-FU nucleotides and nucleosides (5-FUranuc) is very limited. More effective is the catabolism of 5-FU in patients. 5-FU is converted by in several steps into α-fluoro-β-alanine (FBAL), which is renally excreted (cf. Fig. 15.3.4). Figure 15.3.5 demonstrates a 19F MR spectrum after intra-arterial administration of 5-FU in the metastatic liver of a patient. Some of the resonances are resolved and assigned according to their chemical shift to 5-FU, FBAL, DHFU (5-fluoro-5,6dihydrouracil), and 5-FUranuc. Clinical 19F MR spectroscopy is still limited to the above-mentioned application; however, diffusionweighted 19F imaging of the lung using SF6 may be a visible application in vivo.

15.3  Clinical MRS

Fig. 15.3.5  19F MR spectroscopy in vivo. Examination of a patient’s liver with liver metastasis of a colorectal tumor during chemotherapy with 5-fluoruracil (5-FU). Assignment of the resonances: 5-fluoro-uracil nucleosides and nucleotides (5FUranuc), α-fluoro-β-alanine (FBAL, with signal contributions of α-fluoro-β-ureidopropion acid), 5-fluoro-5,6-dihydrouracil (DHFU), not identified (n. i.) possibly free fluoride (F–) (Schlemmer et al. 1994). a 19F MR spectrum acquired during 103 min (sum of 6 spectra) after intravenous (i.v.) application of 5-

FU in a 44-year-old patient, detection of small amounts of the cytotoxic anabolites of 5-FU (5-FUranuc) in liver tissue. b 19F MR spectrum, acquired during 58 min (sum of 13 spectra) after i.v. application of 5-FU in a 56-year-old patient. The detection of the intermediate DHFU makes possible a partial block of the catabolism of a DHFU-producing step with accumulation of 5-FU and in turn, a decreased production rate of FBAL. (TR = 1,000 ms, AC = 256)

1349

1350

15  Clinical Spectroscopy

15.4  Clinical Applications in Selected Organs and Tumors 15.4.1  Brain In the first weeks up to the second year of life, myelinization takes place during brain development and characteristic changes in 1H and 31P MR spectra are observed. During brain development an age-dependent change of the brain metabolite signals (NAA-Cho- [↑] and Cho-(P)Cr intensity ratio [↓]) is observed by means of MRS (Kreis et al. 1993b; Fig. 15.4.1). A quantitative analysis demonstrates an increased mI in newborns and an increased Cho in older children as dominate resonances, whereas the concentrations of (P)Cr and NAA are smaller in newborns than in adults. A extended proton-decoupled 31P MRS study, which included quantitative determination of the metabolites, showed a dominant PE resonance in newborns decreasing rapidly within the first years of life and it is constant until end of life (Bluml et al. 1999); Fig. 15.4.2). The same holds true for the PC resonance; however, the effect is not that pronounced. Both components of PDE—GPE and GPC—have slight increases during brain development. In the case of GPC, the value approaches an asymptotic value at age of 20, whereas GPC increasing up to the end of life. It is assumed that both phospholipids precursors (PC and PE) are important in the myelin synthesis. An increase of PCr is observed in the first weeks and months of life, whereas γ-ATP is almost constant (Holshouser et al. 1997). Signal intensities of PCr and NTP as well as pH values decrease in different tissues, as was demonstrated in animal 31P MRS experiments, whereas the Pi values increases. First relevant clinical results of the human brain metabolism were already obtained by means of 31P MRS in 1983 (Cady et al. 1983). The results of these examinations demonstrate that the prognosis is poor if the PCr–Pi signal intensity falls below a reference value (Hamilton et al. 1986). Evidence was found during asphyxia, demonstrating changes of the metabolism in localized vulnerable areas only, and showing no changes in the rest of the brain (Azzopardi and Edwards 1995). Significant differences are not observed in 31P MR spectra of the brain in normal children compared with those with different diseases (e.g., post-infectious, asphyxia during delivery, small neurological deficiencies) (Boesch et al. 1989; Buchli et al. 1994). Decreased NAA–Cho ratios and increased Lac concentrations are observed in perinatal asphyxia depending on brain areas, the time of onset, and severity of asphyxia (Roelants-Van Rijn et al. 2001; Pavlakis et al. 1999). In general decreased NAA and increased Lac concentrations are correlated with the severity of the asphyxia and are of value for the prognosis. Increased Glx–NAA ratios are found in basal ganglia after severe asphyxia immediately and up to 7 days after delivery (Groenendaal et al.

Fig. 15.4.1  Typical 1H MR spectra in vivo of a subjects with different age. The resonance intensities vary with age. The scale of each spectrum is adjusted to the highest resonance. STEAM TR/ TE = 1,500/30, number of acquisitions AC = 144–256, VOI = 8– 10 cm3 in children, 12–16 cm3 in adults (Kreis et al. 1993b). a Spectrum of a male newborn at age of 4 days (40.5 weeks of gestation); dominant are the mI and Cho resonances. b Spectrum of a male baby 6 months old. c Spectrum of a 4-year-old boy. d Spectrum of a female 31 years of age. In this spectrum of an adult the NAA resonance is dominant

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.2  Age-depended changes of the absolute concentrations of the membrane metabolites PE, PC, GPE, and GPC measured by means of proton-decoupled 31P MR spectros­copy (Bluml et al. 1999). The balanced concentration of the metabolite concentrations was calculated using model functions, which are assumed to be double exponential in the case of PC and PE and single exponential in the case of GPE and GPC (◆)

healthy controls, (*) patients with hydrocephalus, (◊) patients with other pathological findings, but otherwise inconspicuous MRS. If possible, the age was corrected for the gestation age. Gestation age of 40.5 weeks is indicated by a dotted line. PE phosphoethanolamine, PC phosphocholine, GPE glycerophosphorylethanolamine, GPC glycerophosphorylcholine

2001). 1H and 31P MR spectroscopy has a high clinical value for such indications (Cady 2001). Cerebral circulation impairment was mainly investigated by 1H MRS. Acute as well as subacute stroke are characterized by the Lac signal, which can persist for days to weeks (Figs. 15.4.3, 15.4.4; Bruhn et al. 1989; Gillard et al. 1996; Beauchamp, Jr., et al. 1999; cf. also Saunders et al. 1999). The long-lasting persistence of the Lac signal is said due to the ongoing anaerobic glycolysis, the insufficient transport or metabolism of lactate, or to the glycolytic metabolism of the macrophages, which invade the subacute or early chronically infracted brain area (Beauchamp, Jr., et al. 1999). The decrease of NAA concentration—starting slowly in the first few hours and progressing later—is interpreted as a loss of neuronal structures. It is under investigations if the penumbra (the not-yet

severely affected brain tissue), which cannot be imaged using conventional or diffusion imaging, can be identified by means of 1H MRS. Presently, 1H MRS is not the method of choice for the differential diagnosis because of the insufficient spatial resolution and the resulting partial volume effects (Kim et al. 2001). Other brain diseases of vascular origin were also investigated by 1H MRS (Constans et al. 1995; MacKay et al. 1996). The 1H MR spectroscopic investigations of newborn and children after ischemic CNS impairment is useful for long-term prognosis (Holshouser et al. 1997). Decreases of the NAA signal intensity as well as an increase of Cho was observed in plaques of patients with multiple sclerosis by means of 1H MRS (Arnold et al. 1990). Increased Cho signal intensities were observed not only in demyelinating, but also in inflammatory diseases (Ross

1351

1352

15  Clinical Spectroscopy

Fig. 15.4.3  Localized, water signal–suppressed 1H MR spectrum in vivo of the brain (acquisition with stimulated echo pulse sequence, VOI = 3 × 3 × 3 cm3, TR = 1,500 ms, TE = 50 ms, TM = 30 ms, AC = 512, B0 = 1.5 T) (Bruhn et al. 1989). a Spectrum (examination of a healthy volunteer) with resolved resonances of N-acetyl-aspartate (NAA), phosphocreatine and creatine ((P)Cr), choline (Cho), and inositol (Ins). b Spectrum of a infracted brain region of a 42-year-old patient. The Cho–(P)Cr intensity ratio in infracted brain differs significantly compared with the findings in 1H MR spectra of healthy brain. The most prominent resonance is the β-methyl proton doublet of the lactate (Lac) at δ = 1.33 ppm (relative chemical shift to the water protons at δ = 4.7 ppm). The Lac doublet shows a line splitting of JHH ≅ 7 Hz (the β-methyl protons and α-methine proton of the molecule are a scalar coupled homonuclear AX3 spin system with a coupling strength of JHH ≅ 7 Hz). Lactate is the final product of the anaerobic glycolysis and as such, a metabolite of high physiological importance. If present in the tissue, free fatty acids (chemical shifts: δ = 0–2.4 ppm) overlay the lactate doublet and may hide this doublet completely. To receive information about the lactate levels of the tissue in such situations, a selective editing techniques for these metabolites —so-called spectral-editing methods—have to be applied

and Michaelis 1994a; Danielsen and Ross 1999; Arnold et al. 2000; Burtscher and Holtas 2001). Other inflammatory demyelinating diseases (hereditary demyelinization, Binswanger’s disease) yield reduced NAA signals too. 1H MR spectroscopic imaging has been successfully applied in temporal lobe epilepsy for lateralization of the epileptic foci (Ende et al. 1997). A meta-analysis of preoperative 1H MRS in patients with epilepsy demonstrated a connection of ipsilateral MRS abnormality to good outcome (Willmann et al. 2006); however, the authors recommending further prospective MRS studies before using MRS as a routine examination. Hepatic encephalopathy was intensively investigated by Ross et al. (1994) (Fig. 15.4.5). Encephalopathies of other origin, like hypoxic encephalopathy in nearly drowned patients (Kreis et al. 1996), and in patients with Alzheimer’s disease (Fig. 15.4.6). 1H MRS yield important evidence for diagnosis, therapy monitoring and prognosis (Ernst et al. 1997; Ross and Michaelis 1994; Wolf et al. 2000). Furthermore, in these diseases increased concentrations of mI are observed (Ross et al. 1996, 1997; Ross and Bluml 2001; Kreis et al. 1990; Ross and Michaelis 1994; Naegele et al. 2000). Patients with other types of dementia show significant decreased NAA concentrations, but normal values for mI (Shonk et al. 1995). In patients with AIDS encephalopathy, changes in the cerebral metabolism were observed by means of 1H MRS and 1H MR spectroscopic imaging (Ernst et al. 2000; Barker et al. 1995; Laubenberger et al. 1996; Meyerhoff et al. 1993; Chang et al. 1995). Osmotic changes and interferences of the brain membrane and energy metabolism were observed using 1H MRS in combination with proton decoupled 31P MRS (Bluml et al. 1999, 2001). Some brain diseases can be diagnosed by MRS directly. In the 1H MR spectra of patients with Canavan’s disease—a rare autosomal recessive inborn error based on enzyme (aspartoacylase) deficiency—show a dominating NAA-resonance in brain, plasma, and urine (Fig. 15.4.7; Grodd 1990). The present status of proton MR spectroscopy for the diagnosis of neurological diseases is summarized in reference (Lin et al. 2005). 31P MRS studies in patients with chronic cerebral infarctions changes are observed in the concentrations of the phosphometabolites as well as the pH value (Levine et al. 1992; Sappey-Marinier et al. 1992). Examinations in volunteers demonstrate a decrease of the pH value but a constant Lac intensity caused by drug-induced reduction of blood flow. However, the pH and lactate signal intensities rises if the reduction of blood flow is caused by hyperventilation (Van Rijen et al. 1989). Clinical brain MRS examinations with other nuclei than 1H and 31P are very rare. In functional studies after application of 13C labeled substances (e.g., 13C-labeled glucose, Beckmann et al. 1991) the metabolic products

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.4  Distinct left hemiparesis, pronounced dysarthria, and central paralysis of the eighth cerebral nerve in a 56-yearold patient with occlusion or high-grade stenosis of the right middle cerebral artery (MCA). The conventional T2-weighted MR image (b/T2) does not show any hyperintensity or other pathological signs 24 h after the occurrence. At the same time, a multisectional transversal 2D 1H spectroscopic imaging (TR/ TE = 2,300/272, ST = 15 mm, matrix, 32 × 32) was performed (Beauchamp, Jr., et al. 1999). a 1H MR spectra of four regions of the brain, which are each shown transversally. In the spectra

of the lesions 1 and 2, a reduced NAA resonance and a lactate (Lac) are observed, whereas in spectra 3 and 4, the spectra have a normal appearance. b Transversal T2-weighted image and metabolic images of choline (Cho), N-acetyl-aspartate (NAA), and lactate (Lac). A reduced NAA- (red arrow) and an increased lactate signal intensity (curved red arrow) are observed within the area of lesion. In the areas of decreased NAA and increased lactate, a progression was observed and the picture of a brain infarct developed. This is consistent with the appearance of a penumbra

and the kinetics were observed. It is still an open question, if the 13C MRS can contribute to the diagnosis of the hepatic encephalopathy (Lin et al. 1999). Drugs (e.g., neuroleptics) and the kinetics thereof were observed in the brain of patients by means of 19F MRS (Bartels et al. 1991; Komoroski 1989; Bolo et al. 2000). MR studies of the human brain were performed with the quadrupolar nuclei 7Li and 23Na, both possessing nuclear spin 3/2 (nuclei with spin 1/2 [e.g., protons] have only a magnetic dipole moment, but no quadrupolar moment). Lithium compounds are used for prophylaxis and therapy of depressions. After administration of Li

carbonate, Li concentrations in the order of mmol were observed in patients by means of 7Li MRS and Li MR imaging (Renshaw et al. 1985; Renshaw and Wicklund 1988) and even the pharmacokinetics of lithium carbonate during the therapy could be monitored (Komoroski 1990; Soares et al. 2001). In tissue, 23Na has the second strongest MR signal and can be used for fast MR imaging. Because only a single 23Na resonance—the signal of the free Na+ ions—can be observed in vivo, MR spectroscopic applications may yield no additional information. 23Na spectroscopic studies are not verifiable in literature.

1353

1354

15  Clinical Spectroscopy

Fig. 15.4.6  Localized water signal–suppressed 1H MR spectra in vivo of the occipital gray brain matter (Shonk et al. 1995). a Healthy volunteer. b Alzheimer’s disease patient: characteristic spectroscopic findings: increased myo-inositol (mI) and decreased NAA

Fig. 15.4.5  Localized water signal–suppressed 1H MR spectra in vivo of the parietal white matter in patients with liver disease. The hepatic encephalopathy (HE) is MR spectroscopically characterized by reduced concentration of myo-inositol (mI), an increased glutamine/glutamate levels (Glx) in brain tissue (d) (Ross et al. 1994). a Patient with liver disease without HE does not show this findings. b Spectrum of a patient that was spectroscopically classified as subclinical HE (= SCHE). c Spectrum of a patient with HE and less pronounced symptoms (grade 1)

15.4.2  Liver Clinical MRS examinations in vivo were performed mainly be means of 31P and 19F MRS. 31P MRS may be used to diagnose enzyme defects or may help uncovering enzyme functions by means of stress tests. For instance after fructose administration an increase of PME signal intensities and a decrease of the intracellular pH values is observed in the liver of healthy volunteers, whereas the resonances of the other phosphometabolites do not change. In patients with fructose intolerance (fructokinase deficiency), the spectral param-

eters are constant after fructose infusion (Segebarth et al. 1989; Boesiger et al. 1994). Furthermore the inhibition of the enzyme aldolase was assessed after administration of alcohol and fructose (Boesch 1997). In patients with an acute fatty liver disease, increased PME signals are observed in the liver (Cox et al. 1988) and with polychemotherapy increased PME and PDE signals (Steudel et al. 1990). Commonly in patients with alcoholic liver disease, absolute concentrations of phosphometabolites in liver tissue are reduced compared with healthy volunteers; however, the intensity ratios are equal for both patient groups. The correlation of absolute ATP concentrations in liver tissue and the intracellular pH value makes possible differentiation of the alcoholic hepatitis, alcohol-induced cirrhosis, or healthy liver (Fig. 15.4.8) (Meyerhoff et al. 1989). Recent results were reported using proton-decoupled 31P MRS (Schlemmer 2005). Increased PME resonances are observed not only in liver cirrhosis (Meyerhoff et al. 1989; Taylor-Robinson 1997), but also in other diseases such as hepatitis (Kiyono 1998), rejection of transplanted livers (Taylor-Robinson 1998), and liver tumors (cf. Sect. 15.4.6). In summary, increase PME resonances have to be interpreted as nonspecific signs for disease-related changes in the liver. Examinations in healthy volunteers yield changes in postprandial 31P MR spectra among others, increased PME-NTP-ratio (Schilling et al. 1990; Bourdel-Mar-

15.4  Clinical Applications in Selected Organs and Tumors

chasson et al. 1996); this has to be considered when liver spectra are evaluated. Up to now 1H MR spectroscopy of the liver has no

2.2

Cirrhosis

2.0

Control

Fig. 15.4.8  Localized 31P MR spectroscopy of the liver in vivo in patients with liver diseases of different origin. Absolute concentrations of ATP in liver tissue as function of intracellular pH values in patients with alcoholCirrhosis induced cirrhosis ( ),Cirrhosis alcoholHepatitis and healthy ­induced hepatitis ( ) Hepatitis volunteers ( ) Control (acquisition with Control surface coils of 9 and 14 cm ∅, ISIS localization techniques, VOI = 64–120 cm3, TR = 1,000 ms, AC = 500–1,600, B0 = 1.5 T) (Meyerhoff et al. 1989)

Hepatitis

1.8 1.6 2.2 2.2

1.2

2.0 2.0

1.0 2.6 6.95

2.8 7.06

7.20

ATP (mMol/kg wet weight)

1.4

ATP (mMol/kg wet weight)

ATP (mMol/kg wet weight)

Fig. 15.4.7  Localized water signal–suppressed 1H MR spectra in vivo of the brain of an infant (double-spin-echo sequence VOI = 2 × 2 × 2 cm3, TR = 1,500 ms, TE = 135 ms, AC = 256, B0 = 1.5 T (Grodd et al. 1990). a Normal spectrum of a 6-month-old infant with spinal muscle atrophy. b Spectrum of a 6-month-old infant with Canavan’s disease. In comparison with the normal findings, increased NAA signals are observed, and signals of Cho are close to detection limits

significant clinical relevance. This is due to the breath-related motion of the organ and the susceptibility artifacts impeding the examinations. New techniques are able to overcome these limitations (Star-Lack et al. 2000; Tyszka and Silverman 1998). With these techniques, high-resolution spectra were acquired, and a new resonance was discovered in liver spectra (triethylammonium [⇒ betain + choline]) (Star-Lack et al. 2000). 13C MRS of the liver contributes to the understanding of the tissue glycogen metabolism and the glucose consumption. 13C MRS elucidates the normal glucose homeostasis and the pathophysiology of the diabetes mellitus type II (Roden et al. 2001). 19F MRS allows to monitor the chemotherapy of 5-FU and to study the pharmacokinetics of the compound (Glaholm et al. 1990; Gonen et al. 1997; Krems et al. 1994; Li and Gonen 1996; Li et al. 1996; Presan 1994; Schlemmer et al. 1994; Semmler 1990). The cytotoxic active anabolites (5-FUranuc) has been detected only in a few patients (Schlemmer et al. 1994; Semmler 1990). The signal–time curves of 5-FU and its catabolite FBAL have been observed in clinical studies, and the results have been applied to a three compartment model (Port et al. 1991). A nonlinear open three compartment model describes adequately the observed kinetics (Port et al. 1991). Tumor diagnostics and therapy monitoring of liver tumors by means of 19F MRS are mentioned in Sect. 15.3.4 and 15.4.6 too. The application of 5-FU causes an enrichment of metabolites in the gall bladder, which can be detected by means of 19F MRS. The resonance of α-fluoro-β-alanine is shifted due to conjugation with bile acids (Dzik-Jurasz et al. 2000). For selective evaluation of 19F resonance in tumor and liver tissue, special decoupling and localizing methods were developed to increase S/N ratio and spatial resolution (Gonen et al. 1997; Krems et al. 1994; Li and Gonen 1996).

1.8 1.8

1.8

1.6 3.0

3.2

1.6

7.35 1.4

7.54

1.4

3.4 7.82

ppm

3.6

pH

8.52

1.2

1.2

1.0

1.0

2.6 6.95

2.6

2.8 7.06

3.0

2.8

7.20

7.35

3.2

3.0

7.54

3.4

3.2

7.82

3.6

3.4

8.52

ppm 3.6 pH

ppm

1355

1356

15  Clinical Spectroscopy

15.4.3  Heart 31P MRS can detect ischemic conditions as well as changes caused by cardiomyopathies, as was shown in animal experiments. However, patient studies are relatively rare, because of the technical difficulties; in particular the localization of the relatively thin and movable ventricular wall and contamination of the spectra from thoracic wall musculature and blood cannot be completely avoided. Despite of these limitations 31P MRS spectra of the heart were reproducibly acquired in healthy volunteers (Fig. 15.4.9) (Bottomley et al. 1987; Bottomley et al. 1996; Löffler et al. 1998). The concentrations of PCr and ATP were calculated from the saturation corrected PCr–ATP ratio (1.8) to be 11 and 6 µmol per gram wet weight, respectively. The determination of pH value of the ventricular wall is relatively difficult, because the weak signal from the Pi signal do overlap with the (background) signal of the 2,3-diphosphoglycerates (2,3-DPG) of blood. For the healthy heart muscle the pH value was determined to be pH = 7.11 ± 0.05 (Blamire et al. 1999). In healthy volunteers atropine–dobutamine stress causes a reduction of PCr by 21%, of ATP by 9%, and of the PCr– ATP ratio by 14% from 1.42 ± 0.18 to 1.22 ± 0.20 (Lamb et al. 1997). Similar examinations were conducted in patients with hypertensive heart disease; the values were 1.20 ± 0.18 (resting) and 0.95 ± 0.25 (stress), respectively (Fig. 15.4.10) (Lamb et al. 1999). 31P MR spectroscopic imaging was conducted in 41 patients with stenosis (>50%) of the coronal vessels and in 11 healthy volunteers. Absolute quantification of the metabolites was per-

Fig. 15.4.9  Localized 1H-decoupled 31P MR spectrum in vivo of the heart of a 28-year-old healthy volunteer. Single spectrum of the 512 spectra of a 3D CSI dataset (image oriented along the long axis of the heart, ECG triggering, surface coil 11 cm ∅, 31P–(1H) double-resonance technique, 3D chemical-shift imaging (CSI) with 8 × 8 × 8 VOIs, VOI = 2.5 × 2.5 × 5 cm3, TR > 800 ms, B0 = 1.5 T). The resonance of 2,3-diphospho-glycerate (2,3-DPG) of blood is resolved (H. Kolem, Siemens, 2007, personal communication)

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.11  31P MR spectroscopy in vivo of the heart at 1.5 T, surface coil, 1D-CSI with sagittal slice selection suppressing the muscle contribution of the thoracic wall, slice thickness SDsag = 60 or 80 mm, SDinplane = 20 mm. The given PCr–ATP ratios are corrected for saturation and contamination of blood (2,3-DPG) (Yabe et al. 1995). a 31P MR spectra of selected patients from two patient groups (>50% stenosis of the left coro-

nary) with reversible defect (RD+) and irreversible defect (RD–, scar) as classified by 209Tl myocardial scintigraphy in comparison with spectra of healthy volunteers after stress. Obviously the PCr resonance is smaller in the group with irreversible defects (RD–). b PCr in the control and the two patient groups. The PCr concentration in both patient groups is significantly reduced compared to the control group

9 Fig. 15.4.10  31P-MR spectroscopy of the heart at 1.5 T, surface coil, 3D ISIS, VOI = 60 × 70 × 70 mm3, ECG triggering, average TR ≈ 3,700 ms and ≈ 3,400 ms at rest and stress, respectively, depending on the heart frequency. The given PCr/ATP-ratios are corrected for saturation and contamination of blood (2,3DPG). a 31P MR spectra of the anterior wall of the left ventricle of a healthy volunteer before, during, and after atropine–dobutamine stress. b 31P MR spectra of a patient with hypertensive heart disease before and during atropine–dobutamine stress (Lamb et al. 1999)

formed. The patients were divided in two groups according to the result of stress 201Tl scintigraphy; those with reversible defects RD+ and those with fixed defects RD–. In comparison to the volunteers, the PCr concentration was reduced in patients with reversible defect and with fixed defects even stronger reduction of the ratio was observed (Fig. 15.4.11) (Yabe et al. 1995). Investigations using 31P–{1H} double-resonance techniques yield strongly reduced PCr–Pi ratios in patient with heart infarctions (Löffler et al. 1998). In summary, 31P MRS in vivo of the heart has high significance in animal experiments and clinical research, but is nowadays not a clinical routine method. 1H MRS of the heart is rarely used in human studies because of the technical difficulties. Nakae et al. (2004) assessed total creatine in human hearts using cardiac-

1357

1358

15  Clinical Spectroscopy

gated 1H MRS with image-guided PRESS localization. They measured septal creatine in healthy hearts and compared with those with cardiomyopathy. This study demonstrated myocardial creatine being decreased in nonischemic dysfunctional hearts. 15.4.4  Skeletal Muscles The first human 31P MRS examinations were performed at the muscles of extremities. The relatively homogeneous muscle tissue can be investigated using surface coils yielding 31P MR spectra with high-quality and high temporal resolution. The 31P MRS opens a window to observe non-invasively the energy metabolism of muscle cells (cf. Sect. 15.2) and was applied already very early to explore muscle physiology and pathophysiology (Chance 1989; Radda et al. 1989). Using this method, the turnover of the energyrich phosphates (NTP, PCr) can be observed in primary muscle diseases (myopathies, muscle dystrophies) as well as systemic diseases (e.g., diabetes mellitus) and diseases of vascular origin. Stress-induced changes of the energy metabolism of muscle cells cause intensity fluctuations of PCr, NTP, and Pi as well as pH shifts and can be used to diagnose pathological states of muscles (Taylor et al. 1994). In patients with McArdle syndrome—a glycogen storage disease, characterized by a deficiency of glycogen phosphorylase—an increase of the pH value in muscles

is observed under stress. Control examinations in healthy volunteers yield a decrease of the pH value. The analysis of the 31P MRS data in vivo led to the conclusion that a block occurs in the first step of glycolysis (Fig. 15.4.12; Ross et al. 1981). This was one of the first clinical relevant MRS studies in vivo. Furthermore, mitochondriopathies, which are often caused by genetic enzyme defects, were investigated by means of 31P MRS in vivo. The abnormal function of the mitochondria is expressed by increased Pi signal intensity, and reduced PCr–ATP intensity ratio as wells as in a different recovery of PCr and pH after stress (Argov et al. 1987; Taylor et al. 1994). In patients with Duchenne muscle dystrophy reduced PCr–ATP and PCr–Pi intensity ratios, increased pH values, and an intensive PDE signal are observed in resting state. In progressive disease these changes are even more pronounced. A slow increase of pH and PCr signal intensity is observed immediately after the end of stress in peripheral occlusive disease of the lower extremities. A good correlation of PCr–(PCr + Pi) signal intensity ratio and the severity of vessel stenosis determined by means of Doppler sonography and angiography exists. The observed correlation of MR parameters with the invasive measured blood parameters of hypoxanthine, alanine, ammonia, and lactate lead to the assumption that not only the anaerobic glycolysis, but also the purine nucleotide cycle may play a role in the energy supply of the muscle in patients with peripheral arterial occlusive disease (Fig. 15.4.13); Rexroth et al. 1989).

Fig. 15.4.12  31P-MR spectroscopy in vivo of the skeletal muscle. Spectra of a healthy volunteer (a) and a patient (b) with McArdle syndrome, acquired before, during, and after ischemic stress, which lasted 45 s and 1.5 min in the patient and in the volunteer, respectively (Ross et al. 1981)

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.13  31P MR spectroscopy of the calf muscle in patients with arterial occlusion disease (Rexroth et al. 1989). a Series of spectra of the gastrocnemius muscle of a patient with arterial occlusion disease acquired before, during, and after stress (lifting a lead weight). The applied stress set free inorganic phosphate (signal increase of Pi resonance) and a reduction in the concentration of phosphor creatine (PCr) (t = 0: start of stress).

b Kinetics of PCr and Pi before, during, and after stress, quantitative evaluation of the 31P MR spectra in vivo. c Function of time of PCr –(PCr + Pi) signal-intensity ratios. d Intracellular pH after quantitative evaluation of 31P MR spectra in vivo and invasive measured arteriovenous (A.–V.) differences of lactate concentration, before, during, and after stress

In addition to rare primary and neurogenic, secondary muscle diseases (e.g., kidney failure) were investigated by means of 31P MR spectroscopy. These secondary muscle diseases influence the metabolism as well (Sala et al. 2001; Thompson 1997; cf. also Taylor et al. 2000). Investigations of the skeletal muscle of healthy volunteers demonstrate resonances of methyl and methylene protons of free fatty acids, N-methyl and N-methylene protons in (phosphor)creatine, N-trimethyl protons of choline and carnitine, olefinic protons, as well as ring protons of histidine residues of the muscle peptides anserine and carnosine (Fig. 15.4.14) (Bachert et al. 1992a; Bruhn et al. 1991). Muscle examinations were increasingly performed using 1H MRS, e.g., the recovery of the

gastrocnemius muscle after exhausting stress was observed. From these examinations was concluded that the (P)Cr resonance represents the PCr content and not the Cr content (Kreis et al. 1999). The 1H and 31P MRS was used during rest and during stress as well as after creatine administration to determine the PCr–Cr ratio (Trump et al. 2001). In patients with hyperinsulinemia and after infusion of free fatty acids a fast accumulation of lipids in muscle could be observed (Brechtel et al. 2001). Atrophic muscle diseases yield increased content of free fatty acids (Narayana et al. 1988), which can be also observed in MR imaging. Muscle atrophies caused by myalgia show changed (P)Cr–H2O and Cho–H2O intensity ratios (Schick et al. 1994).

1359

1360

15  Clinical Spectroscopy

Fig. 15.4.14  Localized water signal-suppressed 1H MR spectrum in vivo of the calf muscle of a healthy volunteer. The 1H MRS-detectable metabolites are indicated (Measurement with knee resonator, stimulated echo-pulse sequence, VOI = 2 × 2 × 2 cm3, TR = 1,500 ms, TE = 100 ms, AC = 200, B0 = 1.5 T)

In particular 13C MRS substantially contributed to elucidate the glycogen metabolism in muscle. These studies allowed to newly define the roll of the enzyme glycogen synthase (Shulman and Rothman 2001b) and to elucidate the energy supply of the fast-contracting muscle, which is granted by very fast glycogenolysis within milliseconds. From these experiments it was concluded that lactate is produced by this process and not by means of shortage of oxygen supply (Shulman and Rothman 2001a). Although 31P and 1H MR spectra of muscle diseases contain rich biochemical information, MRS could not be established in clinical routine. For the time being, MRS in this field will be mainly used in clinical research. 15.4.5  Urogenital Tract 31P and 1H MRS examinations of the prostate are performed with the aim to differentiate benign and malignant lesions. High-quality spectra can be obtained using endorectal coils, which also guarantee precise localiza-

tion (Hricak 1989). Nowadays, 31P MRS prostate spectra are hardly used for clinical diagnostics; however, 1H MRS has gained clinical importance for the differential diagnosis carcinoma/ benign prostate hypertrophy (BPH) (see also Sect. 15.4.6 and reference therein; cf. Fig. 15.4.16; see also Chap 7, see 7.2.2). It is routinely used and significantly contributes to the radio therapy planning, in particular also of prostate carcinoma (Payne and Leach 2006). Acquisition of localized MR spectra of the kidney in vivo is difficult because of the motion of the organ due to the respiration. Relevant clinical applications are not reported. However, 31P MRS can be used for the monitoring of kidneys that are prepared for transplantation, as well as for the status of the kidneys after transplantation. Phosphometabolite ratios of fully functioning heterotopic transplanted kidneys are equal to those of healthy orthotropic kidneys (Moller et al. 1995; Kugel et al. 2000). If graft-versus-host reaction takes place the Pi–ATP ratio increases, and the pH value drops (Heindel et al. 1997; Vallee et al. 1996). 31P MRS was also used to investigate the function of testes and to test infertility. 31P MR spectra of the normal testes are characterized by an intensive PME resonance. In oligo- and azoospermia an increasingly reduced PMEβ-ATP-ratio is observed (1.71±0.03, 1.41±0.06, and 1.22±0.07) (van der Grond et al. 1991). 31P MRS seems to be helpful for clinical diagnostics (van der Grond 1995). 15.4.6  Tumors Clinical MR spectroscopy, in particular 31P MRS, was used already in its infancy to characterize human tumors. Animal experiments have demonstrated differences of the resonances in tumors compared with those in normal tissue. Furthermore, MR spectroscopy proved valid for monitoring of therapy. In the recent past, 1H MR spectroscopy of the brain gained increasing importance and now surpasses 31P MRS. This is especially due to the pronounced lower spatial resolution of 31P MRS compared with 1H MRS. A recent comprehensive overview on the present status of MRS in cancer is given by Gillies and Morse (2005). Localized 1H MR spectra in vivo of human brain tumors yield no specific resonances (Ott et al. 1993), not considering the inconsistently observed resonance of the amino acid alanine (Preul et al. 1998). Valine, leucine, isoleucine, as well as the resonances of acetate and succinate were identified in abscesses (Kim et al. 1997). Prospective studies document the value of the localized 1H MRS for the differentiation of benign and malignant focal lesions; the latter lesions are characterized by an increase of the relative signal intensity of Cho in comparison with the (P)Cr and NAA resonances (Adamson et al. 1998; Lin et al. 1999). Furthermore, a lac-

15.4  Clinical Applications in Selected Organs and Tumors

tate resonance (β-methyl proton doublet) is often found (Herholz et al. 1992). Similar differences are seen in recurrent tumors and radiation induced tissue changes after radiation therapy of brain tumors (Schlemmer et al. 2001) (cf. Fig. 15.4.15). In patients with AIDS, 1H MRS can contribute significantly to the differential diagnosis of toxoplasmosis versus lymphoma of MR tomographically indistinguishable lesions (Chang et al. 1995). An interrelation of the signal intensity of Cho with proliferating tumor activity was found in glial brain tumors (Shimizu et al. 2000; Tamiya et al. 2000). Yet, localized 1H MRS is not very helpful to distinguish different brain tumors as was found by Ott et al. (1993). Possibly, the relatively large voxels (ca. 8 ml) cause partial-volume effects with different signal contributions of the potentially heterogeneous tumor, the tumor rim area, and the surrounding normal tissue. In contrast, high specificity was reported in studies using 1H MR spectroscopic imaging (Preul et al. 1998). This technique has the advantage to simultaneously investigate the different tumor areas, the tumor rim, as well as the surrounding tissue, and therefore the results can be directly correlated (cf. Fig. 15.4.21). The diagnostics of brain tumors is performed mainly by means of 2D or 3D 1H MR spectroscopic imaging with long echo times (130–270 ms) (Preul et al. 1998; McKnight et al. 2001; Nelson et al. 1997). Improved and standardized analysis of MR spectra may increase the value of 1H MRS (Opstad et al. 2007). New results reported in multicenter studies show added value for the diagnosis of brain tumors (Tate et al. 2006). In combination with cerebral blood volume measurements 1H MRS improves the differentiation of low-grade and anaplastic oligodendroglial tumors (Sibtain et al. 2007). With 31P MRS, no specific resonances were found in human brain tumors. Such resonances are not detected in healthy brain tissue either, but differences are found in the signal intensities of the phosphor metabolites (Heindel et al. 1988; Heiss et al. 1990). The quantitative evaluation of the spectra yield, for instance, significant differences of the signal intensities in the 31P MR spectra of meningiomas in comparison with those of normal healthy brain tissue (Blankenhorn et al. 1999). Less significant are the spectral differences of glioblastomas and astrocytomas in comparison with the findings in normal brain tissue. Because of the big voxels of the 31P MRS (approximately 30 cm3) different cell fractions contribute (vital, hypoxic, anoxic, necrotic) to the spectrum, and therefore the differentiation will be obstructed. In gliomas and meningiomas, higher pH values are observed compared with normal brain tissue (Arnold et al. 1989; Cadoux-Hudson et al. 1989; Hubesch et al. 199). Among others, a comprehensive review of the 31P MRS data of primary brain tumors was published by W. Negendank (1992) in the early days of MRS. The differentiation between and malignant changes of the prostate versus prostate carcinoma is a difficult dif-

ferential diagnostic problem. 31P MRS—but mainly 1H MRS—can significantly contribute to the correct diagnosis. By means of 31P MRS, a differentiation of BPH and normal prostate tissue may be possible (Kurhanewicz et al. 1991; Narayan et al. 1991; Fowler et al. 1992). Normal and malignant tissue can be distinguished by means PME–PCr intensity ratio (Kurhanewicz et al. 1991). 1H MR spectroscopic examination of the prostate is technically difficult, because a sufficient water and fat suppression is mandatory, field homogeneity is low, and the low signal often requires the use of endorectal coils (van der Graaf et al. 1999; Heerschap et al. 1997; Kurhanewicz et al. 1995, 1996; Scheidler et al. 1999; Swanson et al. 2001; Yu et al. 1999). In the prostate gland high amounts of citrate are synthesized, secreted, and stored (Costello and Franklin 1991; Pretlow et al. 1985). Because of the intramolecular proton coupling, citrate shows four resonances in high-field MR spectrometry (Schiebler et al. 1993); however, in vivo only one broad resonance is observed at 2.6 ppm (Kurhanewicz et al. 1996; Scheidler et al. 1999; Menard et al. 2001; Swanson et al. 2006). Figure 15.4.16 (Kurhanewicz et al. 1996) shows the result of MR spectroscopic imaging of the prostate. Characteristically, an increased Cho resonance and a reduced citrate resonance are observed for malignant prostate tissue (see also Sect. 15.4.5 and references thereto). Clinical applications are covered in Chap. 9, Sect. 9.2. Comprehensive reviews discussing the potential of MRS in the diagnosis of prostate cancer were done by Hricak (2005) and Huzjan et al. (2005). High Cho signal intensities are observed by means 1H MRS not only in brain tumors and in prostate carcinomas, but also in breast cancer (Bakken et al. 2001; Thomas et al 2001; Yeung et al. 2001). However, 1H MRS does not play a decisive role in the diagnosis of breast. In most of the studies, 31P MRS was applied. From investigations ex vivo it is known that phospholipid metabolism is increased in breast cancers and in turn, in clinical studies, increased PME resonances are observed (Glaholm et al. 1989; Sijens et al. 1988). However, the observations are nonuniform and may be due to the heterogeneity of tumors (Smith et al. 1991) and/or to partial volume effects caused by the large voxels. Definitive statements to the tumor differentiation are presently difficult to make (Leach et al. 1998). Numerous clinical 31P MRS studies were conducted to examine tumors of the extremities (e.g., osteosarcoma, Ewing sarcoma, malignant melanoma, etc.) (Bachert et al. 1992a; Dewhirst et al. 1990; Redmond et al. 1992a; Semmler et al. 1988a, b; Sostman et al. 1990; Zlatkin et al. 1990). Tumors close to the body surface are easily accessed. A problem of the use of surface coils is the contamination of the spectrum of an invasive growing tumor due to signal contributions of the surrounding tissue. In general, the spatial resolution of 31P MRS is not sufficient to separate these signal contributions.

1361

1362

15  Clinical Spectroscopy Fig. 15.4.15 a-c  1H MRS of brain tumors (PRESS, VOI = 2 × 2 × 2 cm3, TR = 1,500 ms, echo time TE = 135 ms, AC = 200, B0 = 1.5 T). Compared with normal findings in brain tumors the signal intensities of choline (δ = 3.22 ppm) are characteristically increased due to the increased membrane metabolism, NAA (δ = 3.22 ppm) is characteristically decreased due to loss of neurons. The observation of the β-methyl-proton doublet of the lactate (Lac) at δ = 1.33 ppm is not usually possible. Observation of lactate is assumed to be an indication for anaerobic glycolysis. In general resonances of free fatty acids (chemical shift range δ = 0– 2.4 ppm) are not observed in 1H MR spectra in vivo of brain tumors and are indicative for a tumor necrosis. a Patient with a World Health Organization (WHO) grade II oligodendroglioma. b Patient with a cystic astrocytoma. The high lipid content is a spectroscopic indication of necrosis and may be indicative for anaplastic tumor regions. c Patient with a brain metastasis from bronchial carcinoma

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.16  1H MRS of the prostate (modified from Kurhanewicz et al. 1996). a T2-weighted MR images of the prostate (TR/ TE = 5,000/108) in several successive slices. The (Cho + Cr)–citrate ratios are overlaid in red. The threshold for the ratio is set to 0.82. No ratios above the threshold are observed after cryotherapy (not displayed). b Representative 1H MR spectra in vivo of necrotic tumor tissue after cryotherapy, of tissue of benign prostate hyperplasia, and of malignant prostate carcinoma (3D

spectroscopic imaging, voxel size = 0.24 cm3, TR/TE = 1,000/30). Whereas high concentrations of citrate are observed in tissues of benign prostate hyperplasia and of normal prostate, in tissue of prostate carcinoma increased signal intensities of choline and decreased citrate levels are typically measured. In necrotic tissue no resonances of metabolites are anticipated to be observed using 1H MRS. c Corresponding histological samples (hematoxylin staining, original magnification ×400)

1363

1364

15  Clinical Spectroscopy

Fig. 15.4.17  Intracellular pH values of benign and malignant tumors non-invasively measured by means of 31P MRS in vivo. Reference value: pH value of calf muscle (with standard deviation indicated). Typically increased pH values are observed in tumors using 31P MRS

Typically, PME and PDE resonances of tumor spectra and the intratumoral pH (Fig. 15.4.17) are increased (Ng et al. 1989; Oberhänsli et al. 1986). However, a differentiation of tumors by means of these parameters is not possible. The observed pH increase (in contrary to Warburg’s hypothesis) is consistent with results of intensive investigations ex vivo and in vivo yielding alkaline pH values in necrotic tissue (Vaupel et al. 1989). Methods of high-resolution nuclear magnetic resonance increase the spectral information. Application of 1H spin decoupling in 31P MRS allows the resolution of broad PME and PDE resonances (Luyten et al. 1989) (cf. Fig. 15.2.9). Multinuclear MRS investigations (1H, 31P,

Fig. 15.4.18  31P MRS of a malignant melanoma of a 57-year-old patient. Spectrum (a) in situ (B0 = 1.5 T), and b ex vivo after extirpation of the tumor tissue (B0 = 11.7 T) (in cooperation with Dr. W. Hull, German Cancer Research Center). In the high-field spectrum, which was measured at a temperature of 4°C, PME and PDE resonance bands are almost resolved and the resonance lines can be assigned; however, the signals from PCr and NTP are strongly reduced. This is due to the short time between tumor tissue extirpation and the freeze clamping of the sample. This time is sufficient to consume the energy reserve of the tumor cell and to initiate the irreversible decay of the energy rich phosphate compounds (GP phosphorylated glucose, UDPG uridine-5′-diphosphoglucose). The Pi and the energy-rich phosphates are fitted to Lorentzian lines in the MR spectrum in vivo (upper spectrum)

31P) in combination with double-resonance techniques yield better assignment of the resonances of tumor spectra (Bachert et al. 1992b). Further information can be gained by the comparison of 31P MRS spectra ex vivo obtained at high fields (Fig. 15.4.18). Limitations of these comparisons are the immediate and fast disintegration of the energy-rich phosphates during the time between extirpation and freeze clamping of the tumor.

15.4  Clinical Applications in Selected Organs and Tumors

In localized 1H MRS of soft tissue and bone tumors, resonances in the range of methyl and methylene protons are observed, which are assigned to mobile fatty acids (= 13, Bongers et al. 1992a). In many cases signals of aliphatic compounds are detected between the water and lipid signals; however, up to now tumor-specific in vivo 1H resonances are not observed. Spectroscopic investigations of the bone marrow are possible only by means of 1H MRS due to the difficult localization (Schick 1996). Magnetic susceptibility changes due the trabecular bone structure cause field homogeneities, degrading the spectral resolution. Besides resonances of free fatty acids (methyl, methylene, olefinic protons) and of water, no other resonances can be resolved in 1H MR spectra in vivo of the bone marrow (Schick 1996). In patients with leukemia, investigations with single-voxel techniques and spectroscopic imaging were performed to detect and quantify the infiltration of the bone marrow (Bongers et al. 1992b; Gückel et al. 1990a, b; Irving et al. 1987). Clinical studies demonstrate that the relative fat and water fractions—which can be determined non-invasively, spatially resolved, and quantitatively by spectroscopic imaging—depend on the bone marrow infiltration (Gückel et al. 1990a, b). Seminomas are the most frequent malignant lesions in the testis. 31P spectra in vivo yield high Pi and reduced PME–β-NTP signal-intensity ratios. Motivated by animal experiments that were performed already in the 1980s and in which the growth and the therapy of tumors was monitored by means of 31P MRS in vivo (Evanochko et al. 1984; Naruse et al. 1985a, b; Ng et al. 1982; Shine et al. 1989), clinical studies were conducted to monitor the metabolic changes in patients as early as possible after onset of therapy. Correlation of these early spectroscopic parameters with other clinical findings is assumed to be prognostic. Monitoring studies by means of 31P MRS were conducted mainly in tumors, which could be localized with surface coils, e.g., lymphomas, lymph node metastases, malignant melanomas, and bone, soft-tissue, and liver tumors (75, Meyerhoff et al. 1992; 85, Ng et al. 1987; Prescott et al. 1993; Redmond et al. 1992a, b; Semmler et al. 1988a, b, Sostman et al. 1994). Using volume-selective 1H and 31P MRS, patients with brain tumors were examined during and after irradiation (Heesters et al. 1993; Schlemmer et al. 2001), chemotherapy (Arnold et al. 1989), and embolization therapy (Blankenhorn et al. 1999; Jüngling et al. 1993). Short-term monitoring (e.g., during the first infusion of the cytotoxic drug), measures the immediate therapeutic to physiologic and metabolic processes. Long-term monitoring is used to determine spectral changes, which become detectable after days. In patients with soft-tissue tumors and lymphomas

during and after chemotherapy and/or radiation treatment, changes were observed in the tumor metabolism early after therapy onset (Ng et al. 1987; Semmler et al. 1988a, b). In a patient with a recurrent malignant melanoma, locoregional perfusion therapy was monitored by means of 31P MRS. A decrease in the intracellular pH and a total breakdown of the NTP pool, as well as an increase in the concentration of Pi and the intermediates of the membrane phospholipid metabolism (PME, PDE) was observed in the first 2 days after onset of the therapy (Semmler et al. 1988a; cf. Fig. 15.4.19). In the course of the investigation the pH increased. The histological examination yielded a necrotic tumor. This finding is in accordance with observed increase of the pH value. Radiation-therapy monitoring was conducted in patients with brain tumors. In accordance with the observations of other groups, Heesters et al. (1993) measured before irradiation reduced NAA and increased Cho and Lac resonances in patients with high-grade gliomas by means of 1H MRS (Schlemmer et al. 2001; Alexander et al. 2006). After radiation (54–60 Gy) a decrease of Cho and Lac signal intensities and no change in the NAA signal intensity was observed (Figs. 15.4.20, 15.4.21, 15.4.22). Recently, 1H MRS was used not only for radiation therapy monitoring, but also for radiation-treatment planning (Payne and Leach 2006). Before surgery a selective embolization can be performed in patients with intracranial meningiomas to reduce bleeding during surgery. Using 1H and 31P MRS the induced changes of the tumor metabolism can be directly verified; it also makes possible a successful control of the embolization. Initially, a lactate signal was observed, as well as a retarded signal increase of the aliphatic resonances (Jüngling et al. 1993), a reduction of high-energy phosphates, and an increase of PME and PDE signal intensities (Blankenhorn et al. 1999). These effects are characteristic for the transition of vital tumor tissue to necrosis (hypoxic–anoxic state, deliberation of mobile lipids and phospholipids during cell degradation) caused by the devascularization of the tissue. In 31P MRS therapy–monitoring studies of combined radiotherapy and hyperthermia of soft-tissue tumors, a correlation could be established between the parameters PCr–PDE, PME–Pi, and NTP–PME signal-intensity ratios, as well as pH and the response of the tumor to the treatment (Dewhirst et al. 1990; Prescott et al. 1993; Sostman et al. 1994). In particular, a correlation was established between increased pH of the tumor and the degree of necrosis (>90%) in the histological specimen. The difference between the intracellular pH of tumor and normal tissue may be used for tests of the efficacy of tumor therapies (Gerweck and Seetharaman 1996; Ng et al. 1989). Changes in the in vivo 31P MR spectra observed at early times during therapy monitoring are predictive for the therapy success (Prescott et al. 1993; Pretlow et al. 1985).

1365

1366

15  Clinical Spectroscopy

Fig. 15.4.19  Long-term therapy monitoring. A patient with a recurrent malignant melanoma plantar at the foot was examined by means of 31P MRS in vivo before and 1, 2, 3, 7, 10, and 80 days after a locoregional perfusion therapy (Semmler et al. 1988a). a The 31P MR spectrum in vivo displays only the resonance of the inorganic phosphate (Pi), and no energy-rich phosphates can be detected at the first day. This exceptional finding mir-

rors the breakdown of the energy metabolism of the tumor cells and demonstrates the efficacy of the therapeutic intervention. b Spectral parameter as function of time after locoregional perfusion therapy (t = 0). The quantitative analysis of the 31P MR spectra clearly demonstrates the change of the intracellular pH. The tumor diameter—as evaluated from the MR images—changed only marginally during the observation interval of 80 days

Changes of Cho, (P)Cr, and Lip signal intensities were observed in patients with Ewing sarcomas, fibrous histiocytoma, and multiple myelomas by means of 1H MRS (Bongers et al. 1992a). In a patient with a Ewing sarcoma, a significant increase of the Lip signal intensity during chemotherapy is interpreted as a fatty degeneration. The spectroscopic 1H MR imaging of Lip and water protons is a clinically applicable method of systemic neoplastic bone marrow diseases (e.g., leukemia, malignant lymphoma) (Gückel et al. 1990a, b). An increase of the

lipid fraction is indicative for a response to therapy. Clinical MRS studies of the pharmacokinetics are primarily performed during chemotherapy with fluoropyrimidines using the 19F nucleus. The kinetics of 5-FU and FBAL in the liver of patients during treatment can be observed time resolved (Glaholm et al. 1990; Presant et al. 1994; Schlemmer et al. 1994; Semmler et al. 1990; Wolf et al. 1992). The low concentration of the cytotoxic 5-FUranuc (in most patients below the detection limits) may explain in part the low response rate of patients with liver metas-

15.4  Clinical Applications in Selected Organs and Tumors Fig. 15.4.20  T1-weighted scans after contrast media application, a frontal lesion, assumed to be radionecrosis, c left temporal, assumed to be recurrent of tumor. Single-voxel 1H MR spectrum (b) of the VOI (necrosis) indicated in (a) and (d) of the VOI (vital tumor) indicated in (c) (parameters for both spectra: PRESS, VOI = 20× 2 × 2 cm3, TR = 1,500 ms, echo time TE = 135 ms, AC = 200, B0 = 1.5 T) (Schlemmer et al. 2001). a Transversal and frontal T1-weighted images of the contrast-enhancing frontal lesion. A conformal tumor irradiation was performed with 64 Gy 36 months before the image was taken. The tumor was biopsied and the histology confirmed radionecrosis. b The 1H MR spectrum of the VOI (frontal lesion) indicated in a shows lipid resonances being a characteristic feature of necrosis. c Contrast-enhanced transversal and frontal T1-weighted images of left temporal lesion. 18FDG PET proofed strong glucose uptake of the fast-growing tumor. d The 1H MR spectrum of the VOI (left temporal lesion) indicated in b demonstrates strong choline and weak NAA resonances of the progredient tumor

1367

15  Clinical Spectroscopy

1368

Fig. 15.4.21  1H MR spectroscopic imaging of a patient with a glioblastoma 4 weeks after irradiation of an initially histological diagnosed anaplastic astrocytoma (TR = 1,500 ms, echo time TE = 135 ms, B0 = 1.5 T). a T2-weighted imaging of the lesion. b Overlay of T2-weighted-image and the color-coded NAA sig-

nal intensities. The NAA signal intensities are markedly reduced in the area of the tumor and are indicative for a loss of neuronal tissue. c Overlay of T2-weighted-image and the color-coded Cho–NAA signal-intensity ratio. The ratio is higher in the region of the glioblastoma

4

4 4

ItCho / ItChr

3

3

0

2

2

0

ItCho / ItChr

3

ItCho / ItChr

PD

1

3

ItCho / ItChr

4 2

RI SD Contralateral area

2 1

2

3

ItCho / INAA

1

PD

1

4

5 PD RI

Fig. 15.4.22  Relative signal intensity of total choline and NAA ( , RI RI, n = 17), and stable disease ( , SD SD, n = 15). The lines PD Contralateral (ItCho1/INAA) versus total choline and total creatine (ItCho/ItCr) in ce- wereSD calculated by means of a discriminant analysis and differarea RI rebral lesions after stereotactic radiation therapy of progredient entiates between malignant (PD) and non-malignant lesions Contralateral 0 area SD 3 4 of the contralateral 5 intensities healthy brain 0 brain tumors (1H MRS, PRESS, T0R = 1,500 ms,1echo time 2(SD, RI) The relative Contralateral T  = 135 ms, B0 = 1.5 T). The lesions were classified as tumor tissue ( , n = 33) are below the line (Schlemmer et al. 2001) 0E 1 2 3 4 5 /area ItCho INAA PD n = 29), radiation-induced tissue alterations progression ( , PD, 0 ItCho / INAA 0 2 3 4 5 RI 1 SD Contralateral area

3

Cho / INAA

4

5

ItCho / INAA

15.4  Clinical Applications in Selected Organs and Tumors

Fig. 15.4.23  Series of successively acquired 19F MR spectra (6 min/spectrum) during intravenous 5-FU chemotherapy. The 19F MR spectra in vivo originate from cervical lymph nodes of a patient with an oropharyngeal/tongue tumor during and after a 50-min infusion of 5-FU (dose: 1,000 mg/m2 body surface).

The spectra were acquired using a 5-cm surface coil. The intratumoral kinetics of 5-FU, DHFU, and FBAL can be monitored. The chemical shift is given relatively to the external standard trichloroacetic acid (TFA) at 0 ppm. 5-FU, δ = –93.7 ppm, FBAL, δ = –112.7 ppm, DHFU, δ = –26.3 ppm (Schlemmer et al. 1999)

tases of progredient colorectal tumors to the 5-FU chemotherapy. 19F MRS allows follow-up of the intratumoral 5-FUmetabolism. For instance, in patients with oropharyngeal carcinomas treated with a simultaneous radiochemotherapy of 5-FU, the intratumoral increase of the 5-FU concentration could be observed during follow-up (Fig. 15.4.23; Schlemmer et al. 1999). For the first time, this study yielded evidences that interaction of radiation therapy and 5-FU-pharmacokinetics takes place. A trapping of 5-FU is a necessary but not a sufficient condition for the tumor response. The half-life of ≥20 min can be considered to be a trapping of 5-FU in the tumor (Wolf et al. 1998). The modulation of the 5-FU therapy by combined administration of methotrexate or interferon may be observed by 19F MRS as well (Presant et al. 2000; cf. also Wolf et al. 2000). The results of the clinical studies demonstrate that MRS can contribute to the diagnostics of brain and prostate tumors. The 1H MR signals in vivo of NAA, Cho, Lac, and citrate are important metabolites for the diagnostic of tumorous tissue. The resonances of PME and PDE as well as the pH, which are increased in tumors, play a similar roll in the application of the 31P MRS in vivo. The metabolization of fluorinated drugs may be non-invasively observed during treatment.

References 1.

2.

3.

4.

5.

Adamson AJ, Rand SD, Probst RW, Kim TA, Schultz C, Haughton VM (1998) Focal brain lesions: effect of single– voxel proton MR spectroscopic findings on treatment decisions. Radiology 209:73–78 Alexander A, Murtha A, Abdulkarim B, Mehta V, Wheatley M, Murray B, Riauka T, Hanson J, Fulton D, McEwan A, Roa W (2006) Prognostic significance of serial magnetic resonance spectroscopy over the course of radiation therapy for patients with malignant glioma. Clin Invest Med. 29:301–311 Argov Z, Bank WJ, Maris J, Peterson P, Chance B (1987) Bioenergetic heterogeneity of human mitochondrial myopathies: phosphorous magnetic resonance spectroscopy study. Neurology 37:257–262 Arnold DL, Shoubridge EA, Emrich J, Feindel W, Villemure JG (1989) Early metabolic changes following chemotherapy of human gliomas in vivo demonstrated by phosphorus magnetic resonance spectroscopy. Invest Radiol 24:958–961 Arnold DL, Matthews PM, Francis G, Antel J (1990) Proton magnetic resonance spectroscopy of human brain invivo in the evaluation of multiple sclerosis: assessment of the load of disease. Magn Reson Med 14:154–159

1369

1370

15  Clinical Spectroscopy 6.

7. 8.

9. 10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Arnold DL, de Stefano N, Narayanan S, Matthews PM (2000) Proton MR spectroscopy in multiple sclerosis. Neuroimaging Clin N Am 10:789–798 Azzopardi D, Edwards D (1995) Magnetic resonance spectroscopy in neonates. Curr Opin Neurol 8:145–149 Bachert P (1997) Kinetics of heteronuclear cross-relaxation in dipolar coupled-spin systems: applications to in vivo measurements. Concepts Magn Reson 9:227–245 Bachert P (1998) Pharmacokinetics using fluorine NMR in vivo. Prog. NMR Spectrosc 33:1–56 Bachert P, Bellemann ME (1992b) Kinetics of the in vivo 31P-{1H} nuclear Overhauser effect of the humancalf-muscle phosphocreatine resonance. J Magn Reson 100:146–156 Bachert P, Bellemann ME, Layer G, Koch T, Semmler W, Lorenz WJ (1992a) In vivo 1H , 31P-{1H} and 13C-{1H} magnetic resonance spectroscopy of malignant histiocytoma and skeletal muscle tissue in man. NMR Biomed 5:161–170 Bachert-Baumann P, Ermark F, Zabel H-J, Sauter R, Semmler W, Lorenz WJ (1990) In vivo nuclear Overhauser effect in 31P-{1H} double-resonance experiments in a 1.5-T whole-body MR system. Magn Reson Med 15:165–172 Bakken IJ, Gribbestad IS, Singstad TE, Kvistad KA (2001) External standard method for the in vivo quantification of choline-containing compounds in breast tumors by proton MR spectroscopy at 1.5 Tesla. Magn Reson Med 46:189–192 Barker PB, Breiter SN, Soher BJ, Chatham JC, Forder JR, Samphilipo MA, Magee CA, Anderson JH (1994) Quantitative proton spectroscopy of canine brain: in vivo and in vitro correlations. Magn Reson Med 32:157–163 Barker PB, Lee RR, McArthur JC (1995) AIDS dementia complex: evaluation with proton MR spectroscopic imaging. Radiology 195:58–64 Bartels M, Günther U, Albert K, Mann K, Schuff N, Stuckstedte H (1991) 19F magnetic resonance spectroscopy of neuroleptics: the first in vivo pharmacokinetics of trifluoperazine in the rat brain and the first in vivo spectrum of fluphenazine in the human brain. Biol Psychiatry 30:656–662 Bartha R, Drost DJ, Williamson PC (1999) Factors affecting the quantification of short echo in-vivo 1H MR spectra: prior knowledge, peak elimination, and filtering. NMR Biomed 12:205–216 Baslow MH (2000) Functions of N-acetyl-l-aspartate and N-acetyl-aspartylglutamate in the vertebrate brain: role in glial cell-specific signalling. J Neurochem 75:453–459 Beauchamp NJ Jr, Barker PB, Wang PY, vanZijl PC (1999) Imaging of acute cerebral ischemia. Radiology 212:307–324 Beckmann N, Turkalj 1, Seelig J, Keller U (1991). 13C NMR for the assessment of human brain glucose metabolism in vivo. Biochemistry 30:6362–6366

21. Birken DL, Oldendorf WH (1989) N-Acetyl-l-aspartic acid: a literature review of a compound prominent in 1HNMR spectroscopic studies of brain. Neurosci Biobehav Rev13:23–31 22. Blamire AM, Rajagopalan B, Radda GK (1999) Measurement of myocardial pH by saturation transfer in man. Magn Reson Med 41:198–203 23. Blankenhorn M, Bachert P, Semmler W, Ende G, Tronnier V, van Kaick G, Sartor K (1999) Phosphor-31-MR Spektroskopische Bildgebung bei präoperativer Embolisationstherapie von Meningeomen. Fortschr Rontgenstr 170:568–574 24. Bluml S, Seymour KJ, Ross BD (1999) Developmental changes in choline- and ethanolamine-containing compounds measured with proton-decoupled 31P MRS in in vivo human brain. Magn Reson Med 42:643–654 25. Bluml S, Moreno-Torres A, Ross BD (2001) 1–13C glucose MRS in chronic hepatic encephalopathy in man. Magn Reson Med 45:981–993 26. Boesch C, Gruetter R, Martin E, Duc G, Wüthrich K (1989) Variations in the in vivo P-31 MR spectra of the developing human brain during postnatal life. Radiology 172:197–199 27. Boesch C, Elsing C, Wegmuller H, Felblinger J, Vock P, Reichen J (1997) Effect of ethanol and fructose on liver metabolism: a dynamic 31phosphorus magnetic resonance spectroscopy study in normal volunteers. Magn Reson Imaging 15:1067–1077 28. Boesiger P, Buchli R, Meier D, Steinmann B, Gitzelmann R (1994) Changes of liver metabolite concentrations in adults with disorders of fructose metabolism after intravenous fructose by 31P magnetic resonance spectroscopy. Pediatr Res 36:436–440 29. Bolo NR, Hode Y, Nedelec JF, Laine E, Wagner G, Macher JP (2000) Brain pharmacokinetics and tissue distribution in vivo of fluvoxamine and fluoxetine by fluorine magnetic resonance spectroscopy. Neuropsychopharmacology 23:428–438 30. Bongers H, Schick F, Skalej M, Hess CF, Jung WI (1992a) Localized in vivo 1H spectroscopy of human bone and soft tissue tumours. Eur J Radiol 2:459–464 31. Bongers H, Schick F, Skalej M, Jung Wl, Einsele H (1992b) Localized in vivo 1H spectroscopy and chemical shift imaging of the bone marrow in leukemic patients. Eur J Radiol 2:350–356 32. Bottomley PA (1989) Human in vivo NMR spectroscopy in diagnostic medicine: clinical tool or research probe? Radiology 170:1–15 33. Bottomley PA, Foster TH, Darrow RD (1984) Depth resolved surface coil spectroscopy (DRESS) for in vivo 1H, 31P and 13C NMR. J Magn Reson 59:338–342 34. Bottomley PA, Herfkens RJ, Smith LS, Bashore TM (1987) Altered phosphate metabolism in myocardial infarction: P-31 MR spectroscopy. Radiology 165:703–707 35. Bottomley PA, Atalar E, Weiss RG (1996) Human cardiac high-energy phosphate metabolite concentrations by 1 Dresolved NMR spectroscopy. Magn Reson Med 35:664–670

15.4  Clinical Applications in Selected Organs and Tumors 36. Bourdel-Marchasson I, Biran M, Thiaudiere E, Delalande C, Decamps A, Manciet G, Canioni P (1996) 31P magnetic resonance spectroscopy of human liver in elderly patients: changes according to nutritional status and inflammatory state. Metabolism 45:1059–1061 37. Bovée WMM (1991) Quantification of glutamate, glutamine, and other metabolites in in vivo proton NMR spectroscopy. NMR Biomed 4:81–84 38. Bovée W, Creyghton J (1992) Acquisition and processing of the NMR data. In: Certaines JD, Bovée W, Podo F (eds) Magnetic resonance spectroscopy in biology and medicine. Pergamon, Oxford 39. Brechtel K, Dahl DB, Machann J, Bachmann OP, Wenzel I, Maier T, Claussen CD, Haring HU, Jacob S, Schick F (2001) Fast elevation of the intramyocellular lipid content in the presence of circulating free fatty acids and hyperinsulinemia: a dynamic 1H MRS study. Magn Reson Med 45:179–183 40. Brown TR, Kincaid BM, Ugurbil K (1982) NMR chemical shift imaging in three dimensions. Proc Natl Acad Sci USA 79:3523–3526 41. Bruhn H, Frahm J, Gyngell ML, Merboldt KD, Hänicke W, Sauter R (1989) Cerebral metabolism in man after acute stroke: new observations using localized proton NMR spectroscopy. Magn Reson Med 9:126–131 42. Bruhn H, Frahm J, Gyngell ML, Merboldt KD, Hänicke W, Sauter R (1991) Localized proton NMR spectroscopy using stimulated echoes: applications to human skeletal muscle in vivo. Magn Reson Med 17:82–94 43. Buchli R, Martin E, Boesiger P, Rumpel H (1994) Developmental changes of phosphorus metabolite concentrations in the human brain: a 31P magnetic resonance spectroscopy study in vivo. Pediatr Res 35:431–435 44. Burtscher IM, Holtas S (2001) Proton MR spectroscopy in clinical routine J Magn Reson Imaging 13:560–567 45. Cadoux-Hudson TAD, Blackledge MJ, Rajagopalan B, Taylor DJ, Radda GK (1989) Human primary brain tumor metabolism in vivo: a phosphorus magnetic resonance spectroscopy study. Br J Cancer 60:430–436 46. Cady EB (2001) Magnetic resonance spectroscopy in neonatal hypoxic-ischaemic insults. Childs Nerv Syst 17:145–149 47. Cady EB, Costello AM, Dawson MJ, Delpy DT, Hope PL, Reynolds EO, Tofts PS, Wilkie DR (1983) Non-invasive investigation of cerebral metabolism in newborn infants by phosphorus nuclear magnetic resonance spectroscopy. Lancet 14:1059–1062 48. Capizzano AA, Vermathen P, Laxer KD, Ende GR, Norman D, Rowley H, Matson GB, Maudsley AA, Segal MR, Weiner MW (2001) Temporal lobe epilepsy: qualitative reading of 1H MR spectroscopic images for presurgical evaluation. Radiology 218:144–151 49. Chamuleau RAFM, Bosman DK, Bovée WMM, Luyten PR, den Hollander JA (1991) What the clinician can learn from MR glutamine/glutamate assays. NMR Biomed 4:103–108 50. Chance B (1989) What are the goals of magnetic resonance research? NMR Biomed 2:5–6

51. Chang L, Miller BL, McBride D, Conford M, Oropilla G, Buchthal S, Chiang F, Aronow H, Ernst T (1995) Brain lesions in patients with AIDS: H-1 MR spectroscopy. Radio­ logy 197:525–531 52. Cheng LL, Newell K, Le R, Hyman BT, Gonzalez RG (2001) Now it can be said: N-acetylaspartate is a neuronal marker—an HRMAS proton MR spectroscopy and stereo­ logy study. Proc Intl Soc Magn Reson Med 9:1672 53. Choi DW (1988) Glutamate neurotoxicity and diseases of the nervous system. Neuron 1:623–634 54. Constans JM, Meyerhoff DJ, Gerson J, MacKay S, Norman D, Fein G, Weiner MW (1995) H-1 MR spectroscopic imaging of white matter signal hyperintensities: Alzheimer disease and ischemic vascular dementia. Radiology 197:517–523 55. Costello LC, Franklin RB (1991) Concepts of citrate production and secretion by prostate: 1: Metabolic relationship. Prostate 18 : 25–46 56. Cox IJ, Bryant DJ, Collins AG, George P, Harman RR, Hall AS, Hodgson HJ, Khenia S, McArthur P, Spencer DH, et al. (1988) Four-dimensional chemical shift MR imaging of phosphorus metabolites of normal and diseased human liver. J Comput Assist Tomogr. 12:369-376 57. Daly PF, Lyon RC, Faustino PJ, Cohen JS (1987) Phospholipid metabolism in cancer cells monitored by in vivo 31P NMR spectroscopy. J Biol Chem 262:14875–14878 58. Danielsen R, Ross B (eds) (1999) Magnetic resonance spectroscopy diagnosis of neurological diseases. Dekker, New York 59. de Stefano N, Matthews PM, Ford B, Genge A, Karpati G, Arnold DL (1995) Short-term dichloroacetate treatment improves indices of cerebral metabolism in patients with mitochondrial disorders. Neurology 45:1193–1198 60. Dewhirst MW, Sostman HD, Leopold KA, Charles HC, Moore D, Burn RA, Tucker JA, Harrelson JM, Oleson JR. (1990) Soft-tissue sarcomas: MR imaging and MR spectroscopy for prognosis and therapy monitoring. Radiology 174:847-853 61. Dross K, Kewitz H (1972) Concentration and origin of choline in the rat brain. Naunyn Schmiedebergs Arch Pharmacol 274:91–106 62. Du F, Zhu XH, Qiao H, Zhang X, Chen W (2007) Efficient in vivo 31P magnetization transfer approach for noninvasively determining multiple kinetic parameters and metabolic fluxes of ATP metabolism in the human brain. Magn Reson Med 57:103–114 63. Dzik-Jurasz AS, Collins DJ, Leach MO, Rowland IJ (2000) Gallbladder localization of 19F MRS catabolite signals in patients receiving bolus and protracted venous infusional 5-fluorouracil. Magn Reson Med 44:516–20 64. el-Deredy W (1997) Pattern recognition approaches in biomedical and clinical magnetic resonance spectroscopy: a review. NMR Biomed 10:99–124 65. Ende G, Bachert P (1993) Dynamic 13C-{1H} nuclear polarization of lipid methylene resonances applied to broadband proton-decoupled in vivo 13C MR spectroscopy of human breast and calf tissue. Magn Reson Med 30:415–423

1371

1372

15  Clinical Spectroscopy 66. Ende GR, Laxer KD, Knowlton RC, Matson GB, Schuff N, Fein G, Weiner MW (1997) Temporal lobe epilepsy: bilateral hippocampal metabolic changes revealed at proton MR spectroscopic imaging. Radiology 202:809–817 67. Ernst T, Kreis R, Ross BD (1993) Absolute quantitation of water and metabolites in the human brain. I. Compartments and water. J Magn Reson B 102:1–8 68. Ernst T, Chang L, Melchor R, Mehringer CM (1997) Frontotemporal dementia and early Alzheimer disease: differentiation with frontal lobe H-1 MR spectroscopy. Radio­ logy 203:829–836 69. Ernst T, Itti E, Itti L, Chang L (2000) Changes in cerebral metabolism are detected prior to perfusion changes in early HIV-CMC: A coregistered 1H MRS and SPECT study. J Magn Reson Imaging 12:859–865 70. Evanochko W, Ng TC, Glickson JD (1984) Application of in vivo NMR spectroscopy to cancer. Magn Reson Med 1:508–534 71. Federico F, Simone IL, Lucivero V, Giannini P, Laddomada G, Mezzapesa DM, Tortorella C (1998) Prognostic value of proton magnetic resonance spectroscopy in ischemic stroke. Arch Neurol 55:489–494 72. Fowler AH, Pappas AA, Holder JC, Finkbeiner AE, Dalrymple GV, Mullins MS, Sprigg JR, Komoroski RA. (1992) Differentiation of human prostate cancer from benign hypertrophy by in vitro 1H NMR. Magn Reson Med. 25(1):140-147 73. Frahm J, Merboldt KD, Hänicke W (1987) Localized proton spectroscopy using stimulated echoes. J Magn Reson 72:502–508 74. Frahm J, Bruhn H, Gyngell ML, Merboldt KD, Hänicke W, Sauter R (1989) Localized high-resolution proton NMR spectroscopy using stimulated echoes: initial applications to human brain in vivo. Magn Reson Med 9:79–93 75. Fu L, Matthews PM, de Stefano N, Worsley KJ, Narayanan S, Francis GS, Antel JP, Wolfson C, Arnold DL (1998) Imaging axonal damage of non-appearing white matter in multiple sclerosis. Brain 121:103–113 76. Gerweck LE, Seetharaman K (1996) Cellular pH gradient in tumor versus normal tissue: potential exploitation for the treatment of cancer. Cancer Res 56:1194–1198 77. Geurts JJ, Barkhof F, Castelijns JA, Uitdehaag BM, Polman CH, Pouwels PJ (2004) Quantitative 1H MRS of healthy human cortex, hippocampus, and thalamus: metabolite concentrations, quantification precision, and reproducibility. J Magn Reson Imaging 20:366–371 78. Gillies RJ, Morse DL (2005) In vivo magnetic resonance spectroscopy in cancer. Annu Rev Biomed Eng. 7:287–326 79. Gillard JH, Barker PB, van Zijl PC, Bryan RN, Oppenheimer SM (1996) Proton MR spectroscopy in acute middle cerebral artery stroke. AJNR 17:873–886 80. Glaholm J, Leach MO, Collins DJ, Mansi J, Sharp JC, Madden A, Smith IE, McCready VR (1989) In-vivo 31P magnetic resonance spectroscopy for monitoring treatment response in breast cancer. Lancet 10:1326–1327

