Neuroimaging Part II [1st Edition] 9780702045387, 9780444534866

Table of contents :
Content:
Series pagePage ii
CopyrightPage iv
Handbook of Clinical Neurology 3rd SeriesPages v-vi
ForewordPage viiMichael J. Aminoff, François Boller, Dick F. Swaab
PrefacePage ixJoseph C. Masdeu, R. Gilberto González
ContributorsPages xi-xiv
Chapter 32 - Functional anatomy of the spinePages 675-688Nikolai Bogduk
Chapter 33 - Neuroimaging of spine tumorsPages 689-706Nandor K. Pinter, Thomas J. Pfiffner, Laszlo L. Mechtler
Chapter 34 - Vascular diseases of the spinePages 707-716Guillaume Saliou, Timo Krings
Chapter 35 - Infections of the spine and spinal cordPages 717-731Majda M. Thurnher, Richard B. Olatunji
Chapter 36 - Imaging of noninfectious inflammatory disorders of the spinal cordPages 733-746Joshua P. Klein
Chapter 37 - Imaging of trauma of the spinePages 747-767Vahe M. Zohrabian, Adam E. Flanders
Chapter 38 - Hereditary and metabolic myelopathiesPages 769-785Peter Hedera
Chapter 39 - The degenerative spine: pattern recognition and guidelines to image interpretationPages 787-808P.M. Parizel, A.J.L. Van Hoyweghen, A. Bali, J. Van Goethem, L. Van Den Hauwe
Chapter 40 - Peripheral nerve imagingPages 811-826Neil G. Simon, Jason Talbott, Cynthia T. Chin, Michel Kliot
Chapter 41 - Magnetic resonance imaging of skeletal muscle diseasePages 827-842Bruce M. Damon, Ke Li, Nathan D. Bryant
Chapter 42 - Muscle ultrasoundPages 843-853Sigrid Pillen, Andrea Boon, Nens Van Alfen
Chapter 43 - Sudden neurologic deficitPages 857-872Marissa Kellogg, Conrad W. Liang, David S. Liebeskind
Chapter 44 - Neuroendocrine disorders: pituitary imagingPages 873-885Alexander Faje, Nicholas A. Tritos, Brooke Swearingen, Anne Klibanski
Chapter 45 - Visual impairmentPages 887-903Carl Ellenberger
Chapter 46 - Vertigo and hearing lossPages 905-921David E. Newman-Toker, Charles C. Della Santina, Ari M. Blitz
Chapter 47 - Imaging of progressive weakness or numbness of central or peripheral originPages 923-937Joshua P. Klein
Chapter 48 - Gait and balance disordersPages 939-955Joseph C. Masdeu
Chapter 49 - Movement disordersPages 957-969A. Jon Stoessl, Martin J. Mckeown
Chapter 50 - Imaging of neurodegenerative cognitive and behavioral disorders: practical considerations for dementia clinical practicePages 971-984Alireza Atri
Chapter 51 - Neuroimaging of epilepsyPages 985-1014Fernando Cendes, William H. Theodore, Benjamin H. Brinkmann, Vlastimil Sulc, Gregory D. Cascino
Chapter 52 - MyelopathyPages 1015-1026B. Oyinkan Marquis, Patrick M. Capone
Chapter 53 - Low back pain, radiculopathyPages 1027-1033Stephen M. Selkirk, Robert Ruff
Chapter 54 - Brain CT and MRI: differential diagnosis of imaging findingsPages 1037-1054Joseph C. Masdeu, Rajan Gadhia, Alireza Faridar
Chapter 55 - Vascular imaging: ultrasoundPages 1055-1064David Rodriguez-Luna, Carlos A. Molina
Chapter 56 - Diffusion tensor imaging and functional MRIPages 1065-1087Massimo Filippi, Federica Agosta
Chapter 57 - Normal developmentPages 1091-1119Nadine Girard, Meriam Koob, Herv Brunel
Chapter 58 - Congenital malformations of the brain and spinePages 1121-1137Prashant Shankar, Carlos Zamora, Mauricio Castillo
Chapter 59 - CNS and spinal tumorsPages 1139-1158Andre D. Furtado, Ashok Panigrahy, Charles R. Fitz
Chapter 60 - Vascular diseasePages 1159-1171Catherine Amlie-Lefond, Dennis Shaw
Chapter 61 - InfectionsPages 1173-1198Jill V. Hunter, Lee Goerner
Chapter 62 - TraumaPages 1199-1220Thierry A.G.M. Huisman, Andrea Poretti
Chapter 63 - Metabolic, endocrine, and other genetic disordersPages 1221-1259Hisham M. Dahmoush, Elias R. Melhem, Arastoo Vossough
Chapter 64 - Hydrocephalus in childrenPages 1261-1273Harold L. Rekate, Ari M. Blitz
Chapter 65 - Indications for the performance of neuroimaging in childrenPages 1275-1290Fenella Jane Kirkham
Chapter 66 - Endovascular treatment of acute ischemic strokePages 1293-1302Thabele Leslie-Mazwi, James Rabinov, Joshua A. Hirsch
Chapter 67 - Endovascular treatment of intracranial aneurysmsPages 1303-1309Orlando Diaz, Leonardo Rangel-Castilla
Chapter 68 - Endovascular treatment of arteriovenous malformationsPages 1311-1317Orlando Diaz, Robert Scranton
Chapter 69 - Postmortem imaging and neuropathologic correlationsPages 1321-1339Jean C. Augustinack, André J.W. van der kouwe
IndexPages I-1-I-30