81. Glaholm J, Leach MO, Collins D, al Jehazi B, Sharp JC, Smith TA, Adach J, Hind A, McCready VR, White H. (1990) Comparison of 5-fluorouracil pharmacokinetics following intraperitoneal and intravenous administration using in vivo 19F magnetic resonance spectroscopy. Br J Radiol. 63:547-553 82. Gonen O, Murphy-Boesch J, Li C-W, Padavic-Shaller K, Negendank WG, Brown TR (1997) Simultaneous 3D NMR spectroscopy of proton-decoupled fluorine and phosphorus in human liver during 5-fluorouracil chemotherapy. Magn Reson Med 37:164–169 83. Gonen O, Catalaa I, Babb JS, Ge Y, Mannon LJ, Kolson DL, Grossman RI (2000) Total brain N-acetylaspartate: a new measure of disease load in MS. Neurology 54:15–19 84. Gotsis ED, Fountas K, Kapsalaki E, Toulas P, Peristeris G, Papadakis N (1996) In vivo proton MR spectroscopy: the diagnostic possibilities of lipid resonances in brain tumors. Anticancer Res 16:1565–1567 85. Graham GD, Kalvach P, Blamire AM, Brass LM, Fayad PD, Prichard JW (1995) Clinical correlates of proton magnetic resonance spectroscopy findings after acute cerebral infarction. Stroke 26:225–229 86. Grand S, Passaro G, Ziegler A, Estève F, Boujet C, Hoffmann D, Rubin C, Segebarth C, Décorps M, le Bas JF, Rémy C (1999) Necrotic tumor versus brain abscess: importance of amino acids detected at 1H MR spectroscopy—initial results. Radiology 213:785–793 87. Graves EE, Nelson SJ, Vigneron DB, Verhey L, McDermott M, Larson D, Chang S, Prados MD, Dillon WP (2001) Serial proton MR spectroscopic imaging of recurrent malignant gliomas after gamma knife radiosurgery. AJNR Am J Neuroradiol 22:613–624 88. Grodd W, Krägeloh-Mann I, Petersen D, Trefz FK, Harzer K (1990) In vivo assessment of N-acetylaspartate in brain in spongy degeneration (Canavan’s disease) by proton spectroscopy. Lancet 336:437–438 89. Groenendaal F, Roelants-Van Rijn AM, van Der Grond J, Toet MC, de Vries LS (2001) Glutamate in cerebral tissue of asphyxiated neonates during the first week of life demonstrated in vivo using proton magnetic resonance spectroscopy. Biol Neonate 79:254–257 90. Gückel F, Brix G, Semmler W, Knauf W, Ho AD, van Kaick G (1990a) Systemic bone marrow disorders: characterization with proton chemical shift imaging. J Comp Assist Tomogr 14:633–642 91. Gückel F, Brix G, Semmler W, Ho AD, Körbling M, Georgi M, van Kaick G (1990b) Proton chemical shift imaging of bone marrow for monitoring therapy in leukemia. J Comp Assist Tomogr 14:954–959 92. Hagberg G (1998) From magnetic resonance spectroscopy to classification of tumors. A review of pattern recognition methods. NMR Biomed 11:148–156 93. Hamilton PA, Cady EB, Wyatt JS, Hope PL, Delpy DT, Reynolds ED (1986) Impaired energy metabolism in brains of newborn infants with increased cerebral echo densities. Lancet 1:1242–1246

15.4  Clinical Applications in Selected Organs and Tumors 94. Häussinger D, Laubenberger J, vom Dahl D, Ernst Th, Bayer S, Langer M, Gerok W, Hennig J (1994) Proton magnetic resonance spectroscpic studies on human brain myoInositol in hypoosmolarity and hepatic encephalopathy. Gastroenterology 107:145–148 95. Heerschap A, Luyten PR, van der Heyden JI, Oosterwaal LJ, den Hollander JA (1989) Broadband proton decoupled natural abundance 13C NMR spectroscopy of humans at 1.5 T. NMR Biomed 2:124–32 96. Heerschap A, Jager GJ, van den Graaf M, Barentsz JO, Ruijs SH (1997) Proton MR spectroscopy of the normal human prostate with an endorectal coil and a double spin-echo pulse sequence. Magn Reson Med 37:204–213 97. Heesters MAAM, Kamman RL, Mooyaart EL, Go KG (1993) Localized proton spectroscopy of inoperable brain gliomas: response to radiation therapy. J Neurooncol 17:27–35 98. Heindel W, Bunke J, Glathe S, Steinbrich W, Mollevaneer L (1988) Combined 1H MR imaging and localized 31P spectroscopy of intracranial tumors in 43 patients. J Comp Assist Tomogr 12:907–916 99. Heindel W, Kugel H, Wenzel F, Stippel D, Schmidt R, Lackner K (1997) Localized 31P MR spectroscopy of the transplanted human kidney in situ shows altered metabolism in rejection and acute tubular necrosis. J Magn Reson Imaging 7:858–864 100. Heiss WD, Heindel W, Herholz K, Rudolf J, Bunke J, Jeske J, Friedmann G (1990) Positron emission tomograph of fluorine-18-deoxyglucose and image-guided phosphorus31 magnetic resonance spectroscopy in brain tumors. J Nucl Med 31:302–310 101. Herholz K, Heindel W, Luyten PR, denHollander JA, Pietrzyk U, Voges J, Kugel H, Friedmann G, Heiss WD (1992) In vivo imaging of glucose consumption and lactate concentration in human gliomas. Ann Neurol. 31:319-327 102. Holshouser BA, Ashwal S, Luh GY, Shu S, Kahlon S, Auld KL, Tomasi LG, Perkin RM, Hinshaw DB (1997) Proton MR spectroscopy after acute central nervous system injury: outcome prediction in neonates, infants and children. Radiology 202:487–496 103. Hricak H (1989) Phosphorus-31 MRS of the kidney. Invest Radiol 24:993–996 104. Hricak H (2005) MR imaging and MR spectroscopic imaging in the pre-treatment evaluation of prostate cancer. Br J Radiol 78(Spec no. 2):S103–S111 105. Hubesch B, Sappey-Marinier D, Roth K, Matson GB, Weiner MW (1990) P-31 MR spectroscopy of normal human brain and brain tumors. Radiology 174:401–409 106. Husted CA, Duijn JH, Matson GB, Maudsley AA, Weiner MW (1994) Molar quantitation of in vivo proton metabolites in human brain with 3D magnetic resonance spectroscopic imaging. Magn Reson Imaging 12:661–667 107. Huzjan R, Sala E, Hricak H (2005) Magnetic resonance imaging and magnetic resonance spectroscopic imaging of prostate cancer. Nat Clin Pract Urol 2:434–442

108. Irving MG, Brooks WM, Brereton IM, Galloway GJ, Field J, Bell JR, Harris MG, Baddeley H, Doddrell DM. Use of high resolution in vivo volume selected 1H-magnetic resonance spectroscopy to investigate leukemia in humans. Cancer Res. 1987 47(14):3901-3906 109. Jimenez JV, Richards TL, Heide AC, Grierson JR, Shankland EG (1995) Incorporation of a phosphonium analogue of choline into the rat brain as measured by magnetic resonance spectroscopy. Magn Reson Med 33:285–292 110. Jüngling FD, Wakhloo AK, Hennig J (1993) In vivo proton spectroscopy of meningioma after preoperative embolization Magn Reson Med 30:155–160 111. Kanamori K, Ross BD, Chung JC, Kuo EL (1996) Severity of hyperammonemic encephalopathy correlates with brain ammonia level and saturation of glutamine synthetase in vivo. J Neurochem 67:1584–1594 112. Kelm BM, Menze BH, Zechmann CM, Baudendistel KT, Hamprecht FA (2007) Automated estimation of tumor probability in prostate MRSI: pattern recognition vs. quantification. Magn Reson Med 150–159 113. Kim GE, Lee JH, Cho YP, Kim ST (2001) Metabolic changes in the ischemic penumbra after carotid endarterectomy in stroke patients by localized in vivo proton magnetic resonance spectroscopy (1H MRS). Cardiovasc Surg 9:345–355 114. Kim SH, Chang KH, Song IC, Han MH, Kim HC, Kang HS, Han MC (1997) Brain abscess and brain tumor: discrimination with in vivo H-1 MR spectroscopy. Radiology 204:239–245 115. Kinoshita Y, Yokota A (1997) Absolute concentrations of metabolites in human brain tumors using in vitro proton magnetic resonance spectroscopy. NMR Biomed 10:2–12 116. Kiyono K, Shibata A, Sone S, Watanabe T, Oguchi M, Shikama N, Ichijo T, Kiyosawa K, Sodeyama T (1998) Relationship of 31P MR spectroscopy to the histopathological grading of chronic hepatitis and response to therapy. Acta Radiol 39:309–314 117. Komoroski RA (2005) Biomedical applications of 7Li NMR. NMR Biomed 18:67–73 118. Komoroski RA, Newton JEO, Karson C, Cardwell D, Jagannathan NR, Sprigg J (1989) Detection of psychoactive drugs in vivo in human brain using 19F NMR spectroscopy. In: Proceedings of the 8th Annual Meeting SMRM, Amsterdam, vol. 1, p 444 119. Komoroski RA, Newton JEO, Walker E, Cardwell D, Jagannathan NR, Ramaprasad S, Sprigg J (1990) In vivo NMR spectroscopy of lithium-7 in humans. Magn Reson Med 15:347–356 120. Koopmans RA, Li DKB, Zhu G, Allen PS, Penn A, Paty DW (1993) Magnetic resonance spectroscopy of multiple sclerosis: in-vivo detection of myelin breakdown products. Lancet 341:631–632 121. Kreis R, Farrow NA, Ross BD (1990) Diagnosis of hepatic encephalopathy by proton magnetic resonance spectroscopy. Lancet 336:635–636

1373

1374

15  Clinical Spectroscopy 122. Kreis R, Ernst T, Ross BD (1993a) Absolute quantitation of water and metabolites in the human brain. II. Metabolite concentrations. J Magn Reson B 102:9–19 123. Kreis R, Ernst T, Ross BD (1993b) Development of the human brain: in vivo quantification of metabolite and water content with proton magnetic resonance spectroscopy. Magn Reson Med 30:424–437 124. Kreis R, Arcinue E, Ernst T, Shonk TK, Flores R, Ross BD (1996) Hypoxic encephalopathy after near-drowning studied by quantitative 1H-magnetic resonance spectroscopy. J Clin Invest Mar 97:1142–1154 125. Kreis R, Jung B, Slotboom J, Felblinger J, Boesch C (1999) Effect of exercise on the creatine resonances in 1H MR spectra of human skeletal muscle. J Magn Reson 137:350–357 126. Krems B, Bachert P, Schlemmer H-P, Zabel H-J, van Kaick G, Lorenz WJ (1994 2D 19F chemical shift imaging of the metastatic liver of patients with colorectal tumors undergoing 5-fluorouracil chemotherapy. In: Proceedings of the 2nd Annual Meeting SMRM, San Francisco, vol. 3, p 1163 127. Krems B, Bachert P, Zabel H-J, Lorenz WJ (1995) 19F{1H} nuclear Overhauser effect and proton decoupling of 5-fluorouracil and α-fluoro-β-alanine. J Magn Reson B 108:155–164 128. Kugel H, Wittsack H-J, Wenzel F, Stippel D, Heindel W, Lackner K (2000) Non-invasive determination of metabolite concentrations in human transplanted kidney in vivo by 31P MR spectroscopy. Acta Radiol 41:634–641 129. Kurhanewicz J, Thomas A, Jajodia P, Weiner MW, James TL, Vigneron DB, Narayan P (1991) 31P spectroscopy of the human prostate gland in vivo using a transrectal probe. Magn Reson Med 22:404–413 130. Kurhanewicz J, Vigneron DB, Nelson SJ, Hricak H, MacDonald JM, Konety B, Narayan P (1995) Citrate as an in vivo marker to discriminate prostate cancer from benign prostatic hyperplasia and normal prostate peripheral zone: detection via localized proton spectroscopy. Urology 45:459–466 131. Kurhanewicz J, Vigneron DB, Hricak H, Parivar F, Nelson SJ, Shinohara K, Carroll PR (1996) Prostate cancer: metabolic response to cryosurgery as detected with 3 D H-1 MR spectroscopic imaging. Radiology 200:489–496 132. Lamb HJ, Beyerbacht HP, Ouwerkerk R, Doornbos J, Pluim BM, van der Wall EE, van der Laarse A, de Roos A (1997) Metabolic response of normal human myocardium to high-dose atropine-dobutamine stress studied by 31P MRS. Circulation 96:2969–2977 133. Lamb HJ, Beyerbacht HP, van der Laarse A, Stoel BC, Doornbos J, van der Wall EE, de Roos A (1999) Diastolic dysfunction in hypertensive heart disease is associated with altered myocardial metabolism. Circulation 99:2261–2267 134. Laubenberger J, Häussinger D, Bayer S, Thielemann S, Schneider B, Mundinger A, Hennig J, Langer M (1996) HIV-related abnormalities in the brain: depiction with proton MR spectroscopy with short echo times. Radiology 199:805–810

135. Laubenberger J, Häussinger D, Bayer S, Gufler H, Hennig J, Langer M (1997) Demonstration of cerebral metabolic abnormalities in asymptomatic patients with liver cirrhosis by proton MR spectroscopy. Gastroenterology; 112:1610–1618 136. Laubenberger J, Bayer S, Thiel T, Hennig J, Langer M (1998) Klinische Anwendungen der Protonen-Magnetresonanz-Spektroskopie des Gehirns. Fortschr Röntgenstr 168:539–549 137. Leach MO, Verrill M, Glaholm J, Smith TA, Collins DJ, Payne GS, Sharp JC, Ronen SM, McCready VR, Powles TJ, Smith IE (1998) Measurements of human breast cancer using magnetic resonance spectroscopy: a review of clinical measurements and a report of localized 31P measurements of response to treatment. NMR Biomed 11:314–340 138. Levine SR, Helpern JA, Welch KM, Vande Linde AM, Sawaya KL, Brown EE, Ramadan NM, Deveshwar RK, Ordidge RJ (1992) Human focal cerebral ischemia: evaluation of brain pH and energy metabolism with P-31 NMR spectroscopy. Radiology. 185:537-544 139. Li C-W, Gonen O (1996) Simultaneous 3D NMR spectroscopy of fluorine and phosphorus in human liver during 5fluorouracil chemotherapy. Magn Reson Med 35:841–847 140. Li C-W, Negendank WG, Padavic-Shaller KA, O’Dwyer PJ, Murphy-Boesch J, Brown TR (1996) Quantitation of 5fluorouracil catabolism in human liver in vivo by three-dimensional localized 19F magnetic resonance spectroscopy. Clin Cancer Res 2:339–345 141. Lin A, Bluml S, Mamelak N (1999) Efficacy of proton magnetic resonance spectroscopy in clinical decision making for patients with suspicious malignant brain tumors. J Neurooncol 45:69–81 142. Lin A, Ross BD, Harris K, Wong W (2005) Efficacy of proton magnetic resonance spectroscopy in neurological diagnosis and neurotherapeutic decision making. NeuroRx 2:197–214 143. Löffler R, Sauter R, Kolem H, Haase A, von Kienlin M (1998) Localized spectroscopy from anatomically matched compartments: improved sensitivity and localization for cardiac 31P MRS in humans. J Magn Reson 134:287–299 144. Lowry OH, Berger SJ, Chi MMY, Blackshaw A, Outlaw W (1977) Diversity of metabolic patterns in human brain tumors. I. High energy phosphate compounds and basic composition. J Neurochem 29:959–977 145. Luyten PR, Bruntink G, Sloff FM, Vermeulen JWAH, van der Heijden JI, den Hollander JA, Heershap A (1989) Broadband proton decoupling in human 31P NMR spectroscopy. NMR Biomed 1:177–183 146. MacKay S, Ezekiel F, Di Sclafani V, Meyerhoff DJ, Gerson J, Norman D, Fein G, Weiner MW (1996) Alzheimer disease and subcortical ischemic vascular dementia: evaluation by combining MR imaging segmentation and H-1 MR spectroscopic imaging. Radiology 198:537-545 147. Mangia S, Tkac I, Gruetter R, Van De Moortele PF, Giove F, Maraviglia B, Ugurbil K (2006) Sensitivity of singlevoxel 1H MRS in investigating the metabolism of the activated human visual cortex at 7 T. Magn Reson Imaging. 24:343–348

15.4  Clinical Applications in Selected Organs and Tumors 148. Marshall I, Wardlaw J, Cannon J, Slattery J, Sellar RJ (1996) Reproducibility of metabolite peak areas in 1H MRS of brain. Magn Reson Imaging 14:281–292 149. Maudsley AA, Hilal SK, Perman WH, Simon HE (1983) Spatially resolved high-resolution spectroscopy by “fourdimensional” NMR. J Magn Reson 51:147–152 150. McKnight TR, Noworolski SM, Vigneron DB, Nelson SJ (2001) An automated technique for the quantitative assessment of 3D MRSI data from patients with glioma. J Magn Reson Imaging 13:167–177 151. Menard C, Smith ICP, Somorjai RL, Leboldus L, Patel R, Littman C, Robertson SJ, Bezabeh T (2001) Magnetic resonance spectroscopy of the malignant prostate gland after radiotherapy: a histopathologic study of diagnostic validity. Int J Radiat Oncol Biol Phys 50:317–323 152. Meyer RA, Fisher MJ, Nelson SJ, Brown TR (1988) Evaluation of manual methods for integration of in vivo phosphorus NMR spectra. NMR Biomed 1:131–135 153. Meyerhoff DJ, Boska MD, Thomas AM, Weiner MW (1989) Alcoholic liver disease: quantitative image-guided P-31 MR spectroscopy. Radiology 173:393–400 154. Meyerhoff DJ, Karczmar GS, Valone F, Venook A, Matson GB, Weiner MW (1992) Hepatic cancers and their response to chemoembolization therapy. Quantitative image-guided 31P magnetic resonance spectroscopy. Invest Radiol 27:456–464 155. Meyerhoff DJ, MacKay S, Bachman L, Poole N, Dillon W, Fein CM, Weiner M (1993) Reduced brain N-acetylaspartate suggests neuronal loss in cognitively impaired human immunodeficiency virus-seropositive individuals: in vivo 1H magnetic resonance spectroscopic imaging. Neurology 43:509–515 156. Michaelis T, Merboldt KD, Hänicke W, Gyngell ML, Bruhn H, Frahm J (1991) On the identification of cerebral metabolites in localized 1H NMR spectra of human brain in vivo. NMR Biomed 4:90–98 157. Michaelis T, Merboldt KD, Bruhn H, Hänicke W, Math, Frahm J (1993) Absolute concentrations of metabolites in the adult human brain in vivo: quantification of localized proton MR spectra. Radiology 187:219–227 158. Mierisová S, Ala-Korpela M (2001) MR spectroscopy quantitation: a review of frequency domain methods. NMR Biomed 14:247–259 159. Mierisová S, van de Boogaart A, Tkac I, van Hecke P, Vanhamme L, Liptaj T (1998) New approach for quantification of short echo time in vivo 1H MR spectra of brain using AMARES. NMR Biomed 11:32–39 160. Moats RA, Ernst T, Shonk TK, Ross BD (1994) Abnormal cerebral metabolite concentrations in patients with probable Alzheimer disease. Magn Reson Med 32:110–115 161. Moller HE, Gaupp A, Vestring T, Dietl KH, Vermathen P, Buchholz B, Peters PE (1995) Kinetics of metabolism in human kidney transplants measured by dynamic 31P NMR spectroscopy. Z Naturforsch (C) 50:439–450 162. Moon RB, Richards JH (1973) Determination of intracellular pH by 31P magnetic resonance. J Biol Chem 248:7276–7278

163. Naegele T, Grodd W, Viebahn R, Seeger U, Klose U, Seitz D, Kaiser S, Mader I, Mayer J, Lauchart W, Gregor M, Voigt K (2000) MR imaging and 1H spectroscopy of brain metabolites in hepatic encephalopathy: time-course of renormalization after liver transplantation. Radiology 216:683–691 164. Nakae I, Mitsunami K, Matsuo S, Matsumoto T, Morikawa S, Inubushi T, Koh T, Horie M. (2004) Assessment of myocardial creatine concentration in dysfunctional human heart by proton magnetic resonance spectroscopy. Magn Reson Med Sci 3:19–25 165. Narayana PA, Hazle JD, Jackson EF, Fotedar LK, Kulkarni MY (1988) In vivo 1H spectroscopy studies of human gastrocnemius muscle at 1.5 T. Magn Res Imaging 6:481–485 166. Narayana PA, Jajodia P, Kurhanewicz J, Thomas A, MacDonald J, Hubesch B, Hedgcock M, Anderson CM, James TL, Tanagho EA, et al. (1991) Characterization of prostate cancer, benign prostatic hyperplasia and normal prostates using transrectal 31phosphorus magnetic resonance spectroscopy: a preliminary report. J Urol. 146:66-74 167. Naruse S, Horikawa Y, Tanaka C, Higuchi T, Ueda S, Hirakawa K, Nishikawa H, Watari H (1985) Observations of energy metabolism in neuroectodermal tumors using in vivo 31P-NMR. Magn Reson Imaging. 3:117-123 168. Naruse S, Hirakawa K, Horikawa Y, Tanaka C, Higuchi T, Ueda S, Nishikawa H, Watari H (1985) Measurements of in vivo 31P nuclear magnetic resonance spectra in neuroectodermal tumors for the evaluation of the effects of chemotherapy. Cancer Res. 45:2429-2433 169. Negendank W (1992) Studies of human tumors by MRS: a review. NMR Biomed 5:303–324 170. Nelson SJ, Huhn S, Vigneron DB, Day MR, Wald LL, Prados M, Chang S, Gutin PH, Sneed PK, Verhey L, Hawkins RA, Dillon WP (1997) Volume MRI and MRSI techniques for the quantitation of treatment response in brain tumors: presentation of a detailed case study. J Magn Reson Imaging. 7:1146-1152 171. Nelson SJ, Vigneron DB, Dillon WP (1999) Serial evaluation of patients with brain tumors using volume MRI and 3D 1H MRSI. NMR Biomed 12:123–138 172. Ng TC, Evanochko WT, Hiramoto RN Ghanta VK, Lilly MB, Lawson AJ, Corbett TH, Durant JR, Glickson JD (1982) 31P NMR spectroscopy of in vivo tumors. J Magn Reson 49:271–286 173. Ng TC, Vijayakumar S, Majors AW, Thomas FJ, Meaney TF, Baldwin NJ (1987) Response of a non-Hodgkin’s lymphoma to 60Co therapy monitored by 31P MRS in situ. Int J Rad Oncol Biol Phys 13:1545–1551 174. Ng TC, Majors AW, Vijayakumar S, Baldwin NJ, Thomas FJ, Koumoundouros I, Taylor ME, Grundfest SF, Meaney TF, Tubbs RR, et al. (1989) Human neoplasm pH and response to radiation therapy: P-31 MR spectroscopy studies in situ. Radiology. 170:875-878 175. Oberhänsli RD, Hilton-Jones D, Bore PJ, Hands LJ, Rampling RP, Radda GK (1986) Biochemical investigation of human tumours in vivo with phosphorus-31 magnetic resonance spectroscopy. Lancet 2:8–11

1375

1376

15  Clinical Spectroscopy 176. Opstad KS, Ladroue C, Bell BA, Griffiths JR, Howe FA (2007) Linear discriminant analysis of brain tumour 1H MR spectra: a comparison of classification using whole spectra versus metabolite quantification. NMR Biomed (in press). doi:10.1002/nbm.1147 177. Ott D, Hennig J, Ernst T (1993) Human brain tumors: assessment with in vivo proton MR spectroscopy. Radiology 186:745–752 178. Pavlakis SG, Kingsley PB, Harper R, Buckwald S, Spinazzola R, Frank Y, Prohovnik I (1999) Correlation of basal ganglia magnetic resonance spectroscopy with Apgar score in perinatal asphyxia. Arch Neurol 56:1476–1481 179. Payne GS, Leach MO (2006) Applications of magnetic resonance spectroscopy in radiotherapy treatment planning. Br J Radiol 79(Spec no. 1):S16–26 180. Pels P, Ozturk-Isik E, Swanson MG, Vanhamme L, Kurhanewicz J, Nelson SJ, Van Huffel S (2006) Quantification of prostate MRSI data by model-based time domain fitting and frequency domain analysis. NMR Biomed. 19:188–197 181. Port RE, Bachert P, Semmler W (1991) Kinetic modeling of in vivo-nuclear magnetic resonance spectroscopy data: 5fluorouracil in liver and liver tumors. Clin Pharmacol Ther 49:497–505 182. Presant CA, Wolf W, Waluch V, Wiseman C, Kennedy P, Blayney D, Brechner RR (1994) Association of intratumoral pharmacokinetics of fluorouracil with clinical response. Lancet 343:1184–1187 183. Presant CA, Wolf W, Waluch V, Wiseman CL, Weitz I, Shani J (2000) Enhancement of fluorouracil uptake in human colorectal and gastric cancers by interferon or by high-dose methotrexate: An in vivo human study using noninvasive 19F-magnetic resonance spectroscopy. J Clin Oncol 18:255–261 184. Prescott DM, Charles HC, Sostman HD, Dodge RK, Thrall DE, Page RL, Tucker JA, Harrelson JM, Leopold KA, Oleson JR, et al. (1994) Therapy monitoring in human and canine soft tissue sarcomas using magnetic resonance imaging and spectroscopy. Int J Radiat Oncol Biol Phys. 28:415-423 185. Pretlow TG, Harris BF, Bradley E, Bueschen AJ, Lloyd KL, Pretlow TP (1985) Enzyme activities ion prostatic carcinoma related to Gleason grades. Cancer Res 45:442–446 186. Preul MC, Caramanos Z, Leblanc R, Villemure JG, Arnold DL (1998) Using pattern analysis of in vivo proton MRSI data to improve the diagnosis and surgical management of patients with brain tumors. NMR Biomed 11:192–200 187. Provencer SW (2001) Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed 14:260–264 188. Provencer SW (1993) Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med 30:672–679 189. Radda GK, Rajagopalan B, Taylor DJ (1989) Biochemistry in vivo: an appraisal of clinical magnetic resonance spectroscopy. Magn Res Quart 5:122–151

190. Redmond OM, Bell E, Stack IP, Dervan PA, Carney DN, Hurson BJ, Ennis JT (1992) Tissue characterization and assessment of preoperative chemotherapeutic response in musculoskeletal tumors by in vivo 31P magnetic resonance spectroscopy. Magn Reson Med 27:226–237 191. Redmond OM, Stack JP, O’Connor NG, Carney DN, Dervan PA, Hurson BJ, Ennis JT (1992) 31P MRS as an early prognostic indicator of patient response to chemotherapy. Magn Reson Med 25:30–44 192. Renshaw PF, Haselgrove JC, Leigh JS, Chance B (1985) In vivo NMRI of lithium. Magn Reson Med 2:512–516 193. Renshaw PF, Wicklund S (1988) In vivo measurement of lithium in humans by NMRS. Biol Psychiatry 23:465–475 194. Rexroth W, Semmler W, Gückel F, Stadtlander M, Weikker H, Hild R, van Kaick G (1989) Assessment of muscular metabolism in peripheral arterial occlusive disease using 31P nuclear magnetic resonance. Klin Wochenschr 67:804–812 195. Roden M, Petersen KF, Shulman GI (2001) Nuclear magnetic resonance studies of hepatic glucose metabolism in humans. Recent Prog Horm Res 56:219–37 196. Roelants-Van Rijn AM, van der Grond J, de Vries LS, Groenendaal F (2001) Value of 1H MRS using different echo times in neonates with cerebral hypoxia-ischemia. Pediatr Res 49:356–62 197. Ross BD, Radda GK, Gadian GD, Rocker G, Esiri M, Falconer-Smith J (1981) Examination of a case of suspected McArdle’s syndrome by 31P nuclear magnetic resonance. N Engl J Med 304:1338–1342 198. Ross BD (1991) Biochemical consideration in 1H-spectroscopy. Glutamate and glutamine; myoinositol and related metabolites. NMR Biomed 4:59–63 199. Ross BD, Jacobson S, Villamil F, Korula J, Kreis R, Ernst T, Shonk T, Moats R (1994) Subclinical hepatic encephalopathy: proton MR spectroscopic abnormalities. Radiology 193:457–463 200. Ross BD, Michaelis T (1994) Clinical applications of magnetic resonance spectroscopy. Magn Reson Q 10:191–247 201. Ross BD, Danielsen ER, Bluml S (1996) Proton magnetic resonance spectroscopy: the new gold standard for diagnosis of clinical and subclinical hepatic encephalopathy? Dig Dis 14(Suppl 1):30–39 202. Ross BD, Bluml S, Cowan R, Danielsen E, Farrow N, Gruetter R (1997) In vivo magnetic resonance spectroscopy of human brain: the biophysical basis of dementia. Biophys Chem 68:161–172 203. Ross BD, Bluml S, Cowan R, Danielsen E, Farrow N, Tan J (1998) In vivo MR spectroscopy of human dementia. Neuroimaging Clin N Am 8:809–822 204. Ross B, Bluml S (2001) Magnetic resonance spectroscopy of the human brain. Anat Rec 265:54–84 205. Ross B, Lin A, Harris K, Bhattacharya P, Schweinsburg B (2003) Clinical experience with 13C MRS in vivo. NMR Biomed 16:358–369

15.4  Clinical Applications in Selected Organs and Tumors 206. Sala E, Noyszewski EA, Campistol JM, Marrades RM, Dreha S, Torregrossa JV, Beers JS, Wagner PD, Roca J (2001) Impaired muscle oxygen transfer in patients with chronic renal failure. Am J Physiol 280:R1240–R1248 207. Sappey-Marinier D, Hubesch B, Matson GB, Weiner MW (1992) Decreased phosphorus metabolite concentrations and alkalosis in chronic cerebral infarction. Radiology 182:29–34 208. Saunders DE (2000) MR spectroscopy in stroke. Br Med Bull 56:334–345 209. Saunders DE, Howe FA, van den Boogaart A, Griffiths JR, Brown MM (1999) Aging of the adult human brain: in vivo quantitation of metabolite content with proton magnetic resonance spectroscopy. J Magn Reson Imaging 9:711–716 210. Sauter R, Müller S, Weber H (1987) Localization in in vivo 31P NMR spectroscopy by combining surface coils and slice-selective saturation. J Magn Reson 75:167–173 211. Scheidler J, Hricak H, Vigneron DB, Yu KK, Sokolov DL, Huang LR, Zaloudek CJ, Nelson SJ, Carroll PR, Kurhanowicz J (1999) Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study. Radiology 213:473–480 212. Schick F, Duda S, Durk H, Bunse M, Lutz O, Claussen CD (1994) Eosinophilia-myalgia syndrome: findings at MR imaging and proton spectroscopy of the lower leg. Magn Reson Imaging 12:513–522 213. Schick F (1996) Bone marrow NMR in vivo. J Progr NMR Spectroscopy 29:169–227 214. Schiebler ML, Miyamoto KK, White M, Maygarden SJ, Mohler JL (1993) In vitro high resolution 1H-spectroscopy of the human prostate: benign prostatic hyperplasia, normal peripheral zone and adenocarcinoma. Magn Reson Med 29:285–291 215. Schilling A, Gewiese B, Stiller D, Römer T, Wolf KJ (1990) Einfluss der Ernährungslage auf das 31P MR Spektrum der gesunden Leber. RoFo 153:369–372 216. Schlemmer HP, Bachert P, Semmler W, Hohenberger P, Schlag P, Lorenz WJ, van Kaick G (1994) Drug monitoring of 5-fluorouracil: In vivo 19F NMR study during 5-FU chemotherapy in patients with metastases of colorectal adenocarcinoma. Magn Reson Imaging 12:497–511 217. Schlemmer HP, Becker M, Bachert P, Dietz A, Rudat V, Vanselow B, Wollensack P, Zuna I, Knopp MV, Weidauer H, Wannenmacher M, van Kaick G (1999) Alterations of intratumoral pharmacokinetics of 5-fluorouracil in head and neck carcinoma during simultaneous radiochemotherapy. Cancer Res 59:2363–2369 218. Schlemmer HP, Bachert P, Herfarth KK, Zuna I, Debus J, van Kaick G (2001) Evaluation of suspicious brain lesions after stereotactic radiotherapy of brain tumors using proton MR spectroscopy. AJNR Am J Neuroradiol 2001, 22:1316–1324

219. Schlemmer HP, Sawatzki T, Sammet S, Dornacher I, Bachert P, van Kaick G, Waldherr R, Seitz HK (2005) Hepatic phospholipids in alcoholic liver disease assessed by protondecoupled 31P magnetic resonance spectroscopy. J Hepatol. 42:752–759 220. Segebarth CM, Grivegnée AR, Luyten PR, den Hollander AJ (1989) Quantitative 31P NMR spectroscopy of phosphate metabolism in the human liver following fructose administration. In: Proceedings 8th Annual Meeting SMRM, Amsterdam, vol. 1, p 263 221. Semmler W, Gademann G, Schlag P, Bachert-Baumann P, Zabel H-J, Lorenz WJ, van Kaick G (1988a) Impact of hyperthermic regional perfusion therapy on cell metabolism of malignant melanoma monitored by 31P MR spectroscopy. Magn Reson Imaging 6:335–340 222. Semmler W, Gademann G, Bachert-Baumann P, Zabel H-J, Lorenz WJ, van Kaick G (1988b) Monitoring human tumor response to therapy by means of P-31 MR. spectroscopy. Radiology 166:533–539 223. Semmler W, Bachert-Baumann P, Gückel F, Ermark F, Schlag P, Lorenz WJ, van Kaick G (1990) Realtime follow-up of 5-fluorouracil metabolism in the liver of tumor patients by means of F-19 MR spectroscopy. Radiology 174:141–145 224. Shimizu H, Kumabe T, Shirane R, Yoshimoto T (2000) Correlation between choline level measured by proton MR spectroscopy and Ki-67 labeling index in gliomas. AJNR Am J Neuroradiol 21:659–665 225. Shine N, Palladino MA, Patton JS, Deisseroth A, Karczmar GS, Weiner MW (1989) Early metabolic response to TNF in mouse sarcoma. A 31P nuclear magnetic resonance study. Cancer Res 49:2123–2127 226. Shonk TK, Moats RA, Gifford P, Michaelis T, Mandigo JC, Izumi J, Ross BD (1995) Probable Alzheimer disease diagnosis with proton MR spectroscopy. Radiology 195:65–72 227. Shulman RG, Rothman DL (2001a) The “glycogen shunt” in exercising muscle: a role for glycogen in muscle energetics and fatigue. Proc Natl Acad Sci U S A 98:457–461 228. Shulman RG, Rothman DL (2001b) 13C NMR of intermediary metabolism: implications for systemic physiology. Annu Rev Physiol 63:15–48 229. Sibtain NA, Howe FA, Saunders DE (2007) The clinical value of proton magnetic resonance spectroscopy in adult brain tumours. Clin Radiol 62:109–119 230. Sijens PE, Wijrdeman HK, Moerland MA, Bakker CJ, Vermeulen JW, Luyten PR (1988) Human breast cancer in vivo: H-1 and P-31 MR spectroscopy at 1.5 T. Radiology 169:615–620 231. Sijens PE, Levendag PC, Vecht CJ, van Dijk P, Oudkerk M (1996) 1H MR spectroscopy detection of lipids and lactate in metastatic brain tumors. NMR Biomed 9:65–71 232. Slotboom J, Boesch C, Kreis R (1998) Versatile frequency domain fitting using time domain models and prior knowledge. Magn Reson Med 39:899–911

1377

1378

15  Clinical Spectroscopy 233. Smith TA, Glaholm J, Leach MO, Machin L, McCready VR (1991) The effect of intra-tumour heterogeneity on the distribution of phosphorus-containing metabolites within human breast tumours: an in vitro study using 31P NMR spectroscopy. NMR Biomed 4:262–267 234. Soares JC, Boada F, Spencer S, Mallinger AG, Dippold CS, Wells KF, Frank E, Keshavan MS, Gershon S, Kupfer DJ (2001) Brain lithium concentrations in bipolar disorder patients: preliminary 7Li magnetic resonance studies at 3 T. Biol Psychiatry 49:437–43 235. Soher BJ, Young K, Govindaraju V, Maudsley AA (1998) Automated spectral analysis III: Application to in vivo proton MR spectroscopy and spectroscopic imaging. Magn Reson Med 40:822–831 236. Sostman HD, Charles HC, Rockwell S, Leopold K, Beam C, Madwed D, Dewhirst M, Cofer G, Moore D, Burn R, et al. (1990) Soft-tissue sarcomas: detection of metabolic heterogeneity with P-31 MR spectroscopy. Radiology. 176:837-843 237. Sostman HD, Prescott DM, Dewhirst MW, Dodge RK, Thrall DE, Page RL, Tucker JA, Harrelson JM, Reece G, Leopold KA, et al. (1994) MR imaging and spectroscopy for prognostic evaluation in soft-tissue sarcomas. Radiology 190:269-275 238. Stanley JA, Drost DJ, Williamson PC, Thompson RT (1995) The use of a priori knowledge to quantify short echo in vivo 1H MR spectra. Magn Reson Med 34:17–24 239. Star-Lack JM, Adalsteinsson E, Gold GE, Ikeda DM, Spielman DM (2000) Motion correction and lipid suppression for 1H magnetic resonance spectroscopy. Magn Reson Med 43:325–30 240. Steudel A, Harder T, Träber F, Gieseke J (1990) 31P- Spektroskopie der Leber unter Polycheniotherapie von Hodentumoren. In: Lissner J, Doppman JL, Margulis AR (eds) MR ‘89. Deutscher Ärzte, Cologne, pp 261–264 241. Swanson MG, Vigneron DB, Tran TK, Kurhanewicz J (2001) Magnetic resonance imaging and spectroscopic imaging of prostate cancer. Cancer Invest 19:510–523 242. Swanson MG, Zektzer AS, Tabatabai ZL, Simko J, Jarso S, Keshari KR, Schmitt L,Carroll PR, Shinohara K, Vigneron DB, Kurhanewicz J (2006) Quantitative analysis of prostate metabolites using 1H HR-MAS spectroscopy. Magn Reson Med 55:1257–1264 243. Tallan HH (1957) Studies on the distribution of N-acetyl-laspartic acid in the brain. J Biol Chem 224:41–45 244. Tallan HH, Moore S, Stein WH (1956) N-Acetyl-l-aspartic acid in the brain. J Biol Chem 224:41–45 245. Tamiya T, Kinoshita K, Ono Y, Matsumoto K, Furuta T, Ohmoto T (2000) Proton magnetic resonance spectroscopy reflects cellular proliferative activity in astrocytomas. Neuroradiology 42:333–338

246. Tate AR, Underwood J, Acosta DM, Julia-Sape M, Majos C, Moreno-Torres A, Howe FA, van der Graaf M, Lefournier V, Murphy MM, Loosemore A, Ladroue C, Wesseling P, Luc Bosson J, Cabanas ME, Simonetti AW, Gajewicz W, Calvar J,Capdevila A, Wilkins PR, Bell BA, Remy C, Heerschap A, Watson D, Griffiths JR, Arus C (2006) Development of a decision support system for diagnosis and grading of brain tumours using in vivo magnetic resonance single voxel spectra. NMR Biomed 19:411–434 247. Taylor DJ (2000) Clinical utility of muscle MR spectroscopy. Semin Musculoskelet Radiol 4:481–502 248. Taylor DJ, Kemp GJ, Radda GK (1994) Bioenergetics of skeletal muscle in mitochondrial myopathy. J Neurol Sci 127:198–206 249. Taylor-Robinson SD, Sargentoni J, Bell JD, Saeed N, Changani KK, Davidson BR, Rolles K, Burroughs AK, Hodgson HJ, Foster CS, Cox IJ (1997) In vivo and in vitro hepatic 31P magnetic resonance spectroscopy and electron microscopy of the cirrhotic liver. Liver 17:198–209 250. Taylor-Robinson SD, Sargentoni J, Bell JD, Thomas EL, Marcus CD, Changani KK, Saeed N, Hodgson HJ, Davidson BR, Burroughs AK, Rolles K, Foster CS, Cox IJ (1998) In vivo and in vitro hepatic phosphorus-31 magnetic resonance spectroscopy and electron microscopy in chronic ductopenic rejection of human liver allografts. Gut 42:735–43 251. Thomas MA, Binesh N, Yue K, CeBruhl N (2001) Volumelocalized two-dimensional correlated magnetic resonance spectroscopy of human breast cancer. J Magn Reson Imaging 14:181–186 252. Thompson CH, Irish AB, Kemp GJ, Taylor DJ, Radda GK (1997) The effect of propionyl l-carnitine on skeletal muscle metabolism in renal failure. Clin Nephrol 47:372–378 253. Tien RD, Lai PH, Smith JS, Lazeyras F (1996) Single-voxel proton brain spectroscopy exam (PROBE/SV) in patients with primary brain tumors. AJR Am J Roentgenol 167:201–209 254. Trump ME, Hanstock CC, Allen PS, Gheorghiu D, Hochachka PW (2001) An 1H MRS evaluation of the phosphocreatine/creatine pool (tCr) in human muscle. Am J Physiol 280:R889–R896 255. Tunggal B, Hofmann K, Stoffel W (1990) In vivo 13C nuclear magnetic resonance investigations of choline metabolism in rabbit brain. Magn Reson Med 13:90–102 256. Tyszka JM, Silverman JM (1998) Navigated single-voxel proton spectroscopy of the human liver. Magn Reson Med 39:1–5 257. Vallee JP, Lazeyras F, Sostman HD, Smith SR, Butterly DW, Spritzer CE, Charles HC (1996) Proton-decoupled phosphorus-31 magnetic resonance spectroscopy in the evaluation of native and well-functioning transplanted kidneys. Acad Radiol 3:1030–1037

15.4  Clinical Applications in Selected Organs and Tumors 258. van der Graaf M, van den Boogert HJ, Jager GJ, Barentsz JO, Heerschap A (1999) Human prostate: multisectional proton MR spectroscopic imaging with single spin-echo sequence—preliminary experiences. Radiology 213:919–925 259. van der Grond J van der, Laven JS, te Velde ER, Mali WP (1991) Abnormal testicular function: potential of P-31 MR spectroscopy in diagnosis. Radiology 179:433–436 260. van der Grond J van der (1995) Diagnosing testicular function using 31P magnetic resonance spectroscopy: a current review. Hum Reprod Update 1:276–83 261. Vanhamme L, Sundin T, Hecke PV, Huffel SV (2001) MR spectroscopy quantitation: a review of time-domain methods. NMR Biomed 14:233–246 262. Van Rijen, Luyten PR, Berkelbach van der Sprenkel JW, Kraaier V, van Huffelen AC, Tulleken CAF, den Hollander JA (1989) 1H and 31P NMR measurement of cerebral lactate, high-energy phosphate levels, and pH in humans during voluntary hyperventilation: associated EEG, capnographic, and Doppler findings. Magn Res Med 10:182–193 263. Vaupel P, Kallinowski F, Okunieff P (1989) Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review. Cancer Res 49:6449–6465 264. Veech RL (1991) The metabolism of lactate. NMR Biomed 4:53–58 265. Weiss RG, Bottomley PA, Christopher JH, Gerstenblith G (1990) Regional myocardial metabolism of high-energy phosphates during isometric exercise in patients with coronary artery disease. N Engl J Med 323:1593–1600 266. Willmann O, Wennberg R, May T, Woermann FG, Pohlmann-Eden B (2006) The role of 1H magnetic resonance spectroscopy in pre-operative evaluation for epilepsy surgery. A meta-analysis. Epilepsy Res 71:149–58 267. Wolf W, Albright MJ, Silver MS, Weber H, Reichardt U, Sauter R (1987) Fluorine-19 NMR spectroscopic studies of the metabolism of 5-fluorouracil in the liver of patients undergoing chemotherapy. Magn Reson Imaging 5:165–169 268. Wolf W, Presant CA, Waluch V (1992) 19F MRS of 5-fluorouracil in human tumors: clinical correlations of trapping with therapeutic response. In: Proceedings of the 11th Annual Meeting SMRM, Berlin, vol. 1, p 58

269. Wolf W, Waluch V, Presant CA (1998) Non-invasive 19FNMRS of 5-fluorouracil in pharmacokinetics and pharmacodynamic studies. NMR Biomed 11:380–387 270. Wolf W, Presant CA, Waluch V (2000) 19F MRS studies of fluorinated drugs in humans. Adv Drug Deliv Rev 41:55–74 271. Yabe T, Mitsunami K, Inubushi T, Kinoshita M (1995) Quantitative measurements of cardiac phosphorus metabolites in coronary artery disease by 31P magnetic resonance spectroscopy. Circulation 92:15–23 272. Yeung DK, Cheung HS, Tse GM (2001) Human breast lesions: characterization with contrast-enhanced in vivo proton MR spectroscopy—initial results. Radiology 220:40–46 273. Yu KK, Scheidler J, Hricak H, Vigneron DB, Zaloudek CJ, Males RG, Nelson SJ, Carroll PR, Kurhanowicz J (1999) Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic. Radiology 213:481–488 274. Zandt HJA in ’t, van der Graaf M, Heerschap A (2001) Common processing of in vivo MR spectra. NMR Biomed 14:224–232 275. Zlatkin MB, Lenkinski RE, Shinkwin M, Schmidt RG, Daly JM, Holland GA, Frank T, Kressel HY (1990) Combined MR imaging and spectroscopy of bone and soft tissue tumors. J Comput Assist Tomogr. 14:1-10

Suggested Reading 1. 2. 3. 4. 5.

de Graaf Robin A (1999) In vivo NMR Spectroscopy: principles and techniques. Wiley, New York Keeler J (2005) Understanding NMR spectroscopy. Wiley, New York Kimmich, R (1997) NMR tomography, diffusometry, relaxometry. Springer, Berlin Heidelberg New York Liney G (2005) MRI from A to Z. Cambridge University Press, Cambridge Slichter CP (1990) Principles of magnetic resonance. Springer, Berlin Heidelberg New York

1379

Chapter 16

Molecular Imaging

16.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1381 W. Semmler

16.2 

Molecular Imaging Modalities  .. . . . . . .   1384

16.2.1 

Optical Imaging  .. . . . . . . . . . . . . . . . . . . . .   1384

16.2.1.1  Bioluminescence Imaging  .. . . . . . . . . . . .   1385 16.2.1.2  Fluorescence Reflectance/ Transillumination Imaging   . . . . . . . . . . .   1385 16.2.1.3  Optical Tomography   .. . . . . . . . . . . . . . . .   1385 16.2.2 

Computed Tomography  . . . . . . . . . . . . . .   1386

16.2.2.1  Microcomputed Tomography  . . . . . . . . .   1387 16.2.2.2  Flat-Panel Volumetric CT   . . . . . . . . . . . .   1387 16.2.3 

Positron Emission Tomography and Single-Photon Emission Tomography     1387

16.2.4 

Magnetic Resonance Imaging and Spectroscopy  . . . . . . . . . . . . . . . . . . . .   1389

16.2.5 

Ultrasound  . . . . . . . . . . . . . . . . . . . . . . . . . .   1389

16.2.6 

Multimodal Imaging Approaches: PET–CT and PET–MRI  .. . . . . . . . . . . . . .   1390

16.2.6.1  PET–CT  .. . . . . . . . . . . . . . . . . . . . . . . . . . . .   1390 16.2.6.2  PET–MRI  . . . . . . . . . . . . . . . . . . . . . . . . . . .   1390

16.1  Introduction W. Semmler Molecular medicine will play a major role in the near future for prevention, diagnosis, and therapy and will finally lead to a paradigm shift in healthcare. In particular molecular diagnosis will play a central role: In vitro molecular diagnosis will be decisive for early detection and molecular imaging (MI) will contribute to localize lesions and will further characterize the disease in vivo (Fig. 16.1.1). This approach may help to achieve the ultimate goal of an optimized and individualized therapy for each patient.

16

16.3 

Molecular Imaging Probes  . . . . . . . . . . .   1390

16.3.1 

Signal Generators  . . . . . . . . . . . . . . . . . . . .   1392

16.3.1.1  Radioactive Markers  . . . . . . . . . . . . . . . . .   1392 16.3.1.2  Optical Signal Generators  .. . . . . . . . . . . .   1392 16.3.1.3  Ultrasound Signal Generators  .. . . . . . . .   1393 16.3.1.4  Magnetic Imaging Signal Generators  .. .   1394 16.3.2 

Molecular Targeting for Imaging  . . . . . .   1397

16.4 

Applications  .. . . . . . . . . . . . . . . . . . . . . . . .   1399

16.4.1 

Antibodies  . . . . . . . . . . . . . . . . . . . . . . . . . .   1399

16.4.2 

Protein Amplification  . . . . . . . . . . . . . . . .   1399

16.4.3 

Reporters  .. . . . . . . . . . . . . . . . . . . . . . . . . . .   1400

16.4.4 

Smart Probes  . . . . . . . . . . . . . . . . . . . . . . . .   1403

16.4.5 

Cell Tracking  . . . . . . . . . . . . . . . . . . . . . . . .   1403

16.4.5.1  Radioactive Labeling  . . . . . . . . . . . . . . . . .   1404 16.4.5.2  Optical Labeling  . . . . . . . . . . . . . . . . . . . . .   1404 16.4.5.3  Magnetic Labeling  . . . . . . . . . . . . . . . . . . .   1405 16.4.6 

Imaging Using Hyperpolarization  . . . . .   1407 References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1407

Today’s diagnostic imaging approaches are based on displaying anatomy and morphology; furthermore, functional parameters are used increasingly to characterize the diseases and/or lesions. Because most patients show up relatively late with symptoms, often diseases have advanced and therefore therapy is more intricate and cure less probable. Molecular imaging should allow early detection on a molecular level, before morphological and/ or clinical changes can be observed and therefore, stratification will change and can be optimized. Monitoring of therapeutical intervention—e.g., targeted molecular therapies, conventional chemotherapeutic interventions, or others—will help to optimize therapies due to the early detection of the response by means of molecular imag-

1382

16  Molecular Imaging Table 16.1.1  Molecular imaging. Imaging methods used to directly or indirectly monitor and record the spatiotemporal distribution of molecular or cellular processes

Fig. 16.1.1  Strategy for molecular medicine (courtesy of Erich R. Reinhardt, modified)

ing. This approach is based on the assumption that disease are caused by alterations on the genetic or molecular level (e.g., modified receptor expression), which in turn is followed by pathophysiological alterations and ultimately morphological changes (e.g., tumor growth) (Fig.16.1.2). Molecular imaging will influence not only the patient care, but is expected to change strategies for drug development, too. At present the term molecular imaging is defined quite differently. The very strict definition confines molecular imaging to the imaging with disease-specific probes labeled with modality-specific signal molecules, particles, or isotopes (e.g., dyes, superparamagnetic particles, 111In, etc.) and in its broadest sense also high-resolution morphologic, functional imaging as well as metabolic imaging is included (e.g., BOLD imaging, flow and diffusion measurement, magnetic resonance spectroscopy, etc.) (Table 16.1.1). Among others, the following definitions are given in the literature: 1 Molecular imaging is the in vivo characterization and measurement of biologic processes on the cellular and molecular level. In contradistinction to “classical” diagnostic imaging, it sets forth to probe such molecular abnormalities that are the basis of disease rather than to image the end results of these molecular alterations (Weissleder and Mahmoud 2001). 2 Molecular imaging implies the convergence of multiple image-capture techniques, basic cell/molecular biology, chemistry, medicine, pharmacology, medical physics, biomathematics, and bioinformatics into a new imaging paradigm. Molecular imaging usually exploits specific molecular probes as the source of image contrast. This change in emphasis from a nonspecific to a specific approach represents a significant paradigm shift, the impact of which is that imaging can now provide the potential for understanding of integrative biology, earlier detection, and characterization of disease, and evaluation of treatment (Massoud and Gambhir 2003). 3 Molecular imaging techniques directly or indirectly monitor and record the spatiotemporal distribution of

• Imaging surrogate markers • Morphology: – High-resolution imaging – … • Function: – Perfusion – Diffusion – Flow – Oxygenation – … • Cell tracking • Metabolic imaging • Spectroscopic imaging (SI) (Chemical-shift imaging [CSI]) • Imaging using molecular probes • Imaging with labeled probes – Radioactive labeling – Labeling with dyes – Coated and functionalized nanoparticles – Ultrasound bubbles – …

molecular or cellular processes for biochemical, biologic, diagnostic, or therapeutic applications (Thakur and Lentle 2005). These three definitions are the most commonly used ones, put emphasis on different aspects, and are relatively ill defined and therefore leave room for interpretation. The last-mentioned definition—established in a consensus meeting of experts—is the definition that should be used in general. Even in the recommended definition there is still room for interpretation: “…indirectly monitor and record … cellular processes” can include also imaging of physiological parameters as diffusion and flow; surrogate marker imaging, as well as assessment of the metabolism by means of magnetic resonance spectroscopy can be considered to be molecular imaging methods. In this definition, even the high-resolution imaging can be included, because the function of a lesions, for instance that of a tumor, can be assessed by high-resolution CT imaging when the vessel density is determined, which establishes indirect information on the perfusion and turns on the nutrition supply of a tumor, which might influencing the metabolism of tumor cells. Adopting this broad definition, the imaging modalities given in Table 16.1.2 can be considered as molecular imaging modalities. Each of the modalities has its own advantages and disadvantages, and none of the modalities is ideal and a so-

16.1  Introduction Table 16.1.2  Imaging modalities that can be used for experimental and/or clinical molecular imaging • X-ray–computed tomography (CT) • Emission tomography (ET) – Positron emission tomography (PET) – Single photon emission computed tomography (SPECT) • Magnetic resonance imaging (MRI) (including functional MR imaging) • Magnetic resonance spectroscopy (MRS) • Ultrasound (US) – A mode B mode – Doppler – SPAQ • Optical imaging (OI)

called all-round method—in other terms, the methods are mostly complementary and do not compete directly. This is in part due to different sensitivities of the methods of consideration. The given physical sensitivity of the method can be used in different ways—either for a high spatial or temporal resolution—at the end it is a tradeoff. The general features concerning the sensitivity of the modalities are given in Fig. 16.1.3. Because in this chapter the molecular imaging using magnetic resonance methods is in the focus and of utmost interest, it should be emphasized that nuclear magnetic resonance has a sensitivity gap of almost five orders of magnitude compared with nuclear methods (e.g., PET). This is an obstacle for the conversion of MR into a molecular imaging modality and needs special solutions. Achieving this goal for MRI requires passive, or better yet, active accumulation of the specific mole­cules into the target cell to increase the concentration of the imaging agent in the region of interest. Surrogate markers as well as reporter agents can be used to visualize enzyme activity or validate gene therapy approaches. In particular for MRI—but also for the other imaging modalities suited for molecular imaging—the aim is to synthesize and characterize macromolecules, peptide conjugates, oligonucleotides, and other conjugates for specific imaging with high accumulation in the target area or to find mechanisms, which allow the monitoring of enzyme activity or the gene expression in tumors, diseased cells, or tissue. The MRI low-sensitivity hurdle can further be overcome when agents with high relaxivity, e.g., targeted iron oxide nanoparticles are used, as was shown by Kresse et al. (1998). In animal experi­ments Moore et al. (1998, 2001) demonstrated that MRI utilizing iron oxide nanoparticles targeted to an engineered human transferrin receptor enables the imaging of gene expression. However, the relatively high doses of iron oxides used in those experi­ments

Fig. 16.1.2  Development of disease (courtesy of Ralf Schulz, Dept. of Med. Phys., DKFZ and GFS, Neuherberge, modified)

indicated the need for improved MR imaging probes. Högemann et al. (2000) synthesized Tf–S–S–cross-linked iron oxide (CLIO) nanopar­ticles, which considerably improved receptor binding and relaxivity to increase MRI detection. However, we have to keep in mind that when using MR spectroscopy (MRS), MR has the capability to directly visualize the metabolism in vivo, which is why we can designate MR a molecular imaging modality per se. As has been shown with brain tumors, prostate carcinomas, and for the diagnoses of other diseases, e.g., Canavan’s disease (Grodd et al. 1990), MRS has been proven to be of real clinical value (cf. Chap 15). Frequently, cell trafficking as well as measurement of functional parameters using MRI are considered molecular imaging methods because surrogate parameters for molecular processes are being evaluated. For instance macrophages can be labeled in vivo or ex vivo, and in cell trafficking experiments inflammation or tumors can be localized. Contrast-enhanced dynamic MRI provides quantification of perfusion and vessel permeability to complete the characterization of normal, pathological, and neoplastic tissue, which allows indirect assessment of angiogenic activity including its regulating factors (cytokines, angiopoietin, vascular endothelial growth factor [VEGF], etc.) and monitoring of antiangiogenic treatments (Kießling et al. 2003, 2004). The general feature of the above-mentioned modalities will be discussed in the first part of this chapter, the application in the second part. As already stated, emphasis will give to MR methods, optical, CT and nuclear methods will be mentioned as reference only. With the exception of positron emission tomography (PET) and single-photon computed tomography (SPECT) and in part MRI and MRS, molecular imaging is still experimental and clinical applications are rare.

1383

1384

16  Molecular Imaging Fig. 16.1.3  Spectrum of molecular imaging modalities (courtesy of Michael Schäfers, Münster, modified)

16.2  Molecular Imaging Modalities 16.2.1  Optical Imaging Optical imaging techniques are considered to be an ideal modality for molecular imaging because they may provide an easy and reliable way of translating in vitro research to in vivo. For instance, optical imaging techniques enable identification of the pathways of disease and the action of novel therapies in living animals. Besides radioactive methods, in particular optical imaging should make possible translational research “from mouse to man.” It is used because of its high sensitivity and the relatively simple equipment required. This is true for microscopic techniques, where the penetration depth not a major issue. However, most important, for optical imaging, numerous fluorescent markers and optical probe systems are available due to the developments in immunohistology and fluorescence microscopy as well as the discovery of bioluminescent and fluorescent proteins, which enabled simplification of gene expression imaging (Massoud and Gambhir 2003). Furthermore, activatable probes allow monitoring enzyme reactions (Bremer et al. 2001a; Ntziachristos et al. 2002b). Fluorescent proteins nowadays can be engineered. It is favorable if the absorbed and emitted light are in the far-red region because in this region is the so-called “optical window” of tissue—here the unspecific absorption by water and other fluorochromes is minimized (Shaner et al. 2005). However, these advantages are in part compensated by drawbacks inherent to optical imaging. First, in general optical imaging is confined to superficial regions due to

heavy scattering of the optical photons in tissue. Second optical imaging it is non-quantitative. Optical imaging is still highly experimental and yet not used in clinical settings. To overcome these hurdles, optical tomography is under development (see below). Furthermore, optical molecular probes may be prone to photobleaching, are sometimes unstable, and toxic in the amount necessary to yield sufficient contrast. Any optical imaging device requires three fundamental system parts: • Light sources • Filters eliminating background signal • Photon detectors acquiring signals According to Weissleder and Ntziachristos (2003) in the area of macroscopic resolution with molecular contrast agents, three different kinds of optical imaging can be distinguished. This classification is not complete and not generally adopted (Table 16.2.1).

Table 16.2.1  Classification of optical imaging methods • Bioluminescence imaging (BLI) • Fluorescence reflectance imaging (FRI) • Optical tomography (OT) – Optical coherence tomography (OCT) – Diffuse optical tomography (DOT) – Fluorescence molecular tomography (FMT) – Thermoacoustic computed tomography (TCT)

16.2  Molecular Imaging Modalities Fig. 16.2.1a,b  Geometries of the a fluorescence reflectance (FMI) and b fluorescence transmission imaging (FTI) (see text). A modification of FTI leads to tomographic representation if the detector and the light source are rotating around the object and appropriate reconstruction algorithms are applied

16.2.1.1  Bioluminescence Imaging Bioluminescence imaging in whole animals is straight forward, because the bioluminescence molecules inside of the object (e.g., cells in cell cultures, cells within an animal) are the photon source (e.g., green fluorescence protein [GFP], luciferase). Necessary for imaging are only a detector (CCD camera), a light tide chamber, and imaging software for display and evaluation of the acquired imaging data. Nowadays such systems are commercially available (e.g., Xenogen®). Bioluminescence imaging has the advantage of high contrast, because almost no background signals (e.g., autofluorescence signals) are generated; however, the intensity of the bioluminescence signal can be often low, especially if photons from deep-lying structures are under observation (exponential weighting). Tracking of single cells is described (Massoud and Gambhir 2003). With this planar imaging method, quantification is difficult or almost impossible. 16.2.1.2  Fluorescence Reflectance/ Transillumination Imaging Fluorescence imaging is a cheap and simple to handle method too. In comparison with bioluminescence imaging, fluorescence imaging requires a light source exciting the endogenous (autofluorescence) and/or exogenous fluorochromes and filters to eliminate the excitation light. Fluorescence reflectance imaging (FRI) and fluorescence transillumination imaging (FRI/FTI) (Zacharakis et al. 2006) have in common the excitation by an external photon beam of appropriate energy (wavelength), the detec-

tion is done also by a CCD camera; however, in FRI the detector is positioned on the same side of the object (reflectance geometry), in FTI on the opposite side of the imaged object (cf. Fig. 16.2.1). The drawback of FRI is the heavily surface weighting, which makes quantification almost impossible. Furthermore, to get rid of the backscattered excitation light, filtering is necessary because it has almost the same wavelength as the emitted light of the fluorochromes and differs only by the Stokes shift. Excitation light is the major part of the background and has to be reduced as much as possible. FTI is less surface weighted than is FRI because the excitation light is attenuated exponentially in the object while the fluorescence light increases. 16.2.1.3  Optical Tomography 16.2.1.3.1  Fluorescence Molecular Tomography As in FRI/FMT, in fluorescence molecular tomography (FMT) the light directed to the object excites the fluorochromes, yet the emitted light is measured in multiple projections, and the projection images are reconstructed tomographically to obtain the distribution of fluorochromes in tissues. This concept was realized and developed by Ntziachristos et al. (2002) as schematically displayed in Fig. 16.2.2. Absolute quantification of the fluorochromes should be possible also in vivo. In phantoms the quantification error is within 10% of the expected fluorochrome concentration. The spatial resolution is 1 mm near the surface and approximately 2.8 mm in the center of the phantom.

1385

1386

16  Molecular Imaging

Fig. 16.2.2a–j  Schematic of the FMT imager used for the experimental measurements(modified). The imager consisted of a 675 nm laser diode (a) and a beam splitter (b) that divided light to a reference channel (c) and an optical switch (d). 24 source fibers (e) and 36 fiber bundles (g) were employed to il-

luminate and collect light respectively from the optical imaging bore (f). The fiber bundles and reference fiber were arranged on a grid (h) and imaged with a CCD camera (j) using appropriate filters (i). (Ntziachristos et al. (2002) Nat. Med. 8:757, modified)

16.2.1.3.2  Diffuse Optical Tomography

the penetration depth is in the order of a few millimeters (2–5 mm) and therefore the application is limited. It is mainly used in ophthalmology and dermatology, but other special applications are also known (Fig. 16.2.3). Light with broad bandwidths generated by femtosecond lasers is split into two arms (sample and reference arms). An interference pattern is produced by overlaying reflected light from the sample arm and reference light from the reference arm, if light from both arms have travelled the “same” optical distance. A reflectivity profile of the sample can be obtained by moving the mirror, which is the A-scan containing information about the spatial dimensions and location of structures. As in ultrasound, Bscan can be achieved by combining a neighboring series of A-scans.

Diffuse optical imaging (DOI) or diffuse optical tomo­ gra­phy (DOT) uses—similar to high-energy X-ray photons—light photons with much lower energy to generate images of tissue. Whereas in X-ray tomography the scattering coefficient is small, for light photons it is an order of magnitude higher than the absorption coefficient. This implies multiple scattering of light photons and ballistic photons for reconstruction are rare, even in small samples. Reconstruction algorithms other than those in X-ray CT have to be implied to account for the multiple scattering, and therefore the diffusion approach is commonly used. The technique is sensitive to the optical absorption in particular of oxy- and deoxyhemoglobin, but also endogenous fluorochromes, water, and other tissue constituents. To avoid heavy absorption by water molecules and other molecules in the low wavelength region (<400 nm), wavelengths above 600–1200 nm are used (near-infrared light). DOT instruments are relatively low in cost. Attempts were made to bring this method into clinical practice, especially for breast imaging; however, spatial resolution and contrast to noise is not sufficient for clinical use. On the other hand, it was demonstrated that brain activity, brain hemorrhage, and brain oxygen supply in newborns could be diagnosed and monitored (Kurth et al. 1993; Hintz et al. 1999).

16.2.1.3.4  Other Techniques Other optical techniques as confocal microscopy, confocal laser scanning microscopes, stimulated emission depletion microscopy (STED microscopy) (Hell 2003) and thermoacoustic-computed tomography (TCT) (Kruger et al. 2000, 2003) should be mentioned, but at present in molecular imaging, they play only minor roles. For a current review of available techniques, the reader is referred to Gibson et al. (2005) or Hielscher (2005).

16.2.1.3.3  Optical Coherence Tomography

16.2.2  Computed Tomography

For the sake of completeness optical coherence tomography (OCT) should be briefly described. OCT is a method based on interferometric principle, has high-resolution (submicrometer resolution) use of wide bandwidth light sources (emitting wavelengths range ~100 nm); however,

CT is not a molecular imaging modality per se. However, if the definition given above is adopted, then surrogate marker imaging may be included in molecular imaging. In this context, this can be either high-resolution imaging to obtain information about vessel structure, ves-

16.2  Molecular Imaging Modalities Fig 16.2.3  Schematic optical setup of optical coherence tomography (OCT). Light from the sample and the reference light gives rise to an interference pattern. By scanning the reference mirror, a profile of the sample can be obtained, and the x-y–plane can be scanned using the sample mirror (see text)

sel size, and vessel density and/or dynamic CT, making possible measurement of blood flow, blood volume, etc. CT achieves high spatial resolution (clinical scanners: in-plane 250 µm) and new developments with flat-panel detectors in the order of 100–200 µm. Dedicated microCT scanners are designed to achieve spatial resolution of ≤100 µm (Kalender 2005). Two different approaches for small-animal CT imaging exist, with two different design concepts for microCT scanners: • Micro-CT – Rotating object CT-scanners – Rotating gantry scanners • Flat-panel volumetric CT (fpVCT) 16.2.2.1  Microcomputed Tomography Two different designs for micro-CT are on the market, rotating object scanners and rotating gantry scanners—the latter are known from clinical CT. The rotating object scanners have a simple construction, are robust, and cheap; however, the positioning of a living object is problematic, in particular when the object rotates along a vertical axis. In such a case, the mouse is positioned vertically and in turn exhibited to non-physiological vertical forces, resulting—for instance—in a different blood volume distribution, different flow patterns, etc. These drawbacks can be avoided with rotating gantry scanners. In these scanners, the object is typically placed horizontally. These scanners are mechanically more complicated, prone to failures to a higher extent, and more expensive. Both types of microCT have in common low-contrast details that are often obscured by image noise. They can be compensated for by

longer image time and/or higher X-ray power, resulting in an unacceptably high dose. For instance, whole-body mouse CT scan with a spatial resolution of about 50 μm, the dose per scan requires about 0.6 Gy, which is approximately 5 % of the median lethal does (LD50) value for mice (Fig. 16.2.4) (Weissleder 2002). 16.2.2.2  Flat-Panel Volumetric CT Flat-panel volumetric CT (fpVCT) scanners are conventional CT scanners equipped with flat-panel detectors with high resolution down to 100 µm. Presently, these experimental fpVCT with a whole-body gantry are manufactured by two vendors and are used to scan not only small animals (mice, rats), but also primates. The main advantages of these scanners is the relatively low-dose, high-resolution fast scanning appropriate for dynamic contrast media studies, external triggering, and the use of reconstruction software known from clinical scanners; however they are more expensive compared with micro-CTs. 16.2.3  Positron Emission Tomography and Single-Photon Emission Tomography Positron emission tomography (PET) and single-photon emission tomography (SPECT) are molecular imaging modalities per se. Because of their high physical sensitivity down to picomolar range, small concentrations of radioactively labeled specific molecules can be detected. PET is quantitative and has the ability to observe meta-

1387

1388

16  Molecular Imaging

bolic processes over time and track radiolabeled specific markers in vivo (Czernin and Phelbs 2002). The spatial resolution is smaller compared with CT or other MRI. Clinical PET scanners have a resolution down to about 4 mm, whereas dedicated small animal state-of-the-art PET scanners offer a resolution in the range of 1–1.5 mm with sensitivity of up to 11% (Tai et al. 2005; Wang et al. 2006). This excellent image resolution allows for quantitative assessment of tiny lesions even in mice. A major limitation of PET is that it lacks spatial resolution and does not provide sufficient anatomical infor-

mation. This can be overcome by imaging fusion with images of CT or MRI or as described below, by combined PET–CT scanners. Both modalities—PET and SPECT—use radioactive isotopes as high-energy photon sources. SPECT uses long-lived radioisotopes emitting gamma rays, whereas PET facilitates positron emitters. The positron annihilates with electrons of the surrounding material, emitting two photons in opposite directions (Fig. 16.2.5). Construction of the scanner, detection, and the reconstruction algorithms of both modalities are very similar. Fig. 16.2.4a,b  Two microcomputed tomography scanner geometries. a Rotating object scanner and b rotating gantry scanners— the latter are known from clinical CT (see text)

Fig. 16.2.5  Schematics of a PET scanner. The scanner consists of a detector ring, coincidence processor unit, and a reconstruction computer. The short-lived radioactive tracer isotope (e.g., 19F) decays by emitting a positron, and after a short path annihilates with an electron. The annihilation process produces two γ-rays with energy of 511 keV each. Both γ-rays were detected in coincidence, and the reconstruction algorithm allowed calculation of the event location

16.2  Molecular Imaging Modalities

Both use a gantry for the detector mounting, which is stationary for PET with many detectors in the ring, and SPECT with two or three rotating detectors. The detectors are γ-ray detectors, either scintillation or avalanche detectors. The latter are used preferentially in micro-PET systems. In principle, back-projection algorithms with scattering corrections are used for image reconstruction. Both modalities are used routinely for diagnostic imaging in patients in neurology, oncology, and cardiology. 16.2.4  Magnetic Resonance Imaging and Spectroscopy The basic principles and the necessary equipment are already described in Chap. 2 and applications in Chap. 15. Therefore, only a few words are added. Magnetic resonance imaging—as mentioned earlier— has low physical sensitivity, and is hampered significantly by this fact. However, by increasing the magnetic field, signal to noise can be improved significantly (Fig. 16.2.6). If metabolite concentrations are high enough, direct measurement of metabolites is possible as has been shown previously in clinical studies. MR spectroscopy has been used since the mid-1980s in clinical studies and for studies of isolated organs and animal studies, already in the 1960s and 1970s. It is worth mentioning that for animal and cell-culture experiments, scanners up to 17.6 T are available, allowing testing and evaluating of future applications of ultra-high-field clinical MRI. Too, all sequences and methods used in clinical examination can be adopted and used in experimental settings and this also holds true for chemical-shift imaging. 16.2.5  Ultrasound In the past ultrasound was not considered to be a molecular imaging modality; however, with the development of ultrasound contrast media (microspheres)—stimulated acoustic emission (SAE) and sensitive-particle acoustic quantification (SPAQ)—techniques (Tiemann et al. 2000; Reinhardt et al. 2005) were developed, and in turn, disease-specific imaging by ultrasound and quantitative imaging was realized using ultrasound (Table 16.2.2). Because of the bubble size, extravasation is unlikely; however, quantitative intravascular molecular imaging by single-bubble counting is possible not only in experimental settings, but also in clinical settings, if approval can be obtained by the regulating authorities. Because of the increasing attention for molecular imaging, ultrasound, the microbubbles, and as SAE and SPAQ techniques are described in some more detail. The ultrasonic waves are generated by piezoelectric elements. The piezoelectric elements—often arranged in an array—produce elastic vibrations and transmit them

Fig. 16.2.6  Signal to noise (S/N) as function of the magnetic field (schematically)

Table 16.2.2  Classification of ultrasound methods • Ultrasound (US) – Amplitude mode (A mode) – Brightness mode (B mode) – Doppler • Linear • Pulsed wave • Continuous wave • Color flow mapping (CFM) (color Doppler) – Stimulated acoustic emission (SAE) – Sensitive particle acoustic quantification (SPAQ)

to the material under investigation. The reflected sound waves are received again by these elements, electrical signals are generated, and these are analyzed in the ultrasound equipment. The double function of these elements gives rise to term them “transducer.” Due to the differing propagation velocities of the materials under investigation, the sound waves are reflected and/or refracted at the interfaces of two materials, and attenuated to a different extent, depending on the elastic properties of the material. Amplitude scans are termed A-scans, a B-scan (brightness) is an A-mode with the amplitudes converted in grayscale, and several of these neighboring A-mode lines are set together to an image. The Doppler effect can be used to measure the speed of a moving source (e.g., flowing blood in an artery). Linear, pulsed, and continuous-wave Doppler techniques were developed and combined with B imaging (= duplex sonography). The Doppler signals can be color coded. This method is known as color flow mapping (CFM). Using elastic microbubbles as ultrasound contrast agents, other techniques come into play. The contrast

1389

1390

16  Molecular Imaging

behavior is a function of the used irradiated ultrasound frequency and the sound pressure physical and acoustic characteristics of the gas-filled microbubbles. If the irradiation frequency matches the Eigen frequency of the micro bubbles, they will resonate with the same frequency, reflect the ultrasonic wave, and in turn enhance the contrast. If the sound pressure is raised, increasingly asymmetrical oscillations occur and other frequencies— so-called harmonic (twofold, threefold, etc., of the fundamental [= irradiation] frequency) or sub-harmonic (half, third, etc., of the fundamental frequency) frequencies are emitted, with the second harmonic being the frequency with the highest amplitude. These properties of the microbubbles make second harmonic imaging and wideband harmonic imaging possible. These techniques are combined with B-mode and with Doppler technologies. At high sound pressure, the microbubbles are fragmented depending on their wall thickness and elastic, gas-filled microbubbles. During fragmentation a short-lived wideband, nonlinear frequency signal is emitted, which is the SAE, and it can be detected as color pixels in color Doppler mode of the scanner. If surrogate markers measurement (as defined above) are considered to be molecular imaging, quantitative Doppler ultrasonography measuring the flow in blood vessels has to be included as such. This technique is frequently used in clinical routine and—as demonstrated by Krix et al. 2003—monitoring of tumor treatment is possible even in mice. This holds also true for high-resolution ultrasound, the so-called micro-ultrasound. Specifically developed scanners are available for investigations of small rodents. In the future, these scanners will successfully compete with mirco-CT, fpVCT, and micro-MRI. A new method developed only recently by Reinhardt et al. (2005)—SPAQ—facilitates gas-filled microbubbles and the SAE effect during destruction of the bubbles. This technique allows counting even of a single targetspecific-bound microbubble. It was demonstrated that these technique can be used for “real” molecular imaging modality using ultrasound (Ellegal et al. 2003).