Citation preview

HANDBOOK OF CLINICAL NEUROLOGY Series Editors

MICHAEL J. AMINOFF, FRANC¸OIS BOLLER, AND DICK F. SWAAB VOLUME 136

AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO

ELSEVIER Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, USA © 2016 Elsevier B.V. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-444-53486-6 For information on all Elsevier publications visit our website at https://www.elsevier.com/

Publisher: Shirley Decker-lucke Acquisition Editor: Mara Conner Editorial Project Manager: Kristi Anderson Production Project Manager: Sujatha Thirugnana Sambandam Designer: Alan Studholme Typeset by SPi Global, India

Handbook of Clinical Neurology 3rd Series Available titles Vol. 79, The human hypothalamus: basic and clinical aspects, Part I, D.F. Swaab, ed. ISBN 9780444513571 Vol. 80, The human hypothalamus: basic and clinical aspects, Part II, D.F. Swaab, ed. ISBN 9780444514905 Vol. 81, Pain, F. Cervero and T.S. Jensen, eds. ISBN 9780444519016 Vol. 82, Motor neurone disorders and related diseases, A.A. Eisen and P.J. Shaw, eds. ISBN 9780444518941 Vol. 83, Parkinson’s disease and related disorders, Part I, W.C. Koller and E. Melamed, eds. ISBN 9780444519009 Vol. 84, Parkinson’s disease and related disorders, Part II, W.C. Koller and E. Melamed, eds. ISBN 9780444528933 Vol. 85, HIV/AIDS and the nervous system, P. Portegies and J. Berger, eds. ISBN 9780444520104 Vol. 86, Myopathies, F.L. Mastaglia and D. Hilton Jones, eds. ISBN 9780444518996 Vol. 87, Malformations of the nervous system, H.B. Sarnat and P. Curatolo, eds. ISBN 9780444518965 Vol. 88, Neuropsychology and behavioural neurology, G. Goldenberg and B.C. Miller, eds. ISBN 9780444518972 Vol. 89, Dementias, C. Duyckaerts and I. Litvan, eds. ISBN 9780444518989 Vol. 90, Disorders of consciousness, G.B. Young and E.F.M. Wijdicks, eds. ISBN 9780444518958 Vol. 91, Neuromuscular junction disorders, A.G. Engel, ed. ISBN 9780444520081 Vol. 92, Stroke – Part I: Basic and epidemiological aspects, M. Fisher, ed. ISBN 9780444520036 Vol. 93, Stroke – Part II: Clinical manifestations and pathogenesis, M. Fisher, ed. ISBN 9780444520043 Vol. 94, Stroke – Part III: Investigations and management, M. Fisher, ed. ISBN 9780444520050 Vol. 95, History of neurology, S. Finger, F. Boller and K.L. Tyler, eds. ISBN 9780444520081 Vol. 96, Bacterial infections of the central nervous system, K.L. Roos and A.R. Tunkel, eds. ISBN 9780444520159 Vol. 97, Headache, G. Nappi and M.A. Moskowitz, eds. ISBN 9780444521392 Vol. 98, Sleep disorders Part I, P. Montagna and S. Chokroverty, eds. ISBN 9780444520067 Vol. 99, Sleep disorders Part II, P. Montagna and S. Chokroverty, eds. ISBN 9780444520074 Vol. 100, Hyperkinetic movement disorders, W.J. Weiner and E. Tolosa, eds. ISBN 9780444520142 Vol. 101, Muscular dystrophies, A. Amato and R.C. Griggs, eds. ISBN 9780080450315 Vol. 102, Neuro-ophthalmology, C. Kennard and R.J. Leigh, eds. ISBN 9780444529039 Vol. 103, Ataxic disorders, S.H. Subramony and A. Durr, eds. ISBN 9780444518927 Vol. 104, Neuro-oncology Part I, W. Grisold and R. Sofietti, eds. ISBN 9780444521385 Vol. 105, Neuro-oncology Part II, W. Grisold and R. Sofietti, eds. ISBN 9780444535023 Vol. 106, Neurobiology of psychiatric disorders, T. Schlaepfer and C.B. Nemeroff, eds. ISBN 9780444520029 Vol. 107, Epilepsy Part I, H. Stefan and W.H. Theodore, eds. ISBN 9780444528988 Vol. 108, Epilepsy Part II, H. Stefan and W.H. Theodore, eds. ISBN 9780444528995 Vol. 109, Spinal cord injury, J. Verhaagen and J.W. McDonald III, eds. ISBN 9780444521378 Vol. 110, Neurological rehabilitation, M. Barnes and D.C. Good, eds. ISBN 9780444529015 Vol. 111, Pediatric neurology Part I, O. Dulac, M. Lassonde and H.B. Sarnat, eds. ISBN 9780444528919 Vol. 112, Pediatric neurology Part II, O. Dulac, M. Lassonde and H.B. Sarnat, eds. ISBN 9780444529107 Vol. 113, Pediatric neurology Part III, O. Dulac, M. Lassonde and H.B. Sarnat, eds. ISBN 9780444595652 Vol. 114, Neuroparasitology and tropical neurology, H.H. Garcia, H.B. Tanowitz and O.H. Del Brutto, eds. ISBN 9780444534903 Vol. 115, Peripheral nerve disorders, G. Said and C. Krarup, eds. ISBN 9780444529022 Vol. 116, Brain stimulation, A.M. Lozano and M. Hallett, eds. ISBN 9780444534972 Vol. 117, Autonomic nervous system, R.M. Buijs and D.F. Swaab, eds. ISBN 9780444534910 Vol. 118, Ethical and legal issues in neurology, J.L. Bernat and H.R. Beresford, eds. ISBN 9780444535016 Vol. 119, Neurologic aspects of systemic disease Part I, J. Biller and J.M. Ferro, eds. ISBN 9780702040863 Vol. 120, Neurologic aspects of systemic disease Part II, J. Biller and J.M. Ferro, eds. ISBN 9780702040870 Vol. 121, Neurologic aspects of systemic disease Part III, J. Biller and J.M. Ferro, eds. ISBN 9780702040887 Vol. 122, Multiple sclerosis and related disorders, D.S. Goodin, ed. ISBN 9780444520012 Vol. 123, Neurovirology, A.C. Tselis and J. Booss, eds. ISBN 9780444534880

vi

AVAILABLE TITLES (Continued)

Vol. 124, Clinical neuroendocrinology, E. Fliers, M. Korbonits and J.A. Romijn, eds. ISBN 9780444596024 Vol. 125, Alcohol and the nervous system, E.V. Sullivan and A. Pfefferbaum, eds. ISBN 9780444626196 Vol. 126, Diabetes and the nervous system, D.W. Zochodne and R.A. Malik, eds. ISBN 9780444534804 Vol. 127, Traumatic brain injury Part I, J.H. Grafman and A.M. Salazar, eds. ISBN 9780444528926 Vol. 128, Traumatic brain injury Part II, J.H. Grafman and A.M. Salazar, eds. ISBN 9780444635211 Vol. 129, The human auditory system: Fundamental organization and clinical disorders, G.G. Celesia and G. Hickok, eds. ISBN 9780444626301 Vol. 130, Neurology of sexual and bladder disorders, D.B. Vodusˇek and F. Boller, eds. ISBN 9780444632470 Vol. 131, Occupational neurology, M. Lotti and M.L. Bleecker, eds. ISBN 9780444626271 Vol. 132, Neurocutaneous syndromes, M.P. Islam and E.S. Roach, eds. ISBN 9780444627025 Vol. 133, Autoimmune neurology, S.J. Pittock and A. Vincent, eds. ISBN 9780444634320 Vol. 134, Gliomas, M.S. Berger and M. Weller, eds. ISBN 9780128029978 Vol. 135, Neuroimaging Part I, J.C. Masdeu and R.G. Gonza´lez, eds. ISBN 9780444534859

Foreword

We are proud to present the first volumes dedicated to neuroimaging in the Handbook of Clinical Neurology series. Neurologists, not just those in training, may wonder at the kind of medical world that existed before modern imaging. Indeed, the neuroscience community has since its beginning attempted to understand the human mind and brain through imaging. As far back as 1880, the Italian physiologist Angelo Mosso introduced the “human circulation balance,” which could noninvasively measure the redistribution of blood during emotional and intellectual activity. More recently, semi-invasive techniques such as pneumoencephalography (introduced by Dandy in 1918) and arteriography (pioneered by Moniz in 1927) allowed partial visualization of the brain and its surrounding structures. New methods enabling easier, safer, noninvasive, painless, and repeatable imaging have only been developed in the past 50 years or so, starting with computed tomography, some of whose developers won the 1979 Nobel Prize for medicine or physiology. The many subsequent developments in neuroimaging are masterfully presented in these two volumes. The volumes deal with a variety of neuroimaging-related topics. They include thorough descriptions of the involved methods and their application to specific diseases of the brain, spinal cord, and peripheral nervous system. Emphasis is given to the most common disorders, such as tumors, strokes, multiple sclerosis, movement disorders, infections, dementia, and trauma, but less common conditions such as neurocutaneous syndromes are also discussed. The important questions of when and where to image, as well as the differential diagnosis of imaging findings, are discussed in the light of specific syndromes. A separate section covers pediatric neuroimaging. The volumes conclude with sections dedicated to interventional neuroimaging as well as to postmortem imaging and neuropathologic correlations. We have been fortunate to have as volume editors two distinguished scholars, Dr. Joseph C. Masdeu, of the Department of Neurology, Methodist Hospital, Houston, Texas, and Dr. R. Gilberto Gonza´lez, from the Department of Radiology, Massachusetts General Hospital in Boston. Both have been at the forefront of neuroimaging research for many years. They have assembled a truly international group of authors with acknowledged expertise to contribute to the texts and have produced two authoritative, comprehensive, and up-to-date volumes. Their availability electronically on Elsevier’s Science Direct site as well as in print format should ensure their ready accessibility and facilitate searches for specific information. We are grateful to the volume editors and to all the contributors for their efforts in creating such an invaluable resource. As series editors we read and commented on each of the chapters with great interest. We are therefore confident that both clinicians and researchers in many different medical disciplines will find much in these volumes to appeal to them. And last, but not least, it is always a pleasure to acknowledge and thank Elsevier, our publisher – and, in particular, Michael Parkinson in Scotland, and Mara Conner and Kristi Anderson in San Diego – for their unfailing and expert assistance in the development and production of these volumes. Michael J. Aminoff Franc¸ois Boller Dick F. Swaab