Gamma Medica, and Siemens have recently introduced combined PET–CT machines for small animal imaging. 16.2.6.2  PET–MRI Integrating PET and MRI is more challenging. The photomultipliers are sensitive to magnetic fields and cannot be used inside the magnet. Either light guides have to be applied to keep the multiplier outside of the magnetic field or avalanche photo diodes (APDs) are used, which are not affected by the magnetic field. Different approaches are tested presently. The concept using APD detectors seems to be the most promising. With prototype PET–MRI scanners based on these principles, first images of a volunteer were acquired recently (B.J. Pichler, 2007, private communication). 16.3  Molecular Imaging Probes Molecular imaging probes consist of a specific molecule (ligand) with high affinity to a particular target and a signal generator (SG) (= “signal moderator,” “signal molecule,” “signal particle”); these expressions are used to state the difference between markers (the conjugated entire molecule) and signal-generating isotopes, molecule, or particle appropriate for the imaging modality of choice (cf. Fig. 16.3.1); the latter are for PET and SPECT radioactive isotopes, for optical imaging dyes and quantum dots, for magnetic resonance superparamagnetic nanoparticles and Gd chelates, for ultrasound gas-filled microbubbles, etc. (Table 16.3.1). The classification of molecular imaging probes used in this chapter is given in Table 16.3.1. This classification slightly differs from this given by Weissleder (2002), who uses “compartmental probes” instead of “unspecific probes,” which assess physiological parameters (flow, perfusion, etc.) will not be included, because the molecular process is not directly imaged but rather, a sur-

16.2.6  Multimodal Imaging Approaches: PET–CT and PET–MRI 16.2.6.1  PET–CT PET lacks spatial resolution and does not provide sufficient anatomical information and therefore overlaid morphological information is needed in order to accurately localize a lesion (Beyer et al. 2000; Townsend et al. 2004). Image fusion of independently acquired images of MRI or CT is one way to overcome this obstacle and is facilitated by different software fusion techniques. Hardware fusion is an alternative (Beyer et al. 2000). Combined PET CT scanners are on the market and used clinically. GE,

Fig. 16.3.1  Schematics of molecular probes: The signal generator (= signal moderator/signal molecule/signal particle) is coupled via a linker to the specific ligand. The epitope is the key structure for specific binding to a receptor or other specific structure of a cell

16.3  Molecular Imaging Probes Fig. 16.3.2  Selected molecular cellular extra- and intracellular targets

rogate marker. In a strict sense these “compartmental probes” are conventional contrast media and do not have specificity. However, such contrast media may be used for disease-specific imaging too, for instance, when lymphocytes labeled in vivo with unspecific superparamagnetic nanoparticles homing to lymph nodes and thereby delineate lymph node metastasis. This was first demonstrated in experimental animals by Taupitz et al. (1993a, b) and soon also in clinical examinations (Deserno 2004; Nguyen et al. 1999; Memarsadeghi et al. 2006). The key requisite for molecular imaging with the targeted probes—also called disease-specific probes—is the high affinity of ligands toward a molecular target. Targets can be cell-surface receptors, intracellular enzymes, transporters, oligonucleotides, reporter genes, etc., ligands peptides, enzyme substrates, antibodies, or recombinant proteins (cf. Fig. 16.3.2). Activatable probes—also called “smart” probes—are designed to activate exclusively in the presence of the intended target. The advantage of

Table 16.3.1  Classification of imaging molecular probes. “Nonspecific probes” assess physiological parameters (flow, perfusion, etc.)—so-called a surrogate marker—molecular processes are not directly imaged by these probes; these are rather conventional contrast agents. Targeted probes (disease-specific probes) consist of highly specific ligands toward a molecular target and a signal generator appropriate for the imaging modality, activatable probe—also called “smart” probes—is designed to activate exclusively in the presence of the intended target • Nonspecific probes • Targeted probes • Activatable probes

these activatable probes is the fact that background signals are faintly present. High affinity of the ligand to the target is a mandatory but not sufficient for a molecular imaging probe. A set of further conditions have to be fulfilled to synthesize molecular probe applicable in a living organism. Among these are (1) after coupling of the signal generator to the specific ligand, the affinity has to be kept in the same order of magnitude—typically in the nanomolar region and unspecific binding to other structures in the organism must be low; (2) the target concentration in the region of interest must be sufficient and has to match with the sensitivity of the modality of choice (3) in vivo stability has to be guaranteed (e.g., peptides might by digested by peptidases, etc.); and (4) delivery across biological barriers must be unhindered to reach the target—also intracellular. Furthermore, favorable pharmacokinetics (appropriate plasma half-life, favorable excretion to eliminate unbound molecular probes to reduce background. etc.), and low toxicity are additional challenging boundary condition to be fulfilled (Table 16.3.2). To overcome these problems, numerous approaches were tested. Only a few should be mentioned here. For instance, for stabilizing peptides, d-amino acids can be introduced into the peptide sequence at appropriate positions; the same holds for oligonucleotides, which can be stabilized by sugar modifications of nucleoside triphosphates to protect for nucleases found in serum. Changing the 2′-OH groups of ribose to 2′-F or 2′-NH2 groups yields oligonucleotides with long half-life in blood (Brody and Gold 2000). To achieve high concentration in the target area, amplification strategies have been developed as the use of heptahelical transmembrane receptors, which are internalized into the cell after the docking of the ligand to the receptor. This was particularly demonstrated with the transferrin receptor (Kresse et al. 1998). Using

1391

1392

16  Molecular Imaging Table 16.3.2  Mandatory properties of molecular probes • High affinity – Ligand to target (nanomolar range) – High affinity of conjugate after coupling (via linker) • High target concentration – Concentration depending on and appropriate for the given sensitivity of the modality • High in vivo stability – Ligand – Entire conjugate (ligand, linker and signal generator) • Delivery across biological barriers – Vessel walls (extravasation) – Cell membranes (intracellular targets) • Favorable pharmacokinetics – Fast excretion to reduce background – Fast accumulation in the region of interest • Low toxicity – Acute toxicity – Chronic toxicity

Tat peptides, penetration of nanoparticles into the cell was be improved and in turn, high intracellular uptake of transferrin-coated nanoparticles in the cell was obtained (Josephson et al. 1999). However, every modification concerning stabilization of peptides or oligonucleotides, coating of nanoparticles, as well as coupling to a specific ligand requires new optimization steps of all other parameters of the molecular probe. Furthermore, toxicity—and in particular chronic toxicity—is another parameter that may introduce further difficulty and may even prevent the development of a molecular probe for clinical purposes at the end of a long process. These iterative optimization steps as such make development of molecular imaging probes timeconsuming, cumbersome, and expensive. In the following sections, the difficulties encountered in the development process of a molecular imaging probe are mentioned only if appropriate and cannot be elucidated in extenso in this chapter.

mentioned only briefly. This holds also true for the following section concerning experi­ment­al and clinical applications. 16.3.1.1  Radioactive Markers Positron emitters, most frequently isotopes with low-order number (Z), are suitable for use in medical imaging. These are 11C, 13N, 15O, and 18F with half-lives of ~20, ~10, ~2, and ~110 min, respectively. Most commonly used high-order number positron emitters are 68Ga, 82Rb, 94Tcm, and 124I. In contrary to the above mentioned isotopes and 94Tcm as well as 124I, which have to be produced by an accelerator on line; 68Ga and 82Rb can be produced by generators, giving these isotopes easier availability. Low-Z isotopes are main constituents of organic compounds and do not influence structure and properties of a molecule when introduced in the adequate position of a specific molecule intended to be used as a molecular imaging probe. The short half-lives, however, are a major obstacle making synthesis and handling difficult. For SPECT—in addition to others—the most important radioactive isotopes are 99mTc (6 h), 111In (2.8 days), and several iodine isotopes, in particular, the iodine isotopes 123I (13.2 h) and 125I (59.5 days) are frequently used in medical applications and can be easily facilitated for labeling biomolecules for imaging. 16.3.1.2  Optical Signal Generators The key criteria for the optical signal molecules (signal generator) are the photophysical properties of the compounds, as given in Table 16.3.3. Selection of the compound for the intended purpose needs to take into account these basic features. Signal generators or signal modulators, which enhance or modify signals and in turns contrast, are manifold for optical imaging. They can be classified as given in two ways, either by physical action absorbing, fluorescing or luminescing, or by the type of chemical compound (cf. Tables 16.3.4 and 16.3.5, respectively).

16.3.1  Signal Generators Signal generators (SGs) or signal molecules are the part of the molecular probe specifically responsible for the signal generation and/or signal modification, and are different for each modality. The signal generator might either be radioactive isotopes as for PET or SPECT, a dye, a fluorescing protein, or quantum dot as for optical imaging, paramagnetic molecules, or magnetic nanoparticles as for magnetic resonance and gas filled bubbles as for ultrasound. Because magnetic resonance is in the focus of this chapter, the signal generators of the other modalities are

Table 16.3.3  Photophysical properties of suitable optical signal molecules (optical signal generators) • High absorption yield – High molar extinction coefficients • Appropriate absorption wavelengths • Fluorescence emission – High fluorescence quantum yields – Large Stokes shift • Appropriate life times of the intermediate state • Low photobleaching

16.3  Molecular Imaging Probes

Cyanine dyes are perhaps the most commonly used and versatile fluorophores in the area of molecular imaging. Their wavelengths can be tuned from visible light to near-infrared, have short fluorescence lifetimes (~1 ns), high-molar-extinction coefficients, and moderate fluorescence quantum yields (Licha et al. 2000). The disadvantages of these compounds are chemical instability and the tendency for photobleaching. In molecular imaging tetrapyrrol chromophores, lanthanide chelates, and xanthene dyes do not have the some importance as the indocyanine dyes. Quantum dots (QDs) are a completely different kind of optical signal generator; they are semiconductor nanocrystals consisting of atoms such as Cd, Se, Te, S, or Zn (Medintz et al. 2005). These quantum dots have a diameter of about 2–10 nm, a large quantum yield, and high photostability. The wavelength of fluorescence light is dependent on the size of the particle, which can be synthetically controlled (Lim et al. 2003) and also on the doping with other atoms like In, As, and P (Zimmer et al. 2006); with this, emission frequency can be tuned, covering the range from visible light to near-infrared region. Surface modification of the QDs improved solubility and biocompatibility, and reactive groups are added to biogconjugate specific ligands. An unsolved problem is the chronic toxicity of the QDs, which could prevent approval of application in humans. Because of these problems, metal free nanoparticles (Santra et al. 2004) and polystyrene nanospheres loaded with fluorescent dyes (Nakajiama et al. 2005) are under development. A phenomenon termed “quenching” should be mentioned here because this effect is used for the construction of smart probes. Fluorescence signal quenching occurs due to fluorescence resonance energy transfer (FRET). This is a radiation-free energy transfer from a donor to an acceptor fluorophore by means of dipole–dipole interaction. To facilitate this, the emission and absorption spectrum of donor and acceptor must overlap, and donor and acceptor must be in close proximity (in the order of some nanometers). If these conditions are fulfilled, fluorescence emission will be suppressed. Appropriate dye molecules in a molecular probe may be linked to each other by a cleavable linker, short enough that quenching occurs. Upon cleavage of the linker, the dye molecules are set free; the FRET is interrupted because of the greater distance and fluorescence emission takes place; using such a process, fluorescence emission can be switch on. 16.3.1.3  Ultrasound Signal Generators Gas bubbles possess low acoustic impedance, leading to a stronger reflection of acoustic waves compared with, tissue and in turn, the contrast will be enhanced by such micro bubbles. The various formulations of microbubbles are listed in Table 16.3.5. The simplest ultra-

Table 16.3.4  Classification of optical signal molecules (signal generators) by physical action (1) and by compound (2) 1. • Dyes: – Absorbing dyes – Fluorescing dyes • Luminescence peptides – Green fluorescence protein (GFP) – Red fluorescent protein (RFP) – Luciferin/luciferase 2. • Polymethine dyes – Carbocyanine dyes • Tetrapyrroles – Porphyrins – Chlorins – Bacteriochlorins • Rare-earth metal chelates – Terbium and europium complexes • Xanthene dyes, including – Fluoresceins – Rhodamine-type fluorophores – Quantum dots

Table 16.3.5  Classification of ultrasound microbubbles • Free gas bubbles • Encapsulated soft-shell gas bubbles – Soft elastic shell • Lipids • Proteins – Gas: slightly soluble gases • Perfluoro gases • Encapsulated hard-shell gas bubbles – Stable biodegradable shell of polymers • Polylactide • Cyanacrylate • Targeted-encapsulated gas bubbles – Target-specific ligands coupled to sheath‘s surface

sound contrast media are free gas bubbles; however, they dissolve rapidly in blood. The second class is stabilized gas bubbles with soft and elastic lipid or protein shells. These gas bubbles can be filled air or with only slightly soluble gases, leading to a longer half-life in the blood. These gas bubbles can diffuse through lung tissue. A third class of ultrasound agents are hard-shell gas bubbles with a stable shell of biodegradable polymers such as polylactide or cyanacrylate. These microbubbles have an even longer half-life and—very importantly— make possible, the aforementioned SAE (Tiemann et al. 2000), and fur-

1393

1394

16  Molecular Imaging

Fig. 16.3.3  Strategy for actively targeted micro-bubbles as ultrasound contrast agents (Courtesy of P. Hauff, Bayer Schering Pharma, modified)

thermore, the use of the SPAQ technique (Reinhardt et al. 2005). For molecular probe imaging, this kind of bubble can be surface modified and target-specific ligands can be attached (Fig 16.3.3). The diameter of the bubbles range from 1 to 5 µm and therefore, only intravasal application is feasible. 16.3.1.4  Magnetic Imaging Signal Generators As aforementioned, because of the low intrinsic physical sensitivity of MR, the SG has to have strong influence on the contrast. In other words, strong relaxation centers are needed to change the signal of the observed protons. The signal enhancement produced by these substances depends on their relaxivity: r1, r2 and r2*. The basics of the relaxation mechanisms are explained in Chap. 2 and are not repeated here. In principle, two different classes of substances exist as well as the methods for polarization enhancement, which are suitable for relaxation enhancement (Table 16.3.6). 16.3.1.4.1  Paramagnetic Substances For paramagnetic substances, relaxivity r1 and r2 are the main physicochemical parameters influencing the strength of an effective magnetic label. The dynamic relaxation process depends essentially on the size and chemical structure of a molecule and on the accessibility of water molecules to the magnetic center. Paramagnetic substances are well known in clinical MR imaging. The approved and mostly used substances

are gadopentate dimeglumine (Gd-DTPA), gadoterate meglumine (Gd-DOTA), gadodiamide (Gd-DTPA-BMA) (cf. Chap 2.6), gadoteridol (Gd-HP-DO3A), gadobenate dimeglumine (Gd-BOPTA), and gadoxetic acid (GdEOB-DTPA). To facilitate the use of these substances for molecular probe imaging, they have to be functionalized and coupled to a specific ligand. This is straightforward; however, the relaxivities of these paramagnetic substances are rather low: r1 ranges from 3 to 5 ℓ · (mM · s)–1 and r2 from 4 to 6 ℓ · (mM · s)–1 at 1 T (field dependent). This is not sufficient for molecular probe imaging, where at least two to three orders of magnitude higher relaxivities are required. More promising are the labeling of cell for instance with paramagnetic agents, as it was shown by DaldrupLink et al. (2004) using liposomes containing gadophrin2, a metalloporphyrin, as well as Gd-containing lipophilic nanoparticles (Vuu et al. 2005).

Table 16.3.6  Substances and principles for relaxation enhancement in MR • Paramagnetic substances – Gd chelates – …. • Superparamagnetic nanoparticles – SPIO – USPIO – VSPIO – …. • Hyperpolarization

16.3  Molecular Imaging Probes

16.3.1.4.2  Superparamagnetic Nanoparticles The superparamagnetic nanoparticles are composed of iron oxides, usually magnetite, Fe3O4, maghemite, γFe2O3, or other ferrites. The net magnetic dipole of these particles is large compared with the sum of the individual unpaired electrons because of long range interaction of d electrons in solid matter. Due to this large magnetic moment of the nanoparticles, the magnetic field is locally altered and in turn, large magnetic field heterogeneities are created. Water molecules diffuse into these heterogeneous magnetic field areas, and become dephased, resulting strong (=T2*) relaxation (T2 shortening), and finally a canceling of the signal occurs. In principle this is a susceptibility artifact and therefore these contrast agents are also called susceptibility agents. The physicochemical properties strongly influence the relaxivity of these particles. The size (the diameter of the iron particle) is the most important parameter that influences the magnetic properties of the particles. Due to the kind of particles, the diameter varies: for very small paramagnetic iron oxide (VSPIO) and ultra small paramagnetic iron oxide (USPIO) in the order of 5–10 and 10–50 nm, respectively, and above for SPIO (Lawaczeck et al. 2004) (cf. Fig 16.3.4). However, other parameters have to be considered, for instance, the hydrodynamic diameter. To stabilize the particle in aqueous solution, particles are coated with small molecules (citrate, oleate, silane, etc.) or polymers (Dextran, synthetic polymers, starch, etc.) or enwrapped by liposomes. This increases the hydrodynamic diameter and in turn changes relaxivity. Furthermore, in blood fast opsonization occurs after intravenous administration. This changes the relaxivity as well and enhances the uptake by macrophages and in liver by the

Fig. 16.3.4  Size dependence of the relaxivity. r1 and r2 relaxivities as a function of particle size for VSOP (filled diamonds), USPIO (SH U555C, open squares, Supravist®), USPIO (AMI227, open triangles, Sinerem®, Combidex®), USPIO (NC 100 150, filled circle, Clariscan®), SPIO (SH U555A, filled squares, Resovist®), SPIO (AMI-25, filled triangles, Ferride®, Endorem®); 37°C, 0.47 T (Lawczeck et al. (2004) Appl. Organometal. Chem. 18: 506 – 513)

Kuppfer cells. Coated superparamagnetic particles are approved and are used in clinical routine: Sinerem® (Guerbet), Clariscan® (GE Health Care), Endorem® (Guerbet), and Revovist® (Beyer Pharma Schering). Other particles, like monocrystalline iron oxide nanoparticles (MIONs), are in clinical trials and experimental stages for molecular imaging. This is because superparamagnetic particles have the highest probability to bridge the sensitivity gap mentioned above. 16.3.1.4.3  Other Nuclei In the past, 19F imaging has been demonstrated almost background-free MR imaging. 19F has a high magnetic moment, comparable to that of the protons and therefore, almost the same MR frequency (see Chap. 2) lowering time and effort to implement this method. Perfluoropolyether-labeled dendritic cells can be detected in vivo (Ahrens et al. 2003). 16.3.1.4.4  Hyperpolarization Only recently the attention has focused on hyperpolarization for molecular probes. The polarization (difference in the occupation numbers, cf. Chap. 2) can be increased and thereby the MR signal when the magnetic field is higher. This is well known, and therefore the magnetic field of scanners has increased to 3 T and even above 7 T. However, the non-equilibrium distribution of the nuclei can be influenced much more effectively using polarization transfer. Hyperpolarization can be created by different techniques, which allow producing differences of the population probability by several orders of magnitudes. The nuclear Overhauser effect NOE (= dynamic polarization transfer) has been known since the 1950s. Later, hyperpolarization of 3He was achieved by another technique, optical pumping (Colegrove et al. 1963), and imaging with polarized 3He is widely used nowadays. Ende and Bachert demonstrated that such signal enhancement can be achieved in vivo, using NOE (Ende and Bachert 1993). Another technique achieving hyperpolarization is the so-called para-hydrogen–induced polarization, utilizing the fact that by means of a diabatic field cycling the non-equilibrium spin order of para-hydrogen can be converted into nuclear polarization of the molecule of interest (Golman et al. 2003). A breakthrough for molecular imaging using hyperpolarized biosensors was published recently by Schröder et al. (2006). They achieved a signal three orders of magnitude higher than without polarization enhancement. To do so, the 129Xe ion is caught in a cryptophane-A cage to which—via a linker—a specific molecule (ligand) can be attached. Polarized 129Xe atoms freely diffuse in and out of the cage. Due to the chemical shift—which occurs

1395

1396

16  Molecular Imaging

Fig. 16.3.5a,b  Sensitivity-enhanced NMR detection of Xe biosensors. a Chemical structure of the Xe biosensor illustrating the cryptophane-A cage (green), the linker (black), the targeting moiety (biotin in this case, orange), and the peptide chain (purple) that is required for sufficient water solubility. The 129Xe NMR spectrum of this construct at 50 µM bound to avidin agarose beads yields only a broad, weak resonance from encapsulated Xe at δ3, even for 256 acquisitions. Chemical exchange with free Xe outside the cage (resonance δ1) enables sensitivity enhancement by depolarizing the δ3 nuclei and detecting at δ1. b Amplifying the cage-related magnetization using HYPER-CEST. Selective saturation of biosensor-encapsulated Xe (green) and subsequent chemical exchange with the free Xe (blue) allows accumulation of depolarized nuclei (red). This procedure corresponds to continuous depolarization of cage-related magnetization that can be measured indirectly after several cycles by the difference between initial and final bulk magnetization (Schröder et al. (2006) Science 314:446)

16.3  Molecular Imaging Probes

during the millisecond stay of the atom within the cage— these nuclei can be specifically depolarized using a special sequence (HYPER-CEST). In other words, the signal contribution of all 129Xe atoms entering the cage vanishes (Fig. 16.3.5). The longer the irradiation at this specific frequency, the more 129Xe atoms affected. Subtraction of images before and after specific depolarization of the cage atoms yields an image that indicates the location of the cage atoms (the Xe biosensor) by low signal intensity (Fig. 16.3.6). There is still room to improve the enhancement factor, for instance, by using isotopically enriched Xe and by improvement of the sequences. 16.3.2  Molecular Targeting for Imaging Numerous molecular targets are identified. In Fig. 16.3.2, some important cellular targets and their counterparts— the ligands—are schematically represented, e.g., monoclonal antibodies, peptides (VEGF, VGF, somatostatin), adhesion molecules (integrins, selectins, etc.), transferrin

Fig 16.3.6a–d  Molecular imaging of the Xe biosensor (overlay of transverse 129Xe images obtained from CSI data sets with the 1H image shown in Fig. 16.1.2a). a Selective image for the bead signal at δ1 and off-resonant continuous wave saturation. b Selective image of the pure solution signal at δ2. This signal corresponds to the surrounding outlet tube and is not affected by

as an example for a ligand of a heptahelical transmembrane receptor (cf. Fig. 16.4.2), which is internalized into the cell and are recycled, as well as transporters (glucose), gene reporters (tyrosine kinase), and specific enzyme substrates. Oligonucleotides are another class of molecules to be used as specific ligands but not displayed in Fig. 16.3.2. Two types of targeting can be distinguished, active and passive targeting (Table 16.3.7). Furthermore, the molecular targets can be classified as intercellular and cell-surface targets as done by Weissleder and Mahmood (2001) (Table 16.3.8). This list is not complete. In principal, these targets can be used for all the imaging modalities envisioned for molecular imaging. Yet, we have to take into consideration that receptors, enzymes, etc., are not exclusively presented in a diseased state, but are also found in non pathological circumstances. This turns the problem from a qualitative to a quantitative one, which makes things more complicated. For differentiation, absolute quantification has to be required, and not all methods used for molecular imaging are quantitative at present.

any saturation transfer. c On-resonant saturation of the biosensor resonance substantially depletes the δ1 signal in volume 2. d The difference image of the two CSI data sets yields an exclusive mapping of compartment 2, i.e., a molecular image of the Xe biosensor (Schröder et al. (2006) Science 314:446)

1397

1398

16  Molecular Imaging Table 16.3.7  Active and passive targeting • Active Targeting – Ligands • Monoclonal antibodies • Peptides • Polysaccharides • Oligonucleotides (aptamers) • Drugs • Thrombin – Enzymes substrates • Cathepsin • Caspase • MMP • Passive targeting – Cell trafficking

Table 16.3.8  Intracellular and cell surface marker proteins (Weissleder and Mahmood 2001) Marker Protein

Imaging Modality

Ligand or Substrate

Mechanism

Thymidine kinase

Nuclear

FIAU, ganciclovir, penciclovir, FHPG, FMAU

Phosphorylation of prodrug

Cytosine deaminase

Nuclear, MRS

Cytosine fluorinated prodrugs

Deamination of prodrug, fluorine imaging

Tyrosinase

Nuclear, MRI

Tyrosine, dopa

Oxidation of substrates into melanin or metal scavenging

Arginine kinase

MRS

ATP and arginine

Conversion to phosphoarginine

Creatinine kinase

MRS

ATP and creatinine

Conversion to phosphocreatinine

β-Galactosidase

MRI

Galactosylated chelators

Cleavage of galactose, residues changes, relaxivity

Green fluorescent protein

Optical

None required

Fluorescence

Luciferase

Optical

Luciferin

Bioluminescence

Proteases e.g. cathepsin D

Optical

Quenched NIR-fluorescent fluorochromes

Fluorescence activation

Gastrin-releasing–peptide receptor

Nuclear

Bombesin

Affinity binding

Somatostatin receptor

Nuclear

Peptides

Affinity binding

Dopamine-2 receptor

Nuclear

18F spiperone

Affinity binding

Iodine binding

Nuclear

Iodine

Trapping

Fusion proteins

Nuclear

99mTc chelates

Transchelation

Engineered internalizing receptor

MRI

Transferrin

Internalization, relaxivity change

Intracellular

Cell surface

16.4  Applications

16.4  Applications At present, molecular imaging is predominantly used in preclinical research, except—as aforementioned—PET and SPECT and MRS. These methods (in particular PET and nowadays PET–CT) play an important role in neurological and cardiovascular diseases staging of cancer and therapy monitoring. A comprehensive review of PET–CT and references was published only recently by Blodgett et al. (2007). MRS is also used in clinical routine, for instance, for diagnostics of prostate cancer and brain tumors. These applications are described in Chap. 15 and in other relevant chapters of this book. Therefore, experimental and clinical studies using these methods are not covered in this chapter. This holds true also for surrogate marker imaging in general, except cell tracking. In this part, representative example of experimental and clinical applications is introduced and briefly commented on, and those examples chosen and emphasized are given using MR. Other methods are mentioned only if experiments in the field of MR are not reported (Table 16.4.1). 16.4.1  Antibodies The concept of specific targeting with antibodies and using MR imaging was demonstrated as early as 1989 by Cerdan et al. (1989) and later by Bulte et al. (1992) and Tie­ fen­auer et al. (1993), in vitro and by Moore et al. (2001); in vitro and in vivo followed soon. In these early experiments monoclonal antibodies were coupled to Dextran

Fig. 16.4.1  Molecular imaging using antibodies as specific ligands. The Her-2/neu gene is transfected and the Her-2/neu receptor will be overexpressed on cell surface. Her-2/neu receptor is also a target for the humanized monoclonal antibody (mAb) Herceptin, which is used immunotherapy. The expression of

Table 16.4.1  Classification of MR molecular imaging • Probe imaging – Nonspecific probes • Conventional contrast agents Magnevist®, Omniscan®, … – Targeted probes (disease-specific probes) • Antibodies (Her-2/neu, …) • Peptides (somatostatin, …) • Proteins (transferrin, …) • Oligonucleotides • … • Gene expression • Cell tracking • Hyperpolarized gases – Carbon – Xenon

magnetides or polymeric iron oxide particles, however, high uptake of theses probes in liver, spleen, and bone marrow was observed. Yet in an animal model of myositis at the site of inflammation, reduction of intake was observed by Weissleder et al. (1991). Other approaches using antibodies and superparamagnetic particles are reported, for instance (Artemov et al. 2003a,b) (Fig 16.4.1). 16.4.2  Protein Amplification Kresse et al. (1998) demonstrated in 1996 the concept of amplification by receptor-mediated endocytosis. First in

Her-2/neu receptors can be monitored using a two-step labeling protocol: (1) the receptors are prelabeled with biotinylated humanized mAb (Herceptin), and (2) the streptavidin SPIO T2* MR contrast agent is selectively bound to the relabeled receptors (Artemov et al. 2003a, b)

1399

1400

16  Molecular Imaging

vivo experiments using Dextran-coated SPIO particles as signal generators were performed. As aforementioned, it was previously known that intravenously injected SPIO particles accumulate preferentially in macrophages (mainly of the liver and spleen), and a specific uptake in any tumor had not been observed. For these experiments they coupled transferrin to Dextran-coated SPIOs to make these particles tumor specific. This iron-chelating protein transferrin molecule was selected because it is well known that the iron-loaded transferrin is recognized by the transferring receptor and incorporated into the cells by receptor-mediated endocytosis (Fig. 16.4.1). This process starts over and over until a loop back mechanism stops the iron uptake. By this process an amplification of the iron concentration in the cell is achieved. The concentration in the cells—here tumor cells of experimental SMT (rat mammary carcinoma), which overexpress the transferring receptor—is high enough to be detected in MRI in an experimental rat tumor as demonstrated Fig 16.4.2. At least this experiment demonstrated that the sensitivity gap of MR can be bridged.

Moore et al. (2001) used the same transferring system to imaging transgenic expression by using transferrin receptor as a marker gene. Instead of using USPIO they conjugated transferrin and monocrystalline iron oxide.

They used a cell line that overexpresses the so-called engineered human transferrin receptor (cf. Fig. 16.4.4). The increase in receptor expression can be imaged in vivo by using MION–transferrin conjugates as reporters (Fig. 16.4.5). Up to now it is not doubtlessly proven that a linear correlation between receptor expression and changes in MR signal intensity exist. This holds true not only for this particular experiment, but also for other experiments using reporters. Such type of experiment was not reported for MRI in the past. Yet, a very elegant approach for MRI was published by Louie et al. (2000), in which they demonstrated the in vivo visualization of gene expression using magnetic resonance imaging. They prepared a “switchable” MRI contrast agent—called EgadMe—in which the access of water to the paramagnetic ion is blocked with the molecule galactopyranose, and in this state does not strongly modulate T1. β-Galactosidase enzymatically cleaves the galactopyranose from the chelate cage in which the paramagnetic Gd is bound, and allows free access to the paramagnetic relaxation center (cf. Fig. 16.4.6). The measured relaxivities of the two states of the contrast agent differ by a factor of 3. The contrast agent was used to follow β-galactosidase expression in living Xenopus laevis embryos by in vivo MR imaging. At the two-cell stage of the embryo, the two cells were injected with EgadMe. Messenger RNA for β-galactosidase was injected only to one of the two cells, and this was true also for mRNA for nuclear-localized green fluorescent protein (nGFP) used

Fig. 16.4.2  Receptor-mediated endocytosis (example: transferrin). Iron-loaded transferrin is recognized by a glycoprotein of the cell surface (transferrin receptor) and is incorporated into the cell by receptor-mediated endocytosis. The iron is released from the transferrin, recycled, and the receptor is again expressed at the surface. Studies of the transferrin pathways

show that the coupling of transferrin latex minibeads up to 350 nm does not affect the cellular uptake or biological activity of transferrin, which is also true for the coated iron oxide particles. Quantitative investigations have shown that the whole transferrin cycle takes less than 5 min, resulting in a very high turnover rate of about 20,000 transferrin molecules per cell every minute

16.4.3  Reporters

16.4  Applications

Fig. 16.4.3a,b  Frontal view experimental SMT/2A tumorbearing rats (rat mammary carcinoma). MR images showing changes in signal intensities of tumors after intravenous (i.v.) injection of 200 µmol/kg body weight of USPIO (a) and (b), and tumor-specific transferrin USPIO. Precontrast images on the left and postcontrast images at 150 min after injection on the right. (Bruker Biospec 2.35 T, RARE TR/TE = 2,775 ms/20 ms,

RARE factor = 8, NA = 8, FOV = 150 mm2, matrix = 256 × 256). Almost no change in signal intensity after injection of unspecific USPIO (a), the signal difference obtained before and after injection of tumor-specific transferrin USPIO between is color coded in the post contrast image (Kresse (1998) et al, Magn Reson Med.; 40(2): 236–242, modified)

Fig. 16.4.4  A cell line (rat 9L gliosarcoma) was stably transfected with a mutant form of human transferrin receptor. The cells constitutively overexpresses the so-called engineered human transferrin receptor (ETR). The downregulation of the

transferrin receptor is not suppressed when sufficient iron is in the cell. Transferrin was coupled to monocrystalline iron oxide nanoparticles (MIONs). The resulting increase in receptor expression can be imaged using the MION as reporters

1401

1402

16  Molecular Imaging Fig. 16.4.5  In vivo MR imaging of a mouse with engineered transferrin receptor–positive (left) and engineered transferring receptor–negative (right) flank tumors. Composite image of a T1-weighted spin-echo image obtained for anatomic detail with superimposed R2 changes after administration of transferrin coupled monocrystalline iron oxide nanoparticles, as a color map (Weissleder et al. (2000) Nat Med 6:351, modified)

Fig. 16.4.6a,b  Schematic of the transition of EgadMe ({1-[2(ß-galactopyranosyloxy)propyl]-4,7,10-tris[carboxymethyl]1,4,7,10-tetraazacyclododecane}gadolinium[III]). from a weak to a strong relaxivity state. a Schematic diagram representing the site-specific placement of the galactopyranosyl ring on the tetraazamacrocycle (side view). Upon cleavage of the sugar residue by β-galactosidase (at red bond), an inner sphere coordination

site of the Gd3+ ion becomes more accessible to water. b Spacefilling molecular model (top view, from above the sugar residue) of the complex before (left) and after cleavage by the β-galactosidase (right), illustrating the increased accessibility of the Gd3+ ion (magenta) upon cleavage. White atoms hydrogen, red atoms oxygen, blue atoms nitrogen, gray atoms carbon (Louie A et al. (2000) Nat Biotech 18: 321, modified)

16.4  Applications

Fig. 16.4.7  Schematics of β-galactosidase expression. At the two-cell stage of Xenopus laevis embryos, the two cells were injected with EgadMe. Messenger RNA for β-galactosidase (plasmid carrying the lacZ gene) was injected only to one of the two cells, and this was true also for mRNA for nuclear-localized

green fluorescent protein (nGFP) (not shown) to be used for independent tracking. β-Galactosidase and nGFP is expressed only in the left subset of the cells. EgadMe is enzymatically processed only in the left subset of cells, and relaxation enhancement occurs only in this subset of cells

for independent tracking. (Fig. 16.4.7). EgadMe was present in all cells of the embryo, whereas β-galactosidase and nGFP was expressed only in a subset of the cells (in Fig. 16.4.7 shown to the left). Only in those cells where β-galactosidase is present, is EgadMe expected to be enzymatically processed and in turn, the contrast in MR imaging should be enhanced. GFP is co-localized in the same subset of cells, making possible optical tracking of the gene expression (Fig. 16.4.8).

16.4.5  Cell Tracking

16.4.4  Smart Probes Smart probes have been described for use in optical imaging, in which the quenching effect is facilitated to suppress fluorescence as long as activation does not take place (see above). The use of this effect was impressively demonstrated by Bremer et al. (2001a, 2001b). They used specifically designed optical agents activatable by means of proteinase (MMP-2)-induced conversion into fluorescent products for non-invasive in vivo monitoring of enzyme activity (Figs. 16.4.9, 16.4.10).

In progression and defense of various diseases (local infections, autoimmunological disorders, graft rejection, tissue repair, and cancer), cell migration plays an important role. Cellular and humoral immune response is caused by local interplay between lymphocytes, monocytes, and granulocytes. Progenitor cells are involved in tissue remodeling and repair; they used for therapy in two respects: first, cells of the immune system as specific mediators of immune response, and second, progenitor cells can be used for the renewal of stable tissues, for instance, transplantation of progenitor cells into the myocardium can improve contractibility as well as the end systolic volume (Dimmeler et al. 2005). At present it is assumed that cell-based therapies have a high potential; however, most of these treatments are still in an experimental and/or in an early clinical and not approved. For these approaches, it is highly desirable to monitor the noninvasively assess cell migration, to prove correct injection, and to monitor the presence and organization of the transplanted cells. To do so, in vivo imaging techniques are required. Nu-

1403

1404

16  Molecular Imaging

clear medicine techniques, optical imaging, and MRI are considered to be appropriate and are being developed for such cell tracking approaches. In the following, the uses of these three methods are briefly described and selected examples are given. 16.4.5.1  Radioactive Labeling As early as 1984, cell tracking was performed by McAfee et al. (1984) using radiolabeled leucocytes to visualize and localize inflammatory foci in patients. Radiolabeling of cells developed to a standard method and is used nowadays as reference method in experiments, in particular, because quantification is possible.

radioactivity is involved, making handling more comfortable; however, quantification in vivo is still lacking, making inter-individual comparison almost impossible. Furthermore, in the case of proliferating cells a dilution of the dyes occurs, and after several cell doublings, signal is diluted. External labeling can be done by the aforementioned dyes and QDs. Cyanine dyes and other fluorochromes are widely used for in vitro and in vivo experiments, and numerous experiments are published; although only two authors are cited: Bremer et al. (2001a,

16.4.5.2  Optical Labeling Cell labeling for optical imaging can be performed in two different ways, first, by labeling cells with external dyes and second, using transgenic cells, which express heterologous fluorescent proteins or enzymes activating a fluorescent probe. Direct labeling with external dyes can be performed easily. In contrast to nuclear medicine techniques, no

Fig. 16.4.8a,b  MRI detection of lacZ gene expression using the activable contrast agent EgadMe. a MR image of living embryo injected with plasmids carrying the lacZ gene. A plasmid carrying the lacZ gene was injected to one cell at the two-cell stage, and subsequent enzyme expression was on the left side of the embryo as shown. Regions of high signal intensity are found in the bright stripe of endoderm, regions of the head, and ventrally, including two distinct spots (red arrows) found just ventral to the cement gland. b Bright-field image of same embryo fixed and stained with X-gal. Whole-mount cytochemistry shows that regions demonstrating enzyme expression correlate with the regions of high intensity in the MR image (Louie A et al (2000); Nat Biotech 18: 321, modified)

Fig. 16.4.9a,b  Schematics of the design of the MMP-2–sensitive fluorescent probe for in vivo near-infrared fluorescence (NIRF) imaging. Fluorochromes with excitation and emission wavelengths in the near-infrared spectrum were covalently coupled to a poly-l-lysine backbone (—Lys-Lys-Lys…—) sterically protected by methoxy polyethylene glycol (MPEG) side chains by means of a synthetic MMP-2 peptide substrate. a Owing to the proximity of the fluorochromes, fluorescence resonance energy transfer occurs so that almost no fluorescent signal can be detected in the non-activated state. b After MMP-2 cleavage of the peptide spacer, fluorochromes are released from the carrier and become brightly fluorescent (Bremer C et al. (2001) Nat Med 7(6): 743–748)

16.4  Applications

b) and Becker et al. (2000, 2001). Initial in vivo experiments after subcutaneous implantation of QD-labeled tumor cells in mice where performed by Gao et al. (2004) indicated that a subcutaneous deposit of less than 1,000 cells can be visualized, which, in this experiment, was not feasible for green fluorescent protein positive cells. In vitro labeling of proliferating or differentiating cells by heterologous expression with a bioluminescence dye, e.g., GFP, is common practice and is a versatile tool to label cells, which can be used in cell experiments or after injection of the cells in luminescent small animals (Mothe et al. 2005). The use is limited by the by the low tissue penetration of green light, and the quantitative evaluation of the data is also limited, especially in larger animals. These limitations are less serious using red fluorescent protein (RFP), with a longer wavelength and concomitant greater penetration depth. As it was shown by Yamamoto et al. (2004) GFP and RFP can be used for double labeling of cells and even different cell compartments, allowing study mitosis and apoptosis as well as the deformation of cells and cytoplasm during migration in the vasculature (Yamauchi et al. 2005). Transgenic expression of luciferase can be used to visualize such cells in vivo and is frequently used to follow tumor growth and spread in rodents (Contag et al. 2000). Further cell labels frequently used are the enzymes lacZ, which encodes the β-galactosidase, and the human placental alkaline phosphatase (hPAP).

16.4.5.3  Magnetic Labeling When SPIOs were introduced as contrast agents for MR they were used as liver contrast agents. The particles are rapidly phagocytized by the liver’s Kuppfer cells, whereas malignant tumors (hepatocellular carcinomas or metastases) do not contain Kuppfer cells and do not take up particles. The phagocytized particles cause a signal void in normal liver tissue, whereas the tumor tissue keeps precontrast signal intensity is sharply delineated from the dark liver tissue. High uptake of SPIO is observed not only in the liver, but also in the spleen, and bone marrow due to the phagocytosis of the particles by cells of the reticuloendothelial system (RES). Lymph node and plaque imaging is possible on the same principle, and this was demonstrated experimentally and in patients (Taupitz et al 1993a, b; Memarsadeghi et al. 2006); this holds also true also for thrombus imaging in experimental animals (Schmitz et al. 2001) and clinical study (Li et al. 2007). Clinical applications for liver imaging and lymph node imaging as well imaging of atherosclerotic plaques are mentioned in the relevant clinical chapters of this book. In contrast to the labeling in vivo used in the above mentioned studies, labeling in vitro might be more attractive, more versatile, and suitable for monitoring cell therapies and performing disease-specific diagnosis in vivo after injection of in vitro labeled cells. Several groups

Fig. 16.4.10  In vivo near-infrared fluorescence (NIRF) imaging of HT1080 tumor-bearing animals. The top row shows raw image acquisition obtained at 700-nm emission. Untreated (left), treated with 150 mg/kg prinomastat, twice a day, intraperitoneally for 2 days (right). The bottom row shows color-coded tumoral maps of MMP-2 activity superimposed onto white-light images (left no treatment, right prinomastat treatment) (Bremer C et al. (2001)Nat Med 7(6): 743–748)

1405

1406

16  Molecular Imaging Table 16.4.2  Particles for cell labeling (Kießling 2007) Particles

Size

Coating

Cell type

MPIO

0.9-5.8 µm

divinyl benzene/styrene

Hepatocytes Embryonic fibroblasts MSC

Feridex (SPIO)

~ 150 nm

Dextran

Monocytes

Ferumoxides and MION-46L

~ 150 nm 8-20 nm

Dextran

Lymphocytes MSC CG-4 cells Cervix cancer cells

SHU 555ª SHU 555c

45-65 nm <40 nm

Dextran

Progenitor cells Fibroblasts Hepatocarcinoma cells

Magnetic polysaccharides

~ 50 nm

Dextran

Hematopoietic progenitor cells Dendritic cells

CLIO

~ 45 nm

Dextran

Lymphocytes CD34+ prog. cells

Anionic maghemite nanoparticles

~ 35 nm

DMSA

Macrophages HeLa

Ferumoxtran

~ 35 nm

Dextran

Embryonic stem cells Muscle stem cells MSC

AquaMag100 BMS180549

~35 nm

Dextran

T-cells

MION-46L

8–20 nm

Dextran

Oligodendrocytes Neural precursor cells T-cells

MION

8–20 nm

Dextran

C6 tumor cell

MION

12–14 nm

PEG-phospholipid

Fibroblasts Macrophages

VSOP-C125

~ 8 nm

citrate

Macrophages

MD-100

7–8 nm

dendrimers

Stem cells Neural progenitor cells Olfactory ensheathing glia Muscle stem cells

were demonstrating the high potential of this approach by visualizing less than 2,000 labeled cells after injection. These experiments were performed at very high fields at 9.4 T (Höhn et al. 2002) and 17.6 T (Stroh et al. 2005). Other groups claim that even a single cell can be observed at 1.5 T (Foster-Gareau et al. 2003). In vitro studies demonstrated the uptake of superparamagnetic iron oxide particles not only by the cells of the RES, but also other cells. Furthermore, the uptake

was shown to be dependent not only on the type of cell (macrophages, lymphocytes, fibroblast, progenitor cells, etc.), but also on the particle size (SPIO (≥ ~50 nm), USPIO, VSPIO (≤ ~10 nm) etc.), coating, surface charge, opsonization, and other parameters such as pH, etc. An overview concerning labeling of cells is given in Table 16.4.2. Phagocytosis appears to be the dominant uptake mechanism for large particles (SPIO) and pinocytosis for small particle like USPIO and VSOP (Kießling 2007).

16.4  Applications

To increase the uptake of superparamagnetic particles in cells, other approaches were carried out, like the enhanced uptake–mediated by human immunodeficiency virus transactivator transcription (HIV-Tat) protein. Higher uptake of transferrin-coated nanoparticles in the cell could be obtained if particles were labeled with the Tat protein (Josephson et al. 1999). Another approach is to incorporated SPIOs into liposomal structures mediating crossing the cell membrane and releasing particles intracellularly (Frank et al. 2002). To avoid the “negative” contrast produced by the SPIOs, “positive” contrast media were used. Efficient labeling of cells was shown using liposomes containing paramagnetic and fluorescent metalloporphyrins (Daldrup-Link et al. 2004) as well as Gd-containing lipophilic nanoparticles (Vuu et al. 2005), which allows a label in vitro. Yet another elegant method labeling cells was published by Ahrens et al. (2005). Dendritic cells were labeled with perfluoropolyether (PFPE) and injected into mice. Facilitating 19F MRI for the detection of these cells, almost background-free images can be obtained due to the fact that almost no detectable endogenous 19F exists. Only recently, Giesel et al. (2006) and Hennig et al. (2007) used gadofluorine M and could achieve uptake of the agents into cells and a sufficient contrast in experiments in vitro. As mentioned already above for labeling proliferating cells (e.g., stem cells) with dyes in vitro, disadvantages of labeling method as such are the dilution effect of the paramagnetic substances or superparamagnetc particles expected due to cell doubling. Furthermore, the viability as well the functionality of the cells can be influenced negatively by high intracellular load of particles necessary to achieve sufficient contrast. For the time being and the near future, cell tracking with MRI will be still a tool for preclinical research, even though in several review articles clinical application is envisioned to be a clinical imaging modality soon (Graham et al. 2006). This is different using radioactive labeling, where tracking is routine as in leukocyte scintigraphy, or can be easily realized for tracking stem cells injected into hearts to treat myocardial infarction (Bengel 2006).

and perfusion could be measured in animal experiments using 13C (Ishii et al. 2007). However, using hyperpolarized atoms as signal generators, direct imaging with specific probes becomes visible only with 129Xe as suggested by Schröder et al. (2006). The realization of this concept for specific imaging in vivo will be carried out in the near future. References 1.

2.

3.

4.

5.

6.

7.

8. 9.

16.4.6  Imaging Using Hyperpolarization Hyperpolarized gases (e.g., 3He) were firstly used in volunteers to visualize the airways and the ventilation of the lung (Ebert et al. 1996). Clinical studies followed immediately and demonstrated the potential of the method for patients to measure the function (ventilation) (Kauczer et al. 1996), lung development (Altes et al. 2006), and lung disease (cystic fibrosis) and its therapy monitoring (Mentore et al. 2005). After the development of the parahydrogen–induced polarization techniques, 13C was used outside of a physics laboratory (Goldman et al. 2005) to image cerebral perfusion and pulmonary vasculature,

10.

11.

12. 13.

Ahrens ET, Feili-Hariri M, Xu H et al (2003) Receptor-mediated endocytosis of iron-oxide particles provides efficient labeling of dendritic cells for in vivo MR imaging. Magn Reson Med 49:1006–1013 Ahrens ET, Flores R, Xu H, Morel PA (2005) In vivo imaging platform for tracking immunotherapeutic cells. Nat Biotechnol 23:983–987 Altes TA, Mata J, de Lange EE, Brookeman JR, Mugler JP III (2006) Assessment of lung development using hyperpolarized helium-3 diffusion MR imaging. J Magn Reson Imaging. 24:1277–1283 Artemov D, Mori N, Ravi R, Bhujwalla ZM. (2003) Magnetic resonance molecular imaging of the HER-2/neu receptor. Cancer Res. 63:2723-2727 Becker A, Hessenius C, Bhargava S et al (2000) Cyanine dye labeled vasoactive intestinal peptide and somatostatin analog for optical detection of gastroenteropancreatic tumors. Ann N Y Acad Sci 921:275–278 Becker A, Hessenius C, Licha K, Ebert B, Sukowski U, Semmler W, Wiedenmann B, Grotzinger C. (2001) Receptor-targeted optical imaging of tumors with near-infrared fluorescent ligands. Nat Biotechnol. 19:327-331 Bengel FM (2006) Nuclear imaging in cardiac cell therapy. Heart Fail Rev 11:325–332 Beyer T, Townsend DW, Blodgett TM (2002) Dual-modality PET/CT tomography for clinical oncology. Q J Nucl Med 46:24–34 Blodgett TM, Meltzer CC, Townsend DW (2007) PET/CT: form and function. Radiology 242:360–385 Bremer C, Bredow S, Mahmood U, Weissleder R, Tung CH (2001a) Optical imaging of matrix metalloproteinase-2 activity in tumors: feasibility study in a mouse model. Radiology 221:523–529 Bremer C, Tung CH, Weissleder R (2001b) In vivo molecular target assessment of matrix metalloproteinase inhibition. Nat Med 7:743–748 Bremer C, Bredow S, Mahmood U et al (2001) Optical imaging of matrix metalloproteinase-2 activity in tumors: feasibility study in a mouse model. Radiology 221:523–529 Brody EN, Gold L (2000) Aptamers as therapeutic and diagnostic agents J Biotechnol 74:5–13 Bulte JW, Hoekstra Y, Kamman RL, Magin RL, Webb AG, Briggs RW, Go KG, Hulstaert CE, Miltenyi S, The TH, de Leij L (1992) Specific MR imaging of human lymphocytes by monoclonal antibody-guided dextran-magnetite particles. Magn Reson Med 25:148–157

1407

1408

16  Molecular Imaging 14. Cerdan S, Lotscher HR, Kunnecke B, Seelig J (1989) Monoclonal antibody-coated magnetite particles as contrast agents in magnetic resonance imaging of tumors. Magn Reson Med 12:151–163 15. Colegrove FD, Schearer LD, Walters KG (1963) Polarization of 3He gas by optical pumping. Phys Rev Lett 132:2561–2572 16. Contag CH, Jenkins D, Contag PR, Negrin RS (2000) Use of reporter genes for optical measurements of neoplastic disease in vivo. Neoplasia 2:41–52 17. Czernin J, Phelps ME (2002) Positron emission tomography scanning: current and future applications. Annu Rev Med 53:89–112 18. Daldrup-Link HE, Rudelius M, Metz S et al (2004). Cell tracking with gadophrin-2: a bifunctional contrast agent for MR imaging, optical imaging, and fluorescence microscopy. Eur J Nucl Med Mol Imaging 31:1312–1321 19. Deserno WM, Harisinghani MG, Taupitz M, Jager GJ, Witjes JA, Mulders PF, Hulsbergen van de Kaa CA, Kaufmann D, Barentsz JO (2004) Urinary bladder cancer: preoperative nodal staging with ferumoxtran-10-enhanced MR imaging. Radiology. 233:449–456 20. Dimmeler S, Zeiher AM, Schneider MD (2005) Unchain my heart: the scientific foundations of cardiac repair. J Clin Invest 115:572–583 21. Ebert M, Grossmann T, Heil W, Otten WE, Surkau R, Leduc M, Bachert P, Knopp MV, Schad LR, Thelen M (1996) Nuclear magnetic resonance imaging with hyperpolarised helium-3. Lancet 347:1297–1299 22. Ellegala DB, Leong-Poi H, Carpenter JE, Klibanov AL, Kaul S, Shaffrey ME, Sklenar J, Lindner JR (2003) Imaging Tumor Angiogenesis with Contrast Ultrasound and Microbubbles Targeted to αvβ3. Circulation 108:336–341 23. Ende G, Bachert P (1993) Dynamic 13C-{1H} nuclear polarization of lipid methylene resonances applied to broadband proton-decoupled in vivo 13C MR spectroscopy of human breast and calf tissue. Magn Reson Med 30:415–23 24. Foster-Gareau P, Heyn C, Alejski A, Rutt BK (2003) Imaging single mammalian cells with a 1.5 T clinical MRI scanner. Magn Reson Med 49:968–971 25. Frank JA, Zywicke H, Jordan EK et al (2002). Magnetic intracellular labeling of mammalian cells by combining (FDA-approved) superparamagnetic iron oxide MR contrast agents and commonly used transfection agents. Acad Radiol 9(Suppl 2):484–487 26. Gao X, Cui Y, Levenson RM, Chung LW, Nie S (2004) In vivo cancer targeting and imaging with semiconductor quantum dots. Nat Biotechnol 22:969–976 27. Gibson AP, Hebden JC, Arridge SR (2005) Recent advances in diffuse optical imaging. Phys Med Biol 50:R1–R43 28. Giesel FL, Stroick M, Griebe M, Troster H, von der Lieth CW, Requardt M, Rius M, Essig M, Kauczor HU, Hennerici MG, Fatar M (2006) Gadofluorine m uptake in stem cells as a new magnetic resonance imaging tracking method: an in vitro and in vivo study. Invest Radiol 41:868–873

29. Goldman M, Johannesson H, Axelsson O, Karlsson M (2005) Hyperpolarization of 13C through order transfer from parahydrogen: a new contrast agent for MRI. Magn Reson Imaging 23:153–157 30. Golman K, Olsson LE, Axelsson O, Mansson S, Karlsson M, Petersson JS (2003) Molecular imaging using hyperpolarized 13C. Br J Radiol 76: S118–S127 31. Graham JJ, Lederman RJ, Dick AJ (2006) Magnetic resonance imaging and its role in myocardial regenerative therapy. Regen Med 1:347–355 32. Grodd W, Krageloh-Mann I, Petersen D et al (1990) In vivo assessment of N-acetylaspartate in brain in spongy degeneration (Canavan‘s disease) by proton spectroscopy. Lancet 336:437–438 33. Hell SW (2003)Toward fluorescence nanoscopy. Nat Biotechnol 21:1347–55 34. Henning TD, Saborowski O, Golovko D, Boddington S, Bauer JS, Fu Y, Meier R, Pietsch H, Sennino B, McDonald DM, Daldrup-Link HE (2007) Cell labeling with the positive MR contrast agent Gadofluorine M. Eur Radiol 17:1226–1234 35. Hielscher AH (2005) Optical tomographic imaging of small animals. Curr Op Biotechnol 16:79–88 36. Hintz SR, Cheong WF, van Houten JP, Stevenson DK, Benaron DA (1999) Bedside imaging of intracranial hemorrhage in the neonate using light: comparison with ultrasound, computed tomography, and magnetic resonance imaging. Pediatr Res 45:54–59 37. Hoehn M, Kustermann E, Blunk J, Wiedermann D, Trapp T, Wecker S, Focking M,Arnold H, Hescheler J, Fleischmann BK, Schwindt W, Buhrle C (2002) Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance imaging investigation of experimental stroke in rat. Proc Natl Acad Sci U S A 99:16267–16272 38. Högemann D, Josephson L, Weissleder R et al (2000) Improvement of MRI probes to allow efficient detection of gene expression. Bioconjug Chem 11:941–946 39. Ishii M, Emami K, Kadlecek S, Petersson JS, Golman K, Vahdat V, Yu J, Cadman RV, MacDuffie-Woodburn J, Stephen M, Lipson DA, Rizi RR (2007) Hyperpolarized 13C MRI of the pulmonary vasculature and parenchyma. Magn Reson Med 57:459–463 40. Johansson E, Mansson S, Wirestam R, Svensson J, Petersson JS, Golman K, Stahlberg F (2004) Cerebral perfusion assessment by bolus tracking using hyperpolarized 13C. Magn Reson Med 51:464–72 41. Josephson L, Tung CH, Moore A, Weissleder R (1999) High-efficiency intracellular magnetic labeling with novel superparamagnetic-Tat peptide conjugates. Bioconjug Chem 10:186–191 42. Kalender WA (2005) Computed tomography: fundamentals, system technology, image quality, applications, 2nd edn. Publicis, Erlangen, Germany

16.4  Applications 43. Kauczor HU, Hofmann D, Kreitner KF, Nilgens H, Surkau R, Heil W, Potthast A,Knopp MV, Otten EW, Thelen M (1996) Normal and abnormal pulmonary ventilation: visualization at hyperpolarized He-3 MR imaging. Radiology 201:564–568 44. Kießling F, Heilmann M, Vosseler S et al (2003) Dynamic T1-weighted monitoring of vascularization in human carcinoma heterotransplants by magnetic resonance imaging. Int J Cancer 104:113–120 45. Kießling F, Huber PE, Grobholz R et al (2004) Dynamic magnetic resonance tomography and proton magnetic resonancespectroscopy of prostate cancers in rats treated by radiotherapy. Invest Radiol 39:34–44 46. Kießling F (2007) Non-invasive cell tracking. In: Semmler W, Schwaiger M (eds) Handbook of experimental pharmacology, molecular imaging. Springer, Berlin Heidelberg New York 47. Kresse M, Wagner S, Pfefferer D et al (1998) Targeting of ultrasmall superparamagnetic iron oxide (USPIO) particles to tumor cells in vivo by using transferrin receptor pathways. Magn Reson Med 40:236–242 48. Krix M, Kiessling F, Vosseler S et al (2003) Sensitive noninvasive monitoring of tumor perfusion during antiangiogenic therapy by intermittent, bolus-contrast power doppler sonography. Cancer Res 63:8264–8270 49. Kruger RA, Miller KD, Reynolds HE, Kiser WL Jr, Reinecke DR, Kruger GA (2000) Breast cancer in vivo: contrast enhancement with thermoacoustic CT at 434 MHz-feasibility study. Radiology. 216:279–283 50. Kruger RA, Kiser WL, Reinecke DR, Kruger GA, Miller KD (2003) Thermoacoustic molecular imaging of small animals. Mol Imaging 2:113–123 51. Kurth CD, Steven JM, Benaron D, Chance B (1993) Nearinfrared monitoring of the cerebral circulation. J Clin Monit. 9:163–170 52. Lawaczeck R, Menzel M, Pietsch H (2004) Superparamagnetic iron oxid particles: contrast media for magnetic resonance imaging. Appl. Organometal. Chem. 18:506–513 53. Li W, Salanitri J, Tutton S, Dunkle EE, Schneider JR, Caprini JA, Pierchala LN, Jacobs PM, Edelman RR (2007) Lower extremity deep venous thrombosis: evaluation with ferumoxytol-enhanced MR imaging and dual-contrast mechanism—preliminary experience. Radiology 242:873–881 54. Licha K, Riefke B, Ntziachristos V, Becker A, Chance B. Semmler W (2000) Hydrophilic cyanine dyes as contrast agents for near-infrared tumor imaging: synthesis, photophysical properties and spectroscopic in vivo characterization. Photochem Photobiol 72:392–398 55. Lim YT, Kim S, Nakayama A, Stott NE, Bawendi MG, Frangioni JV (2003) Selection of quantum dot wavelengths for biomedical assays and imaging. Mol Imaging 2:50–64 56. Louie AY, Huber MM, Ahrens ET, Rothbacher U, Moats R, Jacobs RE, Fraser SE, Thomas Meade J (2000) In vivo visualization of gene expression using magnetic resonance imaging. Nat Biotech 18:321–325

57. Massoud T, Gambhir SS (2003) Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev17:545–580 58. McAfee JG, Subramanian G, Gagne G (1984) Techniques of leukocyte harvesting and labelling: problems and perspectives. Semin Nuclear Med 2:83–106 59. Medintz IL, Uyeda HT, Goldman ER, Mattoussi H (2005) Quantum dot bioconjugates for imaging, labelling and sensing. Nat Mater 2005 4:435–46 60. Memarsadeghi M, Riedl CC, Kaneider A, Galid A, Rudas M, Matzek W, Helbich TH (2006) Axillary lymph node metastases in patients with breast carcinomas: assessment with nonenhanced versus uspio-enhanced MR imaging. Radiology 241:367–377 61. Mentore K, Froh DK, de Lange EE, Brookeman JR, PagetBrown AO, Altes TA (2005) Hyperpolarized 3He MRI of the lung in cystic fibrosis: assessment at baseline and after bronchodilator and airway clearance treatment. Acad Radiol 12:1423–9 62. Moore A, Basilion JP, Chiocca EA, Weissleder R (1998) Measuring transferrin receptor gene expression by NMR imaging. Biochim Biophys Acta 1402:239–249 63. Moore A, Josephson L, Bhorade RM, Basilion JP, Weissleder R (2001) Human transferrin receptor gene as a marker gene for MR imaging. Radiology 221:244–250 64. Mothe AJ, Kulbatski I, van Bendegem RL, Lee L, Kobayashi E, Keating A, Tator CH (2005). Analysis of green fluorescent protein expression in transgenic rats for tracking transplanted neural stem/progenitor cells. J Histochem Cytochem 53:1215–1226 65. Nakajima M, Takeda M, Kobayashi M, Suzuki S, Ohuchi N (2005) Nano-sized fluorescent particles as new tracers for sentinel node detection: experimental model for decision of appropriate size and wavelength. Cancer Sci 96:353–356 66. Nguyen BC, Stanford W, Thompson BH, Rossi NP, Kernstine KH, Kern JA, Robinson RA, Amorosa JK, Mammone JF, Outwater EK (1999) Multicenter clinical trial of ultrasmall superparamagnetic iron oxide in the evaluation of mediastinal lymph nodes in patients with primary lung carcinoma. J Magn Reson Imaging 10:468–473 67. Ntziachristos V, Tung CH, Bremer C, Weissleder R (2002) Fluorescence molecular tomography resolves protease activity in vivo. Nat Med 8:757–760 68. Reinhardt M, Hauff P, Briel A, Uhlendorf V, Linker R, Mäurer M, Schirner M (2005) Sensitive particle acoustic quantification (SPAQ): a new ultrasound-based approach for the quantification of ultrasound contrast media in high concentrations. Invest Radiol 40:2–7 69. Santra S, Zhang P, Wang K, Tapec R, Tan W (2001) Conjugation of biomolecules with luminophore-doped silica nanoparticles for photostable biomarkers. Anal Chem 73:4988–4993 70. Schmitz SA, Winterhalter S, Schiffler S, Gust R, Wagner S, Kresse M, Coupland SE, Semmler W, Wolf KJ (2001) USPIO-enhanced direct MR imaging of thrombus: preclinical evaluation in rabbits. Radiology 221:237–243

1409

1410

16  Molecular Imaging 71. Schröder L, Lowery TJ, Hilty C, Wemmer DE, Pines A (2006) Molecular imaging using a targeted magnetic resonance hyperpolarized biosensor. Science 314(5798):446–449 72. Shaner NC, Steinbach PA, Tsien RY (2005) A guide to choosing fluorescent proteins. Nat Methods 2:905–909 73. Stroh A, Faber C, Neuberger T, Lorenz P, Sieland K, Jakob PM, Webb A, Pilgrimm H, Schober R, Pohl EE, Zimmer C (2005) In vivo detection limits of magnetically labeled embryonic stem cells in the rat brain using high-field (17.6 T) magnetic resonance imaging. Neuroimage 2005 24:635–645 74. Tai YC, Ruangma A, Rowland D, Siegel S, Newport DF et al (2005). Performance evaluation of the microPET focus: a third-generation microPET scanner dedicated to animal imaging. J Nucl Med 46:455–63 75. Taupitz M, Wagner S, Hamm B, Binder A, Pfefferer D (1993a) Interstitial MR lymphography with iron oxide particles: results in tumor-free and VX2 tumor-bearing rabbits. AJR Am J Roentgenol 161:193–200 76. Taupitz M, Wagner S, Hamm B, Dienemann D, Lawaczeck R, Wolf KJ (1993b) MR lymphography using iron oxide particles. Detection of lymph node metastases in the VX2 rabbit tumor model. Acta Radiol 34:10–15 77. Thakur M, Lentle BC (2005) Report of a summit on molecular imaging. Radiology 236:753–755 78. Tiefenauer LX, Kuhne G, Andres RY (1993) Antibodymagnetite nanoparticles: in vitro characterization of a potential tumor-specific contrast agent for magnetic resonance imaging. Bioconjug Chem 4:347–352 79. Tiemann K, Pohl C, Schlosser T, Goenechea J, Bruce M, Veltmann C, Kuntz S, Bangard M, Becher H (2000) Stimulated acoustic emission: pseudo-Doppler shifts seen during the destruction of non-moving microbubbles. Ultrasound Med Biol 26:1161–1167 80. Townsend DW, Carney JP, Yap JT, Hall NC (2004) PET/CT today and tomorrow. J Nucl Med 45(Suppl 1):4S–14S 81. Vuu K, Xie J, McDonald MA et al (2005). Gadoliniumrhodamine nanoparticles for cell labeling and tracking via magnetic resonance and optical imaging. Bioconjug Chem 16:995–999

82. Wang Y, Seidel J, Tsui BM, Vaquero JJ, Pomper MG (2006). Performance evaluation of the GE healthcare eXplore VISTA dual-ring small-animal PET scanner. J Nucl Med 47:1891–900 83. Weissleder R, Moore A, Mahmood U, Bhorade R, Benveniste H, Chiocca EA, Basilion JP (2000) In vivo magnetic resonance imaging of transgene expression.Nat Med. 6:351-355 84. Weissleder R (2002) Scaling down imaging: molecular mapping of cancer in mice. Nat Rev Cancer 2:11–18 85. Weissleder R, Mahmood U (2001) Molecular imaging. Radiology 219:316–333 86. Weissleder R, Ntziachristos V (2003) Shedding light onto live molecular targets. Nat Med 9:123–128 87. Weissleder R, Lee AS, Fischman AJ, Reimer P, Shen T, Wilkinson R, Callahan RJ, Brady TJ (1991) Polyclonal human immunoglobulin G labeled with polymeric iron oxide: antibody MR imaging. Radiology 181:245–249 88. Willig KI, Kellner RR, Medda R, Hein B, Jakobs S, Hell SW (2006) Nanoscale resolution in GFP-based microscopy. Nat Methods 3:721–3 89. Yamamoto N, Jiang P, Yang M, Xu M, Yamauchi K, Tsuchiya H, Tomita K, Wahl GM, Moossa AR, Hoffman RM (2004) Cellular dynamics visualized in live cells in vitro and in vivo by differential dual-color nuclear-cytoplasic fluorescent-protein expression. Cancer Res 64:4251–4256 90. Yamauchi K, Yang M, Jiang P, Yamamoto N, Xu M, Amoh Y, Tsuji K, Bouvet M,Tsuchiya H, Tomita K, Moossa AR, Hoffman RM (2005) Real-time in vivo dual-color imaging of intracapillary cancer cell and nucleus deformation and migration. Cancer Res 65:4246–4252 91. Zacharakis G, Shih H, Ripoll J, Weissleder R, Ntziachristos V. (2006) Normalized transillumination of fluorescent proteins in small animals. 5(3):153-159 92. Zimmer JP, Kim SW, Ohnishi S, Tanaka E, Frangioni JV, Bawendi MG (2006) Size series of small indium arsenidezinc selenide core-shell nanocrystals and their application to in vivo imaging. J Am Chem Soc 128:2526–2527

Chapter 17

Systems Biology and Nanotechnology

17

17.1 

Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   1411 M. S. Bradbury, H. Hricak, and J. R. Heath

17.5.3 

17.2 

Linking Cancer Biology, P4 Medicine, and Molecular Imaging  . . . . . . . . . . . . . .   1413

17.5.3.1  Biochips for Drug Discovery  .. . . . . . . . .   1424

17.3 

Application of a Systems Biology Approach to Cancer  . . . . . . . . . . . . . . . . .   1414

17.4 

Molecular Diagnostics and Personalized Therapies  for Cancer  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1418

17.4.1 

Cancer Biomarkers for Clinical Use   . . .   1419

17.5 

Role of Cancer Nanobiotechnology in Personalized Medicine  . . . . . . . . . . . .   1420

17.5.1 

Nanodevices as Biosensors  .. . . . . . . . . . .   1420

17.5.2 

Nanoparticles for Cancer Diagnostics and Therapeutics  . . . . . . . . .   1421

17.5.3.2  Dendrimers in Anticancer Therapy  .. . .   1424 17.5.3.3  Nanobodies as Personalized Medicines for Cancer  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1425 17.6 

Role of Radiology in Personalized Medicine  . . . . . . . . . . . .   1425

17.7 

Clinical and Near-Clinical  Molecular Imaging Applications  Using Targeted Probes  . . . . . . . . . . . . . . .   1426

17.7.1 

Oncologic Molecular Imaging for Detection of Nodal Metastases and Cancer Staging  .. . . . . . . . . . . . . . . . . .   1426

17.7.2 

Molecular Imaging of Dendritic Cell Migration to Regional Nodes for Tumor Immunotherapy   .. . . . . . . . . .   1428

17.7.3 

Molecular Imaging of Thrombosis in Atherosclerotic Disease  . . . . . . . . . . . .   1428

17.5.2.1  Magnetic Nanoparticles  . . . . . . . . . . . . . .   1421 17.5.2.2  Quantum Dots  .. . . . . . . . . . . . . . . . . . . . . .   1423 17.5.2.3  Nanoshells  . . . . . . . . . . . . . . . . . . . . . . . . . .   1423

Nanobiotechnology for Drug Discovery  .. . . . . . . . . . . . . . . . . .   1424

References  .. . . . . . . . . . . . . . . . . . . . . . . . . .   1430

17.1  Introduction M. S. Bradbury, H. Hricak, and J. R. Heath Explosive growth in the fields of molecular biology—particularly in the integrated fields of genomics, proteomics, and informatics (collectively known as systems biology)— is transforming our understanding of disease at the molecular level, and will begin transforming general medical practice within the coming decade into a medicine that is personalized, predictive, preventative, and participatory (i.e., “P4” medicine). This emerging molecular picture of disease is already driving the development of new classes of drugs that are targeted at the specific molecular errors that trigger the transformation from health to disease. As a general rule, these new drugs are effective only on patient subpopulations. This limitation is driving the inte-

gration of molecular therapeutics with in vivo and in vitro molecular diagnostics, which can be used for prescreening patients and/or for monitoring therapeutic responses. This, in turn, is driving the development of new in vitro and in vivo (molecular imaging) diagnostic technologies, with oncology turning into the proving ground for many of these new concepts. Within the context of P4 medicine, imaging will play a crucial role in the translation of these developments. The ability to probe molecular changes within tumor cells in preclinical models and translate this to the clinical setting, using imaging, promises to improve tumor detection and staging, facilitate treatment selection and monitoring, and determine prognosis. Consider how cancer was viewed just a few years ago. Most pathology practices were based on phenomenological examination of tissues, and diagnostic measurements were typically (and often still are) pauciparameter, rely-

1412

17  Systems Biology and Nanotechnology

Fig. 17.1.1  No one invariant sequence of genetic alterations can define the genetic paths followed by all esophageal carcinomas. A detailed study of human Barrett’s esophagus, which is a precursor lesion to esophageal carcinoma, has led to the discovery of a number of alternative genetic paths between the initial metaplastic cells that displace the normally present squamous cells of the esophagus (left) through dysplasia to the cancer clones (right) that may eventually arise. Changes listed include changes in chromosome number (including aneuploidy and tet-

raploidy), as well as loss of heterozygosity (LOH), and apparent loss of tumor suppressor gene expression through promoter methylation. Ki67+ positive for the S-phase Ki67 marker and actively proliferating, CDKN2A- mutation or promoter methylation of p16INK4A gene, TTP53- mutation of p53 gene, 9p- LOH of markers on the short arm of chromosome 9, etc., 2N diploid, 4N tetraploid (Reproduced with permission from Weinberg R [2006] The biology of cancer. Garland, New York)

ing on only a few measurements to assess disease. An increased molecular understanding of cancer has demonstrated that a given type of cancer (for example, breast cancer) can be triggered by different genetic mutations, each of which can lead to a different outcome (e.g., aggressive versus non-aggressive esophageal cancer). This has led to the model of cancer pathways (Weinberg 2006; Fig. 17.1.1). In this model, there are multiple pathways of interacting proteins, each constituting a cascade of molecular events. A given pathway, if genetically altered in specific ways, can lead to cancer. Specific molecular therapeutics are targeted against specific pathways, often targeting the genetically altered proteins. Diagnostic tools, such as immunohistochemical staining to identify phosphorylated (activated) proteins associated with particular pathways, or the identification of genetic defects via mRNA analysis, can be employed in certain cases to identify the genetically altered pathway and thus yield

a pathway-dependent diagnosis of the cancer. Such an analysis can potentially indicate an appropriate therapeutic regimen, and in vivo molecular imaging can then be employed as a diagnostic of drug efficacy. Pathway models are significantly limited, as they do not account for the dynamic evolution of cancer, and they dramatically underestimate the degree of interconnectivity amongst the various genes and proteins. Out of systems biology procedures (generally involving comprehensive transcriptome analysis, occasionally coupled with focused proteomic investigations, and integrated together through computational methods) are emerging network models (and dynamic network models) of disease (and disease progression). Such models are typically constructed to reflect how the onset and progression of disease is manifested in differentially expressed genes and perturbed regulatory networks. The current stateof-the-art network models are still fairly awkward and

17.2  Linking Cancer Biology, P4 Medicine, and Molecular Imaging

unwieldy, although the best of them are lending insight into the molecular origins that correlate with the evolving pathophysiology of the disease, including disease signatures that arise prior to clinical symptoms. More importantly, network models have the potential to provide insights into drug targets, as well as in vitro diagnostic markers and in vivo imaging targets. Such models will eventually provide the foundation for P4 medicine—i.e., a dynamic network model, when integrated with an individual’s genomic information, can constitute a predictive hypothesis of that person’s disease. Such a hypothesis will provide candidate in vitro and in vivo biomarkers for disease diagnosis, facilitate identification of drug targets, assess the potential for therapeutic intervention before the development of disease, and establish suitable objective molecular imaging readouts of drug activity. 17.2  Linking Cancer Biology, P4 Medicine, and Molecular Imaging The implementation of molecular imaging platforms that are sensitive to detecting therapeutic responses before clinical evidence of responsiveness can be observed is paramount. Serial anatomic imaging, using computed tomography (CT) or magnetic resonance imaging (MRI), has provided information on tumor responsiveness in terms of lesion size, which has been particularly useful in cases where tumor shrinkage occurs early, typically after administration of traditional cytotoxic agents. In contrast to this, molecularly targeted drugs that tend to cause arrest of tumor cell development and growth and may not be associated with evident short-term tumor shrinkage will require the application of imaging technologies sensitive to detecting information pertaining to gene and protein activity in tumors. In turn, this molecular-level information will be essential for the selection of appropriate therapeutic agents, the detection, and monitoring of treatment response, and the early prediction of drug unresponsiveness for ultimately individualizing clinical protocols. While a number of imaging modalities (namely optical, positron emission tomography [PET] and MRI) can be utilized for these purposes in both pre-clinical and clinical settings, this discussion primarily emphasizes MR imaging applications in the context of diagnosing and treating cancer and other diseases. Before the advent of combined imaging systems (e.g., PET or CT), MR imaging offered two main advantages over modalities utilizing radiolabeled or optical probes, namely higher spatial resolution (micrometers) and the ability to obtain anatomic, physiologic, and metabolic information in a single imaging session, but offering reduced detection sensitivity relative to PET and optical imaging. However, given recent technological advances in instrumentation and the advent of combined clinical imaging systems (i.e.,

PET–CT or PET–MRI), limitations previously associated with the use of PET imaging alone, including the relative lack of available anatomic information and spatial resolution, have been or are continuing to be addressed. In recent years it has become evident that the future of clinical practice will increasingly rest on the use of combined multimodality imaging devices for answering important biological questions that may not be adequately answered by any single modality. The ability to detect early responses to targeted therapies by functional and molecular imaging technologies before responses are observed clinically is paramount for understanding evolving molecular events occurring within the tumor itself. One example of how these many avenues of information are being used to revolutionize the practice of medicine is the area of drug development. The development of drugs that target specific genes of interest and their products for a given individual forms the basis of personalized medicine. The unique genetic profile of each person will prescribe differences in the individual’s medical course and response to the environment. Detectable genetic variations and the resulting differences in gene expression among patients will ultimately determine drug behavior. Variations in drug responses among patients can pose a serious problem in terms of validating those drugs (Betensky et al. 2002). Recent examples exist in which pauciparameter molecular measurements are being employed to identify potential responders to at least two therapeutics (Hughes and Branford 2003; Lamb et al. 2006; Martin 2006; Radich et al. 2006). Single-parameter measurements for such purposes are unlikely to be the norm. Instead, the coupling of molecular diagnostics with molecular therapeutics will eventually require measurements of a multiparametric (e.g., cells, mRNA, and proteins) biomarker panel that can be used to direct patients to appropriate therapies or combination therapies. As a general rule, however, such molecular diagnostic tests do not yet exist. Given this limitation, trial-and-error approaches have been employed in order to optimize a dosage regimen, potentially increasing the number of adverse reactions. Improving this situation will require a knowledge of the spectrum of genes (pharmacogenomics) and gene products (i.e., proteins, or pharmacoproteomics) that influence drug response, as well as differences in drug response at the molecular and genetic levels, thereby allowing for therapies and medical care to be better tailored to the patient. Beyond clinical patient care, however, pharmacogenomics and pharmacoproteomics are being applied as successful strategies in drug discovery and development, including the tailoring of cancer therapies to the individual. Knowledge of the genetic differences between patients responding to a particular drug may offer a survival benefit, particularly in cases of drug toxicity. One specific example might be to test patients to determine whether or not they possess the appropriate enzyme for drug metab-

1413

1414

17  Systems Biology and Nanotechnology

olism. An example of this has been seen in patients using thiopurine medications, such as azathioprine and thioguanine (Thrall 2004), which have been known to induce fatal myelosuppression, a rare side effect. This difference in responses among patients has been attributed to the enzyme thiopurine S-methyltransferase (TPMT) (Thrall 2004), of which 1 in 300 individuals has a complete deficiency. Having a relative deficiency or lack of this enzyme may, therefore, prove fatal in patients undergoing treatment. The association between agent toxicity and enzyme absence and/or deficiency has led to the development of an indirect genetic assay for identifying this mutation. Such an example underscores the evolving translational role of pharmacogenomics in improving patient care in the era of personalized medicine. 17.3  Application of a Systems Biology Approach to Cancer To understand biology at the systems level, the structure and dynamics of the system as a whole needs to be considered, not just the individual components comprising the system (i.e., genes, proteins, metabolites). While current technologies permit the genes and proteins defining a specific cell or organism to be identified, knowledge of how these various components are linked to each other and how they dynamically interact during a particular process will be critical to elucidating how a given system will function (Kitano 2002). The central components of systems biology are gene and protein regulatory elements (i.e., circuits) that comprise cells and cell networks. The creation of an accurate and detailed dynamic network model will dictate the organization and function of cells in response to environmental cues. By adopting such a model, the specific functions defining a given disease process may be added or removed, essentially genetically or environmentally reprogramming cells (Health et al. 2003). The application of a systems-based computational approach, driven by experimental measurements, is proving particularly useful for attempting to systematically integrate the enormous volume of molecular-level data collected in human cancer research. In a now-seminal work, Hanahan and Weinberg (2000) integrated and classified cumulative molecular alterations associated with cancer physiology into a useful conceptual framework according to seven perturbed regulatory systems: self-sufficiency in growth signals, insensitivity to growth-inhibitory signals, evasion of programmed cell death (apoptosis), limitless replicative potential, sustained angiogenesis, and tissue invasion and metastasis (Fig. 17.3.1a). Two additional essential attributes that highly advanced human cancer cells share is their ability to evade elimination by the immune system and the acquisition of genomic instability. Since this work, it has become clear that many of

Fig. 17.3.1a,b  Acquired biological phenotypes of cancer versus cancer cell genotypes. a Most, if not all cancers, have acquired the same set of functional capabilities during their development, albeit through various mechanistic strategies. During the course of tumor progression, human cells may acquire a number of distinct cancer-associated phenotypes by the introduction of a number of distinct genes needed for transformation to a tumorigenic state. b see next page

the phenotypes of cancer cells are achieved through the complex interactions of several genes or genetic alterations (Fig. 17.3.1b), rather than simply by the successive accumulation of individual genetic alterations which, in turn, might lead to one of the corresponding eight specific attributes listed above. Weinberg has noted that this complexity might be further understood by introducing the concept of regulatory subcircuits that act to govern the life of a cell, define its biological phenotype, and, perhaps, result in tumor progression as such subcircuits are deregulated (Fig. 17.3.2). By employing such a systemsbased approach, exciting new insights and mechanistic understandings may be gained that, in turn, can lead to the development of novel diagnostic and treatment strategies. As can be seen in Fig. 17.3.2, such insights will require the integration of existing biologic information (e.g., genes, proteins, and metabolites organized into pathways and network circuits) with large amounts of high throughput data in order to extract unknown patterns and relationships (Khalil and Hill 2005).