Preface

Neuroimaging has become one of the most useful set of tools for understanding and diagnosing diseases of the nervous system. Fittingly, the present two volumes of the Handbook of Clinical Neurology review the extensive advances in the field. In the first volume, discussions of the various techniques used in neuroimaging are followed by reviews of the imaging findings caused by brain diseases. We have chosen not to include a chapter on brain anatomy because it would be quite long and extant atlases are excellent. The second volume begins with a description of the functional anatomy of the spine and of the imaging findings in diseases of the spine and spinal cord. Imaging of peripheral nerve and muscle follows. Then, there is a section on when and how to image the various clinical syndromes produced by diseases of the nervous system. Adequacy in the use of expensive neuroimaging tools has always been a priority, but it is becoming more acute as the application of neuroimaging expands more rapidly than the available resources. The next section is unusual in a book of this type: a description of the various imaging findings that should lead to consideration of the diseases causing them. Such information is particularly important when using techniques like computed tomography and magnetic resonance imaging, which offer a panoply of findings and are extensively used in clinical practice. Next is a section on pediatric neuroimaging, led by a chapter on imaging findings during normal development. After three chapters on the therapeutic use of endovascular imaging, the second volume concludes with a chapter on postmortem imaging as a tool to better define normal brain structure on imaging and its alteration by some disorders. We hope that this book will be useful to all those who work with clinical imaging of the nervous system, such as neurologists, neuroradiologists, neurosurgeons, and nuclear medicine physicians. Some sections, for instance, the sections on the spine, peripheral nerve, and muscle, may be helpful to orthopedic surgeons and rehabilitation specialists. Neuropsychologists may find helpful the chapters on neurodegenerative disorders leading to cognitive impairment. Neuroimaging is used not only clinically, but also by those interested in clarifying the still largely undiscovered landscape and functional intricacy of the brain. While the clinical literature of neuroimaging is extensive, even more extensive and more widely cited is the literature of neuroimaging applied to the study of the healthy human nervous system. Although human disease has traditionally led to a better understanding of normal structure and function, researchers looking primarily for information on the normal nervous system should look elsewhere. We are most thankful to the authors, who have distilled their expertise in superbly written and illustrated chapters. Mr. Michael Parkinson, from Elsevier, has skillfully coordinated the gathering of information for these two volumes. We are also thankful to the three series editors and, particularly, to Dr. Franc¸ois Boller, for their excellent suggestions. Joseph C. Masdeu R. Gilberto Gonza´lez

Contributors

F. Agosta Neuroimaging Research Unit, Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, Italy C. Amlie-Lefond Department of Neurology, Seattle Children’s Hospital, Seattle, WA, USA A. Atri Ray Dolby Brain Health Center, California Pacific Medical Center Research Institute, Sutter Health, San Francisco, CA, USA J.C. Augustinack Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, MA, USA A. Bali Department of Radiology, Antwerp University Hospital and University of Antwerp, Antwerp, Belgium A.M. Blitz Neuro-radiology Division, Johns Hopkins University School of Medicine, Baltimore, MD, USA N. Bogduk Newcastle Bone and Joint Institute, University of Newcastle, Newcastle, Australia A. Boon Department of Physical Medicine and Rehabilitation and Department of Neurology, Mayo Clinic, Rochester, MN, USA B.H. Brinkmann Division of Epilepsy, Department of Neurology, Mayo Clinic, Rochester, MN, USA

H. Brunel Neuroradiology Service, H^opital la Timone, Marseille, France N.D. Bryant Vanderbilt University Institute of Imaging Science and the Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, USA P.M. Capone Medical Imaging and Neurology, Winchester Medical Center, Winchester and Department of Neurology, Virginia Commonwealth University, Richmond, VA, USA G.D. Cascino Division of Epilepsy, Department of Neurology, Mayo Clinic, Rochester, MN, USA M. Castillo Division of Neuroradiology, Department of Radiology, University of North Carolina, Chapel Hill, NC, USA F. Cendes University of Campinas, Department of Neurology, Campinas, SP, Brazil C.T. Chin Department of Radiology, University of California, San Francisco, CA, USA H.M. Dahmoush Department of Radiology, Children’s Hospital of Philadelphia and University of Pennsylvania, Philadelphia, PA, USA B.M. Damon Vanderbilt University Institute of Imaging Science and the Department of Radiology and Radiological Sciences, Departments of Biomedical Engineering and Molecular Physiology and Biophysics, Vanderbilt University, Nashville, TN, USA

xii

CONTRIBUTORS

C.C. Della Santina Department of Otolaryngology, Head and Neck Surgery, Johns Hopkins University School of Medicine, Baltimore, MD, USA O. Diaz Neurovascular Center, Methodist Hospital, Houston, TX, USA C. Ellenberger Lebanon Magnetic Imaging, Lebanon, Pennsylvania, USA A. Faje Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA A. Faridar Department of Neurology, Houston Methodist Hospital, Houston, TX, USA M. Filippi Neuroimaging Research Unit, Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, Italy

P. Hedera Department of Neurology, Vanderbilt University, Nashville, TN, USA J.A. Hirsch Neuroendovascular Service, Massachusetts General Hospital, Boston, MA, USA T.A.G.M. Huisman Division of Pediatric Radiology, Johns Hopkins Hospital, Baltimore, MD, USA J.V. Hunter Department of Pediatric Radiology, Texas Children’s Hospital and Baylor College of Medicine, Houston, TX, USA M. Kellogg Department of Neurology, Oregon Health and Science University, Portland, OR, USA F.J. Kirkham Neurosciences Unit, UCL Institute of Child Health, London; Southampton Children’s Hospital and Clinical and Experimental Sciences, University of Southampton, Southampton, UK

C.R. Fitz Department of Radiology, Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA

J.P. Klein Departments of Neurology and Radiology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA

A.E. Flanders Department of Radiology, Thomas Jefferson University Hospital, Philadelphia, PA, USA

A. Klibanski Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

A.D. Furtado Department of Radiology, Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA

M. Kliot Department of Neurological Surgery, Northwestern Feinberg School of Medicine, Chicago, IL, USA

R. Gadhia Department of Neurology, Houston Methodist Hospital, Houston, TX, USA

M. Koob Pediatric Radiology Imaging Service, Centre Hospitalier Universitaire de Strasbourg, H^opital de Hautepierre and Laboratoire ICube, Universite de Strasbourg-CNRS, Strasbourg, France

N. Girard Neuroradiology Service, H^ opital la Timone and Aix Marseille Universite, Marseille, France L. Goerner The Radiology Group, Honolulu, HI, USA

T. Krings Division of Neuroradiology and Neurosurgery, University of Toronto, Toronto Western Hospital and University Health Network, Toronto, ON, Canada

R.G. Gonza´lez Department of Radiology, Massachusetts General Hospital, Boston, MA, USA

T. Leslie-Mazwi Neuroendovascular Service, Massachusetts General Hospital, Boston, MA, USA

CONTRIBUTORS xiii K. Li P.M. Parizel Vanderbilt University Institute of Imaging Science and Department of Radiology, Antwerp University Hospital the Department of Radiology and Radiological Sciences, and University of Antwerp, Antwerp, Belgium Vanderbilt University, Nashville, TN, USA T.J. Pfiffner Dent Neurologic Institute, Amherst, NY, USA C.W. Liang Neurovascular Imaging Research Core and Department S. Pillen of Neurology, University of California Los Angeles, Department of Sleep Medicine, Kempenhaeghe Los Angeles, CA, USA Expertise Center for Epileptology, Sleep Medicine and Neurocognition, Heeze, The Netherlands D.S. Liebeskind Neurovascular Imaging Research Core and Department of Neurology, University of California Los Angeles, Los Angeles, CA, USA B.O. Marquis Division of Epilepsy, Department of Neurology, State University of New York Downstate Medical Center, Brooklyn, NY, USA J.C. Masdeu Department of Neurology, Houston Methodist Hospital, Houston, TX, USA M.J. McKeown Pacific Parkinson’s Research Centre and Division of Neurology, Djavad Mowafaghian Centre for Brain Health, University of British Columbia and Vancouver Coastal Health, Vancouver, BC, Canada L.L. Mechtler Dent Neurologic Institute, Amherst, NY, USA

N.K. Pinter Dent Neurologic Institute, Amherst, NY, USA A. Poretti Division of Pediatric Radiology, Johns Hopkins Hospital, Baltimore, MD, USA J. Rabinov Neuroendovascular Service, Massachusetts General Hospital, Boston, MA, USA L. Rangel-Castilla Neuroendovascular Surgery, University of Buffalo Neurosurgery, State University of New York, Buffalo, NY, USA H.L. Rekate Department of Neurosurgery, Hofstra Northshore LIJ School of Medicine, Manhasset, and The Chiari Institute, Great Neck, NY, USA