17.3  Application of a Systems Biology Approach to Cancer

Fig. 17.3.1a,b  (continued) Acquired biological phenotypes of cancer versus cancer cell genotypes. b Specific genes introduced to the cell may contribute to cell phenotypes associated with tumorgenicity. Genes, for instance, encoding hTERT, the catalytic subunit of telomerase, affect only the phenotype of immortalization, while genes encoding for p53 affect at least three distinct phenotypes: resistance to growth inhibition, evasion of apoptosis, and immortalization. A widely acting protein,

like Ras, will affect susceptibility to apoptosis, dependence on exogenous mitogens, angiogenesis, and invasiveness/metastatic ability. There is not a simple one-to-one correspondence between genes and cancer-associated phenotype (Reproduced with permission from Weinberg R [2006] The biology of cancer. Garland, New York; Hanahan D, Weinberg RA [2000] The hallmarks of cancer. Cell 100:57–70, MIT Press, Cambridge)

The above approach constitutes a hypothesis of how the system works at the molecular scale (Ideker et al. 2001), and that hypothesis can be tested via a systematic series of perturbations on the system. The perturbed system is reanalyzed, and the results integrated back into the model so as to eventually lead to a self-consistent and predictive network hypothesis. In addition to revealing the molecular signatures that define the nature and progression of disease, such models will ultimately be mined for diagnostic and/or therapeutic targets (biomarkers) for the development of P4 medicine. A systems biology approach has recently been used to understand prostate cancer progression. Prostate cancer is known to progress from a state of androgen sensitivity (early stage) to one of androgen unresponsiveness as the disease advances. Utilizing a systems approach, the molecular mechanisms underlying this progression have recently been studied using both androgen sensitive and insensitive prostate cancer cell lines (Lin et al. 2005). Novel high-throughput technology has been utilized to assess

changes in mRNA and protein levels during this transition. The integrated mRNA and protein expression data revealed functional differences between these cell lines, with a number of pathways found to be up-regulated or down-regulated. By mapping these data differences or perturbations into existing physiologic networks, a preliminary (early stage) network model reflecting cancer progression can be created (Fig. 17.3.3). Ultimately, the success of systems biology network models in providing a framework for the development of P4 medicine will depend on the quality of the models themselves. Multiparametric, quantitative, and time-dependent analysis of tissues, cells, and even single cells will be required in order to generate reliable information from which accurate network hypotheses may be derived. This places tremendous demands on the development of new measurement platforms. For example, the vast amount of mRNA and protein product data derived will require automated, miniaturized, and highly parallel devices for integrating many simultaneous operations. Fur-

1415

1416

17  Systems Biology and Nanotechnology

Fig. 17.3.2  The intracellular signaling circuitry and collaboration between cancer-associated genes. The signal transduction circuitry diagram indicates only a portion of the proteins that play critical roles in modulating the flow of signals through the various circuits operating within cells. As indicated by the various shadings, different subcircuits are involved in regulating distinct cell physiologic properties. Thus, the growth-promoting, mitogenic circuit (light red), the circuit governing growthinhibitory signals (light brown), the circuit governing apopto-

sis (light green), and the circuit governing invasiveness and metastasis (light blue) can be assigned to distinct regions in a map representing the master circuitry of the mammalian cell. The circuit governing mitogenesis overlaps, in part, with that governing cancer cell invasiveness, suggesting that a common set of proteins mediates both biological responses (Reproduced with permission from Weinberg R [2006] The biology of cancer. Garland, New York)

thermore, the computational challenges for integrating this data into predictive network models are also tremendous. The ongoing fabrication of nanodevices for sensitive, real-time detection of genes, mRNA, and proteins (nanotechnologies), as well as advances in microchip/ nanochip technologies (microfluidics) for conducting multiple analyses of tiny patient fluid samples in parallel (i.e., cell sorting DNA purification, and single-cell gene expression profiling) (Heath et al. 2003; Hood et al. 2004; Weston and Hood 2004), are just a few of the technologies being driven by the demands of systems biology.

Such nanotechnology-based labs, or nanolabs, combined with appropriate bioinformatics algorithms and molecular imaging approaches can potentially identify the molecular signatures of disease and disease progression (i.e., critical protein targets) for exposure to candidate pharmaceuticals (Fig. 17.3.4). By selecting the appropriate molecular biomarkers as objective endpoints of treatment efficacy (Hood and Perlmutter 2004; Rudin and Weissleder 2003), drug development may be expedited and directly translatable to mouse models and patients.

17.3  Application of a Systems Biology Approach to Cancer

Fig. 17.3.3  A network describing the difference between androgen responsive and androgen unresponsive prostate cancer. For this drawing, individual genes are indicated by the colored squares, with the color of the squares indicating the level of differential regulation between the two classes of prostate cancers. The orange circles labeled H1–H5 represent five different hypotheses that can be extracted from this network. More importantly, this network, while an awkward representation of the biological system, highlights the degree of interconnectivity of

the various genes, and also highlights the limitations of the pathways model. In the future, models such as this will also contain a dynamic element, and thus provide a “movie” that describes, at a molecular level of resolution, how cancer transforms an organ from a healthy to a diseased state. Such models will eventually provide a wealth of information that can translate into the clinic in the form of in vitro and in vivo diagnostic markers, drug targets, etc.

1417

1418

17  Systems Biology and Nanotechnology

Fig. 17.3.4  Possible methods through which the various tools described might be used within to solve a systems biology problem and to apply that solution to a drug discovery process. In Step 1, the nanolab is utilized to carry out a rapid and informative molecular analysis of a biological system via a global analysis of proteins and mRNA levels in a few cells. The data, coupled with an informatics algorithm, generate a hypothesis of the system. Step 1 also illustrates a means for carrying out an informative molecular-based diagnosis of disease. In Step 2, the

17.4  Molecular Diagnostics and Personalized Therapies for Cancer

hypothesis is tested using top-down perturbations in the form of molecular probes, and drug targets are identified. In Step 3, the most effective of these molecular probes is turned into an imaging probe (and a drug), and applied toward imaging and treating disease within a living patient (Reproduced with permission from Heath JR, Phelps ME, Hood L [2003] NanoSystems biology. Mol Imaging Biol 5:312–325, Springer, Berlin Heidelberg New York)

in proteins, DNA, and RNA within body fluids (see below). A primary reason for the overall poor survival associated with many cancers at the present time is their Improvements in molecular diagnostics will be a key late-stage detection, usually after metastatic spread of factor driving early cancer detection and prevention, ul- disease to distant sites has occurred. Tumor eradication timately leading to the development of a more person- is more difficult under these conditions. For example, the alized medicine and, thereby, decreasing morbidity and 5-year survival for colon cancer is close to 100% if the tumortality. The aim is to facilitate the early detection of mor is confined to a local area, 30–50% for lymphatic inmolecular differences between normal and diseased cells volvement, dropping to about 10% for distant metastatic and tissues, as well as to detect cancer-specific alterations spread (Wagner et al. 2004). The poorer prognosis with

17.4  Molecular Diagnostics and Personalized Therapies for Cancer

metastatic tumor spread has fueled efforts among oncologists and many other specialties to develop techniques that are sensitive to detecting early-stage cancers. Molecular diagnostics are enabling the creation of a more personalized medicine by providing a better understanding of cancer biology and, thereby, ultimately facilitating rational drug discovery (Table 17.4.1). These technologies are yielding improvements in cancer classification, which have been traditionally limited to the tissue of origin, histologic appearance, and degree of invasiveness. However, as cancers are known to vary genetically and phenotypically in different patients having identical tumor types and stages, molecular profiling has been performed in order to refine these classification schemes on the basis of clinically relevant subtypes, as well as for predicting clinical outcomes. Every type of cancer, and its progression from earlyto late-phase disease, is associated with alterations in gene and protein expression and/or modification. Such cellular, biochemical, or molecular changes associated with the presence and severity of disease, defined as biomarkers, can be used to detect cancer, assess prognosis, and monitor disease progression or therapeutic response. An example may be a molecule secreted by the malignant lesion itself or a specific response to the presence of the lesion. The development of accurate and reproducible assays for measuring protein, DNA, RNA, or metabolite levels in early-phase cancers subscribes to the notion that earlier detection may ultimately lead to cancer prevention and enhanced survival. On this basis, metabolic alterations in genes, proteins, or metabolites that are expressed at abnormally high or low levels in cancerous cells, relative to normal cells, can be selected, separated, and identified using a range of novel molecular tech-

Table 17.4.1  Impact of molecular diagnostics on the management of cancer Cancer classification using microarrays Gene expression profiling Role of methylation markers in cancer stratification Cancer prognosis Prediction of response to chemotherapy Genomic analysis of tumor biopsies to predict therapeutic response Diagnostics as a guide to therapeutics Diagnostics for detection of minimal residual disease Molecular diagnostics combined with cancer therapeutics Use of molecular diagnostic technologies for drug discovery and development Design of future cancer therapies Screening for personalized anticancer drugs Pharmacogenomic tests for stratification of clinical trials Adapted with permission from Jain 2005b

nologies, including DNA, transcriptome, and protein microarrays, microfluidic devices/biochips, laser-capture microdissection, and time-of-flight mass spectrometry (Weston and Hood 2004; Weber 2002; Chatterjee and Zetter 2005; Peano et al. 2006; Cummins and Velculescu 2006; Jain 2004a, b). These technologies, for instance, have been used to isolate certain classes of genes responsible for disease initiation or progression. 17.4.1  Cancer Biomarkers for Clinical Use Dynamic network models and their associated databases are beginning to be mined to provide a host of biomarkers for stratifying and staging cancers, directing therapies, and evaluating positive and adverse responses to those therapies (Aebersold et al. 2005). Tissue-based biomarkers, as a general rule, can provide direct measurements of the disease state, while serum biomarkers (i.e., organspecific, secreted proteins) are reflections of the disease state. Certain biomarkers can serve as in vivo molecular imaging targets. What all of this again points toward is that pathology will increasingly incorporate the quantitation of panels of biomarkers—perhaps broad panels for conducting a health-status survey of an individual, and more focused panels for monitoring specific diseases, or therapies, for instance. An appropriately chosen panel of biomarkers can lead to improvements in the sensitivity and specificity of a diagnosis. Initially such panels may be selected for their value in measuring alterations that differentiate two different pathways (network perturbations). One example is the use of prostate-specific antigen (PSA), the currently used biomarker for prostate cancer, which is regulated by androgen. Prostate cancer is caused by the androgen hormone-specific stimulation of cell growth, although more aggressive forms are known to operate through androgen-independent pathways (Lin et al. 2005; Isaacs 1999). In the latter case, therefore, PSA would not serve as a good biomarker for detecting androgen-independent prostate cancer, nor other circumstances in which androgen depletion is present (e.g., chemotherapy). By utilizing several prostate-specific biomarkers that monitor these different pathways, measures used to assess aggressive, androgen-independent prostate cancers may be performed with increased sensitivity. If a biomarker or panel of biomarkers is found that can accurately and reliably detect early-phase cancers, it may be used to monitor high-risk individuals or screen larger populations for such lesions or premalignant disease without signs or symptoms of the disease. However, such an evaluation should be considered in light of several important issues, such as whether the disease is an important health problem on the basis of measured morbidity and mortality levels, whether the disease can be detected in a pre-clinical phase, whether acceptable lev-

1419

1420

17  Systems Biology and Nanotechnology

els of accuracy and cost are met by the screening test, and whether treatment outcomes are better before detection of clinical symptoms compared with later treatment. For screening larger, asymptomatic populations, assay methods may require adaptation to achieve high-throughput analyses. Biomarker panels need to be efficient, easily performed, and cost-effective. A number of flexible technology platforms—chips, beads, and microfluidics technologies—are presently available for analyzing a variety of biomarkers on a single platform. These platforms can simultaneously analyze more than one type of biomarker or a panel of nucleic acids or protein biomarkers, expediting sample analyses to create a more time-efficient, convenient process. Biochips containing microarrays of genetic information, such as cDNA, oligonucleotides, or proteins will play a crucial role in the personalization of medicine, including cancer therapeutics. Genes can be sequenced and analyzed rapidly using the DNA microarray, which will need to be coupled to bioinformatics platforms for analyzing the vast amount of information generated. In comparison with DNA microarrays, protein microarrays/chips offer the possibility of rapidly performing a global analysis of the entire proteome, facilitating protein diagnostics and therapeutics (Jain 2004b). Molecular diagnostics will be the most clinically relevant application of protein chips, which can be used, for example, to distinguish proteins in normal cells from those in early-phase cancers, as well as from late phase metastatic disease. By successively reducing surface feature dimensions from microns to the nanoscale range, the quality, quantity, and speed of the recovery process can be improved. Nanoscale protein analysis and nanocapture methods will prove particularly useful for tissue or biological fluid samples containing less abundant proteins or limited sample volumes. In another case, a single sample containing DNA, RNA, protein, and cellular substances can be analyzed very rapidly using LabChipTM technology (Caliper Life Sciences) coupled with a A2100 bioanalyzer (Agilent Technologies) for purposes of performing gene and protein analyses, gene expression, and differential display analyses. LabChip integrates numerous liquid handling devices and sensors on a miniaturized platform for rapidly automating large, costly experiments, which would, otherwise, be impractical to execute. A large number of measurements may be derived from such procedures that, when linked with bioinformatics algorithms, can be used to generate molecular-level images of the biological process under investigation. Biomarkers can be detected and measured using the abovementioned or other established technologies, physical examination, and clinical imaging devices, such as CT, MRI, or PET. In everyday clinical practice, oncologic biomarkers are typically anatomic, such as tumor size, which can be assessed using structural imaging modalities, or reflect physiologic/metabolic processes, such as serial

changes in tumor glycolysis, as measured by PET imaging in the presence of 18F-labeled fluorodeoxyglucose. A variety of additional molecular imaging technologies in various stages of development (e.g., optical methods) can detect specific molecules in living organisms, and may offer great promise for imaging biomarkers of interest. 17.5  Role of Cancer Nanobiotechnology in Personalized Medicine Advances in molecular diagnostics and proteomic technologies, as well as an increased understanding of the molecular mechanisms of cancer, are being further refined by the emerging availability of nanobiotechnology platforms. In addition to improving the detection of panels of cancer biomarkers by utilizing materials or devices, nanotechnology is playing an increasingly critical role in molecular imaging, drug discovery, drug development, and the delivery of therapeutics. Nanobiotechnology platforms provide highly sensitive, quantitative, and high throughput functional assays that are complementary to genomic technologies for detection of new cancer biomarkers. 17.5.1  Nanodevices as Biosensors Nanomechanical (Backmann et al. 2005) and nanoelectrical (Heath 2007) devices (nanocantilevers and semiconductor nanowires) are highly sensitive ex vivo biomolecular sensors offering real-time detection of genes, mRNA, and proteins. The nanocantilever (Concentris GmbH, Basel, Switzerland), which is coated with a molecular probe, transforms the chemical binding of target proteins to its surface into mechanical deflections, on the order of nanometers, that can be optically detected and quantitated. Both sensor types constitute label-free methods with the potential for real-time detection and quantitative determination of biomolecular concentrations (Bunimovich et al. 2006). Information obtained from such devices relates to the presence and concentration of the bound target substance against a set of reference data. This is particularly useful for drug discovery applications, in which the attachment of specific antibodies to the cantilever surface enables imaging of bound target antigens and the detection of antigen–antibody interactions (Jain 2005a). Single-walled carbon nanotubes and miniaturized silicon nanowires, as sensitive electronic biomolecular sensors, have been used as platforms for detecting DNA (mRNA) oligomers, and for investigating protein-protein and surface-protein binding. Molecular probes (i.e., antibodies, enzymes), bound to the surface of the silicon nanowires, for instance, will bind specific target proteins, resulting in altered, electrically measurable nanowire

17.5  Role of Cancer Nanobiotechnology in Personalized Medicine

conductivity (Ferrari 2005: Fortina 2005). In one competitive inhibition study in which Abl, a protein tyrosine kinase responsible for chronic myelogenous leukemia, was bound to the surfaces of silicon nanowires, changes in nanowire conductance were monitored to evaluate concentration-dependent ATP binding and its inhibition by the antagonist Gleevec® (Jain 2005a). The sensitivity and dynamic range of nanowires for direct, real time detection of biomolecules is unparalleled. For example, specific ssDNA oligomers have been detected, over the concentration range from 10–16 to 10–17 M concentrations (Cheng et al. 2006).

Fig. 17.5.1a,b  Nanochips for rapid, multiparameter in vitro diagnostics measurements. A diagram (a) and scanning electron micrographs (b) of a single region of a nanochip that contains a small nanowire biomolecular sensor array entrained within a microfluidics channel. Currently, each chip contains about eight such regions, allowing for about 24 individual measurements. However, each chip also contains as many as a few thousand nanowires that can potentially be utilized for measurements. Such biomolecular sensors can detect, in real time (typically a couple of minutes) mRNAs and proteins at a lower concentration limit 10–16 and 10–15 M, respectively, with a dynamic range of >106. A number of challenges remain to be solved before this and other promising technologies can find clinical applications (Adapted with permission from Bunimovich et al. 2006)

A major lesson that is being learned now from systems biology is that in vitro clinical measurements in the future will have to be cheap; information will be the most valuable commodity. It is this lesson that is driving much of the nanowire/nanocantilevers development. Nanofabrication protocols are affording very large libraries of such identical devices (e.g., thousands of cantilevers or nanowires) (Melosh et al. 2004; Bunimovich et al. 2005; Lee et al. 2002; Hegner and Arntz 2004; McKendry 2002). These can be integrated with microfluidics platforms (Pantoja 2004) for potentially automating the performance of numerous biochemical operations (i.e., purification, cell sorting), followed by large scale, multiparametric detection (cells, mRNAs, and proteins) (Fig. 17.5.1). The integration of nanosensor arrays with microfluidics permits the efficient extraction of multiple molecular signals (i.e., proteomic profiles) from single cells or small tissue volumes. An additional advantage of microfluidics-based assays is that, under certain operational conditions, the timescale for detection of the biomolecules within a flowing microfluidics environment is limited only by the Langmuir isotherm (Zimmerman et al. 2005). This means that the multiparametric assay (e.g., measurements of a few sandwich-type immunoassays) can be completed within a couple of minutes, as compared with a few hours if such assays are performed in 96-well plates. A number of scientific and engineering issues remain to be solved before these approaches can find routine clinical use. These involve the development of highthroughput manufacturing methods, the development of chemical approaches for selectively encoding each nanosensor with a specific biomolecular capture agent (Rohde et al. 2006), and the integration of such devices with chips that can process complex biological materials, such as blood (Brody et al. 1996; VanDelinder and Groisman 2006) and solid-tissue samples (Kallioniemi 2001). Finally, antibodies comprise the dominant, clinically useful technology for protein detection. Therefore, the development of capture agent alternatives to antibodies is an extremely important challenge. 17.5.2  Nanoparticles for Cancer Diagnostics and Therapeutics 17.5.2.1  Magnetic Nanoparticles The demonstration that clinically occult lymph nodal metastases in prostate cancer patients could only be detected with paramagnetic nanoparticles and MRI (Harisinghani 2003) has resulted in a number of iron- and gadoliniumbased probes for detecting smaller and earlier-stage tumors (Fig. 17.5.2). Molecularly targeted nanoparticles could be used to probe the tumor microenvironment, providing information on the presence, abundance, and/

1421

1422

17  Systems Biology and Nanotechnology

Fig. 17.5.2a–d  Illustration of an experimental pathway from the in vitro investigation of cellular processes to the in vivo utilization of preclinical models and MRI for translation to the clinical setting. a Iron-based nanoparticle probe. b Histologic evaluation of cells without (top) and after iron-labeling (middle) as seen by Giemsa staining, with Prussian blue staining for iron detection (bottom). c MRI in preclinical models demonstrates nearly isointense signal intensity in brain tumors relative to the contralateral hemisphere after intravenous administration of unlabeled cells using either spin echo (SE) or gradient echo (GRE) sequences. A clear ring is seen along the tumor periph-

ery on both SE and GE sequences after injecting iron-labeled cells. d Histopathological examination of a fixed 6-mm tissue section stained for iron using Prussian blue and for activated endothelial cells using von Willebrand factor (VWF), suggesting incorporation of endothelial precursors as part of ongoing tumor angiogenesis. Arrows indicate cells that are both VWF+ and PB+. e Bone marrow–derived stem cells that are endothelial precursors are being investigated as angiogenesis-selective genetargeting vectors for potential clinical use (c and d reproduced with permission Anderson et al. 2005)

or distribution of cancer markers and signatures (Li et al. 2004; Sullivan and Ferrari 2004). As an example, the conjugation of annexin-V to nanoparticles can facilitate detection of apoptotic cells once the conjugated probe binds to phosphatidylserine, a cell surface marker on apoptotic cells. This probe has been used, in conjunction with MRI, to identify camptothecin-induced apoptotic Jurkat cells in vitro (Schellenberger et al. 2002). In a second application, elevated telomerase activity, a marker of limitless replicative potential (Hayflick 1997), was detected in cell assays using MRI and a biologically “smart”

magnetic nanoparticle that switches its magnetic state on annealing to telomerase-synthesized TTAGGG repeats (Grimm et al. 2004). These findings suggest that telomerase, which is elevated in many malignancies, may serve as an important molecular target for cancer diagnosis and therapeutic intervention. Sustained angiogenesis, found in a number of premalignant lesions of the breast, skin, and cervix (Ferrari 2005), is an important marker for the early detection of cancer. In addition, signatures of angiogenesis reflect essential features of solid tumor progression and inflamma-

17.5  Role of Cancer Nanobiotechnology in Personalized Medicine

tory states. In the later stages of neovascularization that occurs in melanomas, there is rapid tumor progression and metastatic extension, which is often fatal (Schmieder et al. 2005). A number of groups have functionalized MR-based nanoparticle probes for specifically targeting ανβ3-integrin on the endothelial surfaces of vessels in experimental models (Schmieder et al. 2005; Winter et al. 2003; Sipkins et al. 1998; Anderson et al. 2000). By additionally loading the particle with thousands of metal ions (e.g., iron), detection sensitivity is enhanced, enabling the detection of sparse biomarkers with in vivo molecular MRI. In one study, small regions of angiogenesis associated with nascent melanoma tumors were detected (Schmieder et al. 2005), and detection thresholds of sparse epitopes targeted by such nanoparticles were found to be very low (picomolar range) (Schmieder et al. 2005; Morawski 2004). In addition to such polymerized nanoparticle probes showing promise for the detection of incipient tumors on clinical MRI scanners, earlier detection may improve treatment efficacy, particularly for the case of melanoma. By additionally utilizing the nanoparticle as a delivery vehicle for transporting anticancer drugs, larger therapeutic doses can be delivered while effectively eliminating systemic toxicity. The combination of earlier tumor detection and treatment efficacy utilizing a combined nanotechnology platform and MRI is the essence of personalized medicine, and would additionally facilitate the radiologic interpretation of such clinical studies.

rendering such imaging studies difficult. By designing QDs with a soluble coat that emits at longer wavelengths (near-infrared light), improved light penetration has been achieved. Such probes have been applied successfully to, for instance, experimental models of sentinel lymph node (SLN) mapping, permitting non-invasive visualization of the target lymph nodes (Kim et al. 2004). This approach offers significant improvements over traditional, less precise radioactivity/organic dye methods, which generally result in a greater excision of the lymph system than necessary. The application of QDs, in contrast, would enable surgical approaches to be better tailored to the needs of the patient. Therapeutic responsiveness in patients can potentially be linked to the molecular characteristics of specific cancer types. Knowledge of the patient population(s) most likely to benefit from a given therapy forms the essence of personalized medicine. One such example is the use of the marker HER-2 for predicting tumor aggressiveness and sensitivity to trastuzumab (Herceptin) in breast cancer patients. While given molecular diagnostic protocols exist for detecting HER-2 overexpression, improving the accuracy of such tests is paramount. In this regard, a second application of QDs has been to create highly specific probes exhibiting enhanced photostability and brightness relative to other probes. This has been achieved by covalently linking QDs to antibodies (Wu et al. 2003) against the HER-2 receptor for assessing HER-2 overexpression using fluorescence techniques. As QDs can be bound with any number of cell- or tumor-specific molecules for detecting the presence of predictive cell/tumor markers, new avenues for study17.5.2.2  Quantum Dots ing drug targets, proteins, and genes in cell cultures and Quantum dots (QDs) are semiconducting fluorescent in vivo models are now possible (Weissleder et al. 2005). nanoparticles that, when attached to specific molecules, MRI-detectable QDs have been created (Mulder et al. can be used diagnostically as molecular targeted probes 2006) by applying a paramagnetic coating to the QD. By to assay biological materials (Smith et al. 2006), namely further functionalizing this probe with peptides specific fluids (e.g., blood) (Hernandez and Thompson 2004; for targeting the vascular endothelium, tumor angiogenGoessl 2003) and tissue biopsy material (Sukhanova 2002, esis can be evaluated using MRI. However, as QDs are 2004) for specific cancer markers. In addition, they can potentially toxic metallic probes (Jain 2005a; Cheng et al. be used to sensitively detect and track live cells (Parak 2006; Ozkan 2004), they have been effectively excluded et al. 2002; Dahan et al. 2003) and tiny tumor deposits as a choice for targeted drug delivery, although their surin vivo by high-resolution imaging techniques (Gao et al. face modification for this purpose is possible. As they 2004, 2005). In terms of cancer detection, QDs offer sig- cannot be used clinically at the present time, there is a nificant advantages. For example, they are highly efficient growing need for developing safer materials for human and non-photobleaching fluorophores with sharp emis- application. sion spectra that can be tuned as a function of QD size. This makes them tools for imaging and tracking multiple molecular targets simultaneously. These multiplexing ca- 17.5.2.3  Nanoshells pabilities underscore such probes as powerful tools for potentially yielding more accurate diagnoses, as well as As a novel technology providing both diagnostic and helping to elucidate complex oncologic gene expression therapeutic capabilities, nanoshells offer a unique couprofiles. pling of an imaging/therapeutic platform for targetOne limitation with QDs is that superficial cancers are ing clinically relevant biomarkers. Nanoshells are goldmore easily imaged than are deeper ones, as light pen- coated nanoparticles containing an inert silica core, etration into the body is reduced with increasing depth, whose optical properties can be controlled by altering its

1423

1424

17  Systems Biology and Nanotechnology

physical properties (i.e., core size, shell thickness). This has important clinical implications for cancer diagnosis and therapy. Appropriately designed nanoshells absorb and scatter light over a broad spectral range, including the near infrared ([NIR] ~800 nm) that can penetrate through tissues. This makes nanoshells excellent candidates for imaging and treatment of deep-seated tumors. These nanoparticles accumulate by actively extravasating at sites of leaky tumor vasculature, and can be used to selectively target and image oncoproteins overexpressed on cell surfaces by binding them to antibodies via a linker molecule. Using this dual optical imaging/therapy approach, cancer cells are initially detected on the basis of light scattered in the NIR. Subsequent light absorption in the NIR then allows selective thermal destruction of targeted cells. Biological applications of these nanoparticles have included thermal ablation of cancerous cells in vitro and in vivo (Hirsch et al. 2003a, b). Proof-of-principle in vitro studies revealing significant cell death have used cultured human breast adenocarcinoma cells (SKBr3) overexpressing HER-2 (Loo et al. 2005). After attaching anti-HER-2 to the nanoshell, nanoparticles were added to cultured cells, which demonstrated significantly increased scatterbased optical contrast by microscopy relative to control populations, as a result of nanoshell binding. After photothermal therapy, localized, irreversible cell death was observed only in cell populations exposed to anti–HER-2 nanoshells and NIR. No toxicity to human cells was observed following incubation with these nanoparticles. Magnetic gold nanoshells have recently been fabricated for multimodal imaging purposes. These consist of gold nanoshells embedded with iron-based nanoparticles, which have also been conjugated with anti–HER2 for targeting SKBr3 cells overexpressing HER-2 (Kim et al. 2006). Cells targeted by these magnetic nanoshells were detected on a 3-T clinical MRI system and rapidly destroyed by short exposures to NIR wavelengths and low power. The introduction of these multifunctional platforms for diagnosing and treating cancers lays the foundation for the future implementation of powerful new individualized treatment strategies which are both targeted and synergistic in their design.

systems are being constructed to facilitate the development of personalized medicine. A few of these promising technologies are discussed below. 17.5.3.1  Biochips for Drug Discovery A potential application of the biochip, as it relates to the discovery of new drugs for a variety of diseases, is the study of synthetic cell membranes (Jain 2005a). This is particularly relevant for the development of chemoresistant tumors. Cell membranes are comprised of many proteins, of which a small number may facilitate efflux of cancer chemotherapies from the tumor cell microenvironment. A biochip designed to contain an enormous number of test chambers (one drug per chamber) could potentially expedite the screening of a large number of drugs. In turn, this could dramatically increase the number of experiments possible with small amounts of available protein. 17.5.3.2  Dendrimers in Anticancer Therapy

Dendrimers, a novel class of three-dimensional, synthetic structures with highly branched architecture, can be precisely synthesized or modified into biocompatible compounds of low toxicity (Ozkan 2004; Jain 2005b; Padilla et al. 200; Ihre et al. 2002) for use in a range of medical applications. Dendrimers have been most useful for improving existing molecular imaging technologies and for drug delivery (Jain 2005a, b, c; Sahoo and Labhasetwar 2003). For instance, gadomer-17 (Schering AG, Germany), a polylysine dendrimer functionalized at its periphery with gadolinium chelates, is used as an intravascular MRI probe for tumor differentiation and vascular assessment (Helms and Meijer 2006), and dendrimerbased nanoprobes complexed with gadolinium have been utilized for in vivo visualization of SLNs (Talanov et al. 2006) and tumor microcirculatory characteristics (Langereis et al. 2006) in preclinical models. Delivery of therapeutic agents has also been achieved by encapsulating guest molecules in dendrimer cages, with the dendrimer structure improving the efficacy, toxicity, and targeting ability of many drugs. For example, cisplatin or carboplatin anticancer drugs (Malik 2006) have been contained 17.5.3  Nanobiotechnology for Drug Discovery within the voids of the dendritic scaffold. These less toxic Nanobiotechnology promises to facilitate discovery of dendrimer-drug conjugates are active against aggressive individualized therapies. Some of the technical achieve- tumor models that are typically chemoresistant at the ments in nanobiotechnology which are applied to drug maximum tolerated intravenous doses. discovery are important for the future of personalized The unique properties of dendrimers relate to their medicine (a number of them are described above). As cavitary structure and globular configuration, rendernoted earlier, molecular diagnostics is an important com- ing them useful for the delivery of therapeutic agents ponent of personalized medicine, which is being refined (Tripathi 2002; Majoros et al. 2006). Additionally, they by the application of nanobiotechnologies. In addition to are known to exhibit antitumoral activity, and have been these achievements, other novel particles, devices, and used in the development of targeted antitumor gene ther-

17.6  Role of Radiology in Personalized Medicine

apies exhibiting novel modes of action (Dufes et al. 2005). Polyvalent dendrimers can simultaneously interact with multiple drug targets (Jain 2004b). They can be conjugated to different bioactive compounds, such as folic acid using cDNA oligonucleotides, for targeting tumor cells overexpressing the high-affinity folate receptor (Dufes et al. 2005; Choi and Baker 2005). 17.5.3.3  Nanobodies as Personalized Medicines for Cancer Nanobodies (Ablynx) are a potentially new class of antibody-based therapeutics, which are the smallest fragments of naturally occurring heavy chain antibodies in the absence of a light chain (Jain 2005c). These fully functioning fragments harbor the full antigen-binding capacity of naturally occurring heavy chain antibodies. They are suitable for targeting antigens in relatively obstructed sites, such as tumor, given their small size and their high affinity and specificity. In addition, they have an extremely low immunogenic potential. Small antibody fragments are potentially good tumor targeting probes, given their rapid penetration, clearance, and high tumorto-background ratios (Revets et al. 2005). Nanobodies have been successfully used in cancer diagnostic tests for PSA detection (Saerens et al. 2004), as well as for in vivo imaging of experimental tumors (Cortez-Retamozo et al. 2004; Roovers et al. 2007). As a therapeutic tool, nanobodies have been isolated that target the brain in vivo and transmigrate across human brain endothelium (Muruganandam et al. 2001), which has important implications for the future delivery of cancer therapeutics. 17.6  Role of Radiology in Personalized Medicine Molecular imaging will play a crucial role in the translation of these key technologies and the transformation of health care, particularly in the management of cancer. It has the potential to improve the understanding of disease in a number of biological systems, particularly in light of the wide range of experimental models that are now currently available for investigating a spectrum of diseases, including cancer, inflammatory states, immune processes, thrombosis, and neurodegenerative disorders (i.e., Alzheimer’s disease). For instance, it has elucidated key metabolic and cellular processes, such as angiogenesis, apoptosis, and cellular migration, that have been reviewed elsewhere (Weissleder 2002; Massoud and Gambhir 2003; Jaffer and Weissleder 2004; Choudhury et al. 2004). The development of suitable imaging strategies for the monitoring of a particular disease process will rely, in part, on the identification of relevant molecular imaging targets. This, in turn, will rest on knowledge of the

critical pathobiological mechanisms defining the disease of interest. The selection of such informative molecular imaging targets may facilitate disease detection and diagnosis at earlier stages than presently possible using conventional imaging methods. Informative targets may also enable the determination of molecular events responsible for cancer initiation and progression, allow for the stratification of patients on the basis of tumor aggressiveness, and can yield pathway-specific information relating to downstream targets in a known systems network model. Another key factor that is critically important in the development of molecular imaging methods for a particular disease is the identification of molecular imaging ligands (e.g., peptides, antibodies) or probes that exhibit highly efficient and specific affinity for the selected target. Highly efficient and specific in vivo target binding will depend on a range of factors, including the abundance of the target, the stability of the target and probe, and target-to-background ratios. In a number of cases, appropriate affinity ligands can be extracted from existing literature sources. However, novel probe development has relied and will increasingly rely on the use of high-throughput screening methods, peptide libraries, phage displays, and nanotechnology, that all incorporate data from genomic, proteomic, or metabolomics screens (Jaffer and Weissleder 2005). On the basis of utilizing such platforms, a sizeable expansion in the number of targeted probes available for clinical molecular imaging can ultimately be expected. Over the past several decades, a number of imaging probes, the majority for nuclear imaging, have been clinically approved for use in a number of medical subspecialties, including oncology, neurology, cardiology, and infectious disease (Table 17.6.1). The design and synthesis of probes having higher specificity and avidity for their in vivo targets will not only fuel the growth of new imaging applications, particularly their translation to the clinical arena, but will also increase the likelihood of earlier disease detection, for instance, in the detection of subcentimeter metastases, as compared with contrast-enhanced imaging techniques (Harisinghani et al. 2003). The in vivo information derived from the use of such probes, in conjunction with data provided by alternative biomarkers and novel molecular tools (e.g., tissue proteomic analyses), may eventually be used for oncologic screening, diagnosis, disease recurrence, and treatment assessment. This, in turn, may significantly impact patient management. For example, alterations in existing therapeutic regimens may be introduced at earlier time points on the basis of the individual’s responsiveness to existing or emerging pharmacologic therapies. Objective molecular imaging readouts of drug activity in vivo, including information related to the presence, biodistribution, and efficacy of promising new therapies (Rudin and Weissleder 2003; Aboagye et al. 2002) for the clinic can additionally be obtained, underscoring the essential role that such imaging strategies will play in drug

1425

1426

17  Systems Biology and Nanotechnology Table 17.6.1  MRI molecular imaging agents (representative list) Disease

Imaging agent

Clinical application

Cancer

Iron oxide (ferumoxide)a,b Iron oxide (ferumoxtran/ ferumoxytol)b Integrin ανβ3c

Liver cancer Nodal staging Angiogenesis

Atherosclerosis (vulnerable plaque)

Iron oxide (ferumoxotran)b Integrin ανβ3c Gadofluorinec Vascular cell adhesion moleculec

Macrophages Angiogenesis Lipid-rich plaques Inflammation

Thrombosis

Fibrin (gadolinium-based agent)b

Acute/subacute thrombi

Neurological

b-Amyloid targeted iron oxide

Alzheimer dementia

Arthritis

Iron oxidec

Inflammation

Infection

Iron oxide

Inflammation

c

c

The agents are approved by the US Food and Drug Administration These agents are used in humans c Animal trials only to date, with clinical trials anticipated (Adapted with permission from Jaffer and Weissleder 2005) a

b

discovery and development. This information can be ies (Table 17.6.1). A significant limitation of MR imaging derived, for example, by labeling the drug (or prodrug) is its intrinsically lower detection sensitivity to an admin(e.g., carbon-11 radiolabel) and selecting the appropriate istered probe relative to other modalities. The latter limimolecular imaging modality (e.g., PET) and strategy to tation has been partially overcome by utilizing signal amevaluate drug activity at a specific endpoint and/or dy- plification strategies that enhance target-to-background namically. The data obtained from such methods may ratios (Jaffer and Weissleder 2004). reduce the need for time-intensive histological analyses, Targeted probes containing gadolinium are associated and permit serial assessments of drug efficacy and phar- with inherently lower relaxivity properties. This limitamacokinetics. tion has been addressed by synthesizing larger nanoparticle constructs, such as dendrimers and polymerized liposomes that can carry higher magnetic payloads. In contrast, superparamagnetic iron oxide nanoparticles are 17.7  Clinical and Near-Clinical Molecular the preferred agents of interest given their strong relaxImaging Applications Using ation properties, which can induce large signal changes Targeted Probes per unit of metal, particularly on T2*-weighted images. Selection of the appropriate imaging modality to address The cross-linking of these particles with a particular Dexspecific biological questions relevant to the disease pro- tran surface coat (cross-linked iron oxide [CLIO]) has cess under investigation will depend on the required sen- enabled these nanoparticles to serve as platforms for tarsitivity, spatial resolution, and depth penetration needed. geting receptors (Ichikawa et al. 2002), integrins (Kang et An additional consideration is whether a desired targeted al. 2002), and specific cell types (Kircher et al. 2003). probe would need to be synthesized for the application of interest. Magnetic resonance imaging offers distinct advan- 17.7.1  Oncologic Molecular Imaging tages of high spatial resolution and depth penetration, for Detection of Nodal Metastases in addition to providing exquisite soft tissue contrast. and Cancer Staging However, it generally does not provide information related to specific biological pathway activities. A number Although a large number of iron oxide nanoparticle of paramagnetic or gadolinium-based chelates, as well as preparations have been constructed exhibiting a range superparamagnetic iron oxide-containing (SPIO) probes of sizes (10–300 nm), as well as magnetic and physihave been utilized in both pre-clinical and clinical stud- cochemical properties, only a few leading iron oxide

17.7  Clinical and Near-Clinical Molecular Imaging Applications Using Targeted Probes

Fig. 17.7.1a–e  Three-dimensional reconstruction of pelvic lymph nodes (a), conventional MRI (b), MRI with lymphotropic superparamagnetic nanoparticles (c), abdominal computed tomography (CT) (d), and histopathological findings (e). a Three-dimensional reconstruction of the prostate, iliac vessels, and metastatic (red) and non-metastatic (green) lymph nodes, to assist in the planning of surgery and radiotherapy. There is a malignant node (thick arrow) immediately adjacent to the normal node (thin arrow), posteromedial to the iliac vessels. b Conventional MRI shows that the signal intensity is identical in the two nodes (arrows). c MRI with lymphotropic superparamag-

netic nanoparticles shows that the signal in the normal node is decreased (thick arrow) but that it is high in the metastatic node (thin arrow). d Abdominal CT fails to differentiate between the two lymph nodes (arrows). e Histopathological examination of the malignant lymph node reveals sheaths of carcinoma cells (hematoxylin and eosin, ×200 magnification) (Reproduced with permission from Harisinghani MG, Barentsz J, Hahn PF et al. (2003) Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 3:2491–2499, Massachusetts Medical Society, Boston)

preparations can be used as clinical imaging probes (Harisinghani et al. 2003; Kooi et al. 2003). In one such recent study, lymphotropic SPIO nanoparticles were used with high resolution MRI in prostate cancer patients to detect occult nodal metastases as small as 2 mm in size (Harisinghani et al. 2003), which could not be identified with the standard detection thresholds of 8–10 mm used with conventional MRI. As these particles target intranodal macrophages, alterations in the expected pattern of signal loss associated with non-cancerous nodes should be detected on spin-echo (SE) or gradient-echo (GRE) imaging sequences with metastatic nodal spread of disease. In a normal node, homogeneous decreases in signal can be seen on MRI due to the accumulation of

lymphotropic particles 24 h after their administration. On the other hand, for a node to be considered malignant, one of the following three criteria needed to be met: signal decreases of less than 30% on T2- or T2*-weighted imaging, heterogeneous signal or discrete focal defects or both associated with the node, or a node with central hyperintensity (excluding a fatty hilum) and peripheral hypointensity (Harisinghani et al. 2003). Using this method, sensitivities and specificities for subcentimeter nodes were found to be greater than 95%, confirmed surgically, and found to be greater than those for PET (Guller et al. 2003). In this case, the combined use of magnetic nanoparticles and high resolution MRI enhanced detection sensitivity to otherwise undetectable nodal metasta-

1427

1428

17  Systems Biology and Nanotechnology

ses and improved clinical cancer staging. In addition to the use of such particle platforms for deriving information on tumor burden, complementary functional data (such as tumor angiogenesis) that are important predictors of clinical outcome can be assessed simultaneously in such studies (Jaffer 2005). The application of sophisticated post-processing imaging analysis tools, including three-dimensional image reconstructions and automatic detection algorithms for identifying nodal metastases, has the potential to greatly facilitate pre-surgical and pretreatment planning efforts (Fig. 17.7.1). 17.7.2  Molecular Imaging of Dendritic Cell Migration to Regional Nodes for Tumor Immunotherapy A second clinical application of high resolution MRI and ultrasmall superparamagnetic iron oxide nanoparticles has been the in vivo monitoring of cellular therapies for treating tumors. The success of tumor immunotherapy is predicated on the accurate delivery of cells to the appropriate target organs. The ability of this technique to precisely target cell delivery, track cell migration, and serially evaluate living subjects over time is enabling its translation from bench to bedside. As dendritic cells are known to play a crucial role in the initiation of an endogenous immune response, they have recently generated considerable interest as a means to enhance the immune response against tumor cells. However, in order to effectively stimulate the immune system, the delivery and subsequent migration of cells to regional lymph nodes is crucial. In addition to providing detailed anatomic information, MR detection and tracking of small numbers of autologous dendritic cells (i.e., 1.5 × 105), magnetically labeled with a clinical SPIO formulation (Endorem), was shown to be feasible in stage III melanoma patients (de Vries et al. 2005). Assessment of intranodal and internodal cell migration patterns using a 3.T magnetic resonance system and a combination of SE and GRE sequence acquisition was found to be superior to scintigraphic detection methods using indium-111 labeled oxine following intranodal co-injection of these probes. A comparison of gradient echo images acquired before and after intranodal injection of SPIO-labeled dendritic cells demonstrated significant decreases in signal intensity at the site of injection, while spin echo images were relatively insensitive to the induced magnetic susceptibility effects (Fig. 17.7.2). On GE imaging, nodes were observed to be hyperintense (white) prior to injection of labeled cells, while the same nodes were seen to be dark gray on SE imaging. The extent of signal loss on SE imaging after cellular injection, however, was not as great as that observed on GE acquisitions. Thus, by comparing relative differences in the magnitude of the signal changes using the iron-sensitive gradient echo and iron-

insensitive spin echo sequences, the presence of iron in a lymph node can be established (Harisinghani et al. 2003). MRI was found to be at least as sensitive as scintigraphic imaging for detecting dendritic cell migration in vivo. Scintigraphic evaluations, however, were unable to provide detailed information on cellular delivery or the intranodal distribution patterns of dendritic cells. 17.7.3  Molecular Imaging of Thrombosis in Atherosclerotic Disease A third successful application of targeted molecular MR imaging has been based on the synthesis of highly specific probes that can carry a large payload of paramagnetic or superparamagnetic agent, in this case, perfluorocarbon lipid emulsions (Yu et al. 2000). However, the steric hindrance imposed by the large size of these particles (≈200 nm) can limit deep tissue penetration. Therefore, while these particle platforms have been used to provide important functional information on biological processes associated with the process of atherogenesis, their application has been largely restricted to cases favoring direct interaction and binding with accessible molecules on cellular surfaces. These applications include imaging ανβ3 integrin expressed in angiogenesis, endothelial cell adhesion molecules associated with inflammation and fibrin and platelet aggregation involved in thrombosis. For instance, derivatizing these perfluorocarbon nanoparticles with a large number of gadolinium ions (90,000 per particle) has enabled excellent detection and delineation of fibrin while, at the same time, limiting the degree of blood-pool contrast. Moreover, the relative lack of particle penetration into dense fibrin clots serves to underscore the excellent T1-based contrast that can be achieved with the binding of only a single particle layer to the periphery of the clot (Yu et al. 2000; Flacke et al. 2001). It has also been possible to target platelet thrombus in injured vessels of pre-clinical models through the glycoprotein aIIbb3 receptors present on the surface of activated platelets. Ultrasmall SPIO nanoparticles, conjugated to the arginine–glycine–aspartic acid (RGD) peptide, have previously been used to facilitate thrombus visualization; however, the use of this probe in the clinical setting will likely be limited by the need for high spatial resolution in order to achieve adequate detection sensitivity (0.2 × 0.2 × 1 mm) (Johansson et al. 2001). Alternatively, a recently developed paramagnetic probe which contains gadolinium diethylenetriamine pentaacetic acid (Gd-DTPA) conjugated to small-molecule peptide derivatives can bind to fibrin (EPIX Medical, Inc.), and has been shown to penetrate thrombi. This probe, which contains four gadolinium molecules per peptide, can be injected at low doses to achieve sufficient enhancement for thrombus visualization using MRI (Fig. 17.7.3) (Sirol et al. 2005).

17.7  Clinical and Near-Clinical Molecular Imaging Applications Using Targeted Probes

Fig. 17.7.2a–n  In vivo 111In scintigraphy and MRI with SPIOs. a–c Monitoring of the delivery of dendritic cells labeled with SPIO and 111In by MRI before and after intranodal injection in patient 1. (a) Gradient echo transversal magnetic resonance image before vaccination showing a right inguinal lymph node with a hyperintense signal area (1). b SE (technique much less sensitive for SPIO) transverse magnetic resonance image obtained from the same lymph node after vaccination. c Gradient-echo transverse magnetic resonance images after vaccination in same position as b, showing a decreased signal intensity of lymph node 1. d–n Monitoring of in vivo migration of SPIO and 111In-labeled dendritic cells with MRI and scintigraphy after injection in a right inguinal lymph node in patient 3. d In vivo scintigraphy 2 days after vaccination, showing migration of the dendritic cells from the injection lymph node (1) to three following lymph nodes (2–4). e–n)Five image pairs of a coronal gradient-echo and SE image 2 days after vaccination, showing

migration of the dendritic cells from the injection lymph node 1 (e, f) to four following lymph nodes (g–n). Open arrows indicate lymph nodes that do not contain SPIO; on the SE images these nodes are dark gray; on gradient echo images they are white. Closed arrows indicate lymph nodes that are positive for SPIO in the gradient-echo magnetic resonance image. On gradient-echo images SPIO-containing lymph nodes have decreased signal intensity compared with SE images. The concentration of SPIO in lymph node 1 was very high, resulting even in decreased signal intensity in the SE image. The lymph node that was identified by the scintigraphy as the injection lymph node (lymph node 1 in d) actually consisted of two distinct lymph nodes as evidenced by MRI (lymph nodes 1 and 5) (Reproduced with permission from de Vries IJ, Lesterhuis WJ, Barentsz JO et al [2005] Magnetic resonance tracking of dendritic cells in melanoma patients for monitoring of cellular therapies. Nat Biotechnol 23:1407– 1413, Nature America, New York)

1429

1430

17  Systems Biology and Nanotechnology

Fig. 17.7.3  Specific nanoparticle targeting fibrin. In vivo magnetic resonance (MR) thrombus imaging using a fibrin-specific contrast agent. An image acquired before contrast injection is shown on the left. Thrombus in the injured left common carotid artery (arrows) demonstrates dramatic T1-weighted contrast enhancement (right). The contralateral, uninjured carotid artery

does not display enhancement (Reproduced with permission from Sirol M, Fuster V, Badimon JJ et al. [2005] Chronic thrombus detection with in vivo magnetic resonance imaging and a fibrin-targeted contrast agent. Circulation 112:1594–1600, Lippincott Williams & Wilkins, Hagerstown)

References

7.

1.

2.

3.

4.

5.

6.

Aboagye EO, Luthra SK, Brady F, Poole K, Anderson H, Jones T, Boobis A, Burtles SS, Price P (2002) Cancer research UK procedures in manufacture and toxicology of radiotracers intended for pre-phase I positron emission tomography studies in cancer patients. Br J Cancer 86:1052–1056 Aebersold R, Anderson L, Caprioli R, Druker B, Hartwell L, Smith R (2005) Perspective: a program to improve protein biomarker discovery for cancer. J Proteome Res Jul-Aug; 4: 1104–1109 Anderson SA, Wickline SA, Kotyk JJ (2000) Magnetic resonance contrast enhancement of neovasculature with ανβ3targeted nanoparticles. Magn Res Med 44:433–439 Anderson SA, Glod J, Arbab AS, Noel M, Ashari P, Fine HA, Frank JA (2005) Noninvasive MR imaging of magnetically labeled stem cells to directly identify neovasculature in a glioma model. Blood 105:420–425 Backmann N, Zahnd C, Huber F, Bietsch A, Plückthun A, Lang H-P, Güntherodt H-J, Hegner M, Gerber C (2005) A label-free immunosensor array using single chain antibody fragments. Proc Natl Acad Sci USA 102:14587–14592 Barrett MT, Sanchez CA, Prevo LJ, Wong DJ, Galipeau PC, Paulson TG, Rabinovitch PS, Reid BJ (1999) Evolution of neoplastic cell lineages in Barrett’s esophagus. Nat Genet 22:106–109

8.

9.

10.

11.

12.

13.

14.

Betensky RA, Louis DN, and Cairncross JG (2002) Influence of unrecognized molecular heterogeneity on randomized clinical trials. J Clin Oncol 20:2495–2499 Brody JP, Osbom TD, Forster FK, Yager P (1996) A planar microfabricated fluid filter. Sens Actuators A Phys 54:704–708 Bunimovich Y, Shin YS, Yeo WS, Amori M, Kwong G, Heath JR (2006) Quantitative real-time measurements of DNA hybridization with alkylated nonoxidized silicon nanowires in electrolyte solution. J Am Chem Soc 128:16323–16331 Bunimovich YL, Ge G, Beverly KC, Ries RS, Hood L, Heath JR (2004) Electrochemically programmed, spatially selective biofunctionalization of silicon wires. Langmuir 20:10630–10368 Chatterjee SK, Zetter BR (2005) Cancer biomarkers: knowing the present and predicting the future. Future Oncol 1:37–50 Cheng MM, Cuda G, Bunimovich YL, Gaspari M, Heath JR, Hill HD, Mirkin CA, Nijdam AJ, Terracciano R, Thundat T, Ferrari M (2006) Nanotechnologies for biomolecular detection and medical diagnostics. Curr Opin Chem Biol 10:11–19 Choi Y, Baker JR (2005) Targeting cancer cells with DNAassembled dendrimers: a mix and match strategy for cancer. Cell Cycle 4:669–671 Choudhury RP, Fuster V, Fayad ZA (2004) Molecular, cellular, and functional imaging of atherosclerosis. Nat Rev Drug Discov 3:913–925

17.7  Clinical and Near-Clinical Molecular Imaging Applications Using Targeted Probes 15. Cortez-Retamozo V, Backmann N, Senter PD, Wernery U, Be Baetselier P, Muyldermans S, Revets H (2004) Efficient cancer therapy with a nanobody-based conjugate. Cancer Res 64:2853–2857 16. Cummins JM, VelculescuVE (2006) Implications of micro-RNA profiling for cancer diagnosis. Oncogene 25:6220–6227 17. Dahan M, Levi S, Luccardini C, Rostaing P, Riveau B, Triller A (2003) Diffusion dynamics of glycine receptors revealed by single quantum dot tracking. Science 302:442–445 18. Dufes C, Keith WN, Bilsland A, Proutski I, Uchegbu IF, Schatzlein AG (2005) Synthetic anticancer gene medicine exploits intrinsic antitumor activity of cationic vector to cure established tumors. Cancer Res 65:8079–8084 19. Ferrari M (2005) Cancer nanotechnology: opportunities and challenges. Nat Rev Cancer 5:161–171 20. Flacke S, Fischer S, Scott MJ, Fuhrhop RJ, Allen JS, McLean M, Winter P, Sicard GA, Gaffney PJ, Wickline SA, Lanza GM (2001) Novel MRI contrast agent for molecular imaging of fibrin: implications for detecting vulnerable plaque. Circulation 104:1280–1285 21. Fortina P, Kricka LJ, Surrey S, Grodzinski P (2005) Nanobiotechnology: the promise and reality of new approaches to molecular recognition. Trends Biotechnol 23:168–173 22. Gao X, Yuanyuan C, Levenson RM, Chung LWK, Nie S (2004) In vivo cancer targeting and imaging with semiconductor quantum dots. Nat Biotech 8:969–976 23. Gao X, Yang L, Petros JA, Marshall FF, Simons JW, Nie S (2005) In vivo molecular and cellular imaging with quantum dots. Curr Opin Biotechnol16:63–72 24. Goessl C (2003) Noninvasive molecular detection of cancer—the bench and the bedside. Curr Med Chem 10:691–706 25. Grimm J, Perez JM, Josephson L, Weissleder R (2004) Novel nanosensors for rapid analysis of telomerase activity. Cancer Res 64:639–643 26. Guller U, Nitzsche E, Moch H, Zuber M (2003) Is positron emission tomography an accurate non-invasive alternative to sentinel lymph node biopsy in breast cancer patients? J Natl Cancer Inst 95:1040–1043 27. Hanahan D, Weinberg RA (2000) The hallmarks of cancer. Cell 100:57–70 28. Harisinghani MG, Barentsz J, Hahn PF et al (2003a) Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 3:2491–2499 29. Hayflick L (1007) Mortality and immortality at the cellular level. Biochemistry 62:639–43 30. Heath JR (2007) Chap. 12. In: Mirkin C, Niemeyer CM (eds) Label-free nanowire and nanotube biomolecular sensors for in vitro diagnostics of cancer and other diseases, nanobiotechnology. Wiley, New York, p 215 31. Heath JR, Phelps ME, Hood L (2003) NanoSystems biology. Mol Imaging Biol 5:312–325 32. Hegner M, Arntz Y (2004) Advanced biosensing using micromechanical cantilever arrays. Methods Mol Biol 2004; 242:39–49

33. Helms B, Meijer EW (2006) Dendrimers at work. Science 313:929–930 34. Hernandez J, Thompson I (2004) Prostate-specific antigen: a review of the validation of the most commonly used cancer biomarker. Cancer 101:894–904 35. Hirsch LR, Stafford RJ, Bankson JA, Sershen SR, Rivera B, Price RE, Hazle JD, Halas NJ, West JL (2003a) Nanoshellmediated near-infrared thermal therapy of tumors under magnetic resonance guidance. Proc Natl Acad Sci USA. 100:13549–13554 36. Hirsch LR, Jackson JB, Lee A, Halas NJ, West JL (2003b) A whole blood immunoassay using gold nanoshells. Anal Chem 75:2377–2381 37. Hood L, Perlmutter RM (2004) The impact of systems approaches on biological problems in drug discovery. Nat Biotechnol 22:1215–1217 38. Hood L, Heath JR, Phelps ME, Lin B (2004) Systems biology and new technologies enable predictive and preventative medicine. Science 306:640–643 39. Hughes T, Branford S (2003) Molecular monitoring of chronic myeloid leukemia. Semin Hematol 40(Suppl 2):62–68 40. Ichikawa T, Hoegemann D, Saeki Y, Tyminski E, Terada K, Weissleder R, Chiocca EA, Basilion JP (2002) MRI of transgene expression: correlation to therapeutic gene expression. Neoplasia 4:523–530 41. Ideker T, Thorsson V, Ranish JA, Christmas R, Buhler J, Eng JK, Bumgarner R, Goodlett DR, Aebersold R, Hood L (2001) Integrated genomic and proteomic analyses of a systemically perturbed metabolic network. Science 292:929–934 42. Ihre HR, Padilla De Jesus OL, Szoka FC, Frechet JM (2002) Polyester dendritic systems for drug delivery applications: design, synthesis, and characterization. Bioconjug Chem 13:443–452 43. Isaacs JT (1999) The biology of hormone refractory prostate cancer. Why does it develop? Urol Clin N Am 26:263–273 44. Jaffer FA, Weissleder R (2004) Seeing within: molecular imaging of the cardiovascular system. Circ Res 94:433–445 45. Jaffer FA, Weissleder R (2005) Molecular imaging in the clinical arena. JAMA 293:855–862 46. Jain KK (2004a) Role of oncoproteomics in the personalized management of cancer. Expert Rev Proteomics 1:49–55 47. Jain KK (2004b) Applications of biochips: from diagnostics to personalized medicine. Curr Opin Drug Discov Dev 7:285–289 48. Jain KK (2005a) The role of nanobiotechnology in drug discovery. DDT 10:1435–1442 49. Jain KK (2005c) Nanotechnology-based drug delivery for cancer. Technol Cancer Res Treat 4:407–416 50. Jain KK (2005d) Personalized medicine for cancer: from drug development into clinical practice. Expert Opin Pharmacother 6:1463–1476

1431

1432

17  Systems Biology and Nanotechnology 51. Jain KK (2005b) Role of nanobiotechnology in developing personalized medicine for cancer. Technol Cancer Res Treat 4:645–650 52. Johansson LO, Bjornerud A, Ahlstrom HK, Ladd DL, Fuji DK (2001) A targeted contrast agent for magnetic resonance imaging of thrombus: implications of spatial resolution. J Magn Reson Imaging 13:613–618 53. Kallioniemi OP (2001) Biochip technologies in cancer research. Ann Med 33:142–147 54. Kang HW, Josephson L, Petrovsky A, Weissleder R, Bogdanov A Jr (2002) Magnetic resonance imaging of inducible E-selectin expression in human endothelial cell culture. Bioconjug Chem 13:122–127 55. Khalil IG, Hill C (2005) Systems biology for cancer. Curr Opin Oncol 17:44–48 56. Kim J, Park S, Lee JE, Jin SM, Lee JH, Lee IS, Yand I, Kim J-S, Kim SK, Cho M-H, Hyeon T (2006) Designed fabrication of multifunctional magnetic gold nanoshells and their application to magnetic resonance imaging and photothermal therapy. Agnew Chem Int Ed 45:1–6 57. Kim S, Lim YT, Soltesz EG, DeGrand AM, Lee J, Nakayama A, Parker JA, Mihaljevic T, Laurence RG, Dor DM, Cohn LH, Bawendi MG, Frangioni JV (2004) Near-infrared fluorescent type-II quantum dots for sentinel lymph node mapping. Nat Biotechnol 22:93–97 58. Kircher MF, Allport J, Graves EE, Love V, Josephson L, Licht­man A, Weissleder R (2003) In vivo high resolution 3D imaging of antigen-specific cytotoxic T-lymphocyte trafficking to tumors. Cancer Res 63:6838–6846 59. Kitano H (2002) Systems biology: a brief overview. Science 295:1662–1664 60. Kooi ME, Cappendijk VC, Cleutjens KB, Kessels AG, Kitslaar PJ, Borgers M, Frederik PM, Daemen MJ, van Engelshoven JM (2003) Accumulation of ultrasmall superparamagnetic particles of iron oxide in human atherosclerotic plaques can be detected by in vivo magnetic resonance imaging. Circulation 107:2453–2458 61. Lamb J, Crawford ED, Peck D, Modell JW, Blat IC, Wrobel MJ, Lerner J, Brunet JP, Subramanian A, Ross KN, Reich N, Hieronymus H, Wei G, Armstrong SA, Haggarty SJ, Clemons PA, Wei R, Carr SA, Lander ES, Golub TR (2006) The Connectivity Map: using gene expression signatures to connect small molecules, genes, and disease. Science 313:1929–1935 62. Langereis S, de Lussanet QG, van Genderen MH, Meijer EW, Beets-Tan RG, Griffioen AW, van Engelshoven JM, Backes WH (2006) Evaluation of Gd(III)DTPA-terminated poly(propylene imine) dendrimers as contrast agents for MR imaging. NMR Biomed 19:133–141 63. Lee KB, Park SJ, Mirkin CA, Smith JC, Mrksich M (2002) Protein nanoarrays generated by dip-pen nanolithography. Science 295:1702–1705 64. Li KCP, Pandit SD, Guccione S, Bednarski MD. Molecular imaging applications in nanomedicine (2004) Biomed. Microdevices 6:113–116

65. Lin B, White T, Lu W, Xie T, Utleg AG, Yan X, Yi EC, Shannon P, Khrebtukova I, Lange PH, Goodlett DR, Zhou D, Vasicek TJ, Hood L (2005) Evidence for the presence of disease-perturbed networks in prostate cancer cells by genomic and proteomic analyses: a systems approach to disease. Cancer Res 65:3081–3091 66. Loo C, Lowery A, Halas N, West J, Drezek R (2005) Immunotargeted nanoshells for integrated cancer imaging and therapy. Nano Lett 5:709–711 67. Majoros IJ, Myc A, Thomas T, Mehta CB, Baker JR 92006) PAMAM dendrimer-based multifunctional conjugate for cancer therapy: synthesis, characterization, and functionality. Biomacromolecules 7:572–579 68. Malik N et al (2006) US patent 7005124 69. Martin M (2006) Molecular biology of breast cancer Clin Transl Oncol 8:7–14 70. Massoud TF, Gambhir SS (2003) Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev 17:545–580 71. McKendry R, Zhang J, Arntz Y, Strunz T, Hegner M, Lang HP, Baller MK, Certa U, Meyer E, Guntherodt J, Gerber C (2002) Multiple label-free biodetection and quantitative DNA binding assays on a nanomechanical cantilever array. Proc Natl Acad Sci USA 99:9783–9788 72. Melosh NA, Boukai A, Diana F, Gerardot B, Badolato A, Petroff PM, Heath JR (2003) Ultrahigh density nanowire lattices and circuits. Science 300:112–115 73. Morawski AM, Winter PM, Crowder KC, Caruthers SD, Fuhrhop RW, Scott MJ, Robertson JD, Abendschein DR, Lanza GM, Wickline SA (2004) Targeted nanoparticles for quantitative imaging of sparse molecular epitopes with MRI. Magn Reson Med 51:480–486 74. Mulder WJM, Koole R, Brandwijk RJ, Storm G, Chin PTK, Strijkers GJ, Mello Donega C, Nicolay K, Griffioen AW (2006) Quantum dots with a paramagnetic coating as a bimodal molecular imaging probe. Nano Lett 6:1–6 75. Muruganandam A, Tanha J, Narang S, Stanimirovic D (2001) Selection of phage-displayed llama single-chain domain antibodies that transmigrate across human bloodbrain barrier endothelium. FASEB J 16:240–242 76. Ozkan M (2004) Quantum dots and other nanoparticles: what can they offer to drug discovery? DDT 9:1065–1071 77. Padilla De Jesus OL, Ihre HR, Gagne L, Frechet JM, Szoka FC (2002) Polyester dendritic systems for drug delivery applications: in vitro and in vivo evaluation. Bioconjug Chem 13:453–461 78. Pantoja R, Nagarah JM, Starace DM, Melosh NA, Blunck R, Bezanilla F, Heath JR (2004) Silicon chip-based patch clamp electrodes integrated with PDMS microfluidics. Biosens Bioelectron 20:509–517 79. Parak WJ, Boudreau R, LeGros M, Gerion D, Zanchet D, Micheel CM, Williams SC, Alivisatos SP, Larabell C (2002) Cell motility and metastatic potential studies based on quantum dot imaging of phagokinetic tracks. Adv Mater 14:882–885

17.7  Clinical and Near-Clinical Molecular Imaging Applications Using Targeted Probes 80. Peano C, Severgnini M, Cifola I, DeBellis G, Battaglia C (2006) Transciptome amplification methods in gene expression profiling. Expert Rev Mol Diagn 6:465–480 81. Radich JP, Dai H, Mao M, Oehler V, Schelter J, Druker B, Sawyers C, Shah N, Stock W, Willman CL, Friend S, Linsley PS (2006) Gene expression changes associated with progression and response in chronic myeloid leukemia. Proc Natl Acad Sci USA 103:2794–2799 82. Revets H, De Baetselier P, Muyldermans S (2005) Nanobodies as novel agents for cancer therapy. Expert Opin Biol Ther 5:111–124 83. Rohde RD, Agnew HD, Yeo W-S, Heath JR (2006) A Non-Oxidative approach towards chemically and electrochemically biofunctionalizing Si(111). J Am Chem Soc 128:9518–9525 84. Roovers RC, Laeremans T, Huang L, De Taeye S, Verkleij AJ, Revets H, de Haard HJ, van Bergen En Henegouwen PM (2007) Efficient inhibition of EGFR signaling and of tumour growth by antagonist-EGFR nanobodies. Cancer Immunol Immunother 56:303–317 85. Rudin M, Weissleder R (2003) Molecular imaging in drug discovery and development. Nat Rev Drug Discov 2:123–131 86. Saerens D, Kinne J, Bosmans E, Wernery U, Muydermans S, Conrath K (2004) Single-domain antibodies derived from dromedary lymph node and peripheral blood lymphocytes sensing conformational variants of prostate specific antigen. J Biol Chem 279:51965–51972 87. Sahoo SK, Labhasetwar V (2003) Nanotech approaches to drug delivery and imaging. DDT 8:1112–1120 88. Schellenberger EA, Bogdanov A, Hogemann D, Tait J, Weissleder R, Josephson L (2002) Annexin V-CLIO: a nanoparticle for detecting apoptosis by MRI. Mol Imaging 1:102–107 89. Schmieder AH, Winter PM, Caruthers SD, Harris TD, Williams TA, Allen JS, Lacy EK, Zhang H, Scott MJ, Hu G, Robertson JD, Wickline SA, Lanza GM (2005) Molecular MR Imaging of Melanoma Angiogenesis with anb3-targeted paramagnetic nanoparticles. Magn Res Med 53:621–627 90. Sipkins DA. Cheresh DA, Kazemi MR, Nevin LM, Bednarski MD, Li KCP. Detection of tumor angiogenesis in vivo by anb3-targeted magnetic resonance imaging. Nat. Med. 1998 May; 4:623–626 91. Sirol M, Fuster V, Badimon JJ et al (2005) Chronic thrombus detection with in vivo magnetic resonance imaging and a fibrin-targeted contrast agent. Circulation 112:1594–1600 92. Smith AM, Dave S, Nie S, True L, Gao X (2006) Multicolor quantum dots for molecular diagnostics of cancer. Expert Rev Mol Diagn 6:231–244 93. Sukhanova A, Venteo L, Devy M, Artemyev M, Oleinikov V, Pluot M, Nabiev I (2002) Highly stable fluorescent nanocrystals as a novel class of labels for immunohistochemical analysis of paraffin-embedded tissue sections. Lab Invest 82:1259–1261

94. Sukhanova A, Devy M, Venteo L, Kaplan H, Artemyev M, Oleinikov V, Klinov D, Pluot M, Cohen JH, Nabiev I (2004) Biocompatible fluorescent nanocrystals for immunolabeling of membrane proteins and cells. Anal Biochem 324:60–67 95. Sullivan DC, Ferrari M (2004) Nanotechnology and tumor imaging. Mol Imaging 3:364–369 96. Talanov VS, Regino CA, Kobayashi H, Bernardo M, Choyke PL, Brechbiel MW (2006) Dendrimer-based nanoprobe for dual modality magnetic resonance and fluorescence imaging. Nano Lett 6:1459–1463 97. Thrall J (2004) Personalized medicine. Radiology 231:613–616 98. Tripathi PK (2002) Dendrimer grafts for delivery of 5-fluorouracil. Pharmazie 57:261–264 99. VanDelinder V, Groisman A (2006) Separation of plasma from whole human blood in a continuous cross-flow in a molded microfluidic device. Anal Chem 78:3765–3771 100. Vries IJ de, Lesterhuis WJ, Barentsz JO et al (2005) Magnetic resonance tracking of dendritic cells in melanoma patients for monitoring of cellular therapies. Nat Biotechnol 23:1407–1413 101. Wagner PD, Maruvada P, Srivastava S (2004) Molecular diagnostics: a new frontier in cancer prevention. Expert Rev Mol Diagn 4:503–511 102. Weber BL (2002) Cancer genomics. Cancer Cell 1:37–47 103. Weissleder R (2002) Scaling down imaging: molecular mapping of cancer in mice. Nat Rev Cancer 2:11–18 104. Weinberg R (2006) The biology of cancer. Garland, New York 105. Weissleder R, Kelly K, Sun EY, Shtatland T, Josephson L (2005) Cell-specific targeting of nanoparticles by multivalent attachment of small molecules. Nat Biotechnol 23:1418–1423 106. Weston AD, Hood L (2004) Systems biology, proteomics, and the future of health care: toward predictive, preventative, and personalized medicine. J Proteome Res 3:179–196 107. Winter PM, Lanza GM, Wickline SA (2003) Molecular imaging of angiogenesis in early-stage atherosclerosis with anb3-integrin-targeted nanoparticles. Circulation 108:2270–2274 108. Wu X, Liu H, Liu J, Haley KN, Treadway JA, Larson JP, Ge N, Peale F, Bruchez MP (2003) Immunofluorescent labeling of cancer marker Her2 and other cellular targets with semiconductor quantum dots. Nat Biotechnol 21:41–46 109. Yu X, Song SK, Chen J, Scott MJ, Fuhrhop RJ, Hall CS, Gaffney PJ, Wickline SA, Lanza GM (2000) High-resolution MRI characterization of human thrombus using a novel fibrin-targeted nanoparticle contrast agent. Magn Reson Med 44:867–872 110. Zimmermann M, Delamarche E, Wolf M, Hunziker P (2005) Modeling and optimization of high-sensitivity, lowvolume microfluidic-based surface immunoassays. Biomed Microdevices 7:99–110

1433

Glossary

Terms and Definitions of Magnetic Resonance Imaging (MRI) and Magnetic Resonance Spectroscopy (MRS) 180° Pulse. (→) Pulse, 180° (π pulse).

Acquisition time. (→) Image acquisition time.

2D MR. (→) Two-dimensional MR.

Acquisition window. Time in the MR pulse sequence during which the MR signal is recorded. The duration can be denoted TAD (for time of analog to digital conversion).

2D Fourier transform imaging. (→) Two dimensional Fourier transform imaging (2DFT). 3D Fourier transform imaging. (→) Three-dimensional Fourier transform imaging. 90° Pulse. (→) Pulse, 90° (π/2 pulse). Absorption mode. Component of the MR signal that yields a symmetric, positive-value line shape.

Active shielding. Magnetic shielding through the use of secondary shielding coils designed to produce a magnetic field that cancels the field from primary coils in regions where it is not desired, e.g., outside the bore of the magnet. These active shielding coils may be located inside the magnet cryostat. Active shielding can be applied to the main magnet or to the gradient magnetic fields. See also magnetic shielding, self-shielding, and room shielding.

Acceleration factor. The multiplicative term by which faster imaging pulse sequences such as multiple-echo im- Active shimming. Shimming is the process of making aging reduce total imaging time compared with conven- the magnetic field more uniform by suitably adjusting the currents in shim coils. tional imaging sequences such as spin echo imaging. Acoustic noise. Vibrations of the gradient coil support structures create sound waves. These vibrations are caused by interactions of the magnetic field created by pulses of the current through the gradient coil with the main magnetic field in a manner similar to a loudspeaker coil. Sound pressure is reported on a logarithmic scale called sound pressure level expressed in decibels referenced to the weakest audible 1,000 Hz sound pressure of 2 × 10–5 Pa. Sound level meters contain filters that simulate the ear’s frequency response. The most commonly used filter provides what is called A weighting, with the letter A appended to the dB units, i.e., dBA.

Adiabatic fast passage (AFP). Technique of producing rotation of the (→) macroscopic magnetization vector by sweeping the (→) frequency of an irradiating RF wave (or (→) strength magnetic field) through (→) resonance (the (→) Larmor frequency) in a time short compared with the (→) relaxation times. Particularly used for (→) inversion of the (→) spins. A continuous wave MR technique.

Aliasing. Consequence of sampling the MR signal in which any components of the signal that are at a higher frequency than the Nyquist limit will be “folded” in the spectrum so that they appear to be at a lower frequency. In Fourier transform imaging this can produce an apparAcquisition matrix. Number of independent data ent wrapping around to the opposite side within the imsamples in each direction, e.g., in 2DFT imaging it is age of a portion of the object, which extends beyond the the number of samples in the phase-encoding and fre- edge of the reconstructed region. quency-encoding directions, and in reconstruction from projections imaging it is the number of samples in time Analog to digital converter (ADC). Electronic device, and angle. The acquisition matrix may be asymmetric converting analogous signals—e.g., (→) MR signals—into and of different size than the reconstructed image or digital values, which can be processed by (→) computers display matrix, e.g., with zero filling or interpolation, or and stored in the computer memory. (for asymmetric sampling) by exploiting the symmetry of the data matrix. For symmetric sampling, the acquisition Angiography. Application of MRI to produce images of matrix will roughly equal the ratio of image field of view blood vessels, for example with flow effect or relaxation to spatial resolution along the corresponding direction time differences. Some common approaches use the (depending on filtering and other processing). washout of saturated spins from a region by blood flow

1436

Glossary

to increase the relative intensity of blood vessels within images or use the variable sensitivity to motion-induced phase shifts provided by adjusting gradient moments to discriminate against the signal from stationary tissue.

Artifact. Image distortions produced by different causes, e.g., ferromagnetic implants, electronic interferences etc.

Angular momentum. A vector quantity given by the vector product of the momentum of a particle and its position vector. In the absence of external forces, the angular momentum remains constant, with the result that any rotating body tends to maintain the same axis of rotation. In the presence of a (→) torque applied to a rotating body in such a way as to change the direction of the rotation axis, the resulting change in angular momentum results in (→) precession. Atomic nuclei possess an intrinsic angular momentum referred to as (→) spin, measured in multiples of Planck’s constant.

Attenuation. Reduction of power, e.g., due to passage through a medium or electrical component. Attenuation in electrical systems is commonly expressed in decibels.

Asymmetric sampling. The collection of more data points on one side of the k-space origin than on the other. Angular frequency (ω). Frequency of oscillation or ro- With fewer k-space data points prior to the center (echo) a tation (measured, e.g., in radians/second) commonly shorter echo time can be attained. Asymmetric acquisition designated by the Greek letter omega. ω = 2πf, where f is in any phase encoding direction followed by partial-Foufrequency in Hertz (Hz). rier reconstruction leads to a reduction in imaging time.

Attenuator. Device which reduces a signal by a specific amount, commonly given in decibels.

Autotuning. A means for optimizing the tuning and matching of RF coils under different loading conditions without operator intervention. For large, high-power coils such as (→) body coils, autotuning involves the adjustment of variable capacitors using electric or hydrauAnnotation. A description of the factors used in creat- lic motors. For low power coils, the tuning elements are ing an MR image. Appropriate annotation should include most often variable capacitance diodes (varactors) fed by the type and times of the pulse sequence, the number of a computer-generated variable voltage. signals averaged or added (NSA), the size of the reconstructed region, the size of the acquisition matrix in each Axial plane. (→) Transverse plane or transaxial plane. direction, and the slice thickness. B0. A conventional symbol for the constant magnetic (inAntenna. Device for transmitting and receiving electro- duction) field in an MR system (units in Tesla). Should magnetic radiation. To produce the nuclear magnetic be distinguished from (→) magnetic field strength H0. resonance signal only the magnetic fraction of the electromagnetic radiation is of interest. (→) Coil and (→) res­ Balanced gradient. A gradient waveform that will act on onator are often used synonymously for antenna. any stationary spin on resonance between two consecutive RF pulses and return it to the same phase it had beApodization. Multiplication of acquired MR data by a fore the gradients were applied. function smoothly tapering off at higher spatial frequencies so as to reduce “ringing” artifacts near edges in the Balanced steady-state free precession. An MR gradicorresponding image or spectrum due to truncation and ent echo pulse sequence designed to produce contrast Gibbs phenomenon. It is a form of filtering. weighted by the T2/T1 ratio, with higher SNR and reduced artifacts compared with SSFP. Typically, TR is set to be as Array coil. RF coil composed of multiple separate elem­ short as possible (short compared with the T2 values of ents that can be used individually (switchable coil) or the tissues of interest), TE is intermediate (approximately used simultaneously. When used simultaneously, the TR/2), and a flip angle of 45° to 90° is used to result in T2/ elements can either be (1) electrically coupled to each T1-weighted SSFP images. Balanced SSFP sequences use other (coupled array coils), through common transmis- specific “balanced” gradients to return the magnetization sion lines or mutual inductance, or (2) electrically iso- to the same phase it had before the gradients were aplated from each other (isolated array coils), with separate plied, thus increasing signal and reducing artifacts. Spetransmission lines and receivers and minimum effective cific vendor names for this sequence include true Flash mutual inductance, and with the signals from each trans- imaging with steady-state precession (trueFISP), fast immission line processed independently or at different fre- aging employing steady-state acquisition (FIESTA), and quencies. balanced fast field echo (balanced FFE). See also: Steadystate free precession. Array processor. Optional component of the computer system specially designed to speed up numerical calcula- Bandwidth. A general term referring to a range of fretions like those needed in MR imaging. quencies (e.g., contained in a signal or passed by a signal processing system).

Glossary

Baseline correction. Processing of the MR spectrum to levels with corresponding energies, E1 and E2, is given by suppress baseline deviations from zero that may be super­ N1/N2 = exp [–(E1 – E2)]/kT, where k is Boltzmann’s conimposed on desired spectral lines. These deviations may stant and T is absolute temperature. For example, in MR be due either to various instrumental effects or to very of protons at room temperature in a magnetic field of broad spectral lines. 0.25 T, the difference in relative numbers of spins aligned with the magnetic field and against the field is about one Baseline. A generally smooth background curve with re- part in a million; the small excess of nuclei in the lower spect to which either the integrals or peak heights of the energy state is the basis of the net (→) magnetization and resonance spectral lines in the spectrum are measured. the (→) resonance phenomenon. Birdcage coil. An RF volume coil designed to produce a homogeneous B1 field by using multiple parallel conductors symmetrically spaced around the surface of a cylinder, connected by end rings. These are turned into low-pass or high-pass filter sections by adding capacitors in each conductor, or between each conductor in the end rings, so that at resonance there is a resulting homo­gen­ eous B1 field. When the B1 field is circularly polarized, the structure can be used as a quadrature coil. Bloch’s equation. Phenomenological “classical” equations of motion for the (→) macroscopic magnetization vector. They include the effects of (→) precession about the (→) magnetic field (static and RF) and the (→) T1 and T2 relaxation times. Blood-oxygen level-dependent effect (BOLD). A change in MRI-measurable signal caused by changes in the amount of oxygenated hemoglobin available in the venous circulation of the brain. Oxygenated hemoglobin has a smaller magnetic susceptibility than deoxygenated hemoglobin does. Neural activity causes replacement of deoxygenated hemoglobin by oxygenated hemoglobin, which has higher T2* due to its smaller magnetic susceptibility. As T2* increases, higher signal is measured on T2*weighted gradient echo images, yielding a positive signal of increased venous circulation.

Bolus tracking. A method of tracking moving (→) spins after tagging them (locally altering their (→) magnetization). The moving spins are then imaged at some time after tagging to see where the bolus (the small tagged volume of spins) has moved in the imaging plane. Cardiac gating. (→) Gating and synchronization, cardiac. Cardiac phase. A particular point in the cardiac cycle. Carr-Purcell (CP) sequence. (→) sequence with a (→) 90° RF pulse and several following (→)180° RF pulses for generating a train of (→) spin echoes; used for (→) T2 meas­ urements. Carr-Purcell-Meiboom-Gill (CMPG) sequence. Modification of (→) Carr-Purcell (→) RF pulse sequence with 90° phase shift in the (→) rotating frame of reference between the 90° pulse and the subsequent 180° pulses in order to reduce accumulating effects of imperfections in the 180° pulses. Suppression of effects of pulse error accumulation can alternatively be achieved by switching phases of the 180° pulses by 180°. Central processor unit. (→) Computer. Cerebral blood flow (CBF). The flow of capillary blood through the cortex, measured in units of flow (milliliters per minute) per unit mass of cortex.

Body coil. The body coil is installed in the magnet and has a large measurement field capable covering the body. The body coil typically functions both as transmit and as Cerebral blood volume (CBV). The volume of blood in a receiver coil as well ((→) transmit/receive coil). Com- a given volume of cerebral cortex, measured in units of pared with special coils (head coil, (→) surface coils), the volume. body coil has low SNR. The body coil often serves as the Chemical shift (δ). The change in the (→) Larmor fretransmit coil for special coils. quency of a given nucleus when bound in different sites Boil-off rate. Rate of cryogen evaporation in supercon- in a molecule, due to the magnetic shielding effects of the ducting magnets, usually measured in liters of liquid per electron orbitals. Chemical shifts make possible the difhour. It increases during ramping of the magnet and with ferentiation of different molecular compounds and difeddy currents induced in the cryoshields by pulsed-field ferent sites within the molecules in high-resolution MR gradients. In calculating cryogen consumption additional spectra. The amount of the shift is proportional to (→) transfer and filling losses have to be considered. magnetic field strength and is usually specified in parts per million (ppm) of the resonance frequency relative to Boltzmann distribution. If a system of particles which a standard. The actual frequency measured for a given are able to exchange energy in collisions is in (→) ther- (→) spectral line may depend on environmental factors mal equilibrium, then the relative number ((→) popula- such as effects on the local magnetic field strength due to tion) of particles, N1 and N2, in two particular (→) energy variations of (→) magnetic susceptibility.

1437

1438

Glossary

Chemical shift artifact. Image artifact of apparent spa- core ((→) solenoid coil). A coil may consist of one or more tial offset of regions with different chemical shifts along turns. The coil is equipped with electrical connection terthe direction of the frequency encoding gradient; a simi- minals (taps), a matching circuit and often a preamplifier lar effect may be found in the slice selection direction. is closely attached. The whole system is often called a coil (See also: RF-coil, saddle coil, birdcage coil). Chemical shift imaging (CSI). A magnetic resonance imaging technique that provides mapping of the regional Coil loading. In (→) MR imaging, the interaction of the distribution of intensity (images) of a specific range of (→) patient with the RF coil, which causes shifts of the res­on­ chemical shifts, corresponding to individual (→) spectral ance frequency and damping of the coil’s resonance and lines or groups of lines. The chemical shift can be treated hence reduction of the quality factor because of (→) magas an additional dimension to be reconstructed. For ex- netic induction and dielectric losses in the patient. ample, all spatial dimensions can be encoded with (→) phase encoding prior to signal acquisition and the signal Complex conjugate. An operation on a complex numthen acquired in the absence of (→) magnetic field gradi- ber which negates the sign of the imaginary component ents; recovery of spatial and chemical shift dimension can of a complex vector. The two vectors then form a combe achieved by appropriate (→) Fourier transformation of plex–conjugate pair. the resulting data set. Composite excitation. (→) Excitation of tissue created by Chemical shift reference. A compound with respect to a series of (→) pulses rather than by a single RF pulse. The whose (→) frequency the chemical shifts of other com- purpose of composite excitation is usually for the sum pounds can be compared. The standard can either be of these to produce net excitation of a target tissue, but internal or external to the sample. Because of the need for phase variations during the intervals between pulses for possible corrections due to differential magnetic sus- to cause destructive (→) interference, and therefore cause ceptibility between an external standard and the sample little or no excitation, of undesired tissue. Applications being measured, the use of an internal standard is gener- include selective excitation of water, or of tissues not subally preferred. ject to magnetization transfer. Chemical-shift spatial offset. (→) Chemical-shift artifact.

Computer. Modern computers consist of four main sections: the arithmetic and logic unit (ALU), the control Cine acquisition. The collection of images (usually at the unit, the memory, and the input and output devices (I/O) same spatial location) covering one full period of mo- all interconnected by busses. These parts are collectively tion or change, but which may be acquired over several known as a central processing unit (CPU). Furthermore, periods to obtain complete coverage. other peripheral devices as mass storage devices, interfaces, etc., are interconnected. These parts and devices Circularly polarized coil. A (→) coil designed to excite are called the computer’s (→) hardware. The computer’s or detect spins using two orthogonal transmit and/or specific feature is that it can be programmed. All instrucreceive channels. As a transmitter coil, there is a factor tions are combined and integrated in the software which of 2 reduction in power required. It theoretically yields a includes all instructions to run the CPU itself, to run the factor of 2 improvement in SNR over a linearly polarized peripheral hardware and specific devices, like an MR coil as a receiving coil. spectrometers interconnected to the computer by specific computer interfaces. Coherence. Maintenance of a constant (→) phase relationship between rotating or oscillating waves or objects. Continuous wave MR (CW). A technique for studying Loss of phase coherence of the (→) spins results in a de- MR by continuously applying RF radiation to the sample crease in the (→) transverse magnetization and hence a and slowly sweeping either the (→) RF frequency or the decrease in the (→) MR signal. In the quantum mechan­ (→) magnetic field through the (→) resonance values; now ical description of magnetic resonance, coherence refers largely superseded by pulse MR techniques. to a transition between different states of the spin system (See also: Multiple quantum coherence). Continuous wave NMR (CW). A technique for studying NMR by continuously applying RF radiation to the samCoherent. A state of a spin sample in which all spins in a ple and slowly sweeping either the RF frequency or the voxel are in-phase. magnetic field through the resonance values; now largely superseded by pulse MR techniques. Coil. An (electromagnetic coil) is an antenna system used to transmit and receive electromagnetic signals (in Contrast. In conventional radiography, contrast is deMR: FID-signal). The coil is manufactured of a conduc- fined as the difference of the signal intensities divided by tor (usually a copper wire) wound around a cylindrical the average signal intensity in two adjacent regions. In a

Glossary

general sense, we can consider image contrast, where the strength of the image intensity in adjacent regions of the image is compared, or object contrast, where the relative values of a parameter affecting the image (such as (→) spin density or (→) relaxation time) in corresponding adjacent regions of the object are compared: C = (SIA – SIB)/(SIA + SIB) where SIA and SIB are signal intensities for tissues A and B. Relating image contrast to object contrast is more difficult in MR imaging than in conventional radiography, as there are more object parameters affecting the image, and their relative contributions are very dependent on the particular imaging technique used. As in other kinds of imaging, image contrast in MRI will also depend on region size, as reflected through the (→) modulation transfer function (MTF) characteristics. The contrast between an object (e.g., lesion) and the background will also depend on the particular choice of designated background (e.g., fat, muscle, etc.).

Coupling constant. Spectral lines are split by spin–spin coupling into multiplets whose frequencies are separated by an amount depending on the coupling constant, J. The magnitude of J is independent of the strength of the applied magnetic field and is given in units of frequency, Hz.

Coronal plane. The plane defined by the head-to-foot and left-to-right directions in the human body. A stack of images acquired in the coronal plane separates images by their anterior-to-posterior locations ((→) orientation).

DC artifact. A bright point created in the image caused by a constant offset in signal intensity of all raw data points.

Coupling. (→) Spin–spin coupling. Crossed coil. RF coil pair arranged with their magnetic fields at right angles to each other in such a way as to minimize their mutual magnetic interaction. Cryogen. Very low temperature liquefied gas (helium or nitrogen) used to maintain superconducting magnets in a superconducting state. Cryomagnet. (→) Superconducting magnet.

Cryoshield. A metal cylinder surrounding a low temContrast agent. Substance administered to subject being perature object (e.g. Helium vessel in a cryomagnet) to imaged to alter selectively the image intensity of a parti­cu­ reduce heat transfer (transfer of thermal energy) from lar anatomical or functional region, typically by altering the room temperature environment to the cold object the (→) relaxation times, including T1, T2, and T2*. (here: Helium vessel) and hence preventing an excessive boil-off. Contrast-to-noise ratio (CNR). Ratio of the absolute difference in MR signal intensities (SIA, SIB) between two Cryoshielding. By cooling a metal cylinder surrounding (→) regions of interest to the level of fluctuations in inten- the Helium vessel in a superconducting magnet, reduced sity due to noise (standard deviation of the signals in the cryogen boil-off can be achieved. image): CNR = (SIA – SIB)/noise). Cryostat. An apparatus for maintaining a constant low Convolution differencing. A method of suppressing temperature (as by means of liquid helium). Requires broad underlying spectral lines in order to emphasize nar- vacuum chambers to help with thermal isolation. rower spectral lines. Strong smoothing of the spectrum (e.g., by severe negative exponential weighting of the time Data system. (→) Computer. data) will suppress the narrow lines but min­imally affect very broad ones; subtracting such a smoothed spectrum dB/dt. The rate of change of the (→) magnetic induction from the original will largely remove the contributions (magnetic field) (B) with time t. Because changing magfrom the broad lines. This provides a means of baseline netic fields can induce electrical currents, this is one area of potential concern for safety limits. correction.

Correlation time (τc). The characteristic time between significant fluctuations in the local magnetic field ex­per­ ienced by a spin due to molecular motions. For values of the correlation time such that the magnetic field as a function of time has large Fourier components near the resonance frequency, the (→) T1 relaxation time will be shortened. Coupled array coils. Array coils, the signals from whose elements are electrically combined prior to processing.

Decibel (dB). A measure of relative power defined as 20 log of the relative amplitude of voltage in an electrical circuit or 10 log of the relative power, e.g., a factor of 10 change in voltage corresponds to 20 dB, and a factor of 100 corresponds to 40 dB. Decoupling. (1) Specific irradiation designed to remove the multiplet structure in a particular resonance due to spin–spin coupling with other nuclei. (2) A means of preventing the interaction by mutual inductive coupling of two (or more) resonant RF coils, e.g., by detuning coils not in use at a particular point in time. Decoupling can take the form of active decoupling where an externally

1439

1440

Glossary

controlled switching circuit is used to detune the non-se- Digitization noise. Noise introduced into digitized siglected coils or passive decoupling where RF energy from nals by the finite voltage resolution of the digitizer. Also the transmitter pulse is used to switch diodes to detune called quantization noise. the appropriate coil. Digitization. Process of conversion of continuous (anaDemodulator. Another term for (→) detector, by analogy log) signals, such as the detected MR signal (voltage), into to broadcast radio receivers. numbers. This is carried out with an (→) analog to digital converter. There are two kinds of discretization involved. Dephasing gradient. (→) Magnetic field gradient pulse The voltage is only measured (sampled) at particular disused to create spatial variation of phase of (→) transverse crete times, and only voltages within a particular range magnetization. For example, it may be applied prior to and separated by a particular minimum amount can be signal detection in the presence of a magnetic field gra- distinguished. Voltages beyond this range are said to exdient with opposite polarity (or of the same polarity if ceed the dynamic range of the digitizer. separated by a refocusing RF pulse) so that the resulting (→) gradient echo signal will represent a more complete Digitizer. (→) Analog-to-digital converter. sampling of the (→) Fourier transform of the desired image. (→) Spoiler gradient pulse. Dipole field. The field pattern produced by a closely spaced positive and negative electric charge or a north Dephasing. The loss of (→) magnetization in the trans- and south magnetic pole. At distances large compared verse plane, typically due to the fact that different mag- with the dipole length, the field falls off as the third netic dipoles of different nuclei are precessing about the power of the distance away from the charges or poles main magnetic field, B0, at slightly different precessional producing it. frequencies and therefore lose phase coherence. Dipole. (→) Magnetic dipole. Depth pulses. Use of multiple (→) RF pulses with an inhomogeneous RF field to enable acquiring data from only Dipole–dipole interaction. Interaction between a (→) selected regions within the field. Provides a “one-dimen- spin and its neighbors due to their (→) magnetic dipole sional” localization along isocontours of the B1 field. moments. This is an important mechanism contributing to (→) relaxation rates. In solids and viscous liquids this Detector. Portion of the (→) receiver that demodulates can result in broadening of the spectral lines due to lonthe RF (→) MR signal and converts it to a lower (→) fre- ger correlation times (τc). quency signal. Most detectors now used are phase sensitive (e.g., (→) quadrature demodulator/detector), and will Dynamic range. Range of signal intensities that may also give phase information about the RF signal. need to be distinguished in an image or spectrum or that can be distinguished by the electronic components. If the Diamagnetic. Property of a material having a small signal dynamic range is too great, the need to keep the negative (→) magnetic susceptibility. When placed in a (→) highest intensities from overloading the (→) digitizer may magnetic field, a slightly decrease the magnetic field oc- result in the weaker features being lost in the digitization curs (its (→) magnetization is oppositely directed to the noise. This can be dealt with by using an (→) analog to applied magnetic field). All materials have diamagnetic digital converter with a larger range of sensitivity or by property in an applied magnetic field; however, for ferro- using techniques to reduce the dynamic range, e.g., (→) magnetic and paramagnetic materials the diamagnetism suppressing the signal from water in order to detect the is completely overpowered. signal from less abundant compounds. Diffusion. The process by which molecules or other particles intermingle and migrate due to their random thermal motion. NMR provides a sensitive technique for measuring diffusion of some substances.

Echo offset. Adjustment of RF (→) spin echo and (→) gradient echo to be non-coincident in time so as to create phase differences between the signals from different (→) spectral lines (e.g., from fat and water). The magnitude of the resulting phase difference between two lines will be Diffusion-weighted imaging (DWI). Imaging tech- equal to the product of the difference in (→) frequency of niques designed to weight the measured MR signal by the lines and the difference in the echo times (∆TE). the amount of (→) diffusion (random thermal motion) of water molecules in the selected (→) voxels. Echo planar imaging (EPI). A single-shot (→) gradient echo or (→) spin-echo imaging technique that colDigital-to-analog converter (DAC). Part of the (→) inter­ lects a complete 2D image data set with Cartesian (→) face that converts digital numbers from the (→) computer k-space coverage from a single excitation. For example, into analog (ordinary) voltages or currents. the (→) FID is observed while periodically switching

Glossary

the y-magnetic field gradient in the presence of a static x-magnetic field gradient. The (→) Fourier transform of the resulting (→) spin-echo train can be used to produce an image of the excited plane. Echo spacing. The time gap between successive echo peaks in a multiple echo imaging pulse sequence. (→) Multiple echo imaging.

Epoch. In (→) functional MRI, a portion of fMRI signal measurement during which the stimulus presentation or response task is similar or unchanged. Ernst angle. The (→) RF excitation angle cos ΘE at which the signal is a maximum for a short-TR steady-state incoherent sequence. The Ernst angle is found from the relation cos ΘE = exp (–TR/T1).

Echo time. Time between middle of exciting (e.g., 90°) Even-echo rephrasing. A rephasing which occurs when RF pulse and middle of spin echo production. For mul- constant velocity spins return to the same starting phase tiple echoes, use TE1, TE2, etc. When the RF spin echo and they had directly after the initial exciting RF pulse, as a gradient echo are not coincident in time, TE refers to the result of the application of an even number of gradient pulses. This may also result from the application of multime of the gradient-spin echo. tiple gradient echo pulses following the RF pulse. Echo train length (ETL). The number of echoes combined into a single image or image set in multiple-echo Excitation, RF. Input of energy into an excitable system, imaging sequences or echo-train techniques such as e.g., NMR experiment. Energy is transferred into the spin rapid acquisition with relaxation enhancement (RARE), system using electromagnetic radiation (RF) possessing fast spin echo (FSE), and turbo spin echo (TSE). In RARE the resonance frequency ((→) Larmor frequency [ω0]). imaging, the ETL typically equals the acceleration factor. Exponential weighting. In spectroscopy, multiplication (See also: Multiple echo imaging). of the time-dependent signal data by an exponential function, exp(t/TC), where t is time and TC is a parameter Echo. (→) Spin echo. called the time constant. The time constant can be choEddy-current compensation. Means of reducing the sen to either improve the (→) signal-to-noise ratio (with a influence of (→) eddy currents on (→) pulsed (→) gradi- negative TC) or decrease the effective spectral line width ent fields by employing an electrical pre-emphasis in the (with a positive TC) in the resulting spectrum. The use of gradient amplifiers. Usually multiple time constants have a negative TC to improve signal-to-noise ratio is equivato be used to correct for eddy current effects in various lent to line broadening by convolving the spectrum with structures of the MR system such as the (→) cryoshield a Lorentzian function of corresponding reciprocal width. and (→) RF shields. Faraday shield. Electrical conductor interposed between Eddy currents. Electric currents induced in a conduc- (→) transmitter and/or (→) receiver coil and patient to tor by a (→) time-dependent magnetic field or by motion block out electric fields. of the conductor through a magnetic field. One of the sources of concern about potential hazard to subjects in Fast Fourier transform (FFT). An efficient computavery high magnetic fields or rapidly varying (→) gradient tional method of performing a (→) Fourier transform. or main magnetic fields. Can be a practical problem in the (→) cryostat of (→) superconducting magnets. Com- Fat suppression. MRI pulse sequence techniques in mon means to reduce the influence of eddy currents which the signal from hydrogen-containing lipids on gradient fields are eddy current compensation and (mostly CH2) is reduced compared with the signal from water-containing tissues. shielded gradient coils (active or passive). Electron paramagnetic resonance (EPR). (→) Electron spin resonance. Electron spin resonance (ESR). Magnetic resonance phenomenon involving unpaired electrons, e.g., in free radicals. The (→) Larmor frequencies are much higher than corresponding NMR frequencies are in the same (→) static magnetic field.

Ferromagnetic. Property of a material, e.g., iron, having a large positive magnetic susceptibility. In ferromagnetic materials long range coupling forces causes the magnetic moments of neighboring atoms to align, resulting in very large internal magnetic fields. Ferromagnetic materials are attracted to magnetic fields and retain magnetization after removal of the externally applied magnetic field.

Field. (→) Magnetic ~; (→) magnetic ~, static, (→) magEnergy level. In a magnetic field, each spin can exist in netic ~, time dependent; (→) magnetic ~ strength (→) (See one of a number of distinct states having different ener- also: Magnetic induction); (→) magnetic ~ constant. gies; this number is determined by the spin quantum Field echo. (→) Gradient echo. number.

1441

1442

Glossary

Field gradient. (→) Magnetic gradient field. Field lock. A feedback control used to maintain the (→) static magnetic field at a constant strength, usually by monitoring the (→) resonance frequency of a reference sample or line in the spectrum.

can be reduced by synchronization of the imaging sequence with the heart cycle (“cardiac (→) gating”), suppression of the blood signal with (→) saturation pulses, or reduction of phase shifts with (→) gradient moment nulling. The flow effects can also be exploited for MR angiography or flow measurements.

Field of view (FOV). The rectangular region superim- Flow void. The occurrence of low signal in regions of posed over the human body over which MRI data are flow. For a spin echo sequence, this is caused in part by acquired. Its dimensions are specified in length in each a lack of refocusing of blood, which is excited by the 90° in-plane direction and are controlled by the application pulse but not by the 180° pulse. For a gradient echo seof frequency-encode and phase-encode gradients. quence, this is caused by the dephasing of blood signal. Filling factor. A measure of the geometrical relationship Flow-related enhancement. The increase in intensity of the (→) RF coil and the object being studied. It affects that may be seen for flowing blood or other liquids with the efficiency of (→) exciting the object and detecting MR some MR imaging techniques, due to the washout of (→) signals, thereby affecting the (→) signal-to-noise ratio and, saturated spins from the imaging region. ultimately, image quality. Achieving a high filling factor requires fitting the coil closely to the object, thus poten- Fourier transform (FT). A mathematical procedure to tially decreasing patient comfort. separate out the (→) frequency components of a signal from its amplitudes as a function of time, or vice versa. Filter. Filtering is any process that alters the relative fre- The Fourier transform is used to generate the (→) specquency content. This can be done with an analog (conven- trum from the FID or (→) spin echo in pulse MR techtional electrical) filter, e.g., to remove higher frequency niques and is essential to most MR imaging techniques. components so as to avoid aliasing in digitizing. Filtering The Fourier transform can be generalized to multiple dican be carried out numerically on the digitized data. mensions, e.g., to relate an image to its corresponding (→) k-space representation, or to include (→) chem­ical shift inFiltered back projection. Mathematical technique used formation in some (→) chemical shift imaging techniques. in (→) projection–reconstruction imaging to create images from a set of multiple (→) projection profiles. The Fourier transform imaging. MR imaging techniques projection profiles are back projected to produce a two in which at least one dimension is (→) phase encoded by (or three) dimensional image. The projection profiles are applying variable (→) gradient pulses along that dimenprocessed by convolving them with a suitable mathemat- sion before “reading out” the MR signal with a (→) magical function (filtered) prior to back projecting them, in netic field gradient perpendicular to the variable gradient. order to improve the image. Widely used in conventional The (→) Fourier transform is then used to reconstruct an computed tomography (CT). image from the set of encoded MR signals. An imaging technique of this type is (→) spin warp imaging. Flip angle. Angle of the (→) macroscopic magnetization vector M in respect to the (→) static magnetic field vector Free induction decay (FID). If (→) transverse magnetizaB0 (z-axis). tion of the spins is produced, e.g., by a 90° pulse, then a transient MR signal will result that will decay toward Flow compensation. Means of reducing (→) flow effects, zero with a characteristic time constant T2 (or T2*); this e.g., (→) gradient moment nulling. decaying signal is the FID. In practice, the first part of the FID is not observable due to residual effects of the Flow effects. Motion of material being imaged, particu- powerful exciting (→) RF pulse on the electronics of the larly flowing blood, can result in many possible effects in receiver, the (→) receiver dead time. the images, including increase in the signal ((→) “flow-related enhancement”), decrease in the signal or displace- Frequency (f, ν). The number of repetitions of a periodic ment of the signal (image misregistration). These effects process per unit time. The old unit, cycles per second can be understood as being caused by (→) time-of-flight (cps), has been replaced by the SI unit, (→) Hertz (Hz). effects (washout or washin due to motion of nuclei be- It is related to (→) angular frequency, ω, by f = ω/2π (See tween two consecutive spatially selective RF excitations, also: Resonance frequency). repeated in times on the order of or shorter than the relaxation times of blood) or (→) phase shifts that can be Frequency encoding. Encoding the distribution of acquired by excited spins moving along magnetic field sources of MR signals along a direction by detecting the gradients. The inconsistency of the signal resulting from signal in the presence of a (→) magnetic field gradient pulsatile flow can lead to (→) artifacts in the image; these along that direction so that there is a corresponding gra-

Glossary

dient of (→) resonance frequencies along that direction. In the absence of other position encoding, the (→) Fourier transform of the resulting signal is a one-dimensional (→) projection profile of the object. Frequency offset. The difference between given signal frequency and a reference frequency. Frequency-selective RF pulse. An RF pulse containing energy only within a specified frequency range. Usually used for slice excitation or for selective saturation pulses. Fringe field. (→) Magnetic fringe field. Frontal plane. (→) Coronal plane, (→) orientation.

Gibbs phenomenon. Artifactual ripples that occur near a discontinuity when reconstructing a mathematical function from only a finite portion of its (→) Fourier transform. In MR imaging, it can be seen as linear artifacts parallel to sharp edges in the object, particularly with the use of (→) zero filling. Also (→) truncation artifact. Gigahertz (GHz). Unit of frequency; 1 GHz = 1,000 MHz. Golay coil. Term commonly used for a particular kind of (→) gradient coil, commonly used to create (→) magnetic field gradients perpendicular to the main magnetic field. Gradient. The amount and direction of the rate of change in space of some quantity, such as magnetic field strength ((→) magnetic field gradient).

Full-width at half-maximum (FWHM). A commonly used measure of the width at half the maximum value of Gradient- and spin-echo imaging (GRASE). A hybrid peaked functions such as (→) spectral lines or slice pro- MR pulse sequence in which both gradient echo and spinfiles. For a spectral line, this will be proportional to 1/T2. echo techniques are combined to acquire multiple lines in (→) k-space during measurement following a single Functional magnetic resonance imaging (fMRI). The spin-echo excitation. use of MRI to study function in addition to anatomy. In the brain, fMRI measures changes in cerebral blood flow Gradient coils. Current carrying coils designed to proand cerebral blood oxygenation as correlates of neuronal duce a desired (→) magnetic field gradient (so that the activity, for example (→) Blood-oxygen level-dependent ef- magnetic field will be stronger in some locations than fect (BOLD). fMRI is also used to study function of the others). Proper design of the size and configuration of heart and other organs. the coils is necessary to produce a controlled and uniform gradient. Gadolinium. Lanthanide element that is paramagnetic in its trivalent state. It has been used as the active compo- Gradient echo. A signal echo produced by reversing the nent of most (→) contrast agents in MR imaging because direction of a magnetic field gradient or by applying balof its effect of strongly decreasing the (→) T1 relaxation anced pulses of magnetic field gradient before and after a times of the tissues to which it has access. Although toxic refocusing (→) RF pulse so as to cancel out the positionby itself, it can be given safely in a chelated form such as dependent phase shifts that have accumulated due to the Gd-DTPA, which still retains much of its strong effect on gradient. In the latter case, the gradient echo is generally adjusted to be coincident with the RF spin echo. When relaxation times. the RF and gradient echoes are not coincident, the time Gating. Synchronization of imaging with a phase of the of the (→) gradient echo is denoted TE and the difference cardiac or respiratory cycles. A variety of means for de- in time between the echoes is denoted (→) ∆TE, while TE tecting these cycles can be used, such as the ECG, periph- refers to the time of the RF spin echo. eral pulse, chest motion, etc. The synchronization can be Gradient echo pulse sequence. A pulse sequence that prospective or retrospective. relies on gradient reversal to rephrase the (→) transverse Gauss (G). A unit of magnetic flux density in the older magnetization. Gradient echo pulse sequences permit (CGS) system. The Earth’s magnetic field is approxi- small flip-angle excitations, which preserve most of the mately one half gauss to one gauss, depending on loca- (→) longitudinal magnetization and therefore reduce or tion. The currently preferred (SI) unit is the Tesla (T) eliminate the time required for recovery of longitudinal magnetization before repeating the pulse sequence. Gra(1 T = 10,000 G). dient echo pulse sequences have gained common use in Gaussian line shape. A line shape characterized by a bell- 2DFT (planar) and 3DFT (volume) imaging, flow imagshaped form, proportional to exp [–(f – f 0)2 / ∆f2], where ing, (→) magnetic susceptibility imaging, and (→) BOLD ∆f is a measure of the line width. imaging. Gaussian noise. Noise distributed in a normal (Gauss- Gradient magnetic field . (→) Magnetic field gradient. ian) pattern. In such a distribution, approximately 65% of all points fall within one standard deviation of the mean.

1443

1444

Glossary

Gradient moment nulling. Adjustment to zero at the time TE of the net moments of the amplitude of the waveform of the (→) magnetic field gradients with time (e.g., 0th moment is the area under the curve, first moment is the “center of gravity,” etc). The aim is to minimize the (→) phase shifts acquired by the (→) transverse magnetization of excited nuclei moving along the gradients (including the effect of refocusing RF pulses), particularly for the reduction of image artifacts due to motion. Gradient pulse. Briefly applied magnetic field gradient. Gradient recalled echo. (→) Gradient echo. Gyromagnetic ratio (γ). The ratio of the (→) magnetic moment to the (→) angular momentum of a particle. This is a constant for a given nucleus. Hahn echo. Production of (→) spin echo by repeated (→) RF pulses. First observed using equal (90°) RF pulses, now commonly used to describe refocusing of (→) transverse magnetization by a 180° RF pulse. Half-Fourier. (→) Partial Fourier.

Image acquisition time. Time required to carry out an (→) MR imaging procedure comprising only the data acquisition time. The total image acquisition time will be equal to the product of the (→) repetition time, TR, the number of signals averaged (NSA), and the number of different signals (encoded for position) to be acquired for use in image reconstruction. The additional image reconstruction time will also be important to determine how quickly the image can be viewed. In comparing (→) sequential plane imaging and (→) volume imaging techniques, the equivalent image acquisition time per slice must be considered as well as the actual image acquisition time. Image processor. (Often array processor)-part of the computer-systems, built in particular for fast numeric calculations as required for fast reconstruction of MR images, e.g., (→) Fourier transformation. Imaginary signal. Out-of-phase component of the signal from a (→) quadrature detector. Imaging contrast. (→) Contrast.

Hardware. Electrical and mechanical components of the spectrometer or computer.

Impedance matching. Adjusting the electrical impedances of two circuits that are to be joined at an interface so that they are equal, e.g., with a matching network.

Helmholtz coil. Pair of current carrying (→) coils used to create uniform (→) magnetic field in the center of the space between them. For circular coils, their separation equals their radius.

Incoherent spins. A state of a set of spins in which the ensemble of spins in a voxel are uniformly distributed with phases between 0 and 2π, reducing the (→) transverse magnetization in a voxel to essentially zero.

Hemodynamic response. Changes in blood flow, blood volume, and blood oxygenation as a result of local neural activity.

Inductance. Measure of the magnetic coupling between two current-carrying loops (mutual) reflecting their spatial relationship or of a loop (such as a coil) with itself. One of the principal determinants of the resonance frequency of an RF circuit.

Hertz (Hz). The standard (→) SI unit of (→) frequency; equal to the old unit cycles per second.

Induction (B). (→) Magnetic induction. Homogeneity. Uniformity. In MR, the homogeneity of the static (→) magnetic field is an important criterion of Inhomogeneity. Degree of lack of homogeneity, for exthe quality of the magnet. Homogeneity requirements for ample, the fractional deviation of the local (→) magnetic (→) MR imaging are generally lower than the homogene- field from the average value of the field. ity requirements for NMR spectroscopy, but for most imaging techniques must be maintained over a large region. In-phase image. An image in which the signals from two spectral components (such as fat and water) add conHomospoil. Use of a magnetic field gradient to effec- structively in a voxel. tively eliminate residual transverse magnetization by producing a strong position dependence of phase within Interface. An interface is a piece of hardware or software providing the communication of an entity to its outside, a resolution element. (See also: Spoiler pulse). in other words, this device separates the external comHybrid magnet. Magnet system employing both current- munication from internal operation of such entity (e.g. carrying coils and permanently magnetized material to user interface: interface between human and a computer [usually a software interface]; hardware interfaces are generate the (→) magnetic field. physical interfaces).

Glossary

Interference. Interference is the (constructive or destruc- relaxation time, there will be only a small effect of T2 diftive) superposition of at least two waves. Usually used for ferences on image intensities; for longer TEs, the effect of interaction of correlated or coherent waves which have T2 may be significant. the same or nearly the same frequency. Inversion time (TI). In (→) inversion recovery, time beInterleaved image acquisition. The joint collection of tween middle of inverting 180° RF pulse and middle of data for two or more separate images such that a subset the subsequent exciting 90° RF pulse to detect amount of of (→) k-space samples for the second image is acquired longitudinal magnetization. immediately after that for the first image. This method avoids misregistration between the two images and al- Inversion transfer. (→) Saturation transfer. lows for accurate subtraction of the two images. Isocenter, magnetic. The position in the magnet which Interleaved k-space coverage. The sequential collection is centered in the x, y, and z direction. At this location the of raw data from multiple excitations such that each ex- (→) static magnetic field is typically highest in uniformity. citation samples multiple lines or curvilinear paths in (→) k-space. Isochromat. A microscopic group of (→) spins which resonate at the same frequency. For example, a set of spins Interpulse times (t). Times between successive RF pulses moving together in the direction of a magnetic field graused in pulse sequences. Particularly important are the dient yet jointly experiencing the same field at any given inversion time (TI) in (→) inversion recovery, and the time time. between (→) 90° pulse and the subsequent (→) 180° pulse to produce a (→) spin echo, which will be approximately Isotopes. Atoms with varying numbers of neutrons one half the (→) spin echo time (TE). The time between rep- but with the same number of protons. For example, 1H, 2 etitions of pulse sequences is the (→) repetition time (TR). H, and 3H are the three isotopes of hydrogen, otherwise known as proton, deuterium, and tritium. VariInverse Fourier transform. Form of the (→) Fourier ous isotopes have different nuclear magnetic moments transform that reverses the process, e.g., if the Fourier and, hence, have quite different (→) resonant frequencies. transform is used to analyze a function of time into its Many isotopes have no magnetic moment and, hence, are equivalent (→) frequency components, the inverse Fou- therefore not observable by NMR. rier transform will synthesize that function of time from Isotropic imaging. Imaging in which voxel dimensions these frequency components. are equal in x, y, and z directions. Inversion. A non-equilibrium state in which the (→) macroscopic magnetization vector is oriented opposite Isotropic motion. Motion which is uniform in all directo the magnetic field; usually produced by adiabatic fast tions. This is generally used in reference to molecular (→) passage or (→) 180° RF pulses. diffusion or rotation that gives rise to (→) relaxation of the spin system through dipole–dipole interactions. Inversion recovery (IR). Pulsed MR imaging sequence wherein the macroscopic (→) magnetization is inverted Isotropic voxel. A voxel with equal physical dimensions at a time on the order of T1 before the regular imaging in x, y, and z directions. pulse-gradient sequences. The resulting partial (→) relaxation of the spins in the different structures being imaged J-coupling. (→) Spin–spin coupling. can be used to produce an image that depends strongly on T1. This may bring out differences in the appearance J-modulation. Changes in the relative (→) phase of the of structures with different (→) T1 relaxation times. Note component lines of a multiplet ((→) spin–spin coupling) that this does not directly produce an image of T1. T1 in caused by differential phase accumulations, dependent a given region can be calculated from the change in the on the particular acquisition parameters employed. For (→) MR signal from the region due to the inversion pulse example, in multiple (→) spin echo sequences the resultcompared with the signal with no inversion pulse or an ing modulation of the net intensity of the multiplet can affect the apparent T2s in a manner dependent on the inversion pulse with a different (→) inversion time (TI). choice of interpulse delays employed to observe the echo. Inversion recovery imaging. (→) Inversion recovery. Keyhole imaging. A form of dynamic imaging that inInversion-recovery spin echo (IRSE). Form of (→) in- creases temporal resolution without degrading spatial version-recovery imaging in which the signal is detected resolution by updating only a portion of k-space rapidly as a spin echo. For TE short compared with the (→) T2 and frequently. The center of k-space which is updated

1445

1446

Glossary

more rapidly forms the “keyhole.” The remainder of k- Line spread function (LSF). A hypothetical point-like space is updated less frequently to provide adequate spa- or delta-function object in one dimension will generally tial resolution. have an extended (blurred) image resulting from the imaging process; this is the line spread function characterKilohertz (kHz). Unit of frequency (= 1,000 Hertz). izing the imaging process. Knowledge of the LSF permits the prediction of how the object will be imaged in that k-Space. Two-dimensional spatial frequency space in one direction, assuming linearity of the imaging process. which the Fourier transform of the image is represented. The Fourier transform of the line spread function is the The data acquired for MR image reconstruction generally modulation transfer function (MTF). correspond to samples of k-space, that is, they represent values of the Fourier transform of the image at a particular Line width. Spread in frequency of a resonance line in a set of locations in k-space. (See also: Spatial frequency). MR spectrum. A common measure of the line width is (→) full-width at half-maximum (FWHM). k-Space filling. The location and order of obtaining data in two-dimensional spatial frequency space (k-space), Line, spectral. (→) Spectral line. the (→) Fourier transform of which comprises the MR image. For example, conventional MR pulse sequences such Linearity. (1) Fidelity of response, e.g., of (→) magnetic as (→) spin-echo and gradient echo imaging fill a single field gradients or RF system, to input. The output of a linline of (→) k-space with each data measurement. A differ- ear system is directly proportional to its input. (2) Spatial ent phase encoding step is used to fill out another parallel uniformity of the (→) magnetic field gradient over the imline of k-space. The full set of measurements completes aging volume. Because of eddy current effects static and a Cartesian grid of points in k-space. Other options for dynamic linearity have to be distinguished. Both together k-space filling include radial filling (back-projection im- with the magnet homogeneity determine the geometrical aging) or spiral filling (spiral imaging). correctness of the images. k-Space trajectory. The path traced in the spatial-fre- Linearly polarized coil (LP coil). A coil designed to exquency domain during data collection as determined by cite or detect spins using one RF transmit and/or receive the applied gradients. channel. The magnetic field has predominately a single direction. Larmor equation. States that the (→) frequency of precession of the nuclear (→) magnetic moment is proportional Liquefier. System for reliquefaction of cryogenic gases; to the (→) magnetic field. ω0 = –γB0 (radians per second) if closely matched with a superconducting magnet, zero or f0 = –γB0/2γ (Hertz), where ω0 or f0 is the frequency, γ net cryogen boil-off can be achieved. is the (→) gyromagnetic ratio, and B0 is the magnetic induction field. The negative sign indicates the direction of Loading. (→) Coil loading. the rotation. Localization techniques. Means of selecting a restricted Larmor frequency (ω0 or f0). The (→) frequency at which region from which the signal is received. These can in(→) magnetic resonance can be excited; given by the (→) clude the use of surface coils, with or without magnetic Larmor equation. By varying the (→) magnetic field across field gradients. Generally used to produce a spectrum the body with a (→) magnetic field gradient, the corre- from the desired region. sponding variation of the Larmor frequency can be used to encode position. For protons (hydrogen nuclei), the Localized magnetic resonance (LMR). A particular Larmor frequency is 42.58 MHz/T. technique for obtaining MR spectra, for example, of phosphorus, from a limited region by creating a sensitive Lattice. By analogy to NMR in solids, the magnetic and volume with inhomogeneous applied gradient magnetic thermal environment with which nuclei exchange energy fields, which may be enhanced with the use of surface in (→) longitudinal relaxation. coils. Line imaging. (→) Sequential line imaging. Line scanning. (→) Sequential line imaging.

Lock. (→) Field lock.

Longitudinal magnetization (Mz). Component of the (→) macroscopic magnetization vector along the static (→) Line shape. Distribution of the relative strength of res­on­ magnetic field. Following excitation by RF pulse, Mz will ance as a function of frequency which establishes a par- approach its equilibrium value M0 , with a characteristic ticular spectral line. Common line shapes are (→) Lorent- time constant T1. zian and (→) Gaussian.

Glossary

Longitudinal relaxation. Return of longitudinal mag- force on a body within it. Although the dangers of large netization (Mz) to its equilibrium value (M0) after exci- magnetic fields are largely hypothetical, this is an area of tation; requires exchange of energy between the nuclear potential concern for safety limits. Formally, the forces spins and the lattice. experienced by moving charged particles, current carrying wires, and small magnets in the vicinity of magnet Longitudinal relaxation rate. (→) Relaxation rate, longi- are due to (→) magnetic induction (B), which includes the tudinal (R1). effect of (→) magnetization, while the magnetic field (H) is defined so as not to include magnetization. However, Longitudinal relaxation time (T1). Also called spin-lat- both B and H are often loosely used to denote magnetic tice relaxation time or T1 relaxation; the characteristic fields. time constant for spins to tend to align themselves with the external magnetic field. Starting from zero magneti- Magnetic field gradient. A (→) magnetic field which zation in the z-direction, the z-magnetization will grow changes in strength in a certain given direction. Such to 63% of its final maximum value in a time T1. fields are used in (→) MR imaging with selective excitation to select a region for imaging and also to encode the Lorentzian line. Usual shape of the lines in an NMR location of MR signals received from the object being spectrum, characterized by a central peak with long tails; imaged. Measured (e.g.) in (→) Tesla per meter. proportional to 1/[(1/T2)2 + (f – f0)2], where f is (→) frequency and f0 is the frequency of the peak (i.e., central Magnetic |H0| field strength (H0). The magnetic field H0 resonance frequency). A Lorentzian function is the Fou- (H0 = |H0|) is an axial vector field (units are ampere per meter). The (→) magnetic induction B0 and the magnetic rier transform of a decaying exponential. field strength H0 are linked by the equation. B0 = μ0 H0, M0. Equilibrium value of the magnetization; directed μ0 being the (→) magnetic permeability (μ0 = 1.257 · 10–6 along the direction of the static magnetic field, B0. Pro- Vs/Am). portional to spin density, N. Magnetic flux. (→) Magnetic induction. Macroscopic magnetic moment. (→) Macroscopic magnetization vector. Magnetic forces. Forces resulting from the interaction of magnetic fields. (→) Pulsed magnetic field gradients Macroscopic magnetization. (→) Magnetization, macro- can interact with the main magnetic field to produce acoustic noise through the gradient coil. Magnetic fields scopic. attract ferromagnetic objects with forces which can be leMacroscopic magnetization vector. Net (→) magnetic thal if one is hit by an unrestrained object in flight. One moment per unit volume (a vector quantity) M of a sample could also be trapped between the magnet and a large in a given region, considered as the integrated effect of all unrestrained (→) ferromagnetic object or the object could damage the MR system. Access control and personnel the individual microscopic nuclear magnetic moments. awareness are the best prevention of such accidents. The Magnet stability. Temporal stability of the (→) magnetic attraction mechanism for ferromagnetic objects is that field. Factors to be considered are field decay of (→) super­ the magnetic field magnetizes the iron. This induced (→) conducting magnets in persistent mode, aging of perma- magnetization reacts with the gradient of the magnetic nent magnet material, temperature dependence of (→) field to produce an attraction toward the strongest area permanent magnet material, and temporal stability of of the field. The details of this interaction are very dependent on the shape and composition of the attracted object. magnet power supplies. There is a very rapid increase of force as one approaches Magnetic dipole. North and south magnetic poles sepa- a magnet. There is also a (→) torque or twisting force on rated by a finite distance. An electric current loop, includ- objects, e.g., a long cylinder (such as a pen or an intracraing the effective current of a spinning nucleon or nucleus, nial aneurysm clip) will tend to align along the magnet’s field lines. The torque increases with field strength while can create an equivalent magnetic dipole. the attraction increases with field gradient. Depending Magnetic field (H). The region surrounding a magnet on the magnetic saturation of the object, attraction is (or current carrying conductor) is endowed with cer- roughly proportional to object mass. Motion of conducttain properties. One is that a small magnet in such a re- ing objects in magnetic fields can induce (→) eddy curgion experiences a (→) torque that tends to align it in a rents that can have the effect of opposing the motion. given direction. Magnetic field is a (→) vector quantity; the direction of the field is defined as the direction that Magnetic fringe field. The region surrounding a magthe north pole of the small magnet points when in equi- net and exhibiting a magnetic field strength which is librium. A (→) magnetic field produces a magnetizing significantly higher than the earth’s magnetic field (typi-

1447

1448

Glossary

cally 0.05–0.1 mT, depending on geographical location). Due to the physical properties of magnetic fields, the (→) magnetic flux which penetrates the useful volume of the magnet will return through the surroundings of the magnet to form closed field lines. Depending on the magnet construction, the returning flux will penetrate large open spaces (unshielded magnets) or will be confined largely to iron yokes or through secondary coils (shielded magnets).

Magnetic resonance spectroscopy (MRS). Use of magnetic resonance to study the MR spectrum of a sample or a tissue region. In addition to the effects of factors such as (→) relaxation times that can affect the MR signal as seen in magnetic resonance imaging, effects such as J-modulation or the (→) magnetization transfer after selective excitation of particular (→) spectral lines can affect the relative strengths of spectral lines. The frequencies of certain lines may also be affected by factors such as the local pH.

Magnetic induction (B). Also called magnetic flux density. The net magnetic effect from an externally applied (→) magnetic field and the resulting (→) magnetization. B is proportional to H, with the SI unit being the (→) Tesla. The magnetic induction B0 and the magnetic field strength H0 are linked by the equation. B = μ0 H, μ0 being the (→) magnetic permeability (μ0 = 1.257 · 10–6 Vs/Am).

Magnetic shielding. Means to confine the region of strong magnetic field surrounding a magnet; most commonly the use of material with high (→) permeability (passive shielding) or by employing secondary counteracting coils outside of the primary coils (active shielding). The high permeability material can be employed in the form of a yoke immediately surrounding the magnet (self-shielding) or installed in the walls of a room as full- or partialroom shielding. Unlike shielding ionizing radiation, for example, magnetic shielding can only be accomplished by forcing the unavoidable magnetic return flux through more confined areas or structures, not by absorbing it.

Magnetic induction (B0), static. The magnetic induction B0 (B0 = |B0|) is an axial vector field producing solenoidal lines of force around a magnet and is also called magnetic flux density. This field is measured in newton per ampere meter and is called (→) Tesla. Commonly the magnetic induction B0 is referred as magnetic field (should be not confused with the magnetic field strength).

Magnetic susceptibility (χ). Measure of the ability of a substance to become magnetized.

Magnetic induction (B1), time dependent. The mag- Magnetization. Also (→) macroscopic magnetization vecnetic induction B1 (B1 = |B1|) is a magnetic vector field in- tor. The magnetic polarization of a material produced by duced by the RF radiation with a certain frequency. If the a magnetic field (magnetic moment per unit volume). frequency of the RF field matches the Larmor frequency ω0, then a maximum energy transfer from the RF field Magnetization, macroscopic (M). The magnetic polarinto the spin system occurs ((→) excitation). ization produced by a magnetic field: = vector sum of all (→) microscopic magnetic moments (µ) in a given volume (magnetic moment per unit volume). Magnetic moment. (→) Moment, magnetic. Magnetic permeability (µ). The degree of (→) magnetiza- Magnetization, microscopic (µ). (→) Magnetic moment. tion of a material responding to an applied magnetic field. Magnetization transfer (MT). The change in (→) magMagnetic permeability. μ0 = 1.257 10–6 Vs/Am (= (→) per- netization within a multicomponent spin system when meability of vacuum) (= magnetic field constant). one of the component peaks is selectively perturbed. This is observed as a change in relative (→) signal intensities. Magnetic resonance (MR). (→) Nuclear magnetic reso- One of the most common forms of perturbation in imnance (NMR). aging is selective saturation. For example, this phenomenon can be exploited as part of an imaging sequence to Magnetic resonance angiography (MRA). Angiography produce image contrast based on differential amounts of using MRI. magnetization transfer, magnetization transfer contrast (MTC). Magnetic resonance imaging (MRI). Use of magnetic resonance to create images of objects such as the body. Magnetization transfer contrast (MTC). Production Currently, this primarily involves imaging the distribu- of change in relative signal intensities by (→) magnetization of mobile hydrogen nuclei (protons) in the body. tion transfer. For example, (→) saturation of broad specThe image brightness depends jointly on the (→) spin den- tral lines may produce decreases in intensity of lines not sity (N[H]) and the (→) relaxation times (T1 and T2), with directly saturated, through exchange of magnetization their relative importance depending on the particular between the corresponding states; more closely coupled imaging technique and choice of interpulse times. Image states will show a greater resulting intensity change. brightness is also affected by any motion such as blood flow, respiration, etc. Also (→) zeugmatography. Magnetogyric ratio. (→) Gyromagnetic ratio.

Glossary

Magnitude calculation. The result of taking the square root of the sum of the squares of the real and imaginary parts of an MR signal. Matching network. An arrangement of reactive elem­ents (inductors and capacitors) used to transform an input impedance of a given value to an output impedance of a second value. Such circuits are used in interfacing high impedance RF coils to low impedance (usually 50 Ω) transmission lines that feed RF energy to the coil or send the MR signal to an MR (→) preamplifier.

Multiplanar reconstruction. The reformatting of a 3D data set into 2D slices of arbitrary thickness at any angle. Multiple-coil array. A set of decoupled (→) RF coils, usually in receive mode, arranged to cover the whole region of interest. It has both the spatial coverage of a large region-of-interest coil and the high (→) signal-to-noise-ratio (SNR) of a (→) surface coil.

Multiple-echo imaging. Spin-echo imaging or echotrain pulse sequence techniques such as rapid acquisition with relaxation enhancement (RARE) techniques Matching. (→) Impedance matching. (fast spin-echo [FSE] or turbo-spin echo [TSE]) in which more than one echo is acquired per excitation pulse. (→) Matrix size. The number of data points collected in one, Carr-Purcell (CP) sequences and (→) Carr-Purcell-Meitwo, or all three directions. Normally used for the 2D in- boom-Gill (CPMG) sequences are examples of multipleplane sampling. The display matrix may be different from echo imaging techniques in which distinct images are the acquisition matrix, although resolution is determined constructed from signal echoes acquired at a different TE values, yielding different T2 weighting to each image by the latter (See also Nx, Ny, Nz.). set. Echo-train or RARE techniques speed image acquisiMaximum intensity projection (MIP). A projection im- tion by applying a different phase-encoding to each echo, age which is obtained from a 3D data set by selecting the speeding image acquisition, but blending echoes with maximum intensity along lines or rays that cut through different T2 weightings into a single image set. the 3D image volume. Multiple line-scan imaging (MLSI). Variations of seMaxwell coil. A particular kind of (→) gradient coil, com- quential line imaging techniques that can be used if selecmonly used to create (→) magnetic field gradients along tive excitation methods that do not affect adjacent lines are employed. Adjacent lines are imaged while waiting the direction of the main magnetic field. for (→) relaxation of the first line toward equilibrium, which may result in decreased image (→) acquisition time. Megahertz (MHz). Unit of frequency, 1,000,000 Hertz. A different type of MLSI uses simultaneous excitation of Meiboom-Gill sequence. (→) Carr-Purcell-Meiboom-Gill two or more lines with different (→) phase encoding followed by suitable decoding. sequence.

Multiple quantum coherence. Excitation by an (→) RF pulse can be considered as creating a transition (or “coMoment, magnetic. A measure of the net magnetic prop- herence”) between different energy levels. Formally, tranerties of an object or particle. A nucleus with an intrinsic sitions are only allowed between states of the (→) spin (→) spin will have an associated (→) magnetic dipole mo- system differing in quantum number by one unit (“singlement, so that it will interact with a magnetic field (as if it quantum coherence”), but multiple RF pulses can act in cascade and produce multiple-quantum (→) coherence. were a tiny bar magnet). Only single quantum coherence produces a directly observable signal, however, requiring indirect observation MR imaging. (→) Magnetic resonance imaging. of multiple-quantum frequencies. MR-Signal. The MR signal is the FID-signal in the time domain and the resonance line in the frequency domain Multiple sensitive point. Sequential line imaging technique utilizing two orthogonal oscillating magnetic field (See also: NMR signal, Raw data). gradients, an (→) steady-state free precession pulse seMR-signal intensity. The MR-signal intensity is the signal quence, and signal averaging to isolate the MR spectrompower and is represented in the frequency domain by the eter sensitivity to a desired line in the body. integral of the resonance line. The area under the curve (integral) is approximately proportional to the number of Multiple slice imaging. Variation of (→) sequential plane excited nuclei in the sample, with other words; it is a rep- imaging techniques that can be used with selective (→) resentation of the concentration of the particular nuclei excitation techniques that do not affect adjacent slices. Adjacent slices are imaged while waiting for relaxation of in the sample, which are detectable using MR. the first slice toward equilibrium, resulting in decreased image acquisition time for the set of slices. MR tomography (MRT). (→) MR imaging (MRI). Microscopic magnetization. (→) Magnetic moment.

1449

1450

Glossary

Multiple spin-echo. A pulse sequence leading to the pro- magnetic field, after excitation by a suitable (→) RF magduction of multiple spin echoes after an initial excitation netic field. The peak resonance frequency is proportional pulse, for example (→) Carr-Purcell (CP) sequences and to the (→) magnetic field, and is given by the (→) Larmor (→) Carr-Purcell-Meiboom-Gill (CPMG) sequences. equation. Only unpaired electrons or nuclei with a nonzero spin exhibit magnetic resonance (= magnetic resoMultiplet. A pattern of multiple resonances (spectral nance [MR]). lines) observed when the initially single (→) Larmor frequency of a given nucleus in a (→) spin system is split by Nuclear Overhauser effect (NOE). A change in the (→) interactions with neighboring spins through the scalar or steady-state magnetization of a particular nucleus due (→) spin–spin interaction. The magnitude of this interac- to irradiation of a neighboring nucleus with which it is tion is independent of the applied magnetic field and is coupled by means of a (→) spin–spin coupling interaction. referred to as J, the (→) spin–spin coupling constant. The This interaction must be the primary relaxation mechan­ specific pattern produced depends on the number of ism of these nuclei. Such an effect can occur during (→) coupled nuclei and their (→) spin quantum numbers. decoupling and must be taken into account for accurate intensity determinations during such procedures. Multiply tuned coil. (→) RF coil designed to operate at more than one resonance frequency, so that MR of more Nuclear spin quantum number (I). (→) Spin quantum than one kind of nucleus can be observed with the same number. coil. Nuclear spin ((→) spin). Property of particular atomic NMR signal. Electromagnetic signal in the (→) radiofre- nuclei which causes an (→) angular momentum and in quency (RF) range produced by the (→) precession of the turns a (→) magnetic moment. (→) transverse magnetization of the spins. The rotation of the transverse magnetization induces a voltage in a (→) Nucleon. Generic term for a nuclear constituent, a neucoil, which is amplified and demodulated by the (→) re- tron or proton. ceiver; the signal may refer only to this induced voltage. Number of excitations (NEX). Number of signals averaged together to determine each distinct position-enNMR imaging. (→) Magnetic resonance imaging (MRI). coded signal to be used in image reconstruction. Noise. That component of the reconstructed image (or spectrum) due to random and unpredictable processes Number of signal averages (NSA). (→) Number of excias opposed to the signal within the image itself which tations (NEX). is due to predictable processes. Not to be confused with artifacts which are non-random errors in the image. It Nutation. A displacement of the axis of a spinning body is commonly characterized by the standard deviation of away from the simple cone shaped figure which would signal intensity in the image of a uniform object (phan- be traced by the axis during (→) precession. In the rotattom) in the absence of artifacts. The measured noise may ing frame of reference, the nutation caused by an (→) RF depend on the particular phantom used due to variable pulse appears as a simple precession, although the motion effects on the (→) quality factor (Q) of the receiver coil. is more complex in the stationary frame of reference. (→) Acoustic ~; (→) digitization ~; (→) Gaussian ~; (→) Nyquist limit. Frequency of a signal beyond which aliasrandom ~; (→) ~ figure. ing will occur in the sampling process. This frequency is Noise figure. A measure of the (→) noise performance equal to one half the sampling rate. of an (→) amplifier or chain of amplifiers such as an MR receiver. In MR systems the preamplifier should have a Off resonance. A state occurring when the (→) Larmor very low noise figure to prevent significant degradation frequency of a spin isochromat is different from that of of the signal-to-noise ratio of the (→) MR signal. Noise the exciting (→) RF field. figure is a ratio in dBs = 20 log [V0/(Vi G)] where Vi is the input thermal noise voltage, V0 is the amplifier out- On resonance. A state occurring when the (→) Larmor put noise level, and G is the voltage gain of the amplifier frequency of a spin isochromat is the same as that of the (when the input and output impedances of the amplifier exciting (→) RF field. are equal). Opposed-phase image. An image in which the signal Nuclear magnetic resonance (NMR). Resonance phe- from two spectral components (such as fat and water) are nomenon resulting in the absorption and/or emission of 180° out-of-phase and lead to destructive interference in electromagnetic energy by nuclei or electrons in a static a voxel.

Glossary

Orientation. The three basic orthogonal slice orientations are transverse (T), sagittal (S,) and coronal (C) (= frontal). The basic anatomical directions are right (R) to left (L), posterior (P) to anterior (A), and feet (F) to head (H), considered as positive directions. The location in the R/L and P/A directions can be specified relative to the axis of the magnet; the F/H location can be specified relative to a convenient patient structure. A standard display orientation for images in the basic slice orientation is (1) transverse, A to top of image and L to right; (2) coronal, H to top of image and L to right; and (3) sagittal, H to top of image and A to left. The orientation of single oblique slices can be specified by rotating a slice in one of the basic orientations toward one of the other two basic orthogonal planes about an axis defined by the intersection of the two planes. For example, a plane tipped from the transverse 30° toward the sagittal would be denoted T ø S 30. Double oblique slices can be specified as the result of tipping a single oblique plane as above toward the remaining basic orientation plane about an axis defined by the intersection of the oblique plane and the remaining basic plane. For example, tipping the single oblique plane above 40° toward the coronal would be denoted (T ø S 30) ø C 40. In double oblique angulations, the first rotation is chosen about the vertical image axis and the second about the (new) horizontal axis. Angles are chosen to have magnitudes less than 90° (for single oblique slices less than 45°); the sign of the angle is taken to be positive when the rotation brings positive axes closer together. For a scan including a family of single oblique angulations in a fan, we keep the same primary slice order to denote all the images, choosing it so the average angle is in the range ±45°. Labeling the four sides of the image according to the direction relative to the center of the image helps clarify anatomical orientation. The basic orientation images will have four single-letter labels, single oblique images will have two single- and two double-letter labels, and double-oblique images will have two double- and two triple-letter labels. The order of the letters in the label should reflect the relative closeness of the primary axes. The slice location can be specified by the location of the point at the center of the slice. In general, the actual displayed image may have a further, in-plane, rotation; this should be indicated either as an angle of rotation (positive, clockwise) or with a graph­ ical icon, as discussed below. An alternate way to specify the orientation of the image plane is with the direction cosines of the normal to the plane (the cosines of the angle between a line perpendicular to the image plane and the basic axes [A, L, H]). This may be convenient when specifying images to be acquired perpendicular to an axis between anatomical landmarks. The labeling of the locations of the sides of the image relative to the image center would be the same as above for specification of plane orientation by rotations relative to the basic orientation planes. If available, some graphic aids can be

helpful to show image orientations. (1) A graphic icon of the labeled primary axes (A, L, H) with relative lengths given by direction sines and orientation as if viewed from the normal to the image plane can help orient the viewer, both to identify image plane orientation and to indicate possible in-plane rotation. (2) In graphic prescription of obliques from other images, a sample original image with an overlaid line or set of lines indicating the intersection of the original and oblique image planes can help orient the viewer. Pacemaker effect. All implanted electronic devices are susceptible to the fields used in MR. The (→) static magnetic field applies force to magnetic materials and both (→) RF fields and (→) pulsed gradients can induce voltages in circuits. The pacemaker’s susceptibility to static field and its critical role in life support have warranted special consideration. Transcutaneous control or adjustment of pacing rate is a feature of many units. Some achieve this control using switches activated by the external application of a magnet to open/close the switch. Others use rotation of an external magnet to turn internal controls. The fringe field around an MR magnet can activate such switches or controls. Such activations are considered to be a risk. Areas with fields higher than 0.5 mT (= 5 G) commonly have restricted access and/or are posted as being a risk to persons with pacemakers. Paradoxical enhancement. (→) Flow-related enhancement. Parallel imaging. The use of multiple (→) receiver coils to collect simultaneously different portions of the image in physical space, or different data points in (→) k-space, which are then used to reconstruct collected images. Parallel imaging speeds data collection and therefore decreases total imaging time, with some loss in signal-tonoise ratios compared with conventional imaging and longer post-acquisition reconstruction times. Particular strategies in parallel imaging include vendor-specific methods such as sensitivity-encoding (SENSE, mSENSE), simultaneous acquisition of spatial harmonics (SMASH), generalized autocalibrating partially parallel acquisition (GRAPPA), integrated parallel acquisition techniques (iPAT), and others. These parallel imaging techniques differ from one another, some requiring collection of a sensitivity map of each receiver coil for use in image reconstruction. Paramagnetic. Property of a material (e.g., aluminum, strontium, platinum, etc.) having a small but positive (→) magnetic susceptibility. Paramagnetism occurs only in the presence of an externally applied magnetic field. Paramagnetic materials are attracted to magnetic fields because of their small positive magnetic susceptibility. Unlike (→) ferromagnetic materials paramagnetic materials do not retain any magnetization in the absence of the

1451

1452

Glossary

externally applied magnetic field. The addition of a small amount of paramagnetic substance may greatly reduce the (→) relaxation times of water. Typical paramagnetic substances usually possess an unpaired electron and include atoms or ions of transition elements, rare earth elements, some metals, and some molecules including molecular oxygen and free radicals. Paramagnetic substances are considered promising for use as (→) contrast agents in MR imaging.

Perfusion-weighted imaging. Image acquisition techniques that highlight fluids moving through arteries, veins, and capillaries.

Partial-Fourier imaging. Reconstruction of an image from an (→) MR data set comprising an asymmetric sampling of k-space. For example, it can be used either to shorten image acquisition time, by reducing the number of phase encoding steps required, or to shorten the (→) echo time, TE, by moving the echo off center in the acquisition window. In either case the (→) signal-to-noise ratio is reduced and the resolution can be improved to corres­ pond to the maximum available resolution in the data.

Phantom. An artificial object of known dimensions and properties used to test aspects of an imaging machine.

Partial saturation (PS). Excitation technique applying repeated (→) RF pulses in times on the order of or shorter than T1. In MR imaging systems, although it results in decreased (→) signal amplitude, there is the possibility of generating images with increased contrast between regions with different relaxation times. It does not directly produce images of T1. The change in MR signal from a region resulting from a change in the (→) interpulse time, TR, can be used to calculate T1 for the region. Although partial saturation is also commonly referred to as (→) saturation recovery, that term should properly be reserved for the particular case of partial saturation in which recovery after each excitation effectively takes place from true saturation.

Permanent magnet. A magnet whose (→) magnetic field originates from permanently Permeability (μ). Tendency of a substance to concentrate magnetic field, μ = B/H.