E.R. Melhem Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland, Baltimore, MD, USA

D. Rodriguez-Luna Stroke Unit, Department of Neurology, Vall d’Hebron University Hospital, Vall d’Hebron Research Institute, Autonomous University of Barcelona, Barcelona, Spain

C.A. Molina Stroke Unit, Department of Neurology, Vall d’Hebron University Hospital, Vall d’Hebron Research Institute, Autonomous University of Barcelona, Barcelona, Spain

R. Ruff Department of Neurology, Case Western Reserve University School of Medicine, Cleveland, OH, USA

D.E. Newman-Toker Department of Neurology and Department of Otolaryngology, Head and Neck Surgery, Johns Hopkins University School of Medicine, Baltimore, MD, USA R.B. Olatunji Department of Radiology, University Hospital Vienna, Vienna, Austria A. Panigrahy Department of Radiology, Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA

G. Saliou Neuroradiology Service, Centre Hospitalier Universitaire Bic^etre, Le Kremlin Bic^etre, France R. Scranton Department of Neurosurgery, Methodist Hospital, Houston, TX, USA S.M. Selkirk Spinal Cord Injury Division, Louis Stokes Cleveland Veterans Affairs Medical Center and Department of Neurology, Case Western Reserve University School of Medicine, Cleveland, OH, USA

xiv

CONTRIBUTORS

P. Shankar Division of Neuroradiology, Department of Radiology, University of North Carolina, Chapel Hill, NC, USA D. Shaw Department of Radiology, Seattle Children’s Hospital, Seattle, WA, USA N.G. Simon St Vincent’s Clinical School, University of New South Wales, Sydney, Australia A.J. Stoessl Pacific Parkinson’s Research Centre and Division of Neurology, Djavad Mowafaghian Centre for Brain Health, University of British Columbia and Vancouver Coastal Health, Vancouver, BC, Canada V. Sulc Department of Neurology, 2nd Faculty of Medicine, Charles University in Prague and Motol University Hospital, Czech Republic B. Swearingen Department of Neurosurgery, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

N.A. Tritos Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA N. van Alfen Department of Neurology, Radboud University Medical Center, Nijmegen, The Netherlands L. van den Hauwe Department of Radiology, Antwerp University Hospital and University of Antwerp, Antwerp, Belgium A.J.W. van der Kouwe Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, MA, USA J. Van Goethem Department of Radiology, Antwerp University Hospital and University of Antwerp, Antwerp, Belgium A.J.L. Van Hoyweghen Department of Radiology, Antwerp University Hospital and University of Antwerp, Antwerp, Belgium

J. Talbott Department of Radiology, University of California, San Francisco, CA, USA

A. Vossough Department of Radiology, Children’s Hospital of Philadelphia and University of Pennsylvania, Philadelphia, PA, USA

W.H. Theodore National Institute of Neurological Disorders and Stroke, Bethesda, MD, USA

C. Zamora Division of Neuroradiology, Department of Radiology, University of North Carolina, Chapel Hill, NC, USA

M.M. Thurnher Department of Radiology, University Hospital Vienna, Vienna, Austria

V.M. Zohrabian Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, USA

Handbook of Clinical Neurology, Vol. 136 (3rd series) Neuroimaging, Part II J.C. Masdeu and R.G. Gonza´lez, Editors © 2016 Elsevier B.V. All rights reserved

Chapter 32

Functional anatomy of the spine NIKOLAI BOGDUK* Newcastle Bone and Joint Institute, University of Newcastle, Newcastle, Australia

Abstract Among other important features of the functional anatomy of the spine, described in this chapter, is the remarkable difference between the design and function of the cervical spine and that of the lumbar spine. In the cervical spine, the atlas serves to transmit the load of the head to the typical cervical vertebrae. The axis adapts the suboccipital region to the typical cervical spine. In cervical intervertebrtal discs the anulus fibrosus is not circumferential but is crescentic, and serves as an interosseous ligament in the saddle joint between vertebral bodies. Cervical vertebrae rotate and translate in the sagittal plane, and rotate in the manner of an inverted cone, across an oblique coronal plane. The cervical zygapophysial joints are the most common source of chronic neck pain. By contrast, lumbar discs are well designed to sustain compression loads, but rely on posterior elements to limit axial rotation. Internal disc disruption is the most common basis for chronic low-back pain. Spinal muscles are arranged systematically in prevertebral and postvertebral groups. The intrinsic elements of the spine are innervated by the dorsal rami of the spinal nerves, and by the sinuvertebral nerves. Little modern research has been conducted into the structure of the thoracic spine, or the causes of thoracic spinal pain.