Phase. In a periodic function (such as rotational or sinus­ oidal motion), the position relative to a particular part of the cycle. Phase correction. (1) Corrective processing of the (→) spectrum so that spectral lines at different frequencies all have the (→) absorption-mode phase. (2) In imaging, adjustment of the signal in different parts of the image to have a consistent phase. Phase cycling. Techniques of signal excitation in which the phases of the exciting or refocusing (→) RF pulses are systematically varied and the resulting signals are then suitably combined in order to reduce or eliminate certain (→) artifacts.

Phase encoding. Encoding the distribution of sources of MR signals along a direction in space with different phases by applying a (→) pulsed magnetic field gradient along that direction prior to detection of the signal. In general, it is necessary to acquire a set of signals with a suitable set of different phase-encoding gradient pulses Partial-saturation spin echo (PSSE). Partial satura- in order to reconstruct the distribution of the sources tion in which the signal is detected as a spin echo. Even along the encoded direction. though a spin echo is used, there will not necessarily be a significant contribution of the T2 relaxation time to im- Phase-encoding order. The temporal order in which the age contrast, unless the echo time, TE, is on the order of (→) phase encoding gradient pulses are applied. The order or longer than T2. can be sequential, centric, reverse centric, random, etc.

Partial volume effect. The loss of contrast between two Phase-sensitive detector. Detector that measures the adjacent tissues in an image caused by insufficient resolu- phase of the signal relative to the phase of a reference ostion so that more than one tissue type occupies the same cillator. (→) Quadrature detector. voxel (or pixel). PIN diode. The PIN-diode is a diode with a undoped Passive shielding. Magnetic shielding through the use of intrinsic semiconductor region covered on top and on high permeability material (also (→) magnetic shielding, the bottom by p-type semiconductor region and n-type (→) self-shielding, (→) room shielding). semiconductor region, respectively. Passive shimming. Shimming by adjusting the position of suitable pieces of (→) ferromagnetic metal within or around the main (→) magnet of an MR system. Peak. (→) Spectral line.

Pixel. Acronym for a picture element; the smallest discrete part of a digital image display. Note that the corres­ ponding size of the pixel may be smaller than the actual (→) spatial resolution.

Glossary

Planar imaging. Imaging technique in which the im- Projection profile. Spectrum of (→) MR signal whose age of a plane is built up from signals received from the frequency components are broadened by a (→) magnetic whole plane. field gradient. In the simplest case (negligible line width, no relaxation effects, and no effects of prior gradients), it Point-spread function (PSF). A hypothetical point ob- corresponds to a one-dimensional projection of the spin ject will generally have an extended (blurred) image re- density along the direction of the gradient; in this form it sulting from the imaging process; this is the point spread is used in (→) projection–reconstruction imaging. function characterizing the imaging process. Considering any object as composed of an assembly of point ob- Projection–reconstruction imaging. MR imaging techjects, knowledge of the PSF permits the prediction of nique in which a set of projection profiles of the body is how the object will be imaged, assuming linearity of the obtained by observing MR signals in the presence of a imaging process. suitable corresponding set of (→) magnetic field gradients. Images can then be reconstructed using techniques analo­ Pole piece (or pole tip). High (→) permeability material gous to those used in conventional computed tomograused to shape the uniformity of the useful volume of a phy (CT), such as filtered back projection. It can be used for volume imaging or, with plane selection techniques, magnet, especially a (→) permanent magnet. for sequential plane imaging. Population. The numbers of nuclei or electrons in different energy levels. At (→) thermal equilibrium, the relative Prospective synchronization. (→) Synchronization, propopulations of the energy levels will be given by the (→) spective. Boltzmann distribution. Proton density (PD). The quantity of hydrogen nuclei in Preamplifier. A device that amplifies very low-level sig- each voxel or volume of tissue. Spin-echo imaging can nals. A preamplifier is generally placed close to its signal generate a proton density-weighted image by using long source and has a very low noise figure as it is the princi- TR and very short TE settings. pal determinant of electronic (→) noise within the system. Preamplifiers used in MR systems usually have a low in- Pulse length (width). Time duration of a pulse. For an put impedance, and require a matching network to inter- (→) RF pulse near the (→) Larmor frequency, the longer face to the (→) RF coil, although preamplifiers with high the pulse length, the greater the angle of rotation of the input (→) impedance may be used with surface coils. Such (→) macroscopic magnetization vector will be (greater devices typically use a field effect transistor (FET) as their than 180° can bring it back toward its original orientation). For an RF pulse of a given shape as a function of input stage. time, the longer the pulse length, the narrower the equi­va­ Precession. Comparatively slow gyration of the axis of a lent range of (→) frequencies in the pulse will be. spinning body so as to trace out a cone; caused by the application of a (→) torque tending to change the direction Pulse(d) NMR. NMR techniques that use RF pulses and of the rotation axis, and continuously directed at right Fourier transformation of the NMR signal; have largely angles to the plane of the torque. The (→) magnetic mo- replaced the older continuous-wave techniques. ment of a nucleus with spin will experience such a torque when inclined at an angle to the (→) magnetic field, result- Pulse programmer. Part of the spectrometer or (→) inter­ ing in precession at the (→) Larmor frequency. Familiar face that controls the timing, duration, phase, and ampliexamples are the effect of gravity on the motion of a spin- tude of the pulses (RF or gradient). ning top, gyroscope, or the rotating earth. Pulse sequences. Set of RF (and/or gradient) magnetic field pulses and time spacings between these pulses; Precessional frequency. (→) Larmor frequency. used in conjunction with magnetic field gradients and MR signal reception to produce MR images ((→) interPresaturation. (→) Saturation. pulse times). A recommended shorthand designation of Probe. The portion of an MR spectrometer comprising inter­pulse times used to generate a particular image is the sample container and the RF coils, with some associ- to list the (→) repetition time (TR), the (→) echo time (TE) ated electronics. The (→) RF coils may consist of separate and, if using inversion-recovery, the (→) inversion time, receiver and transmitter coils in a crossed-coil configura- TI, with all times given in milliseconds. For example, tion, or, alternatively, a single coil to perform both func- 2,500/30/1,000 would indicate an inversion-recovery pulse sequence with TR of 2,500 ms, TE of 30 ms, and TI tions. of 1,000 ms. If using multiple spin echoes, as in CPMG, then the number of the spin echo used should be stated. Progressive saturation. (→) Saturation recovery.

1453

1454

Glossary

Pulse, 180° (π pulse). RF pulse designed to rotate the Quench. Quench is the sudden loss of superconductiv(→) macroscopic magnetization vector 180° in space as re- ity of a magnet coil due to a local temperature increase ferred to the (→) rotating frame of reference, usually about in the magnet. The cryogen used for superconductivity an axis at right angles to the main magnetic field. If the evaporates rapidly, quickly reducing the magnetic field spins are initially aligned with the magnetic field, this strength. pulse will produce inversion. Quenching. Loss of superconductivity of the currentPulse, 90° (π/2 pulse). RF pulse designed to rotate the carrying coil that may occur unexpectedly in a (→) super­ (→) macroscopic magnetization vector 90° in space as re- conducting magnet. As the magnet becomes resistive, ferred to the (→) rotating frame of reference, usually about heat will be released that can result in rapid evaporation an axis at right angles to the main (→) magnetic field. If of liquid helium in the (→) cryostat. This may present a the spins are initially aligned with the magnetic field, this hazard if not properly planned for. pulse will produce transverse magnetization and free induction decay. Radian. Dimensionless unit of angular measure, 360° = 2π radians. Pulse(d), gradient. (→) Gradient pulse. Radiofrequency (RF). Wave frequency intermediate between auditory and infrared. The RF used in MR studPulse, RF. (→) RF pulse. ies is commonly in the megahertz (MHz) range. The RF Quadrature coil. A coil that produces an RF field with used in ESR studies is commonly in the gigahertz (GHz) circular polarization by providing RF feed points that are range. The principal effect of RF magnetic fields on the out of phase by 90°. When used as a transmitter coil a body is power deposition in the form of heating, mainly factor of 2-power reduction over a linear coil results; as a at the surface; this is a principal area of concern for safety receiver an increase in (→) signal-to-noise-ratio (SNR) of limits. up to a factor of √2 can be achieved. Ramp time. Time required for a change in the (→) magQuadrature detector. A phase sensitive (→) detector or netic field strength, usually measured in Tesla per min; demodulator that detects the components of the signal in depends on construction of the magnet and design of the phase with a reference signal and 90° out of phase with magnet power supply. the reference signal. This may be performed by either Ramping. Changing the strength of the (→) magnetic analog or digital means. field of a magnet. Quadrupole moment. A measure of the non-spherical distribution of electrical charge possessed by nuclei with Random noise. Noise whose amplitude follows some a (→) nuclear spin number greater than 1/2. The resulting probability distribution and is uncorrelated. interaction with electric field gradients in the molecule can lead to a shortening of (→) relaxation times and a Rapid-excitation MR imaging. There are several approaches to speeding up the MRI data acquisition broadening of spectral lines. process by repeating the excitation by (→) RF pulses in Quality factor (Q). Applies to any resonant circuit com- times short compared with T1, typically using small (→) ponent; most often the coil Q determines the overall Q of flip angles and (→) gradient echo refocusing. When TR is the circuit. Inversely related to the fraction of the energy also on the order of or shorter than T2, the repeated RF in an oscillating system lost in one oscillation cycle. Q is pulses will tend to refocus (→) transverse magnetization inversely related to the range of (→) frequency over which remaining from prior excitations, setting up a condition the system will exhibit resonance. It affects the (→) sig- of (→) steady-state free precession, and a dependence of nal-to-noise ratio, because the detected signal increases signal strength (and (→) image contrast) on both T1 and proportionally to Q while the noise is proportional to the T2. This can be modified in various ways, particularly square root of Q. The Q of a coil will depend on the cir- (1) “spoil” the tendency to build up a steady-state by recumstances under which it is measured, e.g., whether it ducing (→) coherence between excitations, e.g., by variation of the phase or timing of consecutive RF pulses or of is “unloaded” (no patient) or “loaded” (patient). the strength of (→) spoiler gradient pulses, thus increasing the relative dependence of signal strength on T1; or Quantization noise. (→) Digitization noise. (2) acquire the signal when it is refocusing immediately prior to the next RF pulse, thus increasing the relative Quantum number. (→) Spin quantum number. dependence of signal strength on T2.

Glossary

Raw data. The digital MR signal sampled and stored dur- Relaxation rates. Reciprocals of the relaxation times, T1 ing data acquisition. Also referred to as (→) k-space data. and T2 (R1 = 1/T1 and R2 = 1/T2). There is often a linear relation between the concentration of MR contrast agents Rayleigh noise. The distribution associated with the and the resulting change in relaxation rate. magnitude of the noise amplitude following a (→) Gaussian distribution. The mean value of this distribution is Relaxation rate, longitudinal (R1), transverse (R2). The roughly 1.25 σ0, where σ0 is the standard deviation of the longitudinal (R1) and transverse (R2) relaxation rate (dioriginal Gaussian distribution. mension: s–1) are defined as the reciprocal of the relaxation time T1 and T2, respectively (R1,2 = 1 / T1,2). Readout delay. See (→) TE. Real signal. In-phase component of signal detected with a (→) quadrature detector. Relaxation times. After excitation, the (→) spins will tend to return to their (→) equilibrium distribution, in which Receiver. Portion of the MR apparatus that detects and there is no (→) transverse magnetization and the (→) lonamplifies (→) RF signals picked up by the (→) receiving coil. gitudinal magnetization is at its maximum value and oriented in the direction of the static (→) magnetic field. It is Includes a preamplifier, amplifier, and (→) demodulator. observed that in the absence of applied (→) RF magnetic Receiver coil. (→) Coil of the (→) RF receiver; “picks up” field, the transverse magnetization decays toward zero with a characteristic time constant T2, and the longitudithe (→) MR signal. nal magnetization returns toward the equilibrium value Receiver dead time. Time after exciting RF pulse during M0 with a characteristic time constant T1. which free induction decay (FID) is not detectable due to Relaxivity, longitudinal (r1), transverse (r2). The longi­ saturation of receiver electronics. tudinal (r1), transverse (r2) (dimension: s–1 · mmol–1 · ℓ), Reconstruction from projections imaging. (→) Projec- are defined as the increase of the (→) relaxation rate of tion-reconstruction imaging. water protons produced by 1 mmol per liter (ℓ) of (→) contrast agent. Reference compound. Standard compound used as a standard reference (→) spectral line for defining (→) chem- Repetition time (TR). The period of time between the ical shifts for a given nucleus. Standard reference com- beginning of a (→) pulse sequence and the beginning of pounds is for 1H tetramethylsilane (TMS) and for 31P it is the succeeding (essentially identical) pulse sequence. phosphoric acid, although for practical biol­ogic­al applications water and phosphocreatine (PCr) have been used Rephasing gradient. Magnetic field gradient pulse apas secondary references for hydrogen and phosphorus plied to reverse the spatial variation of phase of (→) trans(→) spectroscopy, respectively. The reference compound verse magnetization caused by a dephasing gradient. For can be in a capsule outside of the subject (external) or example, in selective excitation, it is a (→) magnetic field can be in the subject (internal); internal references are gradient applied for a brief period after a selective excitagenerally preferable where possible, as external refer- tion pulse, in the opposite direction to the gradient used for the selective excitation. The result of the gradient reences may be subject to different conditions. versal is a rephasing of the spins (which will have gotten out of phase with each other along the direction of Refocusing. (→) Spin echo. the selection gradient), forming a (→) gradient echo and Refrigerator. system for actively cooling structures in a improving the sensitivity of imaging after the selective (→) superconducting magnet. If only (→) cryoshields are excitation process. cooled (two-stage refrigerator), no liquid nitrogen will be needed and He boil-off will be reduced. If addition- Resistive magnet. A magnet whose magnetic field origially the superconducting coil support is actively cooled nates from current flowing through an ordinary (non-su(three-stage refrigerator) the He consumption can be es- perconducting) conductor. sentially reduced to zero. Resolution element. Size of smallest spatially resolved Region-of-interest (ROI). A user-defined subset of pix- regions in image. It may be anisotropic, e.g., with an asymmetric acquisition matrix or slice thickness, and els in a planar image. may be larger than the pixel or voxel. Relaxation. The return of an excited system of spinning Resolution, spatial. Defined as the smallest distance be(→) magnetic dipoles (spins) to its equilibrium state. tween two objects which can be separated in an image, usually measured as the distance between line pairs. The

1455

1456

Glossary

spatial resolution is depending on the contrast and the in the stationary frame of reference). The amount of rosignal-to-noise ratio which is very variable in MR, de- tation will depend on the strength and duration of the pending on the tissue and the imaging technique. Test- RF pulse; commonly used examples are (→) 90° (π/2) and ing the resolution with phantoms must take into account (→) 180° (π) pulses. these dependencies. RF shielding. Cage of electrically conducting material Resolution, temporal . Defined as the shortest time du- designed to isolate the complete MR system from its enration between two events. vironment at the resonant frequencies of interest. Resonance. A large amplitude vibration in a mechanical or electrical system caused by a relatively small periodic stimulus with a frequency at or close to a natural frequency of the system; in MR apparatus, resonance can refer to the NMR itself or to the tuning of the RF circuitry.

RF spin echo. (→) Spin echo produced by an RF pulse. RF spoiling. The use of varying phase or timing of the RF pulses to prevent setting up a condition of (→) steadystate free precession, e.g., in rapid-excitation MR imaging.

Resonance frequency. Frequency at which resonance Room shielding. Magnetic shielding through the use phenomenon occurs; given by the Larmor equation of high (→) permeability material in the walls (plus floor (ω0 = –γB0) for NMR; determined by inductance and ca- and ceiling) of the magnet room. Room shielding can be pacitance for RF circuits. complete (e.g., six sides of a box), or partial if the fringe field is to be reduced only in certain areas (See also magResonance offset (β). Either the phase due to an applied netic shielding). field or field inhomogeneity and generated during the time between two RF pulses, or the phase change of the Rotating frame of reference. A frame of reference (with RF pulse from one pulse to the next. corresponding coordinate systems) that is rotating about the axis of the static magnetic field B0 (with respect to Respiratory gating. (→) Gating. a stationary (“laboratory”) frame of reference) at a frequency equal to that of the applied RF magnetic field, B1. Respiratory ordering of phase encoding. Respiratory Although B1 is a rotating vector, it appears stationary in synchronization that acquires image data at regular times the rotating frame, leading to simpler mathematical forindependent of the respiratory cycle, but chooses the se- mulations. quence of phase encoding data acquisition so as to minimize the respiratory motion-induced (→) artifacts in the Rotating-frame zeugmatography. Technique of MR imresulting image. For example, choosing the sequence of aging that uses a (→) gradient of the RF excitation field phase encoding such that adjacent samples in the final (to give a corresponding variation of the flip angle along full data set have minimal differences in respiratory the gradient as a means of encoding the spatial location phase will minimize the spacing of “ghost” artifacts in of spins in the direction of the RF field gradient) in conthe final image. junction with a (→) static magnetic field gradient (to give spatial encoding in an orthogonal direction). It can be Retrospective respiratory gating. The resorting of data considered to be a form of Fourier transform imaging. collected over several acquisitions to create an image where all phase encoding lines are acquired with the ob- Saddle coil. RF coil configuration design commonly ject at the same spatial location of the respiratory cycle. used when the (→) static magnetic field is coaxial with the axis of the coil along the long axis of the body (e.g., (→) Retrospective synchronization. (→) Synchronization, superconducting magnets and most resistive magnets) as retrospective. opposed to solenoid or surface coil. RF coil. (→) Coil used for transmitting (→) RF pulses and/ Safety. Safety concerns in MR include (→) magnetic field or receiving (→) MR signals. Commonly used in (→) bird- strength, RF heating (SAR) induced currents due to rapcage coil, (→) saddle coil, or (→) solenoid coil configura- idly varying magnetic fields (dB/dt), effects on implanted tions for MR imaging. devices such as pacemakers, magnetic (→) torque effects on indwelling metal such as clips and possible “missile RF pulse. Burst of RF magnetic field delivered to object effect” of (→) magnetic forces, and (→) acoustic noise. by (→) RF transmitter. For (→) RF frequency near the (→) Larmor frequency ω0, it will result in rotation of the (→) Sagittal plane. The plane which is defined by the head-tomacroscopic magnetization vector in the rotating frame foot and anterior-to-posterior directions in the human of reference (or a more complicated nutational motion body. A stack of images acquired in the sagittal plane

Glossary

separates images by their left-to-right locations. The midline sagittal plane bisects the left and right half of the human body. (→) Orientation.

per segment after each trigger until 128 lines of k-space are acquired in 16 triggers.

Selective excitation. Controlling the frequency spectrum of an irradiating (→) RF pulse (via tailoring) while imposing a (→) magnetic field gradient on spins, such that only Sampling. Conversion of the continuous (analog) signal a desired region will have a suitable resonant frequency to a series of discrete (digital) values by measurement at to be excited. Originally used to excite all but a desired a set of particular times; this utilizes the (→) analog to region; now more commonly used to select only a dedigital converter. If the rate of sampling is less than twice sired region, such as a plane, for excitation. Used without the highest (→) frequency in the signal, aliasing will occur. sim­ul­tan­eous magnetic field gradients, tailored RF pulses The duration of sampling determines how small a differ- can be used to selectively excite a particular spectral line ence of frequencies can be separated. or group of lines. RF and gradient pulse combinations can be designed to select both spatial regions and specSaturation. A non-equilibrium state in MR, in which tral frequencies. equal numbers of (→) spins are aligned against and with the (→) magnetic field, so that there is no net (→) mag- Selective irradiation. (→) Selective excitation. netization. Can be produced by repeatedly applying RF pulses at the (→) Larmor frequency with interpulse times Self-shielding. Magnetic shielding by attaching a high permeability yoke to the magnet (passive shielding) or by short compared with T1. incorporating additional magnetic field-generating coils Saturation pulses. Sequence of RF (and gradient) pulses designed to reduce the external field (active shielding). designed to produce (→) saturation, typically in a selected (→) Magnetic shielding. region or set of regions, most often by the use of (→) selective excitation followed by a spoiler pulse. Similar to Sensitive plane. Technique of selecting a plane for sesome spectral suppression techniques. Can be used to quential plane imaging by using an oscillating (→) magreduce signal from flowing blood by saturating regions netic field gradient and filtering out the corresponding time dependent part of the (→) MR signal. The gradient upstream from region being imaged. used is at right angles to the desired plane and the magniSaturation recovery (SR). Particular type of partial satu- tude of the oscillating magnetic field gradient is equal to ration pulse sequence in which the preceding pulses leave zero only in the desired plane. the (→) spins in a state of (→) saturation, so that recovery at the time of the next pulse has taken place from an ini- Sensitive point. Technique of selecting out a point for sequential point imaging by applying three orthogonal tial condition of no (→) magnetization. oscillating (→) magnetic field gradients such that the local Saturation transfer (or Inversion transfer). Nuclei can (→) magnetic field is time-dependent everywhere except retain their magnetic orientation through a chemical re- at the desired point, and then filtering out the corres­ action. Thus, if (→) RF radiation is supplied to the (→) ponding time dependent portion of the (→) MR signal. spins at a (→) frequency corresponding to the (→) chem­ ic­al shift of the nuclei in one chemical state so as to pro- Sensitive volume. Region of the object from which duce (→) saturation or inversion, and chemical reactions (→) MR signal will preferentially be acquired because of transform the nuclei into another chemical state with a strong magnetic field inhomogeneity elsewhere. Effect different chemical shift in a time short compared with can be enhanced by use of a shaped RF field that is stronthe relaxation time, the NMR spectrum may show the gest in the sensitive region. effects of the saturation or inversion on the corresponding, unirradiated, line in the spectrum. This technique Sequence time. (→) Repetition time (TR). can be used to study reaction kinetics of suitable molSequential line imaging (line scanning, line imagecules. ing). MR imaging techniques in which the image is built up from successive lines through the object. In various Scalar. A quantity having only magnitude. schemes, the lines are isolated by oscillating (→) magnetic Segmented k-space data acquisition. A set of k-space field gradients or selective excitation, and then the (→) lines collected in a specified order but not constituting a MR signals from the selected line are encoded for posicomplete coverage of (→) k-space. Several segmental ac- tion by detecting the (→) free induction decay (FID) or quisitions may need to be run for complete coverage of k- (→) spin echo in the presence of a magnetic field gradient space. For example, rapidly acquiring eight k-space lines along the line; the (→) Fourier transform of the detected Sampling window. (→) Acquisition window.

1457

1458

Glossary

Signal averaging. The averaging together of signals acquired under the same or similar conditions so as to suppress the effects of random variations or random artifacts. Sequential plane imaging (planar imaging). MR imag- The number of signals averaged together can be abbreviing technique in which the image of an object is built up ated NSA (also (→) number of excitations [NEX]). from successive planes in the object. In various schemes, the planes are selected by oscillating magnetic field gra- Signal intensity. In general, signal intensity is the signal dients or selective excitation. power or signal level at a specified point and with respect to a specified reference level (See also: MR-signal intenSequential point imaging (point scanning). MR imag- sity). ing techniques in which the image is built from successive point positions in the object. In various schemes, the Signal suppression. The elimination or reduction of a points are isolated by oscillating magnetic field gradients particular signal by, for example, the application of a nar(sensitive point) or shaped magnetic fields. row band frequency-selective preparation pulse centered at the (→) resonant frequency of the signal. This can also be accomplished using an (→) inversion recovery techShaped pulse. (→) Tailored pulse. nique to null the signal as it recovers its (→) longitudinal Shielded gradient coils. Current-carrying gradient coils magnetization. with reduced gradient fringe field inside of the magnet cryostat structures like cryoshields and He vessel. The Signal-to-noise ratio (SNR or S/N). Used to describe shielding can be accomplished by secondary actively the relative contributions to a detected signal of the true driven coils or by passive screens which are inductively signal and random superimposed signals (“noise”). One coupled to the gradient coils. In both cases (→) eddy cur- common method to improve (increase) the SNR is to average several measurements of the signal on the exrents outside of the gradient system will be reduced. pectation that random contributions will tend to cancel Shielding. (→) Magnetic shielding, (→) cryoshielding, (→) out. The SNR can also be improved by sampling larger volumes (with a corresponding loss of spatial resolution) RF shielding, (→) Faraday shield. or, within limits, by increasing the strength of the magShift reagents. Paramagnetic compounds designed to in- netic field used. Surface coils can also be used to improve duce a shift in the (→) resonance frequency of nuclei with local SNR. The SNR will depend, in part, on the electrical which they interact. For example, many rare earths have properties of the sample or patient being studied. been used as shift reagents for positive metal ions such as Sinc interpolation. A method of interpolating image sodium and potassium. data by (→) zero filling the high spatial-frequency compoShim coils. Coils carrying a relatively small current that nents of the raw data so that after (→) Fourier transformaare used to provide auxiliary magnetic fields in order to tion the image (→) matrix size has been increased. This compensate for inhomogeneities in the main magnetic method can significantly improve the image display. field of an (→) MR system (also (→) passive shimming). Single-shot imaging. The process of acquiring all data Shimming. Correction of inhomogeneity of the (→) mag- needed to form a two-dimensional image with a single netic field produced by the main magnet of an (→) MR (→) excitation pulse. (→) Echo-planar imaging is an examsystem due to imperfections in the magnet or to the pres- ple of single-shot imaging. ence of external (→) ferromagnetic objects. May involve changing the configuration of the magnet or the addition Skin depth. Time-dependent electromagnetic fields are of (→) shim coils ((→) active shimming) or small pieces of significantly attenuated by conducting media (including the human body); the skin depth gives a measure of the steel ((→) passive shimming). average depth of penetration of the RF field. It may be a SI (International System of Units). The preferred inter- limiting factor in (→) MR imaging at very high frequennational standard system of physical units and measures. cies (high magnetic fields). The skin depth also affects the (→) quality factor Q of the (→) coils. Signal. In general a signal is a time-varying quantity, which may be a scalar valued function of time (wave- Slice profile. The spatial distribution of sensitivity of the forms), vector valued or may be functions of another rele­ imaging process in the direction perpendicular to the plane of the slice. When the profile deviates appreciably vant independent variable (See also: MR-signal). from rectangular, the slice thickness alone may not provide an adequate description. signal then yields the distribution of emitted (→) MR signal along the line.

Glossary

Slice selection. The (→) excitation of spins in a limited planar section of tissue by applying a (→) gradient (the slice-selective gradient) while sending a narrow-band radiofrequency pulse of appropriate frequencies into the subject. Slice thickness. The thickness of a slice. As the slice profile may not be sharp edged, a criterion such as the distance between the points at half the sensitivity of the maximum ((→) FWHM) or the equivalent rectangular width (the width of a rectangular (→) slice profile with the same maximum height and same area) may be useful. Slice. The effective physical extent of the “planar” region being imaged. Solenoid coil. A coil of wire wound in the form of a long cylinder. When a current is passed through the coil, the magnetic field within the coil is relatively uniform. Solenoid (→) RF coils are commonly used when the (→) static magnetic field is perpendicular to the long axis of the body. Solvent suppression. (→) Suppression. Spatial frequency. A dimension of the Fourier transform space (k-space representation of an image), having units of inverse distance. Higher values of spatial frequencies correspond to finer detail in the image. Spatial resolution. (→) Resolution, spatial. Spatially localized spectroscopy. Process by which regions of tissue are selectively sampled to produce spectra from defined volumes in space. These methods may be employed to sample a single region in space (single voxel method) or multiple regions simultaneously (multi­ voxel methods). The spatial selectivity can be achieved by a variety of methods including surface coils, surface coils in conjunction with (→) RF gradient methods, or (→) RF pulses in combination with switched magnetic field gradients, for example, volume-selective excitation. An indirect method of achieving spatial selectivity is the destruction of coherence of the magnetization in regions that lie outside the region of interest. A variety of spatial encoding schemes have been employed for multivoxel localization. (Also (→) chemical shift imaging.)

with about 2.25 ms between each. This will impart 90° excitation to water and no excitation to fat. Specific absorption rate (SAR). Time-varying electromagnetic fields can deposit energy in tissues. This energy is deposited mostly in the form of heat which is considered the primary mechanism of biological effect. The specific absorption rate (SAR) is defined as the energy dissipated in tissue (unit: W/kg). Inhomogeneity of the (→) RF fields leads to a local exposure where most of the power that is absorbed is applied to one body region rather than the entire person, leading to the concept of a local SAR. Averaging over the whole body leads to the global SAR. Spectral editing. Methods of selectively enhancing or suppressing the (→) MR signal from a particular molecular substance by using its spin properties, typically through (→) spin–spin coupling, e.g., J-modulation. Spectral line. Particular distinct (→) frequency or narrow band of frequencies at which resonance occurs corres­ ponding to a particular (→) chemical shift. Spectral width. The overall width in Hertz needed to observe a particular NMR spectrum. This width is generally set using the (→) Nyquist limit, namely, that the temporal sampling rate must be equal to twice the maximum spread in frequencies. Spectrometer. The portions of the MR apparatus that actually produce the NMR phenomenon and acquire the signals, including the magnet, the probe, the RF circuitry, the gradient coils, etc. The (→) spectrometer is controlled by the (→) computer via the interface. Spectroscopic imaging. MR techniques that permit acquisition of an (→) MR spectrum for each (→) pixel or (→) voxel in the (→) MR image. The resulting acquired data can then be presented as an MR spectrum for each pixel or voxel or as an image or set of images that reflects the intensity of a particular spectral peak at each spatial location in two- or three-dimensions. Spectroscopy. (→) Magnetic resonance spectroscopy.

Spectrum. An array of the frequency components of the (→) MR signal according to frequency. Nuclei with different (→) resonant frequencies will show up as values at difSpatial–spectral (or spectral–spatial) excitation. Exci- ferent corresponding frequencies in the spectrum. When tation that is both spatially and spectrally selective. This resonances are relatively isolated they appear as peaks is generally accomplished by using composite excitation, or “lines” in the spectrum. with 180° phase cycling between water and fat (CH2) magnetization during the interpulse intervals. For ex- Spin. The intrinsic (→) angular momentum of an elemenample, selective excitation of water in an image slice can tary particle, or system of particles such as a nucleus, that be accomplished at 1.5 T using a 1-2-1 composite pulse, is also responsible for the (→) magnetic moment; or, a parconsisting of a combination of 22.5, 45, and 22.5° pulses, ticle or nucleus possessing such a spin. The spins of nu-

1459

1460

Glossary

clei have characteristic fixed values. Pairs of neutrons and Spin tagging. A method used to trace the motion or flow protons align to cancel out their spins, so that nuclei with of tissue. Nuclei will retain their magnetic orientation for an odd number of neutrons and/or protons will have a net a time on the order of T1 even in the presence of motion. nonzero rotational component characterized by an inte- Thus, if the nuclei in a given region have their spin orienger or half integer quantum (→) “nuclear spin number” (I). tation changed, the altered spins will serve as a “tag” to trace the motion for a time on the order of T1 of any fluid Spin density (N). The density of resonating spins in a that may have been in the tagged region. given region; one of the principal determinants of the strength of the MR signal from the region. The SI units Spin–warp imaging. A form of (→) Fourier transform would be moles/m3. For water, there are about 1.1 × 105 imaging in which (→) phase-encoding gradient pulses are moles of hydrogen per m3, or 0.11 moles of hydrogen/cm3. applied for a constant duration but with varying ampliTrue spin density is not imaged directly, but must be cal- tude. The spin warp method, as other Fourier imaging culated from signals received with different (→) interpulse techniques, is relatively tolerant of nonuniformities (intimes. homogeneities) in the magnetic fields. Spin echo (SE). The RF pulse sequence where a 90° ex- Spiral k-space coverage. Acquisition of data whose views citation pulse is followed by a 180° refocusing pulse to cover a spiral in (→) k-space. It is accomplished by applyeliminate field inhomogeneity and chemical shift effects ing an oscillatory gradient which increases in amplitude at the echo. as a function of time. Spin-echo imaging. Any of many MR imaging techniques in which the spin echo is used rather than the (→) free induction decay (FID). Can be used to create images that depend strongly on T2 if TE has a value on the order of or greater than T2 of the relevant image details. Note that spin echo imaging does not directly produce an image of T2 distribution. The spin echoes can be produced as a train of multiple echoes, e.g., using the (→) CPMG pulse sequence. Spin-lattice relaxation time. (→) Longitudinal relaxation time (T1). Spin number, nuclear. (→) Spin quantum number. Spin quantum number (I). Property of all atomic nuclei related to the largest measurable component of the angular momentum. The values of the (→) angular momentum unequal to zero are integral or half integral multiples of ħ/2, h (= 2 π ħ) being the Planck’s constant. The number of possible energy levels of a given nucleus in a static (→) magnetic field is given by 2 I + 1. This is often named as “nuclear spin,” generally speaking, being the maximal (minimal) component ħI along the particular axis. Spin–spin coupling. MR spectral lines may consist of groups of lines called multiplets. This (→) multiplet structure is caused by interactions between nuclei that split the NMR energy levels and result in the observation of multiple allowed transitions separated by an amount of energy related to J, the spin–spin coupling constant. These interactions are called spin–spin coupling. Spin–spin relaxation time. (→) Transverse relaxation time (T2).

Spoiler gradient pulse. Magnetic field gradient pulse applied to effectively remove (→) transverse magnetization by producing a rapid variation of its (→) phase along the direction of the gradient. For example, when used to remove the unwanted signal resulting from an imperfect 180° refocusing RF pulse, a corresponding compensating gradient pulse may be applied prior to the refocusing RF pulse in order to avoid spoiling the desired transverse magnetization resulting from the initial excitation. (Also called (→) homospoil pulse.) Steady-state free precession (SFP or SSFP). Method of MR excitation in which strings of (→) RF pulses are applied rapidly and repeatedly with interpulse intervals short compared with both T1 and T2. Alternating the phases of the RF pulses by 180° can be useful. The (→) MR signal reforms as an echo immediately before each RF pulse; immediately after the RF pulse there is additional signal from the free induction decay (FID) produced by the pulse. The strength of the FID will depend on the time between pulses (TR), the (→) magnetization of the tissue, and the (→) flip angle of the pulse; the strength of the echo will additionally depend on the T2 of the tissue. SSFP may be used as a means of rapid-excitation MR imaging. With the use of appropriate dephasing gradients, the signal can be observed as a frequency-encoded gradient echo either shortly before the RF pulse or after it; the signal immediately before the RF pulse will be more highly T2-weighted. The signal immediately after the RF pulse in a rapid series of RF pulses will depend on T2 as well as T1 unless meas­ ures are taken to destroy signal refocusing and prevent the development of steady-state free precession. To avoid setting up a state of SSFP when using rapidly repeated (→) excitation RF pulses, it may be necessary to spoil the phase coherence between excitations, e.g., with varying phase shifts or timing of the exciting RF pulses or varying spoiler gradient pulses between the excitations.

Glossary

Steady-state coherent. A state of spins which leads to an equilibrium magnetization for the (→) longitudinal and transverse magnetization, or, when the magnetization at, or after, each RF pulse is the same as in the previous pulse.

Surface coil. Receiver coil that does not surround the body and is placed close to the surface of the body. Used to restrict the region of the body contributing to the detected signal. As only the region close to the surface coil will contribute to the (→) noise, there may be an improvement in the (→) signal-to-noise ratio for regions close to Stimulated echo. A form of spin echo produced by three- the coil, compared with the use of receiver coils that surpulse RF sequences, consisting of two RF pulses follow- round the corresponding part of the body. ing an initial exciting RF pulse. The stimulated echo appears at a time delay after the third pulse equal to the Surface-coil MR. A surface coil placed over a (→) region interval between the first two pulses. Although classically of interest (ROI) will have an effective selectivity for a volproduced with 90° pulses, any RF pulses other than an ume approximately subtended by the coil circumference ideal 180° can produce a stimulated echo. The intensity and one radius deep from the coil center. Such a coil can of the echo depends in part on the (→) T1 relaxation time be used for simple localization of sites for measurement because the excitation is “stored” as (→) longitudinal mag- of chemical shift spectra, especially of phosphorus, and netization between the second and third RF pulses. For blood flow studies. Some additional spatial selectivity example, use of stimulated echoes with spatially selective can be achieved with (→) magnetic field gradients. excitation with orthogonal magnetic field gradients permits volume-selective excitation for spectroscopic local- Susceptibility artifact. The loss of MR signal in voxels or regions with varying magnetic susceptibility (magnetic ization. non-uniformities) due to greater T2* decay. Susceptibility Superconducting magnet. A magnet whose magnetic artifacts are more obvious in pulse sequences weighted field originates from current flowing through a (→) super- more heavily by T2* effects, such as (→) gradient echo imconductor. Such a magnet must be enclosed in a cryostat. aging. Superconductor. A substance whose electrical resistance essentially disappears at temperatures near absolute zero. A commonly used superconductor in MR imaging system magnets is niobium-titanium, embedded in a copper matrix to help protect the superconductor from quenching.

Susceptibility. (→) Magnetic susceptibility.

Switchable coil. An RF array coil consisting of several separately resonant elements, any one of which can be selected as the receiver coil at a particular time. Coils not in use are decoupled. Applications of switchable coils inSuperparamagnetic. Superparamagnetic behavior oc- clude imaging the whole spine without patient re­pos­ition­ curs in nanoparticles (1–10 nm) and these particle dem- ing (where the coil elements may collectively be known onstrate almost similar behavior like (→) paramagnetic as a ladder coil), imaging of bilateral structures such as materials. However, in superparamagnetic materials a TMJ or the orbit using separate coils, or imaging using a small length-scale phenomenon occurs, where the energy coil with selectable field-of-view. required to change the direction of the magnetic moment of a particle is comparable to the ambient thermal energy, Synchronization, cardiac. Acquiring images of parwhereas in (→) ferromagnetic materials long range cou- ticular phases of the cardiac cycle, through either retro­ pling forces cause the magnetic moments of neighboring spect­ive or prospective synchronization. Also sometimes atoms to align, resulting in very large internal magnetic called (→) cardiac gating. fields. The thermal energy in the nanoparticles, however, is sufficient to change the direction of magnetiza- Synchronization, prospective. Controlling the timing tion of the entire nanoparticle. These fluctuations in the or sequence of image data acquisition according to the direction of magnetization cause the net magnetic field phase of respiratory or cardiac cycles. Also (→) triggering. to average to zero and the material behaves similar to paramagnetic materials, except that instead of each in- Synchronization, respiratory. The respiratory phase can dividual atom being independently influenced by an ex- be used to control imaging either by only acquiring the ternal magnetic field, the magnetic moment of the entire image data during a particular portion of the respiratory cycle (which increases image acquisition time) or by adnanoparticle tends to align with the magnetic field. justing the sequence of image data collection according Suppression. One of a number of techniques designed to to the phase of the respiratory cycle in such a way as to minimize the contribution of a particular component of minimize motion-induced artifacts in the reconstructed the object to the detected signal. For example, commonly image. Also (→) respiratory ordering of phase encoding; used to suppress the strong signal from water in order to (→) respiratory gating. detect spectral line from other components.

1461

1462

Glossary

Synchronization, retrospective. Sorting and possibly adjusting image data acquired asynchronously with the cardiac or respiratory cycle, according to the phase of the cycle at which it was acquired, so as to reconstruct a set of images corresponding to different phases of the cycle. Also (→) retrospective respiratory gating. T1 Relaxation. (→) Longitudinal relaxation time. T1-weighted (T1W). Often used to indicate an image where most of the contrast between tissues or tissue states is due to differences in tissue T1. This term may be misleading in that the potentially important effects of tissue density differences and the range of tissue T1 values are ignored. A T1 contrast state is approached by imaging with a TR short compared with the longest tissue T1 of interest and TE short compared with tissue T2 (to reduce T2 contributions to image contrast). Due to the wide range of T1 and T2 and tissue density values that can be found in the body, an image that is T1-weighted for some tissues may not be so for others.

pulse in a manner determined by the Fourier transform of the pulse. Temporal resolution. (→) Resolution, temporal. Tesla (T). The preferred SI unit of magnetic flux density. One Tesla is equal to 10,000 Gauss (CGS unit). Thermal equilibrium. A state in which all parts of a system are at the same effective temperature, in particular where the relative alignment of the spins with the (→) magnetic field is determined solely by the thermal energy of the system (in which case the relative numbers of spins with different alignments will be given by the Boltzmann distribution).

Three-dimensional Fourier transform imaging (3DFT). A form of (→) Fourier transform imaging in which an entire volume of data is collected simultaneously. This form of Fourier transform imaging requires phase-encoding to be used to separate individual planar sections within the volume, in addition to phase-encoding and freT2* Relaxation. The observed time constant of the (→) quency encoding in the two orthogonal directions. 3DFT free induction decay (FID) due to loss of phase coher- imaging eliminates gaps or low signal areas between inence among spins oriented at an angle to the (→) static dividual “slices” and generally has higher signal-to-noise magnetic field, commonly due to a combination of mag- ratios than 2DFT imaging. (See also: Volume imaging). netic field (→) inhomogeneities, ∆B, and (→) spin–spin transverse relaxation with resultant more rapid loss in Time of flight. When the local (→) magnetization of mov(→) transverse magnetization and MR signal. MR signals ing tissue or fluid is selectively altered in a region, e.g., by can usually still be recovered as a spin echo in times less selective excitation, it will carry the altered magnetizathan or on the order of T2. 1/T2* ≅ 1/T2 + ∆ω/2; ∆ω = γ ∆B. tion with it when it moves, thus tagging the selected reNote that the FID will generally not be an exponential, so gion for times on the order of the relaxation times. This is that T2* will not be unique. the source of several flow effects. T2 Relaxation. (→) Transverse relaxation time.

Tip angle. (→) Flip angle.

T2-weighted (T2W). Often used to indicate an image where most of the contrast between tissues or tissue states is due to differences in tissue T2. This term may be misleading in that the potentially important effects of tissue density differences and the range of tissue T2 values are often ignored. A T2 contrast state is approached by imaging with a TR long compared with tissue T1 (to reduce T1 contribution to image contrast) and a TE between the longest and shortest tissue T2s of interest. A TR greater than 3 times the longest T1 is required for the T1 effect to be less than 5%. Due to the wide range of T1 and T2 and tissue density values that can be found in the body, an image that is T2-weighted for some tissues may not be so for others.

Torque. A vector quantity given by the vector product of the force and the position vector where the force is applied; for a rotating body, the torque is the product of the moment of inertia and the resulting angular acceleration.

Tagging. (→) Spin tagging. Tailored excitation. (→) Selective excitation. Tailored pulse. Shaped pulse whose magnitude (and possibly phase) is varied with time in a predetermined manner. Affects the frequency components of an (→) RF

Transaxial plane. The plane perpendicular to long axis of the human body (head to foot). Acquiring images in the transaxial plane acquires a stack of parallel images in the head-to-foot direction. Sometimes referred to as the “axial plane” or “transverse plane.” (See also: Orientation). Transmit/receive (T/R) coil. An RF coil that acts as both a transmitter (T) producing the B1 excitation field, and as a receiver (R) of the (→) MR signal. Such a coil requires a T/R switching circuit to switch between the two modes. A (→) body coil is typically a T/R coil, but smaller volume T/R coils (head/extremities) are often used at high field as a means of reducing RF power absorption. Transmitter. Portion of the MR apparatus that produces RF current and delivers it to the transmitting coil.

Glossary

Transmitter coil. Coil of the (→) RF transmitter, used in excitation of the spins. Transverse magnetization (Mxy). Component of the (→) macroscopic magnetization vector at right angles to the (→) static magnetic field (B0). Precession of the (→) transverse magnetization at the (→) Larmor frequency is responsible for the detectable (→) MR signal. In the absence of externally applied RF magnetic field, the transverse magnetization will decay to zero with a characteristic time constant of T2 or T2*.

Twister gradient. A (→) gradient pulse designed to dephase low spatial-frequency components in an image. The simplest such design is to choose the gradient strength so that a linear-phase change of – π to π is generated across the image. Two-dimensional Fourier transform imaging (2DFT). A form of sequential plane imaging using (→) Fourier transform imaging.

Two-dimensional MR. Form of MR spectroscopy in which an additional dimension is added to the convenTransverse plane. The plane perpendicular to long axis tional chemical shift dimension by allowing varying of the human body (head-to-foot). Sometimes referred amounts of different interactions between spin systems to as the “transaxial plane” or “axial plane” ((→) orienta- (such as NOE, spin–spin coupling, or exchange). tion). Variable flip angle. The temporal variation of (→) flip Transverse relaxation. The loss of (→) magnetization in angle (from one (→) RF pulse to the next) to enhance (→) signal-to-noise ratio (SNR), and/or equalize the signal inthe plane perpendicular to the static magnetic field, B0. tensity for each phase encoding step. Transverse relaxation rate. (→) Relaxation rate, transverVariable TE. The variation of echo time from one (→) RF sal (R2). pulse to the next. Transverse relaxation time (T2). Also called spin–spin relaxation time or T2-relaxation; the characteristic time Variable TR. The variation of repetition time from one constant for loss of phase coherence among spins ori- (→) RF pulse to the next. ented at an angle to the static magnetic field, due to interactions between the spins, with resulting loss of Vector. A quantity having both magnitude and direction, transverse magnetization and MR signal. Starting from a frequently represented by an arrow whose length is prononzero value of the magnetization in the x–y-plane, the portional to the magnitude and with an arrowhead at one x–y-magnetization will decay so that it loses 63% of its end to indicate the direction. initial values in a time T2, if relaxation is characterized by Velocity compensation. (→) Gradient moment nulling. a simple single exponential decay.

Traveling saturation pulse. A saturation pulse which moves from one spatial location to another for each (→) RF pulse excitation or each MR slice acquired. Trigger delay time. The time after triggering at which data acquisition takes place.

Vessel tracking. An image post processing method to separate vessels from surrounding tissue. Volume coil. (→) RF coil that surrounds a portion of the body.

Volume imaging. Imaging techniques in which (→) MR Triggering. Generation of an electrical pulse, upon de- signals are gathered from the whole object volume to tection of a physiological signal, that can be used to initi- be imaged at once, with appropriate encoding pulse RF and gradient sequences to encode positions of the (→) ate a synchronized data-acquisition pulse sequence. spins. Many sequential plane imaging techniques can be Truncation artifact. Artifactual ripples adjacent to edges generalized to volume imaging, at least in principle. Adin an image or sharp features in a spectrum, caused by vantages include potential improvement in (→) signal-toomission of higher frequency terms in (→) Fourier trans- noise ratio by including signal from the whole volume at form, particularly with the use of (→) zero filling to re- once; disadvantages include a bigger computational task place unsampled higher frequencies. ((→) apodization; for image reconstruction and longer (→) image acquisition times (although the entire volume can be imaged (→) Gibbs phenomenon). from the one set of data). Also called simultaneous volTuning. Process of adjusting the (→) resonance frequency, ume imaging or (→) three-dimensional Fourier transform e.g., of the RF circuit, to the desired value, e.g., the (→) (3DFT) imaging. Larmor frequency. More generally, the process of adjusting the components of the (→) spectrometer for optimal Volume of interest (VOI). A user-selected subset of (→) voxels in a three-dimensional dataset. performance, including matching impedances.

1463

1464

Glossary

Volume-selective excitation. Selective excitation of spins in only a limited region of space. This can be particularly useful for (→) spectroscopy as well as imaging, as spatial localization of the signal source may be achieved through spatially selective (→) excitation and the resulting signal may be analyzed directly for the spectrum corresponding to the excited region. It is usually achieved with selective excitation. Typically, a single dimension of localization can be achieved with one selective RF excitation pulse (and a (→) magnetic field gradient along a desired direction), while a localized volume (3D) can be excited with a stimulated echo produced with three selective RF pulses whose selective (→) magnetic field gradients are mutually orthogonal, having a common intersection in the desired region. Similar “crossed-plane” excitation can be used with selective 180° refocusing pulses and conventional spin echoes. A degree of spatial localization of excitation can alternatively be achieved with depth pulses, e.g., when using (→) surface coils for excitation as well as signal detection. An indirect application of selective excitation for volume-selected spectroscopy is to use appropriate combinations of signals acquired after selective inversion of different regions, in order to subtract away the signal from undesired regions.

Washout effects. (→) Flow void. Water suppression. The elimination or reduction of water signal from the image by application of a narrowband frequency-selective pulse centered around the resonant frequency of the tissue. Yoke. High-permeability structure used to concentrate the magnetic flux inside itself and thus decrease the magnetic fringe field and increase the field strength in the useful volume of a magnet. Zero filling. Substitution of zeroes for unmeasured data points in order to increase the matrix size of the new data prior to Fourier transformation of MR data. This can be equivalent to performing an interpolation in the transformed data, resulting in pixels smaller than the actual resolution of the image. Zeugmatography. Term for MR imaging coined from Greek roots, suggesting the role of the magnetic field gradient in joining the RF magnetic field to a desired local spatial region through magnetic resonance.

Voxel. Volume element; the element of 3D space corres­ This Glossary is based in part on the Definitions given in ponding to a pixel, for a given slice thickness. the ACR-Glossary of MR-terms, JMRI (3) 1993 and MRGlossary published by Siemens Medical Systems. Washin effects. (→) Flow-related enhancement.

List of Abbreviations

1 1

H MRS H NMR

2,3-DPG 2D SE 2D 2DFT 2D-TFLASH 3D MPRAGE 3D TOF MRA 3D 3DFT CISS 3D-TFLASH He 3-HMG-CoA 3

5-FDHU 5-FU 5-FUranuc Li C 13 C-[1H]-MRS 13 C-{1H}-MRS 9

13

F F-[1H]-MRS 19 F-{1H}-MRS 19 19

Na P 31 P-[1H]-MRS 31 P-{1H}-MRS 23

31

K 99m Tc-JAG3

39

I-MIBG 129 Xe AA AAA ABC 123

H magnetic resonance spectroscopy H nuclear magnetic resonance spectroscopy 2,3-diphosphoglycerate 2 dual-echo spin-echo two-dimensional 2D-Fourier-Transformation 2D-Turbo-FLASH (abbreviation: e.g.: 2D-TFLASH 10/4/10°) three dimensional magnetization prepared rapid acquired gradient echo 3D time-of-flight MR angiography three-dimensional three-dimensional fourier transformation constructive interference in steady state 3D-Turbo-FLASH (abbreviation: e.g.: 3D-TFLASH 10/4/10°) helium-3 3-hydroxy-3-methylglutaryl coenzyme A 5-fluoro-5,6-dihydrouracil 5-fluoro-uracil 5-fluoro-uracil-nucleotides and -nucleosides lithium-9 carbon-13 NOE-enhanced 13C-MR-spectroscopy proton decoupled 13C-MR-spectroscopy fluorine-19 NOE-enhanced 19F-MR-spectroscopy proton decoupled 19F-MR-spectroscopy sodium-23 phosphorus-31 NOE-enhanced 31P-MR-spectroscopy proton decoupled 31P-MR-spectroscopy potassium-39 technetium labeled mercapto-acetyltriglycine 123 I-meta-benzylguanine xenon-129 amino acids abdominal aortic aneurysms aneurysmal bone cysts 1 1

ABER ABI AC ACh AChR ACJ ACL ACR ACS ACTH ACVB AD ADC ADEM ADH ADP ADPKD AF AFP AGS AIDP AIDS AIF AL ALARA ALL ALS AMD AMI121 AMI227 AMI25 AML AMP

abduction–external rotation ankle-brachial index acquisition (number of acquisitions: AC = n) (also: adenylate cyclase) acetylcholine acetylcholine receptor acromioclavicular joint anterior cruciate ligament American College of Radiology American Cancer Society adrenocorticotropic hormone aortocoronary venous bypass Alzheimer’s disease apparent diffusion coefficient, (also: analog to digital converter) acute disseminated encephalomyelitis antidiuretic hormone (also: atypical ductal hyperplasia) adenosine diphosphate autosomal dominant polycystic kidney disease atrial fibrillation α-fetoprotein adrenogenital syndrome acute inflammatory demyelinating polyneuropathy acquired immunodeficiency syndrome arterial input function amyloid light chain (amyloidosis) (= primary amyloidosis) (also: annular ligament) “as-low-as-reasonably-achievable” acute lymphatic leucemia amyotrophic lateral sclerosis acid maltase deficiency code of Advanced Magnetics of nanoscaled iron oxide particle (Ferumoxsil) code of Advanced Magnetics of nano­ scaled iron oxide particle (Ferumoxtran = Sinerem®, Combidex®) code Advanced Magnetics of nano­ scaled iron oxide particle (Ferumoxidees) acute myeloid leukemia (also: myelogenous leukemia) adenosine-5'-monophosphate

1466

List of Abbreviations

ANA ANCA ANS Anti-Jo-1 Ao APD APLSA AQ ARCO ARD ARVC ARVD ARX ASD ASH ASL Asp ATM ATN ATP ATTR AV AVF AVM AVN B B1(t) BBB BC BCC BCE B-cell bFFE Bi (i = x, y, z)

BI-RADS® BLI Bloc BMEP BMES Bo BOLD

antinuclear antibody antineutrophil cytoplasmic antibodies anserine antihistidyl transfer RNA synthetase (Jo-1 Syndrome) aorta (also AO) avalanche photo diode Anterior labroligamentous periosteal sleeve avulsion signal acquisition phase of a sequence Association Recherche Circulation Osseuse arrhythmogenic right ventricular dysplasia arrhythmogenic right ventricular cardiomyopathy arrhythmogenic right ventricular dysplasia aristaless-related homeobox atrial septal defect (also: atrium septum defect) asymmetric septal hypertrophy arterial spin labeling aspartate acute transverse myelitis acute tubular necrosis adenosine-5’-triphosphate (also: autoimmune thrombocytopenia) amyloidogenic transthyretin atrio ventricula arteriovenous fistula arteriovenous malformation avascular necrosis magnetic induction (also: diffusion weighting value) time dependent magnetic induction des radio frequency field blood-brain-barrier bronchogenic carcinoma branchial cleft cyst bovine capillary endothelial Burkitt lymphoma cell balanced fast field echo magnetic field [The magnetic induction of this gradient fields changes along the direction in space x, y, z (Bi = Gi·x).] breast imaging reporting and data system bioluminescence imaging local magnetic induction bone marrow edema pattern bone marrow edema syndrome magnetic induction of the external static magnetic field blood oxygen level dependent (contrast)

BOLD-fMRI BOT BPH BRCA1 BRCA2 BS bSSFP bw C c C c C/N C1 C5b-9 CAA CAD CADASIL CAH CAR CASL CAV3 CBC CBD CBF CBFi CBV CC cc CCAM CCD CCD CCF CCT Cd CD CD8+ CDH cDNA CDP-choline CDUS CE FFE CE FLASH2D FS CE MRA CE CE CE-FAST

blood oxygen level dependency-functional magnetic resonance imaging base of tongue benign prostatic hyperplasia breast cancer susceptibility gene 1 breast cancer susceptibility gene 2 basisphenoid level balanced steady state free precession pulse sequence body weight carotid canal (also: clivus, diffusion weighting value) concentration contrast specific heat capacity contrast-to-noise cervical vertebral body complement factors 5b-9 cerebral amyloid angiopathy coronary artery disease cerebral autosomal dominant arterio­ pathy with subcortical infarct and leukoencephalopathy congenital adrenal hyperplasia carnosine continuous arterial spin labeling caveolin-3 gene complete blood count common bile duct cerebral blood flow cerebral blood flow index cerebral blood volume corpus callosum correlation coefficient congenital cystic adenomatoid mal­ formation central-core disease charge-coupled device carotid-cavernous fistula cranial computed tomography (also: cerebral computed tomography) cadmium compact disc cluster (of) differentiation (8+) congenital diaphragmatic hernia complementary deoxyribonucleic acid cytidine 5`-diphosphocholine color-Doppler-ultra sound (= T2 FFE) contrast-enhanced FLASH 2D fat saturation contrast-enhanced MR angiography contrast-enhanced contrast enhancement contrast-enhanced Fourier acquired steady state sequence

List of Abbreviations

CE-FFE-T1 CE-FFE-T2 CE-GRASS CE-MRA CEMRA CE-MRI CENTRA CF CFM CFTR CG CHAOS CHD CHESS Cho Cho/Cr-PCrratio CIDP

CISS FIESTA CISS CK CK-MB Cl CLIO CM CMD CML CM-MRA CM-MRI CMR CMV CNR

contrast-enhanced fast field echo T1-weighted contrast-enhanced fast field echo T2-weighted contrast-enhanced gradient-recalled acquisition in steady state contrast-enhanced MRA (also: CM-MRA) contrast-enhanced magnetic resonance angiography contrast-enhanced MRI contrast-enhanced time robust angiography cystic fibrosis color flow mapping cystic fibrosis transmembrane-regulator (protein) crista Galli congenital high-airway obstruction syndrome common hepatic duct chemical frequency selective fat presaturation choline (also: choline containing substances [e.g., PC, GPC, free Cho, CDP, acetylcholine, Cho-plasmalogen])

CNS CoA COPD

choline to creatine-phosphocreatine ratio chronic immune demyelinating polyneuropathy (also: chronic inflammatory demyelinating polyneuropathy; chronic inflammatory demyelinating polyradiuculoneuropathy) constructive interference in steady state fast-imaging steady state acqui­ sition constructive interference in steady state creatine kinase creatine kinase-myocardial band clivus Tf–S–S–cross-linked iron oxide contrast medium (also: cavernous malformations and congenital myotonia) congenital muscular dystrophy (also: corticomedullary differentiation) chronic myeloid leukaemia contrast media enhanced MRA (also: CE-MRA) contrast media enhanced MRI cardiac MR (also: cardiovascular magnetic resonance) cytomegalovirus (also: cytomegalic virus) contrast-to noise ratio (also: C/N)

CT CTA CTAP CTC CTD CTPH CTV CUG D

CP CPA CPM CPMG CPP CPT CPU CR Cr CRH CRMO CRN CRP Cr-PCr CS CSE CSF CSI

DAC DAVF dB/dt DCE MRI DCE DCIS DCM DCS DD DE MRI DEFT DESS DFOV dGEMRIC DHFU

central nervous system coenzyme A chronic obstructive pulmonary disease constrictive pericarditis cerebello-pontine angle central pontine myelinolysis Carr-Purcell-Meiboom-Gill (spin echo sequence) chronic pelvic pain carnitine palmitoyltransferase central processor unit abdominal and/or pelvic plain film radiography creatine (also: CRN) corticotrophin-releasing hormone chronic recurrent multifocal osteomyelitis see Cr C-reactive protein creatine-phosphocreatine carotid space conventional spin echo cerebrospinal fluid chemical shift imaging (also: spectroscopic imaging (SI) and also: metabolic imaging) (X-ray) computed tomography CT angiography CT (guided) arterial portograpy CT colonography connective tissue disease chronic pulmonary hypertension, clinical tumor volume cystography or cystourethrography diffusion (also: diffusion coefficient, diffusion constant) digital to analog converter dural arteriovenous fistulas time rate of change of magnetic flux dynamic contrast-enhanced MRI (also: DCE-MRI) dynamic contrast-enhanced ductal carcinoma in situ dilated cardiomyopathy, debt collection system differential diagnosis (also: distal muscular dystrophy) delayed-enhancement magnetic reson­ ance imaging driven equilibrium Fourier transform double-echo steady-state diameter of the imaging volume (→ FOV) delayed gadolinium-enhanced MRI of cartilage 5-fluoro-5,6-dihydrouracil

1467

1468

List of Abbreviations

DIA DIC DICOM DIG DISH DISI DM DNA DNET Dnm (n,m = x, y, z) DNs DNT DOI DOMS DORV DOT DPDN DRE DRIVE DS DSA DSC DSCT DSE DSMR DS-MRA DSP DST DT MRI d-TGA DTI dTMP DTPA DUS DVA DVD DW DWI E(r) E(t) ECA ECE ECF ECG EDMD EEG EF EHE Ekin

desmoplastic infantile astrocytoma delayed-onset muscle soreness digital imaging and communication in medicine desmoplastic infantile ganglioglioma diffuse idiopathic skeletal hyperostosis dorsal intercalated segment instability dermatomyositis deoxyribonucleic acid dysembryoplastic neuroepithelial tumor elements of the diffusion tensor dysplastic nodules (→ DNET) diffuse optical imaging delayed-onset muscle soreness double-outlet right ventricle diffuse optical tomography diphospho-dinucleotide digital rectal examination driven equilibrium radio frequency reset pulse digital subtraction digital subtraction angiography dynamic susceptibility-weighted dual-source CT double spin echo (also: dobutamine stress echocardiography) dobutamine stress imaging digital subtraction MRA digital signal processor dural sinus thrombosis diffusion tensor MRI d-transposition of the great arteries diffusion tensor imaging deoxythymidine-5'-monophosphate diethylene-triamine-pentaacedic acid (complex agent) Doppler ultrasonography (also: duplex ultrasonography) developmental venous anomaly digital versatile disc diffusion-weighted (also: Dw) diffusion-weighted imaging electric field at location r time dependent electric field external carotid artery extracapsular extension extracellular fluid space electrocardiogram Emery-Dreifuss muscular dystrophy electroencephalogram ejection fraction epithelioid hemangioendothelioma kinetic energy

ELF Em EMEA EMG EMX2 ENB ENG EPI EPI-DWI EPISTAR ERC ERCP ERF ERP ESCT ESR ESRD ET Etherm ETR EXIT F FA FAIR FAST Fat FBAL FDA FDG FDG-PET FdU FdUMP FdUrd FdUTP Fe3O4 FESS FEV1 FFE FFT FHPG FIAU

extremely low-frequency (magnetic fields) magnetic energy (Em = µ·Bo) (also Emag) European Medicines Agency electromyogram (also: electromyography) empty spiracles-homeobox 2 esthesioneuroblastoma electronystagmography echo planar imaging echo planar imaging-diffusion weighted imaging echo-planar imaging and signal targeting with alternating radio frequency endorectal surface coil endoscopic retrograde cholangio-pancreatography radio frequency energy (energy of photons; ERF = ħ·ωo) (also: Erad) endoscopic retrograde pancreato­ graphy European Carotid Surgery Trial erythrocyte sedimentation rate end-stage renal disease emission tomography thermal energy engineered human transferrin receptor ex utero intrapartum treatment fluorine fractional anisotropy flow-sensitive alternating inversion recovery Fourier acquired steady state fatty infiltration α-fluoro-β-alanine US Food and Drug Adminsitration fluorine-18-fluorodeoxy-glucose fluorine-18-fluorodeoxy-glucose-positron emission tomography 5-fluoro-2’-deoxyuridine 5-fluoro-2’-deoxyuridine-5’-monophosphate 5-Fluoro-2’-deoxyuridine 5-Fluoro-2’-deoxyuridine-5’-triphosphate iron(II,III) oxide (ferrous ferric oxide)(mineral: magnetide) functional endoscopic sinus surgery forced expiratory volume in 1s fast field echo fast Fourier transformation 9-[3-fluoro-1-hydroxy-2propoxymethyl]guanine 1-(2-deoxy-2-fluoro-beta-D-arabinofuranosyl)-5-iodouracil

List of Abbreviations

FID FIESTA FIGO FISP

FL FLAIR FLASH

FLOW COMP FMAU FMD FMPSPGR fMRI FMT FNA FNH F-Nuc(td) FO FOLD FOV fpVCT FRET FRFSE FRI FROGS fron. fs FS FSE FSE-IR FSH FSPGR FTD FUDP FUDPG FUMP FUPA FUra FUrd FUS

free induction decay fast-imaging steady state acquisition Fédération Internationale de Gynécologie et d’Obstétrique fast imaging with steady precession (abbreviation: e.g.: FISP 300/35/45°; repetition time TR = 300 ms, echo time TE = 35 ms, flip angle α = 45°. The field strength may follow: e.g. FISP 300/35/45°, 1.5 T.) Lorentz-force fluid attenuated inversion recovery fast low angle shot (abbreviation: e.g. FLASH 300/35/45°; repetition time TR = 300 ms, echo time TE = 35 ms, flip angle α = 45°. The field strength may follow: e.g. FLASH 300/35/45°, 1.5 T.) flow compensation 2`-fluoro-5-methyl-1-beta-D-arabinofuranosyluracil fibromuscular dysplasia fast multiplanar spoiled gradientrecalled functional magnetic resonance imaging fluorescence molecular tomography fine needle aspiration focal nodular hyperplasia fluorinated nucleotides foramen ovale flow level dependent contrast field of view flat-panel volumetric CT fluorescence resonance energy transfer fast recovery fast spin-echo fluorescence reflectance imaging fast rotating gradient spectroscopy frontal  = coronal fat suppressed (e.g.: fsSE) fat-saturated (also: fat saturation) fast spin echo fast spin-echo inversion-recovery follicle-stimulating hormone fast spoilt gradient-recalled acquisition in steady state (also: multislice fast spoiled gradient-recalled sequence) fronto-temporal dementia 5-fluorouridine-5’-diphosphate 5-fluorouridine-5’-diphospho(1)-α-Dglucose 5-fluorouridine-5'-monophosphate α-fluoro-β-ureidopropanoic acid 5-fluorouracil 5-fluorouridine focused ultrasound

FUTP FWHM g G-6-P GA

5-fluorouridine-5'-triphosphate full width at half maximum g-factor (dimension less) glucose-6’-phosphates read out gradient (also: frequency encoding gradient) γ-amino butyric acid GABA Gadobutrol gadolinium-10-(2,3-dihydroxy-1hydro­xymethylpropyl)-1,4,7,10- tetra­ azacyclododecane-1,4,7-triacedic acid Gadoteridol gadolinium-10-(2-hydroxypropyle)1,4,7,10-tetraazacyclododecane-1,4,7triacedic acid GBM glioblastoma multiformae Gd gadolinium (element of the rear earth) Gd-BOPTA gadolinium-1,4,7,10-tetraazacyclododecan-n,n’,n’’-triacedic acid Gd-DOTA gadolinium-1,4,7,10-tetraazacylcododecane-n,n’,n’’’-tetraacedic acid Gd-DTPA gadolinium-diethylenetriamine pentaacetic acid Gd-DTPA gadolinium-diethylen-triamine-penta­ acedic acid-dimeglumin ( = Magnevist®) Gd-DTPA-BMA gadolinium-diethylene-triamine penta­ acedic acid-bismethylamide Gd-HP-DO3A gadolinium-benzyloxy-propiontetra acedic acid GE gradient echo (also: GRE) GFAP glial fibrillary acidic protein GFP green fluorescence protein GFR glomerular filtration rate GG Gasser ganglion GH genetic hemochromatosis (also: geniohyoid, growth hormone) Gi (i = x, y, z) magnetic field gradient (index: direction of the field gradient) GI gastrointestinal tract GLAD Glenolabral articular disruption Gln glutamine Glu glutamate Glx glutamate complex (glutamine and glutamate) GM gray matter Gmax maximum magnetic field gradient GMR gradient motion rephasing GP phosphorylated glucose (also: phase encoding gradient) GPC glycerophosphorylcholine (also: glycerolphosphocholine) GPE glycerophosphorylethanolamine GRASE gradient and spin echo GRASS gradient recalled acquisition into steady state GRE gradient echo (abbreviation: e.g. GRE 300/35/45°; repetition time

1469

1470

List of Abbreviations

GRE-EPI GRH GS GTD GTV GVHD ħ H H3PO4 HAE HAGL HARDI HASTE HbS HCA HCC HCM HD HE HEME HER-2 HF HIFU HIV HL HLA HLA-DQA1 HLA-DR2 HLA-DR5 HLA-DRw52 HLHS HME HMN HMO

HMPT HMSN HNCM HOCM

TR = 300 ms, echo time TE = 35 ms, flip angle α = 45°. The field strength may follow: e.g. GRE 300/35/45°, 1.5 T.) gradient-recalled echo and echo-planar imaging growth hormone-releasing hormone slice selection gradient gestational trophoblastic disease gross tumor volume Graft-versus-host disease Planck’s constant (ħ = h/2π = 1.05457168(18) · 10-34 Js; h = 6,6260693(18) · 10–34 Js) magnetic field strength (H = B + M) orthoposphoric acid hepatic alveolar echinococcosis humeral avulsion glenohumeral ligament high angular resolution diffusion imaging half Fourier acquisition single shot turbo spin echo hemoglobin S hepatocellular adenomas hepatocellular carcinoma hypertrophic cardiomyopathy Hodgkin’s disease hepatic encephalopathy hemorrhage early MRI evaluation human epidermal growth factor receptor-2 half Fourier highintensity focused ultrasound human immunodeficiency virus Hodgkin’s lymphoma human leukocyte antigen human leukocyte antigen DQA1 human leukocyte antigen DR2 human leukocyte antigen DR5 human leukocyte antigen DRw52 hypoplastic left heart syndrome hereditary multiple exostoses hereditary motor neuronopathy classification has been proposed for MRI, based on the size of the urogenital hiatus (H), the size of the muscular relaxation (M) and the degree of organ prolapse (O) hexamethylphosphorus triamide hereditary motor and sensory neuropathy hypertrophic non-obstructive cardiomyopathy hypertrophic obstructive cardiomyo­ pathy

HP hPAP HRCT HS HSE HSV DNA HSV-1 HSV-2 HT HTLV-1 HU HVS I I i.a. i.v. IAAA IAC IARC IBM ICA ICD ICH ICNIRP ICP IDEM IEC IGEPI IgG i-IBM IJV ILT IMA IMH IMP INN Ins IOF IPAH iPAT IPMT IPS IR

IR-FSE ISIS ISMRM

hard palate human prostatic acid phosphatase (also: human placental alkaline phosphatise (HPAP)) high-resolution computed tomography hemodialysis shunts herpes simplex enzephalitis herpes simplex virus DNA herpes simplex virus type 1 herpes simplex virus type 2 hemorrhagic transformation human T-cell leukemia virus type 1 Hounsfield units hyperintense vessel sign nuclear angular momentum spin quantum number intra-arterial intra-veneous Inflammatory abdominal aneurysm internal auditory canal International Agency for Research on Cancer inclusion body myositis internal carotid artery implantable cardioverter–defibrillator intracranial hemorrhage International Commission on NonIonising Radiation Protection intracranial pressure intradural extramedullary International Electrotechnical Commission interleaved gradient echo planar imaging immunoglobulin G inherited inclusion body myopathies internal jugular vein interstitial laser therapy internal mammary artery intramural hematoma inosine-5’-monophosphate international non-proprietary name inositol infraorbital foramen idiopathic pulmonary hypertension integrated parallel imaging technology intraductal papillary mucinous tumor idiopathic Parkinson’s syndrome inversion-recovery sequence (abbreviation e.g.: IR 1400/400/35; repetition time TR = 1400 ms, inversion time TI = 400 ms, echo time TE = 35 ms) inversion recovery fast spin echo image-selected in vivo spectroscopy International Society of Magnetic Resonance in Medicine

List of Abbreviations

IT ITAC IVIM IVP IVS1 IVSBAT IVU IWMA J(ω) j Jax JB JNA JZ k LA Lac LAD LAVA LBBB LC LCA LCH LCIS LCL LCLC LD50 LDH LED LEMS LGMD LH LHR LIF Lip LIS LITT LL LOH LP LPSVD LRLN LRPN LT L-TGA

inferior turbinate Intestinal-type adenocarcinoma intra-voxel incoherent motion intravenous pyelogram intervening sequence 1 intravascular and sclerosing bronchioloalveolar tumors Intravenous urography stress-inducible wall motion abnormalities spectral density distribution (function) current density (scalar) coupling constant between nucleus A and X jugular bulb juvenile nasal angiofibroma junctional zone Boltzmann-constant (k = 1,3804 · 10-23 J/K) left atrium lactate left anterior descending coronary artery liver acquisition with volume acceleration left branch bundle block longus colli muscle left coronary artery Langerhans cell histiocytosis lobular carcinoma in situ lateral collateral ligament lateral collateral ligament complex lethal dose (a dose at which 50% of animal will die) lactate dehydrogenase light-emitting diode Lambert-Eaton myasthenic syndrome limb-girdle muscular dystrophy luteinizing hormone luteinizing hormone receptor local intraarterial fibrinolysis lipid lissencephaly spectrum Laser (induced) interstitial thermo therapy (also: laser-induced thermal therapy) lactate and lipid loss of heterozygosity lateral pyloric linear prediction and singular value decomposition left recurrent laryngeal neurectomy lateral retropharyngeal nodes lunotriquetral L-transposition of the great arteries (also: l-TGA)

LUCL LV LVAS LVOT LWS m M m MSS  Z Mwxy  M1 Ma MAPCA MAPCAs MARIBS MAS MC MCA MCF MCL MD MDCT Me MECP2 MELAS MEN I MEN IIa MEN MFS M xyf  Mg MG Mi (i = x, y, z)

MI mI MIBG MION MIP MMA MMP-2

lateral ulnar collateral ligament left ventricle large vestibular aqueduct syndrome left ventricular outflow tract Lesser wing of the sphenoid magnetic moment magnetization mass steady-state-z-magnetization water component of the transversal magnetization motor cortex masseter major aortopulmonary collateral arteries mid-aortic pulmonary arterial collaterals magnetic resonance imaging for breast screening (trial (UK)) magic-angle-spinning (also: minimal access surgery, macrophage activation syndrome) Monte Carlo (also: mononuclear cells, medullary carcinoma) middle cerebral artery (also: Medical Control Agency) middle cranial fossa medial collateral ligament muscular dystrophy multidetector computed tomography (macroscopic) magnetization caused by the electrons methyl-CpG-binding protein 2 myopathy, encephalopathy, lactic acidosis, stroke-like episodes Wermer’s syndrome (→ MEN) Sipple’s syndrome (→ MEN) multiple endocrine neoplasms Marfan syndrome fat component of the transversal magnetization manganese myasthenia gravis magnetization [index: direction of the magnetization vector; Mz = longitudinal magnetization, Mx,y = transversal magnetization)] molecular imaging myo-isonitol meta-benzylguanidine (e.g. 131I- or 123I- MIBG monocrystalline iron oxide nanoparticles maximum intensity projection middle meningeal artery matrix metalloproteinase 2

1471

1472

List of Abbreviations

Mn MND Mn-DPDP

manganese motor neuron disease manganese(II)-n,n-dipyridoxylethyldiamine-n,n-diacetate-5,5-diphosphate (=Mangafodipir = Teslascan®) MOD magneto-optical disc MOLD motion level dependent (contrast) MOTSA multiple thin slice acquisition (3D-TOF-MRA-sequence) MP medial pterygoid MPM malignant pleural mesotheliomas (also: muscles converge toward the pterygoid laminae) MPNST malignant peripheral nerve sheath tumor MPR multi planar reconstruction MPRAGE magnetization-prepared rapid acquisition gradient echo MR magnetic resonance MRA magnetic resonance angiography (also: MR-Angiography) MRC MR colonography MRCP magnetic resonance cholangiopancreatography MRI magnet resonance imaging MRKH Mayer-Rokitansky-Kuster-Hauser (syndrome) MRM magnetic resonance mammography mRNA messenger ribonucleic acid MRPN medial retropharyngeal nodes MRS magnetic resonance spectroscopy MRSI Magnetic resonance spectroscopic imaging MRT magnetic resonance tomography MRV magnetic resonance venography MS multiple sclerosis (also: masticator space) MSA multiple system atrophy (also: myositis-specific antibody) MSCT multislice computed tomography MSH melanocyte stimulating hormone MS z steady-state magnetization MT magnetization transfer MTC magnetization transfer contrast MTS magnetization transfer saturation (also: magnetization transfer suppression) MTT mean transit time MUP motor unit potential MUS maximum upslope MVD measurement of permeability Na+/K+-ATPase Na(+)/K(+)-adenosine-triphosphatase NAA N-acetyl-l-aspartate NAAG N-acetylaspartylglutamate NAC number of acquisitions NaCl natrium chloride

NADP+, NADPH nicotine-amide-adenine-dinucleotide NAFLD nonalcoholic fatty liver disease NASCET North American Symptomatic Carotid Endarterectomy Trial NASH nonalcoholic steatohepatitis NAWM normal appearing white matter Nb3Sn niobium-tin-alloy NBCCS Nevoid basal cell carcinoma syndrome NbTi niobium-titanium-alloy NC 100150 Code of Nycomed NC nasal cavity NdBFe neodymium-boron-iron (magnetic material) NDP nucleoside-5’-diphosphate NDPG nucleoside-5’-diphosphate glucose NE number of echos NEX number of excitations NF1 neurofibromatosis type 1 NF2 neurofibromatosis type 2 nGFP nuclear-localized green fluorescent protein NHL non-Hodgkin’s lymphoma NICE National Institute for Health and Clinical Excellence NIH National Institutes of Health NINDS National Institute of Neurological Disorders & Stroke NIRF near-infrared fluorescence NMO neuromyelitis optica NMR nuclear magnetic resonance NOE Nuclear Overhauser effect NP nasopharynx NPC nasaophaneoplasm NPH number of phase encoding steps NPO nothing by mouth NPV negative predictive value nRA normalized (scaled) anisotropy NSA number of steps ascended NSAID nonsteroidal anti-inflammatory drug NSF nephrogenic systemic fibrosis NTP nucleoside-5’-triphosphate (also: α-, β- and γ-nucleoside-5’-triphosphate) Ø diameter OA osteoarthritis OAR organs at risk OB olfactory bulb OC occipital condyles OCD osteochondritis dissecans OCT optical coherence tomography OD osteochondritis dissecans OH hydroxyl (group) OI optical imaging OLT osteochondral lesion of the talus ON optic nerve OPLL ossification of the posterior longitudinal ligament

List of Abbreviations

OPMD OT p.i. P792 PA PACS PAH PANK2 PAOD. PAPVC PAPVR PASC PASL PAT PBC PC PC(-MRA) PCA PC-FIESTA PCI PCL PCr PCR PCS PD FS PD PDA PDE PDFS PDW PDWI PE PET PET-CT PFPE PH PHACES

pHi Pi

oculopharyngeal muscular dystrophy optical tomography post injectionem monodisperse monogadolinated macromolecular compound pulmonary artery picture archiving and communication system pulmonary arterial hypertension pantothenate kinase 2 peripheral arterial occlusive disease partial anomalous pulmonary venous connection partial anomalous pulmonary venous return phased array surface coils pulsed arterial spin labeling parallel imaging techniques primary biliary cirrhosis phase-contrast (also: phosphor choline, phosphorylcholine) phase contrast (MRA) prostatic carcinoma (also: portacaval anastomosis) phase cycled fast-imaging steady state acquisition percutaneous coronary intervention posterior cruciate ligament (also: pubococcygeal line) phosphocreatine (also: creatine phosphate; (P)Cr; [P]Cr) polymerase chain reaction posterior cervical space proton density fat-saturated proton density (also ρp); Parkinson’s disease (also: prion diseases) patent (ductus arteriosus Botalli) phosphodiester pelvic disease-free survival proton density weighted (also: pw [ = ρw]) proton density weighted phosphoryl-ethanolamine positron-emission tomography (combined) positron emission tomography and computed tomography perfluoro-polyether pulmonary hypertension posterior fossa malformations, facial hemangiomas, arterial anomalies, cardiac anomalies and aortic coarctation, eye anomalies and sternal clefting/supraumbilical raphe intracellular pH-value inorganic phosphate

PI PI MRA

parallel imaging parallel imaging magnetic resonance angiography PICA posterior inferior cerebellar artery PICORE proximal inversion with a control for off-resonance effects PID pelvic inflammatory disease PIN(-diode) p-type–intrinsic–n-type-diode (semiconductor) PISI palmar intercalated segment instability Pixel acronym for picture elements pKa, acid dissociation constant PLS primary lateral sclerosis PM polymyositis PMA progressive muscular atrophy PME phosphomonoesters PML progressive multifocal leukoencephalopathy PMS pharyngeal mucosal space PNET primitive neuroectodermal tumor PNH paroxysmal nocturnal hemoglobinuria PNS paranasal sinuses pO2 partial oxygen pressure POPQ pelvic organ prolapse quantification PPA partially parallel acquisition PPF pterygopalatine fossa PPS parapharyngeal space PPV positive predictive value PRE proton relaxation enhancement PRESS point resolved spectroscopy PRESTO principal of echo-shifting with train of observation PRL prolactin PROPELLER periodically rotated overlapping parallel lines with enhanced reconstruction PS Parkinson’s syndrome PS parotid space PSA prostate-specific antigen (also: persistend stapendial artery) PSC primary sclerosing cholangitis Pseudo-TORCH (→ TORCH) PSIF FISP-sequence, in reversed time order (FISP →PSIF) PSIL percentage signal intensity loss PTA(S) percutaneous transluminal angioplasty (and stent) PTCA percutaneous transluminal coronary angioplasty PTEN phosphatase and tensin (homolog) PTLD posttransplant lymphoproliferative disease PTV planning tumor volume PVL periventricular leukomalacia PVNS pigmented villonodular synovitis PVS perivascular space (also: pulmonary valve stenosis)

1473

1474

List of Abbreviations

pw ( = ρw)

PWI PXA q Q QD QRS-complex QUIPSS r RA RAC RACE RADISH RAF RAI RAO RARE RAS RBC RCA rCBV RCC RCL RCM RD– RD+ RELN RES RESTORE RF RFP RGB RGD RI RIME RIND rMTT RNA RNV RODEO ROI

proton density or proton weighted [e.g. images with the following sequence: SE ≥1200/≤35, SE ≤600/≤35, IR 1400/400/≤35, IR 1400/150/≤35] (also: PDW) perfusion weighted imaging pleomorphic xantho-astrocytoma charge quality quantum dots ventricular depolarization of EEC quantitative imaging of perfusion using a single subtraction reticular right atrium (also: rheumatoid arthritis, relative anisotropy) right a. carotis communis real-time acquisition and evaluation of motion technique Renal Artery Diagnostic Imaging Study in Hypertension (trial) retroantral fat radioactive iodine right anterior oblique rapid acquisition with relaxation enhancement renal arteray stenosis red blood cell right coronary artery regional cerebral blood volume (also: relative ~) renal cell carcinoma radial collateral ligament restrictive cardiomyopathy irreversible defect reversible defect reelin gene reticuloendothelial system “restored” magnetization (fast recovery fast spin echo) radio frequency red fluorescent protein red-green-blue recognition sequence arginine-glycineaspartic acid resistive index receptor-induced magnetization enhancement (also: reporter-interpretermanager-educator) reversible ischemic neurological deficit relative mean transit time ribonucleic acid (also: radionuclide angiography) radionuclide ventriculography rotating delivery of excitation off resonance region of interest

ROPE ROST RPS rrCBV rRNA RTP rTTP RV RVOT RVT Rx s S S(t) S/N S1 SA SAE sagit. SAH SAR SC SCC SCHE SCIWORA SCM SD (also S) SDE SDr Se SE

SEER SEGA SEMS

reordered phase encoding (no sequence; method to suppress artifacts) resonant offset acquired steady state retropharyngeal space relative regional cerebral blood volume (see also: rCBV) ribosomal ribonucleic acid radiotherapy treatment planning relative time-to-peak right ventricle right ventricular outflow tract renal vein thrombosis receive (coil) slew rate spin (eigen vector of the angular momentum) (also: signal [see also SI], sulfur) time dependent MR-signal signal-to-noise primary sensory subarachnoidal stimulated acoustic emission (also: subcortical arteriosclerotic encephalopathy (M. Binswanger)) sagittal sub-arachnoid hemorrhage specific absorption rate surface coil squamous cell carcinoma subclinical hepatic encephalopathy spinal cord injury without radiographic abnormality sternocleidomastoid slice thickness (SD = n; units in mm) subdural empyema empirical standard deviation selenium spin echo (sequence) (abbreviation: SE 1600/35,70; repetition time TE = 1600 ms, double echo with echo times TE = 35 ms and TE = 70 ms, SE 1600/35 … 280,35; repetition time TR = 1600 ms, multiple echo with echo times, with 35 ms intervals, starting with an echo time of TE = 35 ms, than TE = 70 ms, TE = 105 ms, TE = 140 ms etc., with a max. echo time TE = 280 ms. The field strength, and number of acquisitions and slice thickness may follow: e.g. SE 300/35,35; 1.5 T.; SD = 3 mm; AC = 4.) (also: signal enhancement) surveillance, epidemiology and end results subependymal giant cell astrocytoma spin echo multi slice

List of Abbreviations

SENSE SFR SG SGE SGOT

sensitivity encoding stable free radicals signal generator spoiled gradientecho serum glutamic-oxaloacetic transaminase SHU 555 A code of Bayer Schering Pharma AG, Berlin, Resovist® SHU 555 C code of Bayer Schering Pharma AG, Berlin, Supravist® SI signal intensity (also: S (signal)); spectroscopic imaging (→ chemicalshift imaging [CSI]) s-IBM sporadic inclusion body myositis SL scapholunate SL slice thickness SLAP superior labrum anterior-posterior SLE systemic lupus erythematosus SLS sublingual space SMA spinal muscular atrophies (also: supplementary motor area) SMASH simultaneous acquisition of spatial harmonics smax maximum slew rate SMG sternocleidomastoid muscle Smg submandibular snapshotFLASH (= turboFLASH; → FLASH) snapshotGRASS (→ GRASS) SNR signal to noise (also S/N) SOF superior orbital fissures SONK spontaneous osteonecrosis of the knee SP soft plate SPAQ sensitive-particle acoustic quantification SPECT single-photon-emission computed tomography SPGR spoilt gradient-recalled acquisition in steady state SPIO superparamagnetic iron oxide particles SPIR spectral presaturation with inversion recovery (also: selective partial inversion-recovery) SPNET supratentorial primitive neuroectodermal tumor spot spotted SQ subcutaneous SR saturation recovery SS steady state (also: sphenoid sinus) SSC steady state coherent SSFP steady-state free precession SSFSE single-shot fast spin echo (also: SS-FSE) SSI steady state incoherent SSPE subacute sclerotic panencephalitis SSRARE single-shot rapid acquisition with refocused echoes

STAR STE STEAM STED STH STIR SV SV40 SVD SVS SWI T1 T1w T2 T2* T2w TA TA TAA tAC TAO TAPVC TAPVR TART Tat Tau TB TBC TBE TBI TCC tCr TCT TD TE Te TEE teff TFA TFC

signal targeting with alternating radiofrequency stimulated echo stimulated echo acquisition mode stimulated emission depletion (microscopy) somatotrope hormone short tau inversion recovery (also: short-TI-inversion recovery) 434, 1085, 1143 single voxel simian virus 40 singular value decomposition susceptibility vessel sign susceptibility weighted imaging T1-relaxation time (longitudinal relaxation time or spin lattice-relaxation time (units in ms (ms)) (also: T1) T1-weighted [e.g.: SE ≤600/≤35] (also: T1w) T2- relaxation time (transversal relaxation time or spin-spin- relaxation time (units in ms (ms)) (also: T2) T2*- relaxation time (effective relaxation time); units in ms (ms)) (also: T2*) T2-weighted [e.g.: SE ≥1200/≥70] (also: T2w) Takayasu’s arteritis truncus arteriosus thoracic aortic aneurysms acquisition times thromboangiitis obliterans total anomalous pulmonary venous connection total anomalous pulmonary venous return testicular adrenal rest tumors transactivation transcription taurine tuberculosis truncus brachiocephalicus tick-borne encephalitis traumatic brain injury transitional cell carcinoma (urothelial carcinoma) total creatine thermoacoustic computed tomography delay time (units in ms (ms)) (also: TD, triple dose) echo time (units in ms (ms)) (also: TE) technetium transesophageal echocardiography effective stimulation time trichloroacetic acid triangular fibrocartilage

1475

1476

List of Abbreviations

TFCC TFE TFL TFT TGA TGDC TGSE TI TIA Tim® TIPSS TIR TIRM TMA TNF TNM TOF TOF MRA TOF-MRA TONE TORCH tP TP53 TPMT TR Tr trans. TREAT TRH t-RNA trueFISP TRUS TSC TSE TSE TSH TTP TUR-B turboFLASH

turboIR

triangular fibrocartilage complex turbo field echo see turboFLASH thin film transistor transposition of great artery thyroglossal duct cyst turbo gradient spin-echo inversion time (units in ms (msec)) transient ischemia attack total imaging matrix (Siemens Medical Solutions, Erlangen, Germany) transjugular intrahepatic portosystemic stent shunt turbo inversion recovery turbo inversion recovery magnitude trimethylammonium tumor necrosis factor tumour node metastasis tetralogy of Fallot time of flight magnetic resonance angiography time-of-flight-MRA tilted optimized non-saturated excitation toxoplasmosis, other agents, rubella, cytomegalovirus, herpes simplex pulse length (on-time of RF-pulses) tumor protein 53 thiopurine S-methyltransferase repetition time (units in ms (msec)) (also: TR) trachea transversal (also: axial) time-resolved echo-shared angiographic technique thyrotropin-releasing hormone transfer ribonucleic acid FISP-sequence (fully symmetric sequence) → FISP transrectal ultrasonography tuberous sclerosis complex transmissible spongiforme enzephalopathie turbo spin echo (abbreviation: e.g. TSE 100/40) thyroid stimulating hormone time to peak (also: tissue transition projection) transurethral electroresection of the bladder turbo fast low angle shot (= TFL; also turbo field echo = TFE) (abbreviation: TFL 10/4/10°; repetition time TR = 10 ms, echo time TE = 4 ms, flip angle α = 45°) turbo inversion recovery (=TIR) (→ turboFLASH)

TUR-P Tx U U1 snRNP U1-RNP UADT UAE UCL UDPG UH UICC UNFAIR US USPIO UTSE v VCI VCN VCS VEGF VENC VET VF VGF VHL VIBE VN VOI Voxel VQ VR VS VS-ASL VSD VSOP VSOP-C184 VT WB-MRA WB-MRI WHO WM WML XeCT X-linked XLIS

transurethral resection of prostate transmit (coil) voltage U1 small nuclear ribonucleoprotein U1-ribonucleoprotein upper aerodigestive tract uterine arterial embolization ulnar (medial) collateral ligament uridine 5`-diphosphoglucose electric voltage induced by electric charge moving in a magnetic field Union Internationale Contre le Cancer uninverted flow-sensitive alternating inversion recovery ultra sound (also: ultrasonography) ultrasmall superparamagnetic iron oxide particles ultra fast turbo spin-echo (also: ultra short turbo spin echo) velocity (also ν) vena cava inferior ventral cochlear nucleus vena cava superior vascular endothelial growth factor velocity encoding ventricular extra-systole volume fraction vascular growth factor von Hippel-Lindau volumetric interpolated breath-hold examination (also: volumetric interpolated brain examination) vidian nerve volume of interest acronym for volume elements ventilation-perfusion scintigraphy (scan) volume ratio visceral space velocity-selective arterial spin labeling ventricular septal (septum) defect very small superparamagnetic iron oxide particles Code of Ferrupharm, Teltow, Germany ventricular tachycardia whole-body magnetic resonance angio­ graphy whole-body magnetic resonance imaging World Health Organization white matter white matter lesion xenon CT cross linked x-linked subcortical band heterotopia and lissencephaly

List of Abbreviations

Zn α-ADP α-ATP α-NDP α-NTP β-ADP β-ATP β-NDP β-NTP γ-ATP γ-NTP α αopt α-G-6-P α-MSH β−HCG γ δ χ Φn ε µ µk

zinc α-adenosine-5’-diphosphate α-adenosine-5’-triphosphate α-nucleoside-5’-diphosphate α-nucleoside-5’-triphosphate β-adenosine-5’-diphosphate β-adenosine-5’-triphosphate β-nucleoside-5’-diphosphate β-nucleoside-5’-triphosphate γ- adenosine -5’-triphosphate γ-nucleoside-5’-triphosphate flip angle optimal flip-angle α-glucose-6-phosphate α-melanocyte stimulating hormone β−human chorionic gonadotropin gyromagnetic ratio chemical shift magnetic susceptibility phase angle dielectric constant nuclear magnetic moment nuclear magneton (5,050824 · 10-27 J/T) (also µm or µn)

µr µz ω ω0 ωRF π- pulse π π/2- pulse ρ-weighting ρ (1H) ρ = N/V ρ ρp σ τ τc

relative permeability z-component of the nuclear magnetic moment precession frequency Larmor frequency frequency of the irradiated radio frequency 180°-pulse pulse distance (time interval between two subsequent RF-pulses) 90°-pulse proton-weighting (also spin density weighting) spin density (e.g. of 1H etc.) spin density (N = number of spins; V = volume) tissue density proton density (1H-spin density) electrical conductivity (of a medium) (also: screening constant) pulse distance (used to denote different time delays between RF-pulses) (also: equilibration time) correlation time (also: rheobase)

1477

Subject Index

A ABC see aneurysmal bone cyst abdomen/abdominal – imaging technique  864 – lymphoma  106 abdoscan  106 ABER position  1140 ABI see ankle-brachial index ablation catheter  1285 Ablynx  1425 abscess  354, 668 – amebic  884 – Candida  365 – cerebral  358 – intramuscular  1194 – of the pituitary gland  422 – paravertebral  1082 – pyogenic  361, 883 – tuberculous  363 Acanthamoeba  370 acetazolamide  179, 183 acetylcholine  1205, 1344 ACh see myasthenia gravis Achilles tendon  1149 – contractures  1200 acid maltase deficiency  1202 ACJ see acromioclavicular joint ACL – rupture  1122, 1124 – tear  1121 acoustic – neurinomas  287 – noise  84 – schwannoma  287, 288 acquisition time  44 acromioclavicular joint  1131 acromion  1133 acromioplasty  1136 active – shielding  80, 84, 1273 – tracking coil  1263 acute – arterial embolism  841 – disseminated encephalo­ myelitis  645

– inflammatory demyelinating polyneuropathy  1206 – ischemic stroke, see also stroke/ ischemic stroke  317 – MR findings  311 – lymphocytic leukemia (ALL)  602 – myelogenous leukemia (AML)  885 – renal failure  1204 acyl-coenzyme A (CoA)  1202 acylcarnitine transferases  1202 Adamkiewicz artery  831 ADC see apparent diffusion coefficient (ADC) adenocarcinoma  468, 684, 1243 – of the bladder  1007 – of the colon  909 – of the stomach  906 – pancreatic ductal  886 adenoidal hypertrophy  491 adenoid cystic carcinoma  468, 470, 494 adenoma  527 – ampullary  896 – gonadotroph  283 – hepatocellular  867, 870 – invasive  283 – papillary  896 – parathyroid  528 – pituitary  281, 285 – pleomorphic  516 – sebaceum  219 adenomyosis  970, 972, 996 – diffuse  970 – focal  970 adenosine  716 – diphosphate (ADP)  1292 – monophosphate (AMP)  1202, 1345 – phosphate  1210 – triphosphate (ATP)  1292, 1344 adenylate deaminase  1202 adequate adenosine triphosphate (ATP)  1201 adiabatic inversion pulse  117

adipose tissue – quantitative analysis  1246 adnexal lesions  986 – characterization  986 – morphologic criteria  988 ADP see adenosine phosphate ADPKD see autosomal dominant polycystic kidney disease adrenal gland  687, 943 – adenomas  945 – Conn’s syndrome  945 – Cushing’s syndrome  945 – carcinoma  946 – in-phase  943 – metastasis  946 – opposed-phase  943 adrenoleukodystrophy – X-linked  233 adrenomyeloneuropathy  233 adventitia  837 Ag/AgCl  89 agyria  378 – pachygyria complex  200 Ahlbäck’s disease  1114 Aicardi-Goutières syndrome  237 Aicardi syndrome  205 AIDP see acute inflammatory demyelinating polyneuropathy AIDS (acquired immunodeficiency syndrome) – aspergillosis  365 – cholangiopathy  895 – encephalopathy  376, 1352 – lymphoma  268 – toxoplasmosis  370 ALARA principle  1272 albumin  93 alcohol/alcoholic – exposure  1207 – hepatitis  1354 – liver disease  873 alcoholism  1113 aldolase  1204, 1354 Alexander disease  238 ALL see acute lymphocytic leukemia Alpers disease  241

1480

Subject Index

ALS see amyotrophic lateral sclerosis alveolar proteinosis  680 Alzheimer’s disease  343, 381, 1352 AMD see acid maltase deficiency amebiasis  370 ammonia  1202 AMP see adenosine monophosphate ampulla of Vater carcinoma  897 amyloid  744 amyloidosis  739, 744 – ATTR  744 – cardiac  745 amyotrophic lateral sclerosis (ALS)  389, 1208 ANA see antinuclear antibody anaplastic astrocytoma  251 anatomy  817 Anderson-Fabry disease  749 androgen-independent pathway  1419 aneurysm  121, 424, 635 – aortic  830, 836 – atherosclerotic  830 – congenital  831 – intracerebral  341 – intracranial  341, 814, 817 – mycotic  831 – peripheral  837 – saccular  341, 830 – screening  342 aneurysmal bone cyst  565, 1166 angiogenesis  100, 1428 angiography unit  1260 angiolipoma  569 angioma  224, 337 angiomyolipoma  928 – of the liver  866 angioplasty  835 angiosarcoma  480, 783, 876 AngioSURF technique  1239 anhydramnios  1228 anisotropy  132, 173 – indice  134 – of diffusion  134 ankle  1147, 1149 – brachial index (ABI)  838 – capsuloligamentous pathologies  1149 – complete rupture  1149 – stress fractures  1147 – tendon pathology  1149 – traumatic sprain  1149 ankylosing spondylitis  1105 annexin-V  1422 annular – disruption  542

– ligament  1153 – pancreas  886 anomalous pulmonary venous return  807 anorchism  1043 anorectal angle  1072 anserine (ANS)  1210 Anson and Spetzler classification  631 anterior – and posterior median spinal vein  637 – cord syndrome  654 – cruciate ligament (ACL)  1082, 1121 – injury  1117 – median spinal vein  626 – spinal artery  626, 627, 629, 636 – talofibular ligament  1149 antinuclear antibody  1189 aorta/aortic – and renal arteries  803 – aneurysm  830 – anomalies  729 – arch anomalies  830 – arch interruption  829 – coarctation  729, 829 – congenital anomalies  829 – dissection  832 – penetrating atheromatous ulcer  834 – regurgitation  757 – Salmonella infections  831 – stenosis  755, 757 – stent-graft imaging  836 – traumatic rupture  836 aortitis  835 – syphilitic  836 aortoiliac Y prosthesis  1237 APD see avalanche photo diode aphasia  174, 1317 aplastic anemia  1093 apparent diffusion coefficient (ADC)  7, 87, 132, 298, 318, 358 – maps  136, 137, 145, 174, 229, 231 – encephalitis  376 – herpes  376 arachnodactyly  834 arachnoid – cysts  198, 207, 288, 414, 604, 611, 613 – IDEM compartment  603 – mater  591 arachnoiditis  554, 607, 610, 652 – aseptic  610

arcuate ligament syndrome  845 arginine–glycine–aspartic acid peptide  1428 Arnold-Chiari malformations  212 Arnold’s nerve  521 arrhythmia  1285 arrhythmogenic right ventricular dysplasia  740 arterial – dissection  324 – emboli  841 – input function (AIF)  179 – portography  101 – spin labeling (ASL)  179 – acetazolamide  183 – CBF maps  182 – dubbed velocity selective  181 – perfusion maps  182 – pseudo-continuous method  181 – vessel  837 arteria lisoria  830 arteriolosclerosis  837 arteriosclerosis Mönckeberg  837 arteriovenous – fistulas (AVF)  612, 843 – malformation (AVM)  334, 530, 612, 631, 808, 814, 842 – intratesticular  1047 – pulmonary  809 arteritis  835 – temporal  843 artery of Adamkiewicz  626, 627, 629, 636 arthralgia  1191 arthritis  1087 – rheumatoid  1104 – septic  1087, 1097, 1100 arthrofibrosis  1125 arthrographic effect  1099, 1127 arthrogryposis  1197 – multiplex  198 arthropathy – diabetic  1096 – hemophilic  1128 articular disk  1146 – displacement  1146 – perforation  1146 artifact  1302 – correction  1303 – geometric  1323 – in diffusion MRI  140 ARVD  740 – fatty  740 – fibro-fatty  740

Subject Index

– major/minor diagnostic criteria  741 – sudden death  740 ARX gene  200 asbestos  689 ascites  904 ASD see atrial septal defect aser-capture  1419 ASL see arterial spin labeling aspartate aminotransferase  1209 aspergillosis  365, 366, 678 asphyxia  1350 aspiration needle  1266 asplenia syndrome  732, 897 assistance system  1266 astroblastoma  262 astrocytal marker  1344 astrocytes/astrocytic  245 – fibrillary  251 – gemistocytic neoplastic  251 – neoplasms  255 astrocytoma  217, 219, 261, 289, 409, 621 – anaplastic  251, 253, 254, 1368 – diffuse  251 – giant cell  256 – low-grade  252, 296 – neurofibromatosis type 1  409 – of the brain  248, 1361 – pilocytic  248, 286, 410 – protoplasmic  251 astrogliosis  293 ataxia telangiectasia  225 atheroembolism  841 atherogenesis  1428 atherosclerosis  748, 810, 830, 837, 846 – ankle-brachial index (ABI)  838 – digital subtraction angiography (DSA)  838 – of the lower extremities  839 – WB-MRI  1235 atherosclerotic – disease – molecular imaging  1428 – renal artery stenosis  818 – comprehensive stenosis grading on MRI  818 – cross-sectional vessel diameter  818 – perfusion measurement  819 – sensitivities and specificities  818 atomic nuclei  5 atrial septal defect (ASD)  719 atrioventricular (AV) canal  721

atrophic muscle diseases  1359 atrophy – cerebellar  214 – spinal muscular  1208 atropine–dobutamine stress  1356 attenuation  136 autofluorescence signal  1385 automatic slice tracking  1269 autoregulation  1312 autosomal dominant polycystic kidney disease (ADPKD)  865 avalanche photo diode (APD)  1390 avascular necrosis  1106 – of the hip  1107, 1109 AVF see fistulas/arteriovenous fistulas AVM see arteriovenous malformations/malformation azathioprine  1414 azoospermia  1360 azygos vein  732 B b-matrix  142 b-value  136 – optimal choice  144 – range  138 backprojection  1261 bacterial meningitis  348, 349 – CNS complications  353 – concomitant arteritis  353 – drug fever  353 – empyema  353 – pneumonia  353 – subdural effusions  353 bacteroides  355 balanced steady state free precession  1268 balloon cells  198, 204, 220 balun  1273 band heterotopia  200 banding artifact  118, 1268 Bankart lesion  1139 Barrett’s esophagus  1412 basal – cell nevus syndrome  225 – ganglia  376 basiocciput  452, 464 basisphenoid  450, 461, 464 Batson’s plexus  599 Becker muscular dystrophy  1199 Behçet’s disease  845 benign – ovarian epithelial tumor  991 – Brenner tumor  993

– endometrioid cystadenoma  992 – mucinous cystadenoma  992 – serous cystadenoma  991 – prostatic hyperplasia (BPH)  1019, 1036, 1252 benzodiazepine  1214 Bernoulli equation  754, 757, 829 berry aneurysm  341 betaglucosidase  1091 BI-RADS  706 biceps tendon  1136, 1138, 1152 – bursitis  1154 – lesions  1136 biconcave fibrocartilaginous disk  1142 bicuspid aortic valve  755 bile – ducts  893 – ampullary adenoma  896 – benign diseases  895 – calculi  895 – cystic diseases  896 – cysts  896 – papillary  896 – intraperitoneal  905 biliary hamartomas  865 bilirubin encephalopathy  231 Binswanger’s disease  385 biochip  1419, 1420, 1424 – drug discovery  1424 bioluminescence – dye  1405 – imaging  1385 biomarker  1413, 1415, 1416, 1425 – panel  1420 – serum  1419 – tissue-based  1419 biomolecular sensor  1420 biomolecule  1421 biopsy  1259 – endomyocardial  747 – gun  1266 Biot-Savart law  1342 bitemporal hemianopia  281 black-blood – contrast  118 – preparation  713 – technique  668 bladder – adenocarcinoma  1007 – cancer – squamous cell epithelial  1007 – staging  1010 – distension  999

1481

1482

Subject Index

– – – – – – – – – –

diverticula  1012 exstrophy  1003 flat tumor  1009 function  999 infection  1003 leiomyoma  1007 malignant tumor  1003 normal anatomy  1002 outlet obstruction  1226 transurethral electroresection of the (TUR-B).  1011 – wall thickening  1002 Blalock-Taussig shunt  723, 724 Blalock Hanlon operation  730 Bloch-Sulzberger syndrome  226 blood – breakdown products  816 – flow  161 – intraperitoneal  904 – magnetization  181 – oxygenation  728, 1292 – oxygen level dependent (BOLD) effect  1209, 1210, 1294, 1295, 1298 – contrast  1292, 1293, 1296 – functional MRI (fMRI)  194 – oxygen saturation  1293 – relative cerebral blood flow (rCBF)  1292 – relative cerebral blood volume (rCBV)  1292 – saturation effects  313 – sponge  339 – stasis  125 – velocity  181 – profiles  753, 754 – vessel  1293 – cylinders  1293 blood-bolus tagging  1295 blood-pool agents  93, 107 – macromolecular Gd-based  94 blood–brain barrier  97, 245, 434, 1284 blood–ocular barrier  434 blue tumor  266 blurring artifact  125 BMES see bone marrow edema syndrome Bochdalek’s hernia  903 body coil  85 BOLD effect see blood oxygen level dependent bolus timing  795 bone  1174 – bruise  1169 – contusions  1171 – cyst  1165

destruction  1175 ethmoid  450 infarct  1113, 1173 island  569 marrow  1086, 1090, 1093, 1174 – aplastic anemia  1093 – diffuse infiltration  1089 – edema-like signal alteration  1117, 1123, 1171 – edema syndrome  1110 – focal infiltration  1088 – hematopoietic  1087 – infarction  1113 – neoplasia  1088 – non-Hodgkin’s lymphoma  1090 – transient edema-like abnormalities  1148 – metastases  1174, 1239 – sclerosis  1097 – sphenoid  450, 451, 464 – tumor  1155, 1174, 1175 border zone  323 botulinum clostridium  1205 botulism  1205 Bourneville-Pringle disease  219 bowel loop  1222 BPH see benign prostatic hyperplasia brachytherapy  1284 – MRI-targeted  1284 bradykinin  245 brain  802, 1292, 1350 – abnormalities – chemical shift imaging (CSI)  293 – magnetic resonance spectroscopy (MRS)  293 – abscess  174, 357, 359, 370 – entamoebal  371 – acute viral encephalitis  373 – aneurysms  814 – arteriovenous malformations  814 – cerebritis  357 – congenital/aquired diseases in children  193 – diseases of vascular origin  1351 – disorders of cortical development  197 – edema  245 – embryonal tumors  264 – endothelium  1425 – ependymal tumors  261 – hemorrhagic metastases  341 – herniation  312 – – – – –

– high local magnetic field homogeneities  1342 – hypoxic–ischemic injuries  226 – infarction  174 – ischemia  174 – magnetic resonance spectroscopy (MRS)  183 – meningeal tumors  273 – metabolic diseases  172, 231 – metabolic energy  185 – metabolite signal  1350 – metastasis  247, 269 – microlissencephaly  197 – MR spectra  1350 – necrotic metastasis  174 – neoplasm  182, 194 – nonglial tumors  264 – normal intrauterine development  194 – normal postnatal myelination  195 – parenchyma  222, 245, 333, 352, 357, 1293, 1297 – pediatric tumors  194 – pineal parenchymal tumors  267 – stimulation  1292 – surgery  1284 – MRI-guided  1284 – Thallium-201 SPECT  373 – toxoplasmosis  373 – tumor  243, 1281, 1284 – angiogenesis  245 – astrocytic  248 – blood–brain barrier (BBB)  245 – classification  243 – contrast enhancement  245 – FLAIR images  244 – functional imaging  290 – ¹H MRS  1360, 1361, 1362 – interstitial laser therapy  1281 – perfusion-weighted imaging (PWI)  293 – practical aspects of MR imaging  244 – thermal coagulation  1284 – white matter pathways  177 brain–air–bone interfaces  176 brainstem ischemia  812 branchial cleft cysts  514, 530 BRCA1/2  701 – carriers  705 – mutation – BOADICEA  703 – BRCAPRO  703

Subject Index

– genetic testing  704 – tumors  708 breast – biopsy  1266 – cancer  763, 1240, 1252, 1361 – atypical ductal hyperplasia (ADH)  704 – atypical lobular hyperplasia (ALH)  704 – DCIS  701, 707 – false-positive finding  707 – family history  703 – Gail model  704 – high-risk population  701 – Hodgkin’s disease (HD)  704 – lobular carcinoma in situ (LCIS)  704 – mammography  700 – MRI – interpretation  707 – screening  705 – technique  708 – screening guideline  706 – imaging  1386 – tumors  100 breathing belt  90, 1272 broadband ¹H decoupling  1338 Broca’s area  1310, 1316 Brodie’s abscess  1097 bronchial – arteries  804 – atresia  1220 bronchiectasis  682 bronchogenic – carcinoma  1239, 1252 – cyst  692 Brown-Sequard syndrome  654 Brownian motion  130 Brunn’s cell nests  1007 buccal space  518 Buck’s fascia  1062 bucket handle tear  1119 Budd-Chiari syndrome  882, 883 Buerger’s disease, see also thromboangiitis obliterans  843 Buford complex  1133, 1138 bulb, olfactory  450 bulbus spongiosum  1057 bursitis  1135, 1154 butylscopolamide  1018 C CAA see cerebral amyloid angiopathy CAD see coronary artery disease CADASIL  385 café-au-lait spots  216, 219

CAH see congenital adrenal hyperplasia Caisson’s disease  1113 calcification, CT detection  438 calcinosis  1192 calculus disease  498, 514 calf hypertrophy  1200 callosal – agenesis  205 – interhemispheric cyst  206 – hypogenesis  206 calpain  1199 camptothecin-induced apoptotic Jurkat cell  1422 canal – hypo­glossal  452, 462 – vidian  451, 461 Canavan disease  238 cancer – biology  1419 – biomarker  1419 – cell  1414 Candida – albicans  885, 901 – infection  365 candidiasis – hepatosplenic  885 – visceral  885 capillary – hemangioblastoma  285 – hemangioma  440 – telangiectasia  339, 341 capitulum humeri  1154 capsulolabral tear  1137 carbon – carbon-13  1331 – isotope  1342 carboplatin  1424 carcinoma/cancer – ampullary  897 – colorectal  909 – fibrolamellar  874 – gastric  906 – hepatocellular  873, 875 – intrahepatic  876 – of the renal pelvis  932 – of the urachus  1012 – periampullary  897 – peripheral bile-duct  876 cardiac – anomalies  718 – dynamic  714 – electrophysiology ablation  1285 – imaging  106 – malformations  718 – malignomas  782 – normal anatomy  719

pseudotumor  785 synchronization  120 tamponade  763 tumor  778 – benign  779 – malignant  782 cardiomyopathy  734, 735, 744, 745 – arrhythmogenic right ventricular  734 – dilated  734 – hypertrophic  734 – iron-overload  748 – primary  734 – restrictive  734, 764 – secondary  734 cardiovascular – disease  846 – oncologic  1250 – WB-MRI  1235 – magnetic resonance (CMR)  748 – system  1233 – WB-MRA  1233 Carney complex  779 carnitine  1202 – palmitoyltransferase (CPT) deficiency  1201 carnosine (CAR)  1210 Caroli’s disease  896 carotid – artery disease  810, 817 – body tumor  520 – circulation  810 – endarterectomy  183, 814 – space (CS)  519 carotid–cavernous fistula  424 carpal tunnel syndrome  1207 cartilage  1126 – calcified  1127 – focal blistering  1127 – total thickness  1128 – uncalcified  1127 catheter – angiography  99 – motion  1264 cauda equina  591, 593 – demyelinating inflammatory neuropathy  609 caudate lobe hypertrophy  882, 883 cavernoma  340 cavernosography  1055 cavernous – malformations (CM)  337, 638 – sinus  402, 451, 459, 462, 466, 470, 471 cavum scroti  1042 – – – –

1483

1484

Subject Index

CBV see cerebral blood volume CCAM see congenital cystic adenomatoid malformation CDH see congenital diaphragmatic hernia cell – membrane  158 – migration  1403 – optical labeling  1404 – tracking  1403 cellulitis, oral cavity  497 central – cord syndrome  654 – nervous system (CNS) – angiomatosis  222 – aspergillosis  365 – astrocytic tumors  248 – embryonal tumors  264 – ependymal tumors  261 – focal lesions  246 – fungal infection  364 – glioma  294 – hemangiopericytoma  278 – hypomyelination  234 – infections  348 – lymphoma  294, 295, 373 – meningeal tumors  273 – mucormycosis  365 – neoplasms  243, 599 – neuronal–glial tumors  263 – neuronal mixed  263 – nonglial tumors  264 – parasitic infections  365 – pediatric patients  193 – tuberculoma  363 – tuberculosis  351 – tumors  243 – tumors of uncertain histogenesis  285 – vasculitis  816 central-core disease (CCD)  1197 cephaloceles  210, 454, 455, 485 – frontoethmoidal  211 – occipital  211 cerebellar – atrophy  214 – hypoplasia  214 cerebellum  343 cerebral – abscess  359, 360 – amyloid angiopathy (CAA)  343 – blood flow  296 – index (CBFi)  179 – blood volume (CBV)  179, 296 – ischemia  311 – lipomas  290

– metastases  269 – neoplasms  244 – spinal fluid (CSF) – lac  186 – venous sinus thrombosis  815 cerebritis  311, 349, 354, 357 cerebrospinal fluid (CSF)  173 – arachnoid cysts  289 – chronic leak  212 – CSF-filled cyst  210 – flow artifacts  333 – shunting  213 cerebrovascular disease  182 – MR imaging technique  310 cervical – artery dissections  811 – carcinoma  965, 966, 980 – diagnosis  980 – dynamic contrast-enhanced MRI  983 – parametrial invasion  980 – recurrence  982 – staging  966, 980 – stromal invasion  980 – TNM/FIGO classification  980 – treatment selection and follow-up  982 – tumor location  980 – tumor size  980 cervicocystoptosis  1073 CFM see color flow mapping CHAOS see congenital high-airway obstruction syndrome Charcot-Marie-Tooth type I  608 chelate  122 chelation therapy  748 chemical – frequency selective fat presaturation  1084 – reaction  154 – shift  24, 46, 52, 1343 – artifact  29, 30, 78 – imaging (CSI)  184, 687, 1334, 1336 chemically selective saturation (CHESS)  184, 1246, 1335 Chiari malformations  205, 211 childhood ataxia  234 children – acute stroke  194 – anesthesia with intubation  193 – diagnostic imaging of the brain  193 – FLAIR sequences  193 – neoplasm of the brain  194

– sedation  193 – temporal lobe epilepsy  193 chloroma  1088 Cho see choline (Cho) choice of the b-value  143 cholangiocarcinoma  895, 896, 898 – intrahepatic  876 – peripheral  876 cholangiopathy – AIDS  895 cholangitis – infectious  895 – primary sclerosing  895 cholecystitis  893, 895 – acalculous  893 cholecystolithiasis  893 choledochoceles  896 choledocholithiasis  893, 895 cholesteatomas  289 cholesterol  289 – polyps  894 choline (Cho)  293, 1342 – plasmalogen  1344 – resonance  186 – signal intensity  1351 chondroblastic osteosarcoma  1245 chondrocalcinosis, meniscal  1093 chondroid tumor  1164 chondromalacia  1142, 1147 – patellae  1128 chondrosarcoma  480, 570, 1173 chordomas  281, 428, 570 – chondroid  428 choreoathetosis  231 choriocarcinoma  270, 979 choroid – glioma  262 – plexitis  353 – plexus papilloma  280 choroidal hemangioma  437 chronaxie  158, 159 chronic – inflammatory demyelinating polyradiculoneuropathy (CIDP)  608, 1207 – liver disease  874 – lymphocytic leukemia (CLL)  602 – myeloid leukemia (CML)  602 – pelvic pain  996 – pulmonary embolism  807 – pulmonary hypertension (CTPH)  806 – recurrent multifocal osteomyelitis (CRMO)  1097, 1246, 1248 chylothorax  688

Subject Index

CIDP see chronic inflammatory demyelinating polyradiculo­ neuropathy cingulate gyrus  376 circle of Willis  813 circulatory arrest  231 cisplatin  1424 clivus  450, 452, 462, 467, 470 CLL see chronic lymphocytic leukemia cloaca  1012 cloacal exstrophy  1003 clostridium – baratii  1205 – butyricum  1205 CM see cavernous malformations CML see chronic myeloid leukemia CMR see cardiovascular magnetic resonance coarctation  729 – aortic  829 coaxial cable  1263 Cobb’s syndrome  636 cobblestone lissencephaly  201 cocaine abuser  459 cochlea  453, 474 Cockayne syndrome  234 coil  18 collagen–vascular diseases  325 colloid cyst  527 colon – cancer  872, 909 – polyp  1242 color – anisotropy  177 – flow mapping (CFM)  1389 colorectal cancer/carcinoma  1240, 1252 colpocystodefecogram  1069 communicating hydrocephalus  391 compact – bone  1158 – disk  87 compartment syndrome  1203 comprehensive imaging  807 computed tomography  1258, 1386 – angiography (CTA)  318 – supra-aortic and intracranial vasculature  809 – colonography  1240 – colonoscopy  1252 computer-assisted surgery  1278 congenital – adrenal hyperplasia (CAH)  1047, 1052

– atresia of the pulmonary valve  728 – cystic adenomatoid malformation  1220, 1221 – diaphragmatic hernia (CDH)  1220, 1222 – liver-down  1223 – liver-up  1223 – heart disease (CHD)  713, 718 – high airway obstruction syndrome  1218, 1219 – muscular dystrophy  1197 – myotonia  1197 congenitally corrected transpositions  731 congestive heart failure  746 Conn’s syndrome  944 connectivity  1285 consolidation  678 constrictive pericarditis  763 continuity equation  754 contrast – agents/media  7 – blood-pool and intravascular agents  93 – bowel  105 – different field strength  247 – dynamic susceptibilityweighted (DSC)  97 – fibrin-specific  1430 – for MRA  793 – for soft-tissue lesions  99 – Gd chelate-based  97, 246 – hepatocyte-selective  869 – liver-specific  101 – lymphotropic  104 – magnetic field strength  94 – magnetic resonance angiography  98 – necrosis-specific  107 – neuroimaging  97 – nonallergic adverse reactions  95 – orbital and ocular imaging  434 – safety  95 – unspecific extracellular fluid space agents  92 contrast-enhanced MR angiography (CE MRA)  447, 791 contrast-to-noise ratio  93 control acquisition – ASL  182 conus medullaris  590, 593 cord – compression  656

– concussive injury  656 – infarction  635 – tethering  658 coronary – arteries  107, 767 – artery disease (CAD)  766, 839, 846 – bypass grafts  768 – MR angiography  767 corpora cavernosa  1055, 1056, 1057 corpus  1057 – callosum  193, 194, 197, 300 – agenesis  204 – hypogenesis  204 – spongiosum  1055 correlation analysis  1300, 1302 cortical – development  194 – dysplasia  197 – with balloon cells  198 – reorganization  1315 – tubers  220 – vein thrombosis  330 corticomedullary differentiation (CMD)  914 corticotherapy  1209 cor triatriatum  727 Cowden syndrome  215 coxitis fugax  1113 CPMG condition  139 CPT see carnitine palmitoyltransferase Cr-PCr see creatine–phospho­ creatine cramps  1201 craniopharyngiomas  281, 283, 284, 285, 408 – adamantinomatous  409 – squamous-papillary  409 craniotomy  1282 creatine (CRN)  1210 – creatine–phosphocreatine (Cr-PCr)  293 – deficiency syndromes  186 – resonance  185 crescent sign  1108 Creutzfeldt-Jakob disease  174, 175, 386 cribriform plate  450, 461, 469 cricoid cartilage  504 crista galli  450 CRMO see chronic recurrent multifocal osteomyelitis Crohn’s disease  106, 905, 907, 908, 1003, 1061

1485

1486

Subject Index

cruciate ligaments  1122 cryoablation  1282 cryogen  1281 cryolesion  1281 cryostat  79 cryosurgery  1281, 1282 cryotherapy  1271 cryptococcosis  364 cryptococcus  901 cryptophane-A cage  1395, 1396 cryptorchidism  1039, 1043 CSF see cerebrospinal fluid (CSF) CSI see chemical shift imaging CT see computed tomography CTA see CT angiography CTC see CT colono­graphy cubital tunnel syndrome  1207 current  157 Cushing’s disease/syndrome  944, 1107, 1116 cutaneous neurofibromas  216 cyanacrylate  1393 cyanine  1404 cyanosis  721, 724, 728, 841 cyclops lesion  1124, 1125 cyst – bronchogenic  1220 – mesenteric  903 – of the adrenal glands  950 – of the ejaculatory ducts  1029 – parenchymal  365 – splenic  899 – utricle  1029 cystadenocarcinoma – biliary  865 – mucinous  890 – serous  889 cystadenoma – biliary  865 – mucinous  890 – serous  889 cystectomy  1011, 1012 – palliative  1012 cystic – fibrosis  886, 1407 – hygroma  1218 – nodal metastasis  509 – teratoma  1218 cysticercosis  365 – bunch-of-grapes appearance  367 – subarachnoid-intra­ ventricular  367 cystitis – cystica  1007, 1008 – follicularis  1007 – glandularis  1008

– of the urinary bladder  1003 cysto-prostato-vesiculectomy  1012 cystocele  1073, 1074, 1075 cystogram  1069 cystoptosis  1073 cystoscopy  999, 1011, 1013, 1015 cystourethrography  1066 cytomegalovirus  649 – encephalitis  378 cytoplasm  1345 cytotoxic edema  376 D d-amino acid  1391 d-penicillamine  1191 Dandy-Walker – complex  213, 226 – malformation  205, 213 – variant  213 DAVF see dural arteriovenous fistulas – posterior fossa  337 DCE-MRI see dynamic contrastenhanced magnetic resonance imaging DCS/XLIS gene  200 deep – vein thrombosis (DVT)  842 – white matter – metabolic disorders  233 defecogram  1069 DEFT  1268 degeneration  971 degenerative disc disease  1246 Dejerine-Sottas disease/syndrome  608, 609 delayed-enhancement MRI  771 delayed-onset muscle soreness  1203 demodulator  87 demyelinating disease  640 dendrimer  1424 dendritic cell, SPIO-labeled  1428 Denonvilliers’ fascia  1028 deoxyhemoglobin  224, 313, 327, 330, 331, 332, 333, 341, 904, 1292, 1293 depth information  127 De Quervain’s chronic stenosing tendinitis  1145 dermatomyositis  1184, 1189, 1192, 1209 – childhood-onset  1189 dermoid  441, 443 – cyst  289, 531, 990 – Rokitansky nodules  990 – tumor  291

desmoid tumor  903 desmoplastic infantile – astrocytoma (DIA)  263 – ganglioglioma (DIG  263 deterministic radiation effect  1272 detuning circuitry  86 developmental – anomalies  914 – horseshoe kidney  914 – persistent fetal lobulation  914 – renal fusion  914 – venous anomalies (DVA)  337, 338 Devic’s disease  641, 643 dextran-coated SPIOs  1400 dextran surface coat  1426 dextroscoliosis  219 DIA see desmoplastic infantile astrocytoma diabetes – insipidus  404 – mellitus  851, 1195, 1207, 1355 diabetic soft tissue infections  843 diagonalization of the tensor  134 diamagnetism  22 diastolic heart failure  764 DICOM  87 diffuse – alveolar damage  679 – astrocytoma  251 – cerebral anoxia  326 – idiopathic skeletal hyperostosis  548 – optical imaging (DOI)  1386 – optical tomography (DOT)  1386 diffusion  130 – anisotropy  134, 173, 175, 177, 299 – coefficient  131, 136, 173 – contrast  136 – directions  144 – distance  131 – gradient  136 – optimal directions  144 – optimal number  144 – spectrum imaging  144 – tensor  132 – cuboid visualization  147 – ellipsoid visualization  146 – imaging (DTI)  130, 175, 194, 298, 301, 348 – measurement  142, 143 – visualization  145 – time  131 – trace imaging  141

Subject Index

– tractography  148 – weighted imaging  655 diffusion-weighted – CE-FAST sequence  140 – EPI sequence  138 – fast spin-echo sequence  139 – imaging (DWI)  173, 447, 618, 628, 655, 1209 – acute ischemic stroke  318 – data processing  173 – intracranial infections  348 – muscle tissues  1209 – trace signal intensity  174 – MRI (DWI)  130 – brain ischemia  813 – PSIF sequence  140 – SSFP sequence  140 diffusional kurtosis imaging  144 DIG see desmoplastic infantile ganglioglioma DiGeorge syndrome  724 digital – rectal examination (DRE)  1031 – signal processor  88 – subtraction angiography (DSA)  1258 – extracranial assessment  317 – versatile disk  87 dihydropyrimidine dehydrogenase deficiency  236 dipole–dipole interaction  1339 dipyridamole  716 disease-specific probe  1391 disk – bulging  542 – extrusion  543 – herniation  543, 545 – protrusion  542 – sequestration  543 diskitis  556 disorder – cortical development  197 – neuromuscular junction  1204 disruptive technology  1284 dissection  627 disseminated intravascular coagulation  1204 diverticulitis  905 – colonic  1003 diving ranula  498 DNA microarray  1420 DNET see dysembryoplastic neuroepithelial tumors DOI see diffuse optical imaging DOMS see delayed-onset muscle soreness

doppler – effect  1389 – ultrasonography (DUS) – supra-aortic and intracranial vasculature  809 DOT see diffuse optical tomography double – aortic arch  729 – cortex  200 double-duct sign  886 double-outlet right ventricle (DORV)  721, 722 double-resonance  1341 – coil  87 – technique  1338 double spin-echo (DSE)  1335 Down’s syndrome  721 draining vein  1314 DRE see digital rectal examination drug – discovery  1424 – toxicity  1413 DSA see digital subtraction angiography DSC see dynamic susceptibility contrast imaging DSC-PWI see dynamic-susceptib­ ility contrast perfusion-weighted imaging DSE see double spin-echo DST see dural sinus thrombosis DTI see diffusion tensor imaging – eigenvalues  176 – fiber tractography  175 – parametric maps  176 – peritumoral values  300 – tractography  176 Duchenne muscular dystrophy  1198 ductal carcinoma in situ (DCIS)  701 ducts of Rivinus  496 duplex ultrasonography (DUS)  318, 1389 dura – mater  591 – trauma  610 dural – arteriovenous fistulas (DAVF)  335 – sinus thrombosis (DST)  327 – tail  277 DUS see duplex ultrasonography DVA see developmental venous anomalies DVT see deep vein thrombosis

DWI see diffusion-weighted imaging Dy, see dysprosium dynamic – contrast-enhanced magnetic resonance imaging (DCE-MRI)  297, 399, 408, 447, 966 – susceptibility contrast imaging (DSC)  178 dynamic-susceptibility contrast perfusion-weighted imaging (DSC-PWI)  321 – mismatch  322, 323 – reverse match  322 dynamically adaptive imaging  1280 dynamic susceptibility contrast imaging – parametric maps  179 – perfusion maps  179 dysembryoplastic neuroepithelial tumor (DNET)  248, 264, 265 dyslipidemia  1246 dysplasia  199, 845 – of the cerebral and extracranial vessels  219 – urothelial  1008, 1009 dysplastic – cerebellar gangliocytoma  215 – nodules (DNs)  874, 878, 879, 880 dysprosium (Dy)  92 dysprosium oxide –   1262 dystrophy  1200 – Markesbery-Griggs  1200 – Miyoshi  1200 – muscular  1197 – Becker  1199 – distal  1200 – Duchenne  1199 – Emery-Dreifuss  1200 – facioscapular  1199 – limb-Girdle  1199 – oculopharyngeal  1199 – Nonaka  1200 – Welander  1200 E earplugs  1273 ear protection  1261, 1273 Ebstein’s anomaly  724, 725 ECF see extracellular fluid space ECG see electrocardiography echinococcal disease  885 echinococcosis  367

1487

1488

Subject Index

Echinococcus  367 – alveolaris  885 – granulosus  367, 885 – multilocularis  367 echo – planar imaging (EPI)  55, 175, 178, 1295, 1304, 1312 – readout gradients  176 – technique  1298 – FID  1298 – optimal echo tim  1298 – spectroscopy  184 echocardiography (ECG)  88, 711, 719, 745 – transesophageal  754, 762 – transthoracic  751, 752 ectocyst  367 ectopia lentis  834 ectopic – pancreas  1220 – posterior pituitary  408 eddy current  84, 141 edema – bone marrow  1106 – focal  1002 – hepatic  883 – periarticular  1099 – perifocal  1161, 1171 – periportal  878 – stress-related  1114 EEG  90 EgadMe  1400, 1403, 1404 EHE see epithelioid hemangio­ endothelioma Ehlers-Danlos syndrome  342, 845 eigenvalue  134 eigenvector  134, 145, 177 – visualization  145 Eisenmenger syndrome  721 elbow  1151 electric – conductivity  157 – field  157 electrocardiography – cable  89 – gating  711 – triggering  116, 127, 711 electron – magnetic field  92 – magnetization  23 – shell  22 – spin resonance  79 electrophysiology catheter  1265 ellipsoid  146 elliptical scanning  125

embolism – arterial  841 – pulmonary  842 embolotherapy  809 Emery-Dreifuss muscular dystrophy  1200 empty – delta sign  327 – sella  404 – sulcus  1136 empyema  668 – cranial epidural  355 – subdural  355 EMX2 gene  202 ENB see esthesioneuroblastoma encephalitis  373 – cytomegalovirus  378 – hemorrhagic  376 – HIV  376 – HSV-1  374 – HSV-2  376 – necrotizing  376 encephaloceles, occipital  211 encephaloclastic porencephaly  227 encephalomalacia  228, 376 encephalopathy  376 – tuberculous  363 enchondroma  1164, 1173 enchondromatosis  1165 encysted organisms  373 endocardial cushion defects  721 endocarditis  760 endocyst  367 endoleak  837 endometrial – cancer/carcinoma  964, 970, 973, 978 – cervical stroma invasion  965, 976 – diagnosis  973 – dynamic contrast-enhanced MRI  973 – myometrial invasion  965, 974, 978 – parametrial invasion  965 – recurrence  978 – staging  965, 974 – subendometrial enhancement  974 – subendometrial zone  965 – TNM/FIGO classification  974 – treatment selection and follow-up  976 – hyperplasia  973 – adenomatous  973 – atypical  973

– cystic  973 – polyp  973 – tamoxifen  973 endometrioma  903, 990 endometriosis  990, 996 – peritoneum  903 – urinary bladder  1008 endomyocardial fibrosis  749 endorectal – coil imaging  909 – surface coil (ERC)  999, 1018 Endorem  1428 endoscope  1265 energy dissipation  161 engineered human transferrin receptor (ETR)  1400, 1401 enostosis  569 entamoeba histolytica  370, 884 entrapment nerve  1207 enzymatic deficiency syndrome  172 eosinophilia–myalgia syndrome  1196 eosinophilic – fasciitis  1248 – granuloma  569, 1246 ependymal cell  619, 620 ependymoblastoma  264 ependymoma  261, 598, 619, 620, 640 – anaplastic  262 – in the spine  597 – myxopapillary  599, 600 EPI see echo planar imaging epicondylitis  1152 epidermal nevus syndrome  226 epidermoid  290, 613 – cyst  211, 289, 531, 606, 611 epididymitis  1042, 1045, 1046 epidural – abscess  559 – empyema  355 – hematoma  334, 585 – lipomatosis  553 epiglottis  504 epilepsy  231 epispadia  1059 epithelioid hemangioendothelioma (EHE)  877 epoxy resin  83 Epstein-Barr virus  649 equilibration – length  161 – time  161 equilibrium-phase MRA  127 Erlenmeyer-flask deformity  1091 Ernst angle  48

Subject Index

erysipela  1102 erythrocyte sedimentation rate (ESR)  349 Escherichia coli  348, 357 esthesioneuroblastoma (ENB)  373, 487, 490 ethanol  1207, 1338 – splitting pattern  1338 ethmoid sinuses  485 ETR see engineered human transferrin receptor Eustachian tube  460, 475, 476, 489, 490 Ewing sarcoma  480, 571, 1159, 1366 EXIT procedure see ex utero intrapartum treatment procedure expanded gallbladder fossa sign  878 exponential signal decay  1297 extended inner-conductor antenna  1264 external – genitalia  1217 – gradient mounting  1262 – magnetic field  6 extracapsular extension (ECE)  1025, 1031 extracellular fluid space – Gd chelates  97 extrahepatic malignoma  103 extralobar sequestration  1219 exudate  688 ex utero intrapartum treatment procedure (EXIT)  1218 eye muscles  440, 441 F facial – dysmorphism  239 – nerve  454, 478 – nevus flammeus  224 – paralysis  520 Faraday – cage  76, 85, 88, 1260 – effect  1265 – law  156 fascia penis profunda  1062 fasciitis – eosinophilic  1196 – plantar  1150 fast low-angle shot  1267 fat – saturation  56, 668 – suppression  1272 fatigue fracture  1169 fatty liver  881, 882

Fe see iron (Fe) femoral head  1110 – subchondral insufficiency fractures  1110 ferromagnetic joint prostheses  1081 ferucarbotran  101 ferumoxides  95, 101, 104 ferumoxsil  106 fetus/fetal  1217, 1218, 1227 – abdominal disease  1226 – ascity  1218, 1224 – body  1213, 1215 – normal finding  1215 – external genitalia  1217 – gastrointestinal tract anomaly  1227 – genitourinary abnormality  1227 – hydrop  1218 – hydrothorax  1224 – integumentary edema  1224 – internal genitalia  1217 – lung maturity  1222 – MRI technique  1214 – pulmonary sequestration  1219 – upper-airway obstruction  1218 – upper-quadrant mass  1226 – with giant neck mass  1218 fiber – optic  1264 – tracking  148 – algorithm  1209 fiberglass  1273 fibrillary astrocytoma  251 fibrinoid leukodystrophy  238 fibroadenoma  708, 1240, 1282 – of the breast  1282 fibrolamellar carcinoma  874 fibrolipoma  605 fibroma  779, 782 fibromatosis  1246 – aggressive  903 – colli  529 – plantar  1150 fibromuscular dysplasia  823 fibrosing pericarditis  763 fibrosis – cystic  886 – of the penile corpora  1060 – periaortic retroperitoneal  832 fibrotic tumor  990 – cystadenofibroma  991 – fibroma  990 – fibrothecoma  990 – thecoma  990 fibula, non-Hodgkin’s lymphoma  1091

Fick principle  173, 759 FID see free induction decay field – homogeneity  84 – inhomogeneity  21 – strength  153, 1294 filtered back projection  140 filter plate  85 filum terminale  591 – fibrofatty infiltration  605, 606 finger tapping  1296, 1297 FISP  50 – 3D sequence  1325 – signal intensity  50 fistula  441 – urethral  1047 FLAIR see fluid attenuated inversion recovery imaging FLASH image  45, 47, 48, 1294, 1295, 1296, 1297, 1304 – optimal echo time  1296 – signal intensity  48 flat-panel volumetric CT (fpVCT)  1387 flexible coil  86 flexor retinaculum  1144 – palmar bowing  1144 flipped-meniscus sign  1119 flow – compensation  116, 120 – void  623, 634, 635, 752 fluid-attenuated inversion recovery imaging (FLAIR)  42, 172 – acute ischemic stroke  312 – brain tumor  244 – meningitis  349 fluorescence – imaging  1385 – light increase  1385 – molecular tomography  1385 – resonance energy transfer (FRET)  1393 – signal quenching  1393 fluorine-19  1331 fluorochrome  1385, 1386, 1404 fluorodeoxyglucose – ¹⁸F-labeled  1420 fluorophore  1393 – non-photobleaching  1423 fluoroscopy  1258 5-fluorouracil (5-FU)  1348 – metabolism  1369 – pharmacokinetic  1369 fMRI see functional MRI FNH see focal nodular hyperplasia focal – cord atrophy  612

1489

1490

Subject Index

– cortical dysplasia without balloon cells  204 – nodular hyperplasia (FNH)  102, 869 – sparing  881 focused ultrasound (FUS)  1277, 1282, 1283, 1286 – fibroid ablation  1283 – MR-guided  1282 Foix-Alajouanine syndrome  613, 631 FOLD (flow level dependent contrast)  1294, 1295, 1305 foot  1147, 1149 – complete rupture  1149 – deformity  1149 – stress fractures  1147 – tendon pathology  1149 foramen/foramina – lacerum  451, 452, 470 – Luschka  286 – Magendie  286 – magnum  452 – Monroi  221 – of Monroe  256 – ovale  451, 462, 470 – rotundum  451, 462 foregut cysts  865 foreign body  442 fossa, olfactory  450 Fourier – 2D method  33 – 3D method  33 – interpolation  54 – space  1303 – transformation  14, 29 fovea ethmoidalis  450, 469, 485 fractional anisotropy (FA)  135, 145, 300, 1279 – map  145 fragment-in-notch  1119 frataxin  750 free – diffusion  133 – induction decay (FID)  6, 1333 Freiburg’s disease  1147 frequency – encoding  28 – spectrum  18 FRET see fluorescence resonance energy transfer Friedreich’s ataxia  750 frontal sinus  463 fructokinase deficiency  1354 Fukuyama muscular dystrophy  201

full width at half maximum (FWHM)  1332 functional MRI (fMRI)  655, 1292, 1302, 1308 – data acquisition  1298, 1299 – image artifact  1302 fundamental law of electrostimulation  158 fungal infections  363 fungi – pathogeni  363 – saprophytic  363 FUS see focused ultrasound FWHM see full width at half maximum G g-cell tumors  888 gadobutrol  92, 97 gadodiamide  92, 95, 247 gadofosveset  93, 95, 99 gadolinium (Gd)  92, 315, 341 – chelates  97, 1426 – compounds  93 – diethylenetriamine pentaacetic acid (Gd-DTPA)  246, 590, 1428 – Gd-BMA  122 – Gd-BOPTA  122 – Gd-BT-DO3A  122 – Gd-DOTA  122 – Gd-DTPA  122 – Gd-HP-DO3A  122 gadomer  94 – gadomer-17  1424 gadopentate  92 – dimeglumine  95, 97, 102 gadoterate meglumine  92 gadoteriol  92 gadoversetamide  95 gadoxetic acid  102 galactosemia  236 β-galactosidase  234, 1400, 1403, 1405 gallbladder – acute inflammation  893 – adenomyomatosis  894 – carcinoma  894 – fossa  878 – non-neoplastic disease  893 – polyps  894 – wall thickening  894 gallstone disease  893 gamma-gandy bodies  902 ganglia  1143 gangliocytoma  263

ganglioglioma  263 – anaplastic  263 ganglion cyst  546 ganglioneuroblastoma  266 gangliosidoses  239 Garré ’s chronic sclerosing osteomyelitis  1097 gas – bubble  1393 – exchange  680 Gasserian ganglion  373, 451 gastrinoma  888 gastrointestinal – imaging  105 – tract  906 Gaucher’s disease  1091, 1107 Gaussian distribution of SI changes  1299 Gd see gadolinium Gehrig’s disease  389 gene therapy  1383 genetic hemochromatosis (GH)  880 genic edema  312 genitourinary tract trauma  1003 geometric – decoupling  86 – warping artifacts  176 geometrical distortion  1324 germ cell tumor  416, 570, 1047 – α-fetoprotein (AFP)  417 – β-human chorionic gonadotropin (HCG)  417 – choriocarcinoma  416 – embryonal carcinoma  416 – germinoma  416 – mixed germ cell tumor  416 – non-germinomatous germ cell tumor  416 – teratoma  416 – yolk sac tumor  416 germinoma  268, 614 GFAP see glial fibrillary acidic protein ghost image  116 giant – aneurysm  341, 342, 814 – cell – arteritis  812, 836, 843 – astrocytoma  258 – glioblastoma  253 – tumor  221, 566 – tendon sheath  1162 Gibbs artifact  618 glans penis  1064 glenoid labrum  1131 – lesion  1140

Subject Index

glial fibrillary acidic protein (GFAP)  255 glioblastoma  253, 263, 294, 296, 298 – brain  1361 – multiforme – butterfly appearance  254 – multiforme (GBM)  186, 187, 253, 256, 257, 299 glioma  217, 245, 247, 253, 269, 290, 293, 295 – brain  1361 – edema  277 – high-grade  277, 293 – low-grade  297 – mixed type  261 – optic pathway  217 gliomatosis – cerebri  261, 262 – diffuse leptomeningeal  602 gliosarcoma  253 gliosis  357, 359 glomus – jugulare  521 – tympanicum  521 – vagale  520 glottic carcinoma  508 glucagonoma  889 glucocerebroside hydrolase  1091 glucose  1397 – ¹³C-labeled  1332, 1348 glutamate (Glx)  186, 1342, 1344 glutamine  186, 1342, 1344 glutaric aciduria  238, 239 glycerophosphorylcholine (GPC)  1344, 1345 glycerophosphorylethanolamine (GPE)  1345 glycine  1344 glycogen – metabolism  1201 – in muscles  1360 – storage disease  1202 glycogenoses  1202 glycolysis  1345 – anaerobic  1351 glycosphingolipid  749 goiter  527 Gorlin – formula  752 – syndrome  225, 782 gracilis muscle  1186 gradient – echo (GRE)  44, 178 – technique  44 – noise  1261 – refocusing  47, 50

– strength  83 gradient-induced force  1267 graft-versus-host disease (GVHD)  1191 granular cell tumor  281, 417 – choristoma  417 – myoblastoma  417 – neuroma  417 – pituicytoma  417 granulocytic sarcoma (chloroma)  481 granuloma, eosinophilic  1162 granulomatous disease  652 graphical interface  1269 GRAPPA  55 gray matter  232 – metabolic disorders  236 GRE see gradient-echo green fluorescent protein  1403 Guillain-Barré syndrome  608 Gusher’s syndrome  474 Guyon’s canal  1144 gyromagnetic ratio  6, 1339, 1342 H HAE see hepatic alveolar echinococcosis Haemophilus influenzae  348, 355, 357 Hahn spin echo  1336 half-Fourier technique  53 Hall – effect  89 – probe  1264 Hallervorden-Spatz disease  237 hamartoma  221, 292, 293, 683 – biliary  865 – splenic  899, 901 harmonic frequency  1390 HASTE  139 HCA see hepatocellular adenomas head-mounted display  1261 headache  816 head and neck – cancer  508 – epidermoid cysts  531 – MRI imaging  483 – mucosal diseases  484 – non-mucosal diseases  509 – tumors  817 – vascular lesions  529 head motion  1313 heart  1357 – MR spectroscopy  1356, 1357 – valves  751 – flow-sensitive imaging  753

helium – helium-3  671 – liquefier  80 hemangioblastoma  222, 273, 285, 622 – in the spine  597 – pilocytic  286 hemangioendothelioma, epithelioid  877 hemangioma  513, 529, 530, 563, 843 – capillary  441, 442, 443 – cavernous  425, 442 – cotton-wool-like paddling in  103 – hepatic  866 – splenic  899, 900 hemangiomatosis of the bones and soft tissue  1162 hemangiopericytoma  273, 278 hematocele  1043 hematoma  330, 331, 344, 765, 811, 1060 – intracerebral  331 – intramural  834 – intramuscular  1184 – of the scrotum  1045 – parenchymal  327, 1046 – penis  1060 – periadventitial  836 – periaortic  831 – pericardial  765 – testicular  1046, 1053 – type B  834 hematopoiesis  1087 hematopoietic – cell  1085 – system  1086 hematuria  1014 hemiatrophy  226 hemichromes  327 hemihypertrophy  198 hemimegalencephaly  198, 226 hemitongue  497 hemochromatosis  422, 739, 748, 881, 1093 – congenital  1217, 1228 – genetic  880 – idiopathic  880 hemodialysis shunts (HS)  845 hemoglobin  327, 332, 1094, 1292 hemoglobinopathy  899 hemoptysis  682 hemorrhage  244, 310, 339 – anatomical location  332 – epidural  334 – germinal matrix zone  229

1491

1492

Subject Index

– hypertension  343 – hypertensive  343 – intracerebral  327, 332, 334 – intracranial  311 – intramural  325 – lobar  343 – non-traumatic spontaneous  342 – periventricular infarctions  229 – prostatic  1029, 1036 – subarachnoid  332, 613 – subdural  333 – tumoral  341 hemorrhagic – effusion  836 – transformation (HT)  344 hemosiderin  244, 332, 334, 339, 341, 1105 hemosiderosis  1093 hemothorax  688 Henderson-Hasselbalch equation  1345 hepatectomy  864 hepatic – alveolar echinococcosis (HAE)  885 – cysts  864 – encephalopathy  1352 – metastasis see liver metastasis – tuberculosis  885 hepatitis  873, 1354 – autoimmune  878 – viral  878 hepatobiliary imaging  101 hepatoblastoma  877 hepatocellular – adenomas (HCA)  867, 869 – carcinoma (HCC)  102, 873, 879, 882 – diffuse  874, 876 – hypovascular  874 – isovascular  874 – nodules  104 hepatocyte  102 hepatomegaly  239 hepatosplenomegaly  1091 heptahelical transmembrane receptor  1391 HER-2 overexpression  1423 hernia  1216 – bilateral congenital diaphragmatic  1216 – congenital diaphragmatic  1220 – esophageal hiatal  903 herpes – encephalitis  373 – viruses  373, 649 heterotaxia syndrome  731

heterotopia  197, 198, 207, 221 – focal subcortical  199 – subependymal  198 HF excitation  1295 HIFU  1271 high-angular resolution diffusion imaging (HARDI)  144 high-intensity focused ultrasound  1271 Hill-Sachs lesion  1137, 1139 hindbrain  194 Hippel-Lindau disease  921 HIV see human immunodeficiency virus 3-HMG-CoA lyase deficiency  236 Hodgkin’s disease/lymphoma  602, 690, 874, 1090 – of the spleen  902 – skeletal involvement  1090 Hoffa’s fat pad  1126 holoprosencephaly  208 – alobar  208 – lobar  209 – semilobar  209 homocysteinuria  235 Horner syndrome  521, 628, 684 Horton disease  812 Hǿyeraal-Hreidarsson syndrome  214 hPAP see human placental alkaline phosphatase HT see hemorrhagic transformation human immunodeficiency virus (HIV)  650 – encephalitis  377 – encephalopathy  376, 378 – leukoencephalopathy  378 – nasopharyngeal carcinoma (NPC)  491 – transactivator transcription protein  1407 human placental alkaline phosphatase  1405 Huntington’s disease  237, 387 HVS see hyperintense vessel sign hydatid disease  367, 368 hydatidiform mole  979 hydranencephaly  227 hydroceles  1049 hydrocephalus  207, 212, 213, 261, 348, 389 – bacterial meningitis  353 – TB meningitis  353 hydrogen  183 hydronephrosis  938, 1227 hydrophone probe  1283 hydrops fetalis  1221

21-hydroxylase deficiency  1052 hyperacute stroke  322 hyperammonemia  186 hyperbilirubinemia  231 hypercholesterolemia  838 hypereosinophilia  749 hyperglycemia  1250 hyperinflation  679 hyperinsulinemia  1246, 1359 hyperintense vessel sign (HVS)  313 hyperkalemia  1204 hyperlipidemia  1250 hypernephroma  270, 688 hyperplasia – adrenogenital syndrome (AGS)  944 – Conn’s syndrome  944 – Cushing’s disease/syndrome  944 – focal nodular  867, 871 – of the adrenal gland cortex  944 hyperpolarization  671, 1395, 1407 hypertension  1246, 1250 hypertensive heart disease  1356 hypertrophic – cardiomyopathy  736 – obstructive cardiomyopathy (HOCM)  737 hypertrophy  517 – capsular  1133 – infundibular  727 – of the myocardium  745 – ventricular  729 hypointense thrombus  327 hypomelanosis Ito  225 hypoperfusion  321 hypopharyngeal carcinoma  502 hypopharynx  500, 503 – squamous cell carcinoma  500, 502 hypophysis  451 hypoplasia  1216 – cerebellar  214 – X-linked form  214 – of the left lobe  864 – pontocerebellar  214 – pulmonary  1221, 1222 hypoplastic – aplastic sinus  816 – left heart syndrome  721, 722, 723 hypospadia  1059 hypotension  231 hypothalamic hamartoma  417 – tuber cinereum  417 hypotonia  1196