INTRODUCTION

CERVICAL SPINE

In writing a chapter on anatomy for neurologists the risk arises of being arcane or banal. Neurologists will already be familiar with the precepts of classic anatomy, and would not be inclined to read a chapter that repeats boring, undergraduate material. For these reasons, the present chapter has been cast in a different manner. Although conventional elements of anatomy are reprised, they are permeated by several themes. New facts are provided, stemming from modern research into the structure of the spine, along with new perceptions about design and function. Throughout, the focus is on clinical relevance, particularly with respect to the mechanisms of spinal injury and spinal pain. In that regard, certain structures – ignored in conventional undergraduate curricula – are promoted to epidemiologically significant, clinical importance.

The cervical spine serves as a mobile support for the sensory platform of the head. It allows the sensory apparatus for vision, hearing, and smell to be elevated or depressed in the sagittal plane, and to scan the environment in the horizontal plane. In order to subserve these functions, the cervical spine has to be mobile, yet sufficiently strong to support the weight of the head. Its vulnerability, to either minor or major injuries, lies in being long, slender, and carrying the large mass of the head at its summit. Both for descriptive purposes and functionally, the cervical spine can be divided into three zones: the suboccipital zone, centered on the C1 vertebra; a transitional zone formed by the C2 vertebra; and the typical zone, encompassing the C–7 vertebrae (Bogduk and Mercer, 2000) (Fig. 32.1). These zones differ both in structure and in function.

*Correspondence to: Nikolai Bogduk, PO Box 128, The Junction, New South Wales 2291, Australia. E-mail: nbogduk@bigpond. net.au

676

N. BOGDUK

Fig. 32.1. Sagittal magnetic resonance images of the cervical spine, showing its structure and zones. (A) Median scan, showing the vertebral bodies and interverterbral discs. The white dots mark the mean location of the axes of rotation for flexion-extension of the vertebra above. The odontoid process (op) projects rostrally from the body of C2, to lie behind the anterior arch of the atlas (C1). (B) Lateral scan, showing the occipital condyle (oc), the lateral mass (lm) of the atlas (C1), the articular pillars (ap), and the zygapophysial joints (zj) that they form, at the segments labeled. (Courtesy of Dr. Tim Maus, Mayo Clinic, Rochester, MN.)

Suboccipital zone

Upper transition zone

The C1 vertebra (the atlas) shares none of the features of typical cervical vertebrae, and should never have been considered cervical. In structure and in function it is more like an occipital vertebra. In structure it resembles the occipital bone, as can be seen in axial scans. In function, it more closely operates with the head, rather than with the remainder of the cervical spine. The classic description of the atlas as a ring vertebra belies its design and function. The critical components of the atlas are its two lateral masses (Fig. 32.2). Superiorly, these present superior articular processes that receive the occipital condyles, and thereby cradle the skull. Inferiorly, the lateral masses present inferior articular processes that rest on the C2 vertebra, and thereby transmit the load of the head to the remainder of the cervical spine. The anterior and posterior arches of the atlas serve little function other than holding the two lateral masses both apart and together, while the latter do the mechanical work of the atlas. Upon receiving the occipital condyles into their deep sockets, the superior articular processes of each lateral mass form the atlanto-occipital joints (Figs 32.2 and 32.3). These synovial joints constitute the only direct connection between the skull and C1. They allow a small range of flexion-extension, but the depth of their sockets precludes axial rotation. Therefore, as the head rotates (in the transverse plane) the atlas is obliged to move with it. In that respect, the atlas behaves like a passive washer, between the skull and C2.

The upper half of the C2 vertebra (the axis) is designed to support the atlas. Superiorly and laterally, it presents superior articular processes that slope caudally and laterally, and act like sloping shoulders on which the lateral masses of the atlas rest (Fig. 32.2). The inferior articular processes of the atlas have a reciprocal, caudal and lateral slope. The apposed articular processes on each side form the lateral atlantoaxial joint (Figs 32.2 and 32.3). The caudolateral slope of the lateral atlantoaxial joint helps stabilize the atlas in the coronal plane, but also underlies the mechanism of Jefferson fractures. Severe axial loads, applied to the skull, will drive the atlas caudally, but its lateral masses will also spread laterally down the lateral slope of the lateral atlantoaxial joints, resulting in burst fractures of the anterior and posterior arches. Centrally, the axis presents a long odontoid process (the dens) that projects behind the anterior arch of the atlas, with which it forms the median atlantoaxial joint (Figs 32.1, 32.3, and 32.4). The anterior arch is held against the odontoid process by the transverse ligament, which spans like a belt between the two lateral masses of the atlas, behind the odontoid process (Figs 32.4 and 32.5). Posterior displacement of the atlas is prevented by impaction of the anterior arch against the odontoid process, at the median atlantoaxial joint. Anterior displacement is prevented by tension in the transverse ligament (Fielding et al., 1974). The ligament allows up to 3 mm normal range of separation between the odontoid

FUNCTIONAL ANATOMY OF THE SPINE

677

Fig. 32.2. Coronal magnetic resonance images of the cervical spine, showing the structure of its components. (A) Anterior scan, showing the occipital condyles (oc) resting in the sockets of the lateral masses (lm) of the atlas, and forming the atlanto-occipital joints (aoj); and the lateral masses bracketing the odontoid process (op), and resting on the “shoulders” of the axis (C2), where they form the lateral atlantoaxial joints (laaj). The vertebral bodies of C2–7 form the anterior column of the cervical spine. (B) Posterior scan, through the synovial joints of the cervical spine. Note the wedge shape of the lateral mass (lm) between the atlanto-occipital joint (aoj) and the lateral atlantoaxial joint (laaj). The zygapophysial joint at C2–3 slopes caudally and medially, but those at successive levels are essentially horizontal. The dotted line illustrates the ellipsoid shape depicted by the C2–3 zygapophysial joints and the C2–3 disc, into which the atlas (C2) nestles on to the typical cervical spine.

process and the anterior arch in adults, and 5 mm in children. In the past, the magnitude of the interval between the anterior arch and the odontoid process has been used as a measure of atlantoaxial instability, but as a predictor of neurologic compromise the posterior atlantodental interval (Fig. 32.3) has greater sensitivity and specificity (Wasserman et al., 2011).

Severe forces delivered anteriorly to the head can fracture the odontoid process. Such fractures threaten the sagittal stability of the atlas. In turn, anterior or posterior displacement of the atlas can threaten the spinal cord. Rheumatoid arthritis of the atlantoaxial joints can weaken the transverse ligament of the atlas, resulting in anterior subluxation of the atlas (Wasserman et al., 2011).

Fig. 32.3. Close-up views of sagittal magnetic resonance images of the suboccipital joints. (A) Median scan through the odontoid process (op) and vertebral body of C2. With the front of the odontoid process, the anterior arch (aa) of the atlas forms the median atlantoaxial joint (maaj). The transverse arrow marks the posterior atlantodental (pa) interval. (B) Lateral scan through the lateral mass (lm) of the atlas. With the superior sockets of the lateral mass of the atlas, the occipital condyle (oc) forms the atlanto-occipital joint (aoj). With C2, the lateral mass forms the lateral atlantoaxial joint (laaj). Note the bi-convex shape of the lateral atlantoaxial joint. The triangular, white signals anteriorly and posteriorly within the joint are the fibroadipose meniscoids that it contains. (Courtesy of Dr. Tim Maus, Mayo Clinic, Rochester MN.)

678

N. BOGDUK

Fig. 32.4. A sketch of an axial (top) view of the atlas sitting on the axis. The lateral masses of the atlas bracket the odontoid process (op) of the axis, and are themselves joined by the anterior (aa) and posterior (pa) arches of the atlas. With the anterior arch, the odontoid process forms the median atlantoaxial joint (maaj). The transverse ligament (tl) spans like a belt behind the odontoid processes, between the two lateral masses.

Although the osseous articular processes of the lateral atlantoaxial joint are flat, their articular cartilages are convex (along the sagittal plane) (Koebke and Brade, 1982) (Fig. 32.3). As a result, the atlas perches somewhat precariously on the axis, with its convex inferior articular cartilages balancing on the convexities of the superior articular cartilages of the axis. The spaces anteriorly and posteriorly between the convex cartilages are filled by wedge-shaped fibrocartilaginous meniscoids (Mercer and Bogduk, 1993). Although a small degree of flexion and extension is possible between the atlas and the axis (Bogduk and Mercer, 2000), the cardinal movement between these two vertebrae is axial rotation. During this movement the atlas pivots at the median atlantoaxial joint, while its lateral masses slide backwards or forwards, circumferentially, at the lateral atlantoaxial joints. However, because of the convexity of the articular cartilages in these joints, the lateral masses also descend, down the posterior or anterior slope of the cartilages, as they move backwards or forwards, respectively. As a result, the atlas settles (lowers or screws down) during axial rotation, and rises when the movement is reversed (Bogduk and Mercer, 2000). During these displacements, the meniscoids of the joints cover the exposed surfaces of the subluxating articular cartilages. The total range of axial rotation of the atlas is considerably large. It has been measured as 43  5.5° which effectively amounts to 50% of the range of axial rotation of the head and neck (Bogduk and Mercer, 2000). At the extremes of this range, very little of the articular

Fig. 32.5. A sketch of the suboccipital joints and ligaments, as view from behind with the posterior arch of the atlas resected. The lateral mass (lm) of the atlas supports the occipital bone, and rests of the axis, forming the atlanto-occipital joint (aoj) above, and the lateral atlantoaxial joint (laaj) below. The transverse ligament (tl) holds the atlas (al) against the odontoid process (op). The alar ligaments bind the odontoid process to the margins of the foramen magnum, thereby connecting the skull to C2 but bypassing C1.

cartilages of the lateral atlantoaxial joints remain opposed; the joint is almost dislocated. In order to accommodate this large range of motion, the capsules of the lateral atlantoaxial joint are loose, and serve little to hold the atlas on the axis. That service is provided by the alar ligaments. On each side, these ligaments pass essentially transversely from the upper end of the odontoid process to the margin of the foramen magnum (Fig. 32.5). In doing so, they bypass the atlas