Subject Index

hypovolemia  1204 hypoxia  231 hypoxic–ischemic injury  186, 188 – premature infant  228 – term infant  231 hypoxic vasoconstriction  680 hysterectomy  1070, 1071, 1075 hysteresis curve  78 hysteroptosis  1073, 1074 I IAAA see inflammatory abdominal aortic aneurysm IBM see inclusion body myositis ibuprofen  1196 idebenone  750 IDEM see intradural extramedullary (IDEM) idiopathic – inflammatory myopathy  1192 – pulmonary hypertension (IPAH)  806 IIM see idiopathic inflammatory myopathy ILT see interstitial laser therapy image – contrast  36 – determinants and optimization  36 – database  88 – distortion  26 – compensation  1259 – reconstruction  30 – registration  1303 imaging  711 – cardiac  711 – cine imaging  715 – dark band artifact  716 – delayed enhancement imaging  716 – myocardial perfusion imaging  715 – plane  1325 – z-component  1325 – sequence  36, 37 – technique  537 immature teratomas  990 IMP see inosine monophosphate impingement – coracoidal  1133 – syndrome  1133, 1134 implant  52 incisura glenoidalis  1137, 1138 inclusion body myositis  1193 incontinentia pigmenti  226 – achromian  225 indicator dilution theory  179

induction decay  18 induratio penis plastica  1061 infarct, splenic  902 infarcted core  321, 322 infection  1174 – abscess-like  1102 – bone  1174 – fungal  885 – joint  1174 – mycobacterial  885 – soft-tissue  1174 – superficial versus deep phlegmonous  1102 inferior vena cava filter  1262 inflammatory – abdominal aortic aneurysm (IAAA)  832 – myopathy syndrome  1188 – pericarditis  763 – pseudotumor  442 inflatable cuff  125 infrahyoid neck  525 inner volume imaging  1268 inorganic phosphate  1345 inosine-5’-monophosphates (IMP)  1345 inosine monophosphate  1202 instrument tracking  1258 insufficiency fracture  1169 insulinomas  888 insulin resistance  1246 integrated parallel imaging technology  1232 interhemispheric cysts  206 internal carotid artery (ICA)  313 see internal carotid artery (ICA) – atherosclerotic plaque  323 – stenosis  325 internal carotid artery stenosis  814 interpolation algorithm  1303 interstitial laser therapy (ILT)  1281 – MRI-guided  1281 intima  837 intra-articular loose body  1154 intra-voxel dephasing  116, 119 intraaxial tumors and meningiomas  245 intracerebral – hemorrhage  327 – lipoma  291 intracranial – aneurysm  814 – hemorrhage  310 – infection  348 – lipoma  207 – neoplasms  244 – pressure (ICP)  355

– tumors, see also brain tumors  243 intraductal papillary mucinous tumor (IPMT)  889, 890 intradural extramedullary (IDEM) – compartment  607 – vascular malformations  611 – infection  607 – neoplasms  595 – space  591 – spinal compartment  590 – tumors  593 intralobar sequestration  1219 intramural hemorrhage  325 intraoperative cortical mapping  1314 intrauterine growth retardation  1215 intravascular contrast agent gadofosveset trisodium  122 invasive mole  979 inversion recovery sequence  20, 39, 41, 119 – signal intensity  40 ionizing radiation  153, 1272 iPAT see integrated parallel imaging technology IPMT see intraductal papillary mucinous tumor iron (Fe)  92 – overload  880 – transfusional  881 – oxide (CLIO) – cross-linked  1426 – nanoparticle  1383, 1426 – yoke  78, 79 ischemia  312, 748 – hemodynamic  324 ischemic – brain insult  810 – cardiomyopathy  735, 766 – heart disease  766 – parenchyma  312 – penumbra  322 – stroke  310, 311 – diffusion-weighted imaging  318 – hemorrhagic transformation  343 Iselin’s disease  1147 islet-cell tumor – ACTHoma  889 – gastrinomas  888 – glucagonoma  889 – insulinomas  888 – VIPoma  889 isomerism  721, 732

1493

1494

Subject Index

isotope  1390, 1392 – low-Z  1392 isotropic diffusion  133 – weighting  136, 141 Ivemark syndrome  731, 897 J Jacobsen’s nerve  521 Jansky-Bielschowsky disease  236 J coupling  1336 JNA see juvenile nasal angiofibroma Jo-1 syndrome  1190 joint – assessment of internal derangement  1174 – capsule  1131 – effusion  1105, 1110 – glenohumeral  1131 – internal derangement  1117 – normal structures  1117 – radiocarpal  1142 – radioulnar  1142 jugular foramen  462 jumper’s knee  1126 juvenile nasal angiofibroma (JNA)  487 juxta-articular cyst  546 K k-space  810, 817, 849 – filling  712 – segmented  712 – single-shot  713 – trajectories  799 – Cartesian readout  799 – centric  799 – elliptic-centric  799 – radial imaging  799 – spiral imaging  799 kaposiform hemangio­ endothelioma  480 Kearns-Sayre syndrome  240 keratin  289 kernicterus  231 kidney  1360 – transplantation  1360 Kienboeck’s disease  1113 Klatskin tumor  897, 898 Klebsiella  348 Klippel-Trenaunay-Weber syndrome  634 knee  1117, 1126 – extensor apparatus  1126 – injury  1117 Köhler’s disease  1147 Kommerell’s diverticulum  730 Krabbe disease  234

Kupfer cell  101, 1395, 1405 – density  102 – function  102 kurtosis  144 L LabChipTM technology  1420 labeling standard  1273 Lac see lactate lacrimal gland  440 lactate (Lac)  186, 1344 – acidosis of the tissue  1345 lacunar infarct  319, 321 Lambert-Eaton myasthenic syndrome  1205 laminin 2  1198 Langerhans cell histiocytosis  418 Langmuir isotherm  1421 language  1308 – system  1310 large cell carcinoma  684 Larmor frequency  6, 10, 87 laryngocele  505, 508 larynx  502 – cartilaginous lesions  505 – chondroma  505 – chondrosarcoma  505 – nerve loops  505 – trauma  505, 508 – vocal cord paralysis  504 laser-induced thermal therapy  1271 latency  1260 – period  648 lateral – collateral ligament complex  1153 – retropharyngeal lymph node (LRPLN)  494 – retropharyngeal node (LRPN)  493, 494, 523 – ulnar collateral ligament  1153 LAVA see liver acquisition with volume acceleration LCLC see lateral collateral ligament complex lead-time bias  1253 lecithin  1222 Ledderhose’s disease  1150 Leigh syndrome  240 leiomyoma  970, 971, 972, 979 – calcification  971 – degeneration  979 – calcific  971 – fatty  971 – hemorrhagic  971 – hyaline  971

– myxoid  971 – necrotic  971 – sarcomatous  971 – embolization  971 – hysterectomy  972 – intramural  971 – MR-guided focused ultrasound  971, 972 – myomectomy  971, 972 – submucosal  971 – subserosal  971 – urinary bladder  1007 – uterine arterial embolization (UAE)  972 leiomyosarcoma  979 LEMS see Lambert-Eaton myasthenic syndrome Lenz’s law  23 leptomeningeal – carcinomatosis  333 – disease  590 leptomeninges  591 leptomeningitis  363 leukemia  573, 602, 1087, 1365 – acute lymphoblastic  1088 – chronic lymphocytic  900 – single-voxel technique  1365 – spectroscopic imaging  1365 leukodystrophy – globoid cell  234 – metachromatic  233 – with trichothiodystrophy  236 leukoencephalopathy – megalencephalic with cysts  235 – with macrocephaly and mild clinical course  235 leukoplakia  1007 levator ani muscles  1069 LGMD see limb-girdle muscular dystrophy Lhermitte-Duclos syndrome  215, 216 Lhermitte’s sign  648 Li-Fraumeni syndrome  280 ligamentum flavum hypertrophy  546 light-emitting diodes (LEDs)  1278 Liliequist’s membrane  403 – ventriculostomy  403 limb-girdle – dystrophy  1180, 1185 – muscular dystrophy  1199, 1202 limits for workers  1274 line-scan diffusion imaging  140 linear – correlation coefficient  1300 – diffusion  134

Subject Index

– orthogonal field gradient  1323 – prediction and singular value decomposition (LPSVD) method  1333 line of Blumensaat  1121, 1124 linitis plastica carcinoma  907 lipid  186, 1344 – metabolism  1201 – storage disease  1091, 1202 lipoma  779, 781 – intracranial  207 – intradural  605, 606 – peritoneum  903 lipomatous hypertrophy of the interatrial septum (LHIS)  781 lipomyelomeningocele  605 lipoproteins  838 liposarcoma of the kidney  936 Lisch nodules  216 LIS gene  200 lissencephaly  198 – cobblestone malformations  201 – spectrum  200 Lister’s tubercle  1145 lithium  1353 lithotomy  1020 LITT  1271 liver – acquisition with volume acceleration (LAVA)  102 – amebic abscesses  884 – angiomyolipomas  866 – angiosarcoma  876 – cell carcinoma  874 – cholangiocarcinoma  876 – chronic inflammation  878 – cirrhosis  104, 873, 877, 878, 881, 1354 – diffuse parenchymal diseases  877 – fatty  881 – focal lesions  102 – hemangiomas  866 – herniation  1224 – Hodgkin’s and non-Hodgkin’s lymphoma  874 – imaging technique  864 – infectious parenchymal diseases  883 – metastases  102, 104, 869, 881, 887, 889 – ¹H MR spectroscopy  1355 – multifocal microabscesses or granulomas  885 – parenchyma  101, 102, 867 – polyp  1242 – tumor  1354

lobar emphysema  1220 Löffler’s endocarditis  749 Lorentz force  81 loss of heterozygosity (LOH) chromosome –   253 low-pass filtering  88 low–flip angle excitation  44, 46 Lowe syndrome  234 LUCL see lateral ulnar collateral ligament lumbosacral spine imaging  590 lumirem  106 lung  804 – cancer  1240 – fibrosis  679 – neoplasms  270 – perfusion  805 lunotriquetral (LT) ligament  1141 lusory artery  729 lymphadenopathy  491, 498, 508, 1023 – carcinoma of the penis  1063 lymphangioma  440 lymphangiosis carcinomatosa  687 lymphatic malformations  530 lymph node  673 – classification  510 – cystic change  508 – imaging  104 – intercostal  673 – lymphatic drainage  105 – metastasis  104, 954, 1013, 1239, 1391 – necrotic  508 – para-aortic  673 – paraesophageal  673 – paratracheal  673 – peripancreatic  891 – subcarinal  673 – tracheobronchial  673 lymphocytic hypophysitis  420 – adenohypophysitis  420 – infundibuloneurohypo­ physitis  404, 420 lymphoepithelial lesions  514 lymphoid hypertrophy  495 lymphoma  268, 271, 417, 440, 518, 573, 602, 763, 930 – bladder  1007 – cardiac  784 – hepatic  875 – mucosal  487 – Rosai-Dorfman disease  932, 933 – sinus histiocytosis  932 – skeletal  1090

M machine classifier  1334 macroadenoma  281 – of the hypophysis  282 macrocephaly  207 macromolecule  116 macrophage  98, 104, 332, 838 macroscopic magnetization  12 Maffucci syndrome  473, 1165 magic-angle phenomenon  1085 magnetic – labeling  1405 – nanoparticle  1421 magnetic field  26, 156, 1332 – compensation  28 – exposure limits  160 – external  1332 – gradient  6 – H  23 – static  1332 magnetic flux density  153 – B  23 magnetic induction  1264 magnetic resonance – cholangiopancreaticography (MRCP)  106 – compatible applicator  1271 – experiment  18 – fluoroscopy  123 – guidance  1272 – iron oxide–enhanced lymphography  509 – lymphangiography  105 – spectra  1333, 1334 – ¹H  1342 – urography  106 – user interface  87 – venography  327 magnetic resonance angiography (MRA)  98 – 3D TOF  313 – acute ischemic stroke  313 – contrast-enhanced  315 – intracranial infections  348 – M1 branch occlusion  316 – moving-during-scan technique  847 – of the whole-body  846, 849 – phase-contrast  315 – rolling table platforms  847 magnetic resonance imaging (MRI)  1277, 1278, 1280, 1284, 1389 – basic principles  5 – breast  700 – cardiac  106 – coronary arteries  107

1495

1496

Subject Index

– diffusion-weighted  298 – diffusion tensor imaging  298 – dynamic contrastenhanced  297 – female breast  100 – interventional  1277, 1278 – intraoperative  1277 – multi-resolution wavelet encoded  1280 – scrotum  1052 – signal generator  1394 – SVD-encoded  1280 – thermometry  1284 – thermotherapeutic tool  1280 magnetic resonance mammography  100 – Gd chelates  100 magnetic resonance spectro­ scopy (MRS), see also spectro­ scopy  1331, 1358, 1383, 1389 – ¹H  1332 – absolute quantification of meta­ bolite concentration  1342 – basic principles  5 – brain tumor  293, 1399 – carbon-13  1332 – clinical examination  1352, 1354 – fluorine-19  1332, 1348 – in muscle diseases  1358 – intracranial infections  348 – multinuclear investigation  1364 – NAA signal  1344 – of the calf muscle  1359 – of the heart  1356 – of the human brain  1350 – of the prostate  1360 – phosphorus-31  1332, 1345 – tumors  1360 – physical limitation  1339 – prostate cancer  1399 magnetic screening  1272 magnetic susceptibility  23, 78 – artifact  399 magnetism – diamagnetic  153 – ferromagnetic  154 – paramagnetic  154 magnetite  92 magnetization  56, 153 – contrast (MTC)  116, 1085 – longitudinal  13 – suppression (MTS)  247 – transverse  13 magnetization-prepared rapidacquisition gradient-echo (MP RAGE)  864

magneto-hydro-dynamic (MHD) effect  711 magneto-mechanical interaction  154 magneto-optical disks  87 magnetophosphenes  159 magnitude averaging  140 main magnetic field  1323 malacoplakia  938, 1003 malformation  1220 – anorectal  1003 – capillary  842 – congenital cystic adenomatoid  1220 – genitourinary – epispadia–exstrophy complex  1060 – of the spine  219 – vascular  842 malignant melanoma  271, 272, 1051, 1365, 1366 – long-term therapy monitoring  1366 – ³¹P MRS  1364 mammography  700 – false-positive finding  707 mandibular nerve  451 mangafodipir  102 manganese (Mn)  92 mannitol  106 MAPCAs see mid-aortic pulmonary arterial collaterals maple syrup urine disease  188, 189, 235 Marfan syndrome (MFS)  342, 830, 831, 834, 845, 851 MARIBS study  705, 706 Marinesco-Sjögren syndrome  214 marker protein  1398 Markesbery-Griggs dystrophy  1200 mastectomy  703, 707 masticator space (MS)  517 – odontogenic infection  518 – rhabdomyosarcoma  518 matrix-coil systems  849 matrix representation of the tensor  132 mature cystic teratoma  990 – Rokitansky nodules  990 maxillary – nerve  451 – sinuses  485 maximum intensity projection  88 Maxwell – coil pair  83 – terms  157

MCA see middle cerebral artery McArdle disease  1201, 1358 McCune-Albright syndrome  1191 MCL see medial collateral ligament MDCT  696 mean diffusivity  133, 145 Meckel’s cave  451, 459, 462, 471 meconium  1217, 1227 – peritonitis  1227 – pseudocyst  1227 MECP2 gene  237 medial – collateral ligament (MCL)  1125 – tear  1126 – retropharyngeal nodes (MRPN)  523 – temporal sclerosis  395 median nerve  1144 mediastinitis  692 mediastinum  1042 medulla oblongata  212 medulloblastoma  266, 267, 268, 599, 601 megacisterna magna  213 melanin  244 – paramagnetic effect  434 melanocytoma  279 melanocytosis  279 melanoma  270, 436, 442, 688 – mucosal  487 melanomatosis, diffused leptomeningeal  603, 604 MELAS  240 melphalan  745 membrane lipid  1344 membranous labyrinth  453, 476 memory  1309, 1311 meningioma  219, 273, 274, 275, 295, 296, 297, 412, 466, 467, 614 – anaplastic  273, 277 – atypical  273 – brain  1361 – cavernous  439 – intracranial  1365 – of the spine  592, 596, 597 – orbital  439 – parasellar  281 – perioptic  439, 442 – sellar  281 – skull base  276, 278 meningitis  333, 348, 364 – abscess  354 – aseptic  349 – bacterial  348 – cerebritis  354 – CNS complications  353 – cranial nerve dysfunction  353

Subject Index

– dura-subarachnoid type  351 – epidural empyema  355 – FLAIR images  349 – hydrocephalus  353 – influenza  353 – myelomatous  602 – neuroimaging  349 – pia-subarachnoid type  350 – subdural empyema  354, 355 – tuberculous  351, 352 – viral  349 meningomyelocele  212 meniscal tear  1118, 1120 meniscectomy  1120 meniscus – discoid  1120 – lesions  1117 – posterior horn  1118 – sign  590 merosin  1198 MERRF  241 mesencephalon  194 mesothelioma, diffuse malignant  903 metabolic – acidosis  1204 – storage diseases  749 metabolite concentration  1342 metachondromatosis  473, 1165 metameric arteriovenous malformation  636 metaphyseal spur  1110 metaplasia  1007 metastases  414, 573, 623, 930 – avascular  872 – choroidal  442 – hematogenous  1013 – hypervascular  872 – hypovascular  872 – isovascular  872 – leptomeningeal  599, 614 – of the liver  869 – pericardial  766 – peritoneal  904 metencephalon  194 methacholine  683 methemoglobin  270, 327, 331, 332, 334, 341, 811, 904 methylene resonance  1336, 1338 MG see myasthenia gravis micro-ultrasound  1390 microadenomas  281, 283 microbleeds  343 microbubble  1389, 1393 microcephalies  197 microcomputed tomography  1387 microcyst  594

microdissection  1419 microfluid  1416, 1419, 1421 microlissencephaly  197, 198 micrometer  1413 micropolygyria  378 microvascular density (MVD)  297 mid-aortic pulmonary arterial collaterals (MAPCAs)  726 midbrain stroke  320 middle – cerebral artery (MCA)  313 – interhemispheric variant holoprosencephaly  210 midline facial dysmorphism  208 migraine  311 miliary spreading  687 Miller-Dieker syndrome  200 miniaturized silicon nano­wire  1420 MION see monocrystalline iron oxide nanoparticle missile effect  164 mitochondrial disorder  240, 1202 mitochondriopathy  1358 mitral – regurgitation  757 – stenosis  757 – valve prolapse  834 Miyoshi dystrophy  1200 Mn, see manganese MND see motor neuron diseases modified receptor expression  1382 molar tooth midbrain–hindbrain malformation  215 MOLD (motion-level-dependent contrast)  1303 molecular – biology  1411 – diagnostic  1418, 1420 – diffusion  130 – imaging (MI)  1381, 1411, 1413, 1425 – antibody  1399 – cancer staging  1426 – dendritic cell migration  1428 – detection of nodal metastases  1426 – imaging method  1382 – of thrombosis  1428 – probe  1390, 1395 – hyperpolarization  1395 – molecular target  1397 – paramagnetic substance  1394 – superparamagnetic nanoparticle  1395 – signal generator  1392

– smart probe  1403 – informative target  1425 – mobility  318 – motion  130 – targeting  1397 monocrystalline iron oxide  1400 – nanoparticle (MION)  1395 mononeuropathy  1207 monophosphate  1344 monorchism  1043 Morton neuroma  1147, 1150, 1151 motion – artifact  138 – compensation  1272 – correction algorithm  1303 – sensitivity  140 motor – cortex  1292 – neuron disease (MND)  1208 – unit potential  1197 MOTSA  116 moyamoya disease  817 MPRAGE  446, 448 MR see magnetic resonance MRA see magnetic resonance angiography MRI see magnetic resonance imaging MRS see magnetic resonance spectroscopy (MRS), see also spectroscopy MS see multiple sclerosis MSA see myositis-specific antibody MSCT coronary angiography  767 mucopolysaccharidose  239 mucor  355, 365 mucormycosis  365 mucus plugging  682 Müllerian duct anomalies  964, 967 – bicornuate uterus  969 – MRKH syndrome  967, 969 – septate uterus  969 – transverse vaginal septum  969 – unicornuate uterus  967 – uterine agenesis  967 – uterine hypoplasia  967 – uterus didelphys  967 – vaginal agenesis  967 multi-echo sequence  22 multi-shot EPI  139 multi-voxel techniques  184 multifocal osteomyelitis (CRMO)  1248 multiminicore disease  1197 multiphase MRA  124

1497

1498

Subject Index

multiple – aorticopulmonary collateral arteries (MAPCAs)  728 – brain metastases  270 – endocrine neoplasia (MEN)  947 – myeloma  571, 1088, 1243, 1244 – salt-and-pepper pattern  1089 – overlapping thin slab acquisition  116 – receiver channels  849 – sclerosis (MS)  98, 178, 641 multiple-infarct dementia  384 multiple-slice technique  34, 35 multiple-system atrophy  388 multivane  34 MUP see motor unit potential muscle  1186, 1188, 1204, 1206 – acute traumatic injury  1206 – atrophy  1144, 1184, 1186, 1206, 1209 – contusion  1203 – denervation  1206 – direct biopsy  1188 – disease  1177, 1178, 1180 – contrast enhancement  1180 – dystrophy  1185, 1186 – edema  1178, 1184, 1186, 1194 – fatty infiltration  1206 – fiber  1184 – infection  1194 – inflammation  1209 – pain  1191 – polymyositis  1186 – rhabdomyolysis  1204 – signal intensity  1178, 1209 – strains  1203 – tissue  1181 – fatty infiltration  1184 – trauma  1203 muscle–eye–brain disease  201 muscular dystrophies (MDs)  1197 musculoskeletal – imaging  1084 – fat suppression  1084 – system  1081, 1082, 1178 – coil selection  1082 – examination technique  1081 – inflammatory disease  1094 MVD see microvascular density myasthenia gravis  1188, 1204, 1205 mycobacterium tuberculosis  363, 885

myelin – fiber  173, 174 – synthesis  1350 – vacuoles  216, 217 myelination  195, 196, 230 – terminal zones  196 myelography  590 myelolipoma  950 myelomalacia  657, 1105 myelonencephalon  194 myo-inositol (mI)  1342 myocardial – function  769 – cine cardiac MR tagging  771 – dobutamine stress imaging  770 – stress cine MRI of cardiac function  770 – infarction  748, 772, 833, 839 – gadolinium enhancement  746 – ischemia  833 – perfusion  106, 773 – tagging  771 – tissue  107 – viability  107, 716, 771 myocarditis  748 myocardium  106, 107, 763 myoglobinuria  1201 myoinositol  186 myometrial contraction  970 myonecrosis  1184, 1185, 1194, 1195, 1204 – diabetic  1195 myopathy  1177, 1188 – congenital  1196 – endocrine  1203 – familial idiopathic  1191 – granulomatous  1194 – inflammatory  1188 – malignancy-associated necrotic  1191 – metabolic  1185, 1201 myositis  441, 1188, 1191, 1194 – drug-induced  1191 – focal  1194 – ossificans  1184 – circumscripta  1196 – specific antibody  1189 myxoma  779 – prolapse  779 myxopapillary ependymoma  261, 592, 619, 621

N N-acetyl-aspartate (NAA)  185, 293, 1342 – signal intensity  1351 N-acetyl-aspartylglutamate (NAAG)  1344 NAA see N-acetyl-l-aspartate NAAG see N-acetyl-aspartyl­ glutamate Naegleria fowleri  370 NAFLD see nonalcoholic fatty liver disease nanobiotechnology  1420, 1424 – drug discovery  1424 nanobody  1425 nanocantilever  1420 nanofabrication protocol  1421 nanolab  1416 nanoparticle – molecularly targeted  1421 – nanoshell  1424 – paramagnetic  1421 – polymerized  1423 – quantum dot  1423 nanoscale protein analysis  1420 nanosensor array  1421 nanoshell  1423 nanotechnology  1416 nanowire  1421 nasal – cavity (NC)  486 – cephaloceles  485 – glioma  455 NASH see nonalcoholic steatohepatitis nasopharyngeal carcinoma (NPC)  425, 491, 492, 647, 648 – perineural extension  425 nasopharynx  489, 490 – adenoidal hypertrophy  491 – parotid space (PS)  513 – pharyngeal mucosal space (PMS)  512 navigation  1258, 1278 navigator  666 – echo  90, 127 – correction  138, 140 NAWM see normal appearing white matter Nb3Sn  80 NbTi  79 NC see nasal cavity NdBFe  78 near-infrared fluorescence (NIRF) imaging  1404 neck, pediatric  529

Subject Index

necrotizing fasciitis  1102, 1195 needle holder  1265 negative – contrast  1262 – predicting value (NPV)  296 Neisseria meningitidis  348, 353 neo-tendon  1125 – patellar  1125 neoaorta  731 neoplasm  765, 778 – cardiac  778 – primary  779 – cauda equina  597 – cystic pancreatic  889 – filum terminale  597 – hemorrhagic  341 – neuronal–glial  264 – of the spine  592 – papillary epithelial  890 – solid  890 – ventricular  765 nephrogenic systemic fibrosis (NSF)  95 nerve – compression syndrome  1143 – hypoglossal  452, 462 – root avulsion  610 – sheath tumors  218 – of the spine  592, 593 – stimulation  158 – vidian  451 neural foraminal stenosis  551 neurilemoma, see also schwannoma  286, 521 neurinoma  425 – Gasserian ganglion  425 – Meckel’s cave  425 neuroblastoma  266, 693, 950, 1226 – congenital  1220 – olfactory  468 neurocutaneous melanosis  226, 279 neurocysticercosis  365 neurocytoma  263, 265 neuroendocrine carcinoma  1243 neuroenteric cyst of the IDEM compartment  605 neurofibroma  218, 462, 693 – of the spine  593, 594 – plexiform  596 neurofibromatosis, see also von Recklinghausen disease  342, 593, 845 – type 1 (NF1)  216, 249 – type 2 (NF2)  219, 273 – urinary bladder  1007

neurofibrosarcoma  218 neuroimaging  243 – tuberculomas  361 neuromyelitis optica  643 neuron  185 neuronal – ceroid lipofuscinoses  236 – marker  1344 neuroradiology  172 neurosurgery  1277, 1284 – intraoperative guidance  1284 – planning  1314 – stereotactic  1323 neutron  183 neutropenic fever  678 nevoid basal cell carcinoma syndrome (NBCCS)  266 NF2 gene  286 nicotine-amide-adeninedinucleotide  1345 Niemann-Pick disease  237 Nijmegen breakage syndrome  225 Nitinol  1262 NMR see nuclear magnetic resonance NOE see nuclear Overhauser effect noise cancellation  1262 non-communicating hydrocephalus  391 non-Fourier imaging  1280 non-Hodgkin’s lymphoma (NHL)  269, 458, 460, 468, 487, 523, 602, 690, 891, 1090 – testicles  1051, 1053 – urinary bladder  1007 non-ionizing radiation  153 non-mono-exponential diffusion attenuation  144 non-proton coil  87 Nonaka dystrophy  1200 nonalcoholic – fatty liver disease (NAFLD)  881 – steatohepatitis (NASH)  881 nonglial tumors  264 nonketotic hypergylcinemia  236 normal – anatomy  538, 578 – appearing white matter (NAWM)  178 – pressure hydrocephalus  394 NPC see nasopharyngeal carcinoma NPV see negative predicting value NSF see nephrogenic systemic fibrosis nuclear – localization  1403

– magnetic – moment  8 – resonance – sensitivity  1339 – signal  1331 – Overhauser effect (NOE)  1338, 1395 – spin  8 nucleoside – diphosphate (NDP)  1345 – triphosphate (NTP)  1345, 1391 nucleus  8 – MR-relevant properties  9 – X  1337 number of diffusion gradients  144 Nyquist scanning theorem  124 O obliterative endarteritis of the vasa vasorum  836 obstruction  677 occluder  1262 occlusion, tracheal  1222 occult fracture  1171 OCT see optical coherence tomography oculocerebrorenal syndrome  234 oculopharyngeal muscular dystrophy  1199 odontogenic infection  518 olfactory groove  466 oligoastrocytoma  259, 261 oligodendrocyte  594 oligodendroglial tumors  256 oligodendroglioma  248, 260, 261 – anaplastic  259 oligohydramnios  1227 oligonucleotide  1391 oligospermia  1360 olivopontocerebellar atrophy  389 Ollier disease  473, 1165, 1166, 1246, 1247 OLT see osteochondral lesion of the talus oncocytoma  927 oncology, WB-MRI  1237 oncoprotein  1424 oncospheres  365 open-lip schizencephaly  203, 204 operating modes  153 – controlled  153 – experimental  153 – normal  153 operation room  1272 ophthalmoplegia plus  240

1499

1500

Subject Index

OPMD see oculopharyngeal muscular dystrophy opposed solenoid coil  1264 optical/optic – cable  89 – chiasm  451 – hypothalamic gliomas  249 – coherence tomography (OCT)  1386 – glioma  250 – hypoplasia  210 – imaging technique  1384 – labeling  1404 – microphone  1262 – nerve  438 – signal molecules  1392 – quantum dot  1393 – window  1384 opticus meningioma  275 optimal echo time  1296 optimized implementations for renal 3D CE MRA  827 – 3 T  827 – parallel imaging (PI)  827 oral cavity  495, 496 – cellulitis tends  497 – mucosa  496 – squamous cell carcinoma  498 orbital anatomy  434 – delineation  434 orbit surface coil  433 orchiectomy  1046, 1047, 1049, 1052 – cryptorchid  1043 organ metastase  1239 organoaxial volvulus  1224 organogenesis  1214 orientation  155 oropharynx  489, 493, 495, 500 – parapharyngeal space (PPS)  512 – parotid space (PS)  513 – squamous cell careinoma  494 orthogonal coil  1265 orthophosphorus acid  1345 ortical fibers  299 Osgood-Schlatter disease  1126 Osler-Weber-Rendu syndrome  634, 636 osseous lesions  219 ossification of the posterior longitudinal ligament  548 osteitis  1097 – postoperative  1097 osteoarthritis (OA)  1098, 1126, 1134

osteoarthropathy, diabetic  1174 osteoblast  1107 osteoblastoma  563 osteochondral lesion of the talus  1147 osteochondritis  1100, 1116 – dissecan (OCD)  1114, 1128, 1147, 1154 – erosive  1101 osteochondroma  569, 1166 osteochondrosis  1246 osteoclast  1107 osteocyte  1107 osteoid osteoma  563, 1162, 1165, 1166 osteomyelitis  556, 1097, 1174 – acute  1094, 1095 – chronic  1096 – non-diabetic  843 – postoperative  1097 – posttraumatic  1097 – subacute  1097 osteomyelofibrosis  1093 osteonecrosis  1087, 1115, 1116, 1147, 1154 – aseptic  1173, 1174 – of the hip  1174 osteopenia  1087 osteophytes  1136 osteoporosis  1246, 1249 – juxta-articular  1097 osteosarcoma  570, 784, 1155 – cardiac  784 – left distal femur  1157 – of the lower leg  1160 – of the proximal tibia  1167 osteosclerosis  1087 osteotomy  1081 otitis  1094 Oubert syndrome  215 ovarian – cancer/carcinoma  966, 993 – ascites  994 – characterization  993 – clear cell carcinoma  993 – endometrioid carcinoma  993 – mucinous cystadeno­ carcinoma  993 – non-resectable  995 – peritoneal implant  994 – peritoneal metastases  994 – recurrence  996 – resectable  995 – serous cystadeno­ carcinoma  993 – staging  966, 993

– TNM classification  993 – treatment selection and follow-up  995 – cyst  989, 1226 – corpus luteum  989 – follicular  989 – metastasis  996 – Krukenberg tumors  996 oxygen  1292 oxygen-enhanced MRI  671 oxygenation state  114 oxyhemoglobin  327, 330, 331, 1292 P pachygyria  201, 202, 378 pachymeninx  591 PACS  87 Paget’s disease  1246 palatine tonsil  521 pampiniform plexus  1047 Pancoast tumor  684 pancreas  886 – adenocarcinoma  887 – cancer  886, 892 – divisum  886 – gastrinoma  888 – inflammatory disease  891 – lymphoma  891 – metastases  891 – microcystic serous cystadenoma  889 – parenchyma  887 pancreatitis  886, 892, 897, 1107 – acute  891 – chronic  891 – tumor-associated obstructive  887 pancytopenia  1091 PANK2 mutation  237 Panner’s disease  1154 panniculitis, mesenteric  905 pannus, inactive fibrous  1105 pantothenate kinase–associated neurodegeneration  237 PAOD see peripheral arterial occlusive disease papillary – cystadenomas of the endo­ lymphatic sac  223 – fibroelastoma  781 – muscle dysfunction  727 – thyroid carcinoma  527 papilloma  280, 485, 488 PAPVR see partial anomalous pulmonary venous return

Subject Index

para-hydrogen–induced polarization  1395 paradigm selection  1309 paraganglioma  477, 520, 521, 592 – cauda equina  600 – cervical  812 – of the spine  599 parallel – acquisition technique  138 – imaging  55, 86, 122, 138, 667, 800 parallelization  88 paramagnetic substance  1394 parametric map  36, 179 paramyxovirus  1193 paranasal sinuses (PNS)  486 – cephaloceles  485 paraneoplastic syndrome  602 parathyroid glands  528 paravertebral MRI sequences  1082 parenchyma – hepatic  864 – pancreatic  893 – splenic  900 paresthesia  841 Parkinson’s disease  388 parotid malignant tumors  517 parotitis  514 paroxysmal nocturnal hemoglobinuria  941 Parry-Romberg syndrome  226 pars defect  549 partial – anomalous pulmonary venous return (PAPVR)  726, 807, 808 – k-space sampling  122 – oxygen pressure  89 PASCs see phased-array surface coils passive – shielding  80 – tracking  1262 patella  1172 – dislocation  1172 – flake fracture  1128 patent – ductus arteriosus (PDA)  721, 730 – foramen ovale  730 patient – accessibility  76 – burn  89 – implants  165 – metallic objects  165 – pregnant  164 – preparation  1312 – table  1266

– tattoos  165 pauciparameter molecular measurement  1413 PBC see primary biliary cirrhosis PCL tear  1125 PCr see phosphocreatine PDA see patent ductus arteriosus PDE see phosphodiesters peak integral  87 pearl-sign irregularity  811 pectus excavatum  1220 PEEK  1273 Pelizaeus-Merzbacher disease  235 pelvic – congestion syndrome  996 – varices  996 – floor  1070 – anatomical landmarks  1070 – anatomy  1069 – complex congenital anomaly  1005 – congenital malformations  1003 – dysfunction  1069 – insufficiency  1000 – girdle  1189 – inflammatory disease  996 – adnexal abscess  996 – hydrosalpinx  996 – pyosalpinx  996 – pain  1014, 1029 pelvis – male  999, 1018, 1039, 1055 – prolapse  1073 penile – corpora  1064 – fibrosis  1060 – urethra – cancer  1062 – transitional cell carcinoma  1062 penis  1055, 1056, 1057, 1059, 1060 – cancer  1062 – congenital anomalies  1059 – diagnostic algorithm  1067 – epispadia-exstrophy complex  1059 – fracture  1060 – imaging planes  1056 – infection  1061 – inflammatory disease  1061 – metastasis  1062 – MRI examinations  1055 – normal anatomy  1057 – prosthesis  1060

– squamous cell carcinoma  1063, 1066 – trauma  1060 penumbra  1351 percentage signal intensity loss (PSIL)  102 perfluoropolyether (PFPE)  1407 – -labeled dendritic cell  1395 perfusion  127, 805 – CT (PCT)  179 – imaging – adenosine-stress  773 perfusion-weighted imaging (PWI)  293 – brain metastases  297 – ischemic stroke  321 – neuro-oncology  294 periarteriitis nodosa  843 pericardial – adipose tissue  761 – cyst  765 – defect  762 – foregut duplication  1220 – diverticula  765 – effusion  763, 764, 835 – hematoma  765 – inflammation  763 – masse  765 – metastases  766 – sac  761 – tamponade  783 – tumors  765 pericarditis – constrictive  763, 764 – inflammatory  763 pericardium  761 – imaging  762 pericyst  367 perimyositis  1195 perineum – descending  1073, 1075 – infection  1061 periodic signal variation  116 periorchitis  1053 periostitis  1087 peripheral – arterial occlusive disease (PAOD)  840, 845, 846 – Fontaine’s classification  839 – nerve stimulation  84, 1273 – veins  842 – vessel  803 periportal halo sign  878 perisylvian syndrome  202 peritendinitis  1149 peritoneal inclusion cyst  989

1501

1502

Subject Index

peritoneocele  1073, 1074, 1075 peritoneum – abscess  905 – cysts  903 – hernias  903 – inflammation  905 – lipomas  903 – mesenteric lipomatosis  903 – metastatic tumors  903 peritonitis  905 periventricular leukomalacia (PVL)  197, 205, 228, 229, 230 perivertebral space (PVS)  523 personnel  1272 perthes disease  1110, 1111, 1174 PET see positron emission tomography Peyronie’s disease  1061 PHACES syndrome  226 phagocytosis  1405 phakomatoses  216 phantom measurement  1323, 1325 – 2D phantom  1323 – 3D phantom  1325, 1327 pharmacogenomics  1413 pharmacologic stress testing  716 pharmacoproteomic  1413 pharyngitis  523, 1094 phase – angle  30 – encoding  30 – noise  120 phase-contrast – flow measurement  1269 – MRA  788 – velocity-encoding gradient (VENC)  788 phased-array surface coil (PASC)  86, 1040 – body-oil  1019 phenylketonuria  235 phenytoin  1191 pheochromocytoma  947 phleboliths  339 phlegmon of the oral cavity –   497 phosphatidylcholine  1222, 1344 phosphatidylserine  1422 phosphocholine  186 phosphocreatine (PCr)  1201, 1339, 1342, 1345 – resonance  185 – signal  1339 phosphodiesters (PDE)  1210, 1345 phosphoethanolamine (PE)  1345 phospholipids precursor  1350

phosphometabolity  1352 phosphomonoester (PME)  1210, 1345 phosphorus  1345 – 31  1331 – MR  1210 physiologic monitoring  164 pia mater  591 Pick’s disease  383 piezoelectric element  1389 piezo motor  1266 pigmented villonodular synovitis (PVNS)  1128, 1162 pilocytes  249 pilocytic astrocytoma  248, 250 pilomyxoid astrocytoma  411 pinal cord ischemia  831 pincushion  1323, 1325 PIN diode  86 pinealis cyst  268 pinealoblastoma  269 pineoblastoma  267, 614 pink Fallot  728 pituitary – adenocarcinoma  408 – adenoma  281, 405 – cavernous sinus invasion  408 – dynamic contrast-enhanced imaging  405 – dynamic gadoliniumenhanced MR image  407 – GH-producing  406 – macroadenoma  405, 407 – microadenoma  405, 407 – multidetector-row CT  408 – postoperative study  408 – prolactinoma  405, 406 – signal intensities  406 – apoplexy  281, 408 – carcinoma  283 – dwarfism  422 – ectopic posterior lobe  423 – pituitary stalk interruption  423 – transection of the pituitary stalk  423 – gland  399 – adenohypophysis  400, 401 – microadenoma  283 pixel  6 pixel-by-pixel correlation  1300 planar diffusion  134 Planck’s constant  6 planimetry  754

plantar  1150 – aponeurosis  1171 – fasciitis  1150 – fibromatosis  1150 planum sphenoidale  450, 466, 467 plasma cell dyscrasia  602 plasmacytoma  480 plasticity  1316 pleomorphic – adenoma  442, 516 – xantho-astrocytoma (PXA)  248, 255 pleura  676 pleural effusion  1220 plexiform neurofibromas  218 plexopathy  610 plexus papilloma  280 PLS see primary lateral sclerosis PMA see progressive muscular atrophy PME see phosphomonoester PML see progressive multifocal leukoencephalopathy PNET see primitive neuroecto­ dermal tumor pneumatically driven arm  1266 pneumonia  678 – atypical bacterial  678 – bacterial  678 – fungal  678 PNS see paranasal sinuses point-resolved spectroscopy (PRESS)  184, 1024, 1336 – technique  1336 point-spread function, PSF  124 polarization enhancement  1394 pole shoe  78 poliomyelitis  1208, 1209 poliovirus  1208 – paralytic  1209 polycystic – kidney  918, 1228 – acquired  918 – autosomal dominant form  918 – ovarian disease  990 polyhydramnion  1221 polylactide  1393 polylysine dendrimer  1424 polymicrogyria  197, 198, 202, 226 – bilateral syndromes  202 polymyalgia rheumatica  836 polymyositis  1184, 1189, 1202, 1246, 1249 – multinodular  1192

Subject Index

polyneuropathy  608 – inflammatory demyelinating  1206 – metabolic  1207 – toxic  1207 polynomial  1325 polyorchidism  1043 polypectomy  1252 polyradiculopathy  607 polysplenia syndrome  732, 897 pontocerebellar hypoplasia  214 pontosubicular necrosis  229 popliteal artery entrapment  845 POPQ (pelvic organ prolapse quantification)  1073 porencephaly  227 portal – hypertension  879, 880 – varices  880 – vein thrombosis  882 portosystemic shunts  880 port wine stain  224 position encoding  26 positive predictive value (PPV)  296 positron emission tomography (PET)  1387, 1390 – CT  1390 – MRI  1390 post-poliomyelitis syndrome  1209 post-polio syndrome  1209 posterior – cervical space (PCS)  525 – cord syndrome  654 – cruciate ligament (PCL)  1117, 1121 – injury  1117 – fossa  176, 212, 213 – hemangioblastoma  223 – median spinal vein  626 – spinal artery  626, 628 posttraumatic – cyst  658 – syrinx  658 potassium-39  1331 PPF see pterygopalatine fossa PPV see positive predictive value precession  10 pregnancy – acoustic damage  1214 – MRI safety  1213 – teratogenesis  1214 premature infants – hypoxic-ischemic injury  228 PRESS see point-resolved spectroscopy

PRESTO see principles of echo-shifting with a train of observations priapism  1060, 1064 primary – biliary cirrhosis (PBC)  877 – lateral sclerosis  1208 – sclerosing cholangitis (PSC)  895 primitive neuroectodermal tumor (PNET)  266, 614 principles of echo-shifting with a train of observations (PRESTO)  178 Probst’s bundle  206 processing  1278 productivity  653 profiling coil  1264 progressive – multifocal leukoencephalopathy (PML)  376 – muscular atrophy  1208 projection  28 – imaging  1262 – reconstruction method  31 prolapse, pelvic  1073 propeller  34 – diffusion sequence  140 prosencephalon  194, 208 prospective – gating  88 – motion correction  90 – triggering  120 prostate  1018, 1361 – abscesses  1030 – benign tumorous lesions  1030, 1036 – cancer/carcinoma  1030, 1240, 1242, 1252, 1277, 1361 – brachytherapy  1277 – cell line  1415 – extracapsular extension (ECE)  1028 – staging  1018 – TNM classification  1032 – chronic inflammation  1030 – congenital abnormalities  1029, 1034 – cystic lesions  1029 – dynamic, contrast-enhanced MRI (DCE MRI)  1027 – hemorrhage  1023, 1029, 1036 – imaging  1019 – infectious disease  1029, 1036 – inflammatory  1029, 1036 – magnetic resonance imaging  1020

– malignant tumorous lesions  1030, 1037 – MRI-guided biopsy  1018 – MR spectroscopy  1025 – ³¹P MRS spectra  1360, 1361 – ¹H MRS  1363 – multiple proton spectra  1024 – normal anatomy  1028 – of adenocarcinoma  1030 – pseudocapsule  1028, 1034 – punch biopsy  1033, 1037 – trauma  1029 – TRUS-guided biopsy  1034 – zonal anatomy  1022, 1023 prostate-specific antigen (PSA)  1031, 1419 prostatectomy  1000 prostatitis  1029, 1252 prosthetic valves  760 protein – amplification  1399 – chip  1420 – microarray  1419, 1420 proteomic technology  1420 protocol  964 proton  5, 1331, 1332 – decoupling  1340 – density  36 – magnetic field  92 – nuclear magnetic resonance spectroscopy  183 – of the hydroxyl (-OH) group  1336 – resonance frequency (PRF) technique  1272, 1339 – spectrum of ethanol  1336 – spin system  1339 – β-methyl doublet  1345 prune belly syndrome  1227 PSA see prostate-specific antigen psammoma bodies  596 PSC see primary sclerosing cholangitis pseudo-leukodystrophy  235 pseudo-TORCH syndromes  236 pseudoaneurysm  811 – saccular  834 – ventricular  765 pseudocapsule, prostatic  1028 pseudocyst  364, 905 – pancreatic  893 pseudodisease  1253 pseudohypertrophy  1199 pseudomeningocele  554, 611 Pseudomonas  348, 357

1503

1504

Subject Index

pseudotumor  443 – orbital  440 PSIL see percentage signal intensity loss psoas muscle  958 psoriatic arthritis  1105 PTEN gene  253 pterygoid process  451 pterygomandibular raphe  499, 500 pterygopalatine fossa (PPF)  451, 461, 485, 486 pubococcygeal line (PCL)  1070 Pulley lesion  1131 pulmonary  809 – arteriovenous malformation (AVM)  808 – artery  728, 804 – embolism  804, 805, 806 – hypertension  806, 807 – regurgitation  760 – sequestration  1219 – sling  728 – stenosis  760 – thromboendarterectomy  806 – vasculature  804 – vein  804, 807, 808 – stenosis  807 – vessel  802 pulsatility  114 pulsation artifact  116 pulsed-gradient spin echo (PGSE) technique  136 pulsed arterial spin labeling  117 pulseless disease  835 pulse modulation  28 PVL see periventricular leukomalacia PWI see perfusion-weighted imaging PXA see pleomorphic xanthoastrocytoma pyarthrosis  1099 pyelonephritis  936 – acute  936 – chronic  936 – xanthogranulomatous  937 pyodermia  1094 pyogenic – abscess  883 – brain abscesses  361 pyothorax  688 pyriform sinus  502 Q q-ball imaging  144 q-space diffusion imaging  144

quadriceps tendon  1126 quadrupolar nuclei  1353 quantum dot (QD)  1393, 1423 quenching  79, 1393, 1403 R radial – acquisition  34 – collateral ligament  1153 radiation – dose  647 – myelitis  647 – myelopathy  647 – therapy  118 – treatment planning  1316 radiculitis  607 radiculomedullary – artery  626, 629 – vein  626 radiculopathy  596 radioactive – iodine (RAI)  526 – isotope  1388, 1392 – labeling  1404 – marker  1392 radiofrequency (RF)  6, 1333 – antenna  1333 – cabin  85, 1273 – electromagnetic field  161 – field  10, 12, 18, 76, 161 – exposure limits  163 – hot spot  76 – heating  90 – induced heating  1273 – matching  85 – multi-channel coils  176 – power amplifier  85 – pulse  13 – 180°  13 – 90°  13 – spectrum  13 – spoiling  44, 1267 – system  1333 – tuning  85 radiolabeling  1404 radionuclide ventriculography  748 radiosurgery, interstitial  1323 radiotherapy  1323 RAI see radioactive iodine RARE  139, 1268 Rashkind maneuver  730 Rathke’s cleft cyst  285, 413 Rathke’s pouch  283, 413 Raynaud’s phenomenon/ syndrome  841, 1189, 1191 RCL see radial collateral ligament

real-time – MRI  1258 – operating system  88 receive – (Rx) coil  85 – chain  85 receptor-induced magnetization enhancement (RIME)  94 receptor-mediated endocytosis  1399 rectal – cancer  908 – carcinoma  1238 rectocele  1073, 1074, 1075 red–green–blue (RGB) color model  145 red fluorescent protein (RFP)  1405 reduced echo time  1297 reference point  1328 reflectance geometry  1385 region-growing  127 registration  1279, 1285 – non-rigid  1279, 1285 – rigid  1279 regurgitation, pulmonary  760 Reiter’s disease  1105 relative – anisotropy (RA)  135, 145 – map  145 – mean transit time (rMTT)  321 – time-to-peak (rTTP)  321 relaxation  13 – enhancement  1394 – longitudinal  13 – phenomenological description  15 – physical model  14 – rate  121 – time  13 – biological tissue  17 – T1/T2  13 – transverse  13 relaxivity  122 renal – adenoma  927 – arteries  817 – cell carcinoma  509, 891, 920, 1243 – multiple metastasis  272 – Robson classification  920 – TNM classification  920 – tumor infiltration into the renal vein or the inferior vena cava  920 – whole-body staging  926

Subject Index

– cyst  916 – Bosniak classification  917 – complex  916 – simple  916 – transplantation  939 – acute tubular necrosis (ATN)  939 – lymphoceles  939 – lymphoproliferative disease (PTLD)  939 – urinoma  939 – trauma  941 – vein thrombosis  824 Rendu-Osler-Weber syndrome  808 resonance  185 – condition  10 – excitation  12 – frequency  1293 – gradient system  84 respiratory – artifact  712 – breath-hold imaging  712 – real-time imaging  712 – distress syndrome  1222 – gating  127 restrictive cardiomyopathy  739, 765 rete testis  1042 reticuloendothelial system (RES)  101, 1091, 1405 retinal angiomas  223 retinoblastoma  438 retrobulbar space anesthesia  433 retromolar trigone mucosa  499 retroperitoneal – fibrosis (Ormond’s disease)  954 – tumor  953, 954 retroperitoneum – lymphangioleiomyomatosis  953 – tuberous sclerosis  953 – Whipple’s disease  953 retropharyngeal space (RPS)  523 retrospective gating  88, 120 Rett syndrome  237 RF see radiofrequency rhabdoid tumor of the kidney  936 rhabdomyolysis  1203, 1204 rhabdomyoma  779, 782 rhabdomyosarcoma  518, 522, 783, 1003 rheobase  158, 159 rheumatic disease  755 rheumatoid – arthritis (RA)  1094, 1104, 1135 – disease  1102 Rhizopus  365 rhombencephalon  194

rhombencephalosynapsis  214, 215 Rich’s foci  351 right ventricle  680 – failure  806 RIME, see receptor-induced magnetization enhancement rMTT see relative mean transit time roadmap  1263 Robitom  1266 robotic arm  1266 rolling platform system  849 Rosenthal fibers  238, 249 rotator cuff  1134 – degeneration  1134 – inflammation  1134 – lesions  1133 – tear  1131, 1135, 1136 rTTP see relative time-to-peak S saccular aneurysms  341 saccule  453, 454 sacrococcygeal teratoma  570 sacroiliitis  1105 SAE see stimulated acoustic emission safety regulations  153 Salla disease  236 salmonella  355 salpingo-oophorectomy  703 salt-and-pepper pattern  812, 1089 – bone marrow  1089 Sandhoff disease  239 Santavuori-Haltia disease  236 Santorini drains  886 sarcoidosis  419, 456, 650, 652, 691, 739, 901 – cardiac involvement  746 – Churg-Strauss syndrome  419 – IDEM  607 – muscular  1194 – of the heart  747 – Rosai-Dorfman disease  419 – Wegener’s granulomatosis  419 sarcolemma  1202 sarcoma  688, 1088 – granulocytic  1088 saturation – recovery  1271 – sequence  19 – slab  1269 scalar coupling constant JAX  1337 scaphoid bone  1114 – fracture  1114, 1141 – necrosis  1114 scapholunate (SL)  1141

SCC see squamous cell carcinoma schistosoma haematobium cancer  1007 schistosomiasis of the urinary bladder  1003 schizencephaly  197, 202, 203 Schmorl’s node  545 Schwann cell, spindle-shaped  594 schwannoma  219, 220, 286, 439, 462, 477, 478, 521, 592, 693 – acoustic  288 – of the spin  593 – trigeminal  288 – vestibulocochlear  287 scimitar syndrome  676, 808 SCIWORA  659 scleroderma  1191 sclerosing cholangitis  895 sclerosis – amyotrophic lateral  1208 – primary lateral  1208 scolex  365 scrotum  1039 – congenital anomalies  1043 – diagnostic algorithm  1054 – imaging  1052 – inflammatory disease  1047 – MRI examinations  1039 – normal anatomy  1041 – sarcoidosis  1047 – trauma  1043 scyllo-inositol  1344 SDE see subdural empyema SE see spin-echo (SE) secondary cardiac malignoma  785 SEGA see subependymal giant cell astrocytoma segmental artery  626 segmentation  127 segmented – EPI  139 – k-space sampling  55, 56 selective excitation  26 self-diffusion  130 self-navigated diffusion MRI  140 sellar region tumors  281 semicircular canal  453, 454 semiconductor nanowire  1420 seminoma  688 SENSE  55, 590 sensitive-particle acoustic quantification (SPAQ)  1389 sensorimotor cortex  1292 sensory–motor system  1308 sentinel lymph node (SLN) mapping  1423

1505

1506

Subject Index

septic arthritis  562 septicemia  348 septo-optic dysplasia  209, 210 septum pellucidum  204, 210, 264 sequence optimization  1296 seronegative spondylo­ arthropathy  1246 serum – albumin  123 – glutamic-oxaloacetic transaminase  1204 servicing costs  76 Sever’s disease  1147 SGOT see serum glutamicoxaloacetic transaminase shielding constant  24 shim coils  184 short-term patient motion  1261 short tau inversion recovery (STIR)  42, 590 shoulder  1130 – capsuloligamentous  1130 – dislocation  1139 – muscular element  1130 SH U 555 C  94 shunt – intracardiac  719 – vascular  719 Shy-Drager syndrome  389 sickle cell anemia  748, 899, 1094, 1107, 1113, 1116 sigmoid colon  1018 signal – generator (SG)  1392 – intensity of pixel  1299, 1300 – targeting with alternating radiofrequency  117 – void  753, 758, 1263 signal-recognition particle (SRP) antibody syndrome  1191 signal-to-noise ratio (SNR)  86, 93, 794 Sinding-Larsen-Johansson disease  1126 single-parameter measurement  1413 single-photon emission tomography (SPECT)  1387 single-voxel technique  184, 1334 single-walled carbon nanotube  1420 sinonasal tumor  490 sinus – endoscopy  1284 – MRI-guided  1284 – of Morgagni  489 – sphenoid  451, 452, 461, 462

– tarsi syndrome  1149 – thrombosis  329 sinusitis  1094 situs ambigous  719, 731 skeletal muscle  1179, 1181, 1184, 1358 – hypertrophy  1184 – ³¹P MRS examination  1358 – normal anatomy  1181 SLAP lesions  1140 slew rate  83 slice – position  27 – width  27 slice-selective excitation  27 sliding multislice (SMS) technique  1235 SMA see spinal muscular atrophy small cell carcinoma  684 smart probe  1391, 1403 SMASH  590 SNR see signal-to-noise ratio soft-tissue – infection  1102 – lesions  99 – neoplasm  1184 – pseudotumor  1194 – tumor  1155, 1159, 1161, 1174 – hemangioma  1161 – liposarcoma  1161 – malignant fibrous histiocytoma  1161 – ³¹P MRS  1365 – pseudocapsule  1159 solitary pulmonary nodule  687 somatostatinoma  889 – islet-cell tumors  889 SONK see spontaneous osteo­ necrosis of the knee SPAQ see sensitive-particle acoustic quantification spastic diplegia  228 spatial – encoding  28 – field homogeneity  76 – presaturation  116 – resolution  99, 1413 specific absorption rate (SAR)  76, 161 spectra analysis  1333 spectral density function  14 spectroscopic – 3D imaging  1336 – imaging  1334 – ¹H MR  1368 – investigation – bone marrow  1365

spectroscopy, see also magnetic resonance spectroscopy  1332, 1334 spermatic cord  1042 sphenopalatine foramen  461, 487 sphingolipid metabolism  749 Spielmeyer-Vogt disease  236 spin-echo (SE)  22, 178 – cone precession  11 – sequence  21, 37 – signal intensity  37, 38 spin-phase phenomenon  7 spin–lattice relaxation time  1342 spin–spin coupling  1084, 1336, 1337 spin–spin relaxation time  1293 spina bifida  1216 spinal – canal stenosis  551, 553 – cord – contusion  656 – infarction  626 – dysraphism  605 – meningioma  596 – meningitis  652 – MR angiography (MRA)  617, 637 – muscular atrophy  1208 – neoplasms  592 – neurofibroma  594 – vascular malformations  611 spine  536 – degenerative  536 – imaging techniques  537, 556, 562, 575 – infection  523 – intradural extramedullary  590 – normal anatomy  538, 578 SPIO see superparamagnetic particles of iron oxide spiral acquisition  34 spirochetes  836 spleen – accessory  897 – congenital diseases  897 – cysts  899 – hamartomas  899 – hemangiomas  899 – Hodgkin’s  899 – infarcts  902 – metastases  900 – non-Hodgkin’s lymphomas  899 – parenchyma  897 – trauma  902 splenomegaly  900 SPNET see supratentorial primitive neuroectodermal tumor

Subject Index

spoiled gradient-echo (SGE) – non-fat-suppressed  864 spoiler gradient  44 spondylitis  1100, 1101 spondyloarthropathy, sero­ negative  1094, 1105 spondylodiskitis  556, 1100, 1101 spondylolisthesis  549 spongiform leukodystrophy  238 spontaneous osteonecrosis of the knee  1114 sports injury  1171 squamous cell carcinoma (SCC)  290, 461, 468, 484, 512, 521, 684, 1243 – hypopharynx  500, 502 – oral cavity  498 – oropharynx  494 SSFP readout techniques  752 Stanford A dissection  833 Staphylococcus – aureus  355, 1194 – epidermidis  355 STAR (signal targeting with alter­ nating radiofrequency)  117, 1295, 1298, 1305 – optimal echo time  1298 – technique  1294 static – magnetic field  9 – adverse effects  155 – B0  18 – exposure limits  156 – inhomogeneity  24, 46, 51 – tissue  115 statistical – parametric map  1298, 1300 – test  1302 steady-state – free precession  713 – magnetization  46 steal syndrome, HS-induced  845 STEAM see stimulated echo acquisition mode steatosis  881 Stejskal-Tanner – diffusion gradient  136 – pulse sequence  1272 stenogyria  213 stenosis  121, 324 – aortic  755 – mitral  757 Stenson’s duct  513, 517 stent  1262 stent-grafts – aortic  836 – failure  836

– thrombosis  836 stereotactic – coordinate  1325, 1328 – procedure  1259 – zero point  1325, 1327 stereotaxy  1323, 1325 sternocleidomastoid muscle (SCM)  529, 531, 532 steroid  245 stimulated – acoustic emission (SAE)  1389 – echo acquisition mode (STEAM)  184, 1334, 1343 – emission depletion microscopy (STED microscopy)  1386 stimulation – effect  1301 – quantification  1301 – threshold  158, 160 STIR see short tau inversion recovery Stokes shift  1385 strength-duration expression  160 Streptococcus pneumoniae  357 stress fracture  1169, 1170 striatonigral degeneration  388 stroke  174, 179, 183, 813, 1316 – acute ischemic  311 – atherosclerotic  322 – blood-brain barrier  317 – cryptogenic  322 – dissection  324 – embolic pathogenesis  323 – embolism  323 – etiology  322 – FLAIR images  312 – hemodynamic compromise  324 – perfusion-weighted imaging  321 – vascular territories  323 – ventricular volume  759 – volume  754 Student’s test  1299, 1300, 1301, 1303 Sturge-Weber syndrome  223, 224 sub-harmonic frequency  1390 subacute – combined degeneration  659 – hemorrhage  434 subarachnoid hemorrhage – FLAIR  333 – proton density (PD) images  333 subchondral insufficiency fracture  1110 subclavian steal syndrome  812 subdural – empyema (SDE)  355, 356

– hematomas  334, 335 – hemorrhage  333 subendocardium  748 subependymal – germinolytic cysts  239 – giant cell astrocytoma (SEGA)  256 subependymoma  262, 264 submandibular gland (SMG)  532 subscapularis muscle  1135 subsegmental pulmonary  806 subtraction method  1298, 1300, 1301 superficial siderosis  333 superior orbital fissure  450, 451, 462 superparamagnetic particles of iron oxide (SPIO)  101, 104 – probe  1426 supplementary motor area (SMA)  1308 supra-aortic vessels  802 suprahyoid neck  512, 525 supraspinous muscle  1135 – atrophy  1136 supratentorial primitive neuroectodermal tumor (SPNET)  266 surface coil  86, 1333, 1334, 1335 surrogate – endpoints  708 – marker imaging  1386 Surveillance, Epidemiology, and End Results (SEER)  243 susceptibility  153 – artifact  139, 1312 – effect  52 – vessel sign (SVS)  313, 315 – weighted imaging (SWI)  330 SVS see susceptibility vessel sign SWI see susceptibility-weighted imaging 18q-syndrome  236 synovial – cyst  546 – proliferation  1105 synovitis  1154 syntelencephaly  210 synthesizer  85, 87 syringohydromyelia  219 syrinx  657 systemic lupus erythematosus (SLE)  1116, 1192 systems biology  1411, 1412, 1414 – network model  1414 – prostate cancer progression  1415

1507

1508

Subject Index

T T1/T2 relaxation time  6, 248 – T1 measurement  1271 T2 shine-through effect  136, 173, 174, 318, 321 TA see Takayasu’s arteritis tachyzoites  373 Taenia solium  365 Takayasu’s arteritis (TA)  835, 843, 851 tamoxifen  703 TAO see thromboangiitis obliterans tape  87 TAPVR see total anomalous pulmonary venous return target-to-background ratio  1426 TART see testicular adrenal rest tumors taurine (TAU)  1210 Tay-Sachs disease  239, 1208 Taylor – series  120 – type dysplasia  198 TB see tuberculosis TBI see traumatic brain injury TCC see transitional cell (urothelial) carcinoma TEE see transesophageal echocardiography telencephalon  208 telomerase activity  1422 temozolomide  259, 261 temperature rise  161 temporal – arteritis  812 – blurring  121 – data interpolation  124 – lobe epilepsy  193, 1352 – ¹H MR spectroscopic imaging  1352 temporomandibular joint  1146 tendinitis  1134, 1136 tendinopathy  1145, 1152 tendinosis  1134 tendovaginitis  1149 tennis elbow  1152 tenosynovitis  1145 tension hydrothorax  1220 tensor diagonalization  134 teratogenesis  1214 teratoma  289, 692 terbium-gallium-garnet  1265 test bolus  123 testicle – benign neoplasms  1047 – malignant neoplasms  1049 – non-Hodgkin lymphoma  1051

– secondary tumors  1051 – undescended  1044 testicular – adrenal rest tumors (TART)  1047, 1048 – contrast enhancement at MRI  1046 – epidermoid cyst  1048 – hydrocele  1049 – infarction  1046 – metastasis  1051 – torsion  1046 – tuberculosis  1047 – varicoceles  1047 testis – ectopic  1043 – low-lying undescended  1043 tetralogy of Fallot  721, 727, 728, 730 tetramethylsilane  1348 TFCC see triangular fibrocartilage complex TFT monitor  88, 1260 TGA see transposition of the great arteries TGDC see thyroglossal duct cyst thalassemia  748 thermal – ablation  1259, 1281, 1282 – conductivity  161 – effect  161 – equilibrium  6, 11 – motion  130 thermoacoustic-computed tomography (TCT)  1386 Thiemann’s disease  1147 thioguanine  1414 thiopurine, S-methyltransferase (TPMT)  1414 thoracic – outlet syndrome  841 – spine imaging  590 thoracoamniotic shunting  1220 thoracocentesis  1220 three-dimensional (3D) phantom  1323, 1327 thresholding  88 thromboangiitis obliterans (TAO)  87, 843 thrombolysis  344 thrombophlebitis  493, 842, 1251 thrombosis  314, 339, 348, 842, 845 – hepatic venous  882 – of the portal vein  880, 882, 895 thrombus – atrial  781 – cardiac  786

thymic rebound  692 thymoma  1205 thymus  1205 – abnormality  1205 – tumor  1205 thyroglobulin  527 thyroglossal duct cyst (TGDC)  530 thyroid – arteries  526 – cancer  508, 509, 527 – gland  526 – nodules  526 – ophthalmopathy  440, 441 tilted optimized non-saturating excitation  115 time-of-flight (TOF) MRA  7, 446, 790 – mass spectrometry  1419 – multiple overlapping thin slab acquisition (MOTSA)  790, 809 – supra-aortic and intracranial vasculature  809 – tilted optimized non-saturating excitation (TONE)  809 time-resolved MRA  124, 805, 806, 807, 808 time-resolved non-contrastenhanced MRA (tagging)  791 time constant T2*  19, 51 tissue heating  161 Tolosa-Hunt syndrome  424 tomosynthesis  708 tongue – base tumor  498 – cancer  494 – squamous cell carcinoma  501 – tumor  501 tonsillar – ectopia  212 – hypertrophy  493 tonsillectomy  493 torque  1267, 1273 torsion – extravaginal  1046 – intravaginal  1046 – testicular  1046 tortuous vessel  115 total – anomalous pulmonary venous return (TAPVR)  726, 807 – imaging matrix  86 toxoplasma  650 – encephalitis  370 – gondii  370 toxoplasmosis  269, 372, 373

Subject Index

TPMT see thiopurine S-methyltransferase trace – imaging  134 – of the diffusion tensor  134 tracking  1278, 1279 – coil  87 – electromagnetic  1278 – image  1279 – sensor  1278 tract, olfactory  450 tractography  148, 175 transcriptome  1419 transducer  1389 transesophageal echocardiography (TEE)  833 transferrin  1400 – receptor  1391 Fourier  1333 transformer  1273 transient ischemia attack (TIA)  183 transitional cell (urothelial) carcinoma (TCC)  1009 – of the bladder  1008 transit time  123 transmantle cortical dysplasia  198, 220 transmit – (Tx) coil  85 – chain  85 – receive switch  87 – transmit/receive (TxRx) coil  85 transplantation  824 transposition of the great arteries (TGA)  730, 731 transrectal – MR-guided biopsy  1262 – prostate biopsy  1266 – ultrasonography (TRUS)  1018, 1031, 1037 transthyretin  744 transudate  688 transurethral electroresection of the bladder (TUR-B)  1011 transversal relaxation rate  1293 transverse myelitis  646 trapezius muscles  525 trastuzumab  1423 trauma – aortic  836 – dural  610 – pelvic  1029, 1034 – penis  1060 – prostate  1034 – spleen  902 – to the scrotum  1043

– to the urinary tract  1003 traumatic brain injury (TBI)  178 – in children  193 TREAT  812 tree-in-bud  682 triacylglycerides  1348 triangular fibrocartilage complex  1141, 1142 tricuspid – atresia  724, 725 – regurgitation  753, 760 trigger delay  123 triggering  88 triquetrum  1143 trismus  518 trueFISP  48 truncated-meniscus  1119 truncus arteriosus  724, 726 TRUS see transrectal ultra­sono­ graphy tuberculoma  652 – caseating  361 – non-caseating  361 – parenchymal  363 tuberculoprotein  363 tuberculosis (TB)  361, 650 – encephalopathy  361 – hepatic  885 – mastoiditis  351 – meningeal  351 – parenchymal  361 – urinary bladder  1003 tuberculous – abscess  363, 364 – encephalopathy  363 – meningitis  652 tuberous sclerosis  197, 219, 256, 258, 782 – giant cell tumors  221 – subependymal nodules  221 tubers  221 tumor – extraosseous extension  1158 – immunotherapy  1428 – intraosseous extension  1155 – paravertebral  1082 – pericardial  765 – probability map  1334 tumoral cyst  619, 622, 623 tunica – albuginea testis  1042, 1043, 1060 – rupture  1046 – dartos  1041 – vaginalis testis  1042 – vasculosa  1043

TUR-B see transurethral electro­ resection of the bladder turbo-fast low-angle shot (turboFLASH) sequences  864 turbo spin-echo (TSE) technique  1295, 1298, 1304 – optimal echo time  1298 turbulent jet  120 Turner’s syndrome  729 twin gradient  84 twister gradient  1263 two-dimensional (2D) phantom measurement  1323 typical crazy paving  680 tyrosine kinase  1397 U UADT see upper aerodigestive tract UCL see ulnar (medial) collateral ligament ulna – (medial) collateral ligament  1153 – impaction syndrome  1172 ulnocarpal impaction  1143 ultrasmall paramagnetic iron oxide (USPIO)  93, 94, 671, 1395, 1406 – esophageal or gastric cancer  104 – Gd chelates  100 – lymph node–specific  104 – neuroimaging  98 – rectum cancer  105 – unclear breast tumors  100 ultrasmall particles of iron oxide  966 ultrasonography  674, 1258, 1389 – microbubbles  1389, 1393 – penis  1065 – signal generator  1393 undifferentiated – carcinoma  468 – sarcoma  783 univentricular heart  724 unsaturated magnetization  114 unspecific extracellular fluid space agents  92 upper – aerodigestive tract (UADT)  508 – airway obstruction  1218 uremia  1207 ureteropelvic junction obstruction  1227 urethra, prostatic  1029 urethropubic  1072

1509

1510

Subject Index

urinary bladder, see also bladder  999, 1028 – cervical cancer  1006 – chronic granulomatous pseudotumor  1008 – complex congenital anomaly  1005 – cystitis  1003 – diagnostic procedures  1015 – endometriosis  1008 – HASTE  1000 – inflammatory pseudo­ tumor  1006 – MRI  999, 1001 – non-Hodgkin’s lymphoma  1007 – non-neoplastic disorders  1003 – pseudosarcomatous myofibroblastic (fibromyxoid) tumor  1008 – SSFSE  1000 – therapy regimens  1011 – tumorous lesions – differential diagnosis  1014 urinary obstruction  938 urogenital – diaphragm  1028 – hiatus  1072 USPIO see ultra small paramagnetic iron oxide uterine – fibroid  1282 – lymphoma  973, 985 – cervical  985 – sarcomas  979 utricle  453, 454 V vacuum – chamber  1262 – phenomenon  542 vaginal – carcinoma  985 – TNM/FIGO classification  985 – vault – descent  1073 – prolapse  1075 Valsalva maneuver  502, 1047 valve incompetence  781 valvular – disease  748 – heart diseases  751 – pulmonary  760 – regurgitation  753 – stenosis  727, 751, 754

van der Knaap – disease  235 – leukencephalopathy  234 vanishing white matter disease  234 Varicella – infection  1196 – zoster  649 vascular – anatomy of the spinal cord  625 – bands, webs, pouches  806 – diseases of the spinal cord  625 – dissection  587 – endothelial growth factor (VEGF)  297 – malformation  530, 631 – periphery  121 vasculitis  348, 353, 365, 836, 843, 845 – CNS  817 vasculopathy  325, 354, 817 vasogenic edema  311, 341, 373 VEGF see vascular endothelial growth factor velocity – encoding (VENC)  120, 809 – sensor  1271 velocity-sensitized acquisition  120 velocity encoding (VENC)  120 venetian blind artifact  116 venous – angioma  337 – contamination  124 – hypertension  335, 635 – infarct  327 – ischaemia  630, 632 – malformation  530 – thrombosis  354 – varice  635 – vessel draining  1293 ventricular – septal defect (VSD)  721 – volume  753 ventriculomegaly  215 vertebral artery dissection  325, 328 verumontanum  1028 very small paramagnetic iron oxide (VSPIO)  1395, 1406 vessel-to-background contrast  1233 vestibule  453 vestibulocochlear nerve  454, 474, 478 VHL see von Hippel-Lindau VIBE see volumetric interpolated breathhold examination

video – camera  1278 – projector  1260 view sharing  122 viral – encephalitides  373 – infection, splenomegaly  901 – meningitis  349 – myelitis  649 Virchow-Robin space  239, 364, 652 visceral space (VS)  525 visual – acuity  210 – cortex  1292 visualization  1278, 1279 vitamin B12  650, 651 – deficiency  659 vitreous hemorrhage  436 vocal cord paralysis  504, 507 volume – element  26 – fraction (VF)  135 – ratio (VR)  135 – resonator  86 volumetric interpolated breathhold examination (VIBE)  102, 446, 447, 448, 449, 471 von Hippel-Lindau (VHL)  623 – disease  285, 599, 886, 889 – syndrome  597 – tumor suppressor gene  286 von Meyenburg complexes  865 von Recklinghausen disease of the urinary bladder  1007 voxel  6 VSD see ventricular septal defect VSOP-C184  94 VSPIO see very small paramagnetic iron oxide vulvar cancer  986 W Wada test  1318 Waldeyer’s ring  512 Walker-Warburg syndrome  201 Warthin’s tumor  513, 516 wash-in/wash-out effect  7 water – cooling  79 – excitation  1084 watermelon phantom  1327 watershed  629, 630 – area  323 Waterston Cooley anastomosis  728

Subject Index

waveguide  85 WB-MRI see whole-body MRI weighted image  36 Welander dystrophy  1200 Wernicke’s area  1311, 1316 Wharton’s duct  496, 499 whisker plot  146 white-marker phenomenon  1263 white matter  172, 196, 232 – abnormalities  216 – injury of prematurity  228 – lesions  222 whole-body MRI  848, 1231, 1232, 1235, 1250 – cardiovascular screening  1235, 1250 – cost–benefit relationship  1249 – examination protocol  1252 – multiple phased-array surface coil  1232

– musculoskeletal system  1233 – postprocessing/reading  1235 – radiofrequency (RF) signal  1232 – sequence protocol  1232 – skeletal muscle  1179 – tumor screening  1250 – viscera  1233 Wilms’ tumor (nephroblastoma)  933 wrap around artifact  120 wrist  1141 – ganglia  1143 – inflammatory joint diseases  1141 X X-ray – angiography  752 – contrast agents  7

– fluoroscopy  1285 x–y-plane distortion  1324 xanthoastrocytoma  257 – pleomorphic  255 Xe biosensor  1396, 1397 xenon-129  671 Xenopus laevis embryos  1400 XMR system  1260 Z z-dephaser  1263 Z-score statistic  1300 Zeeman levels  9 Zellweger syndrome  239 zero – crossing  119 – velocity  315 Zollinger-Ellison syndrome  888 zoonosis  885

1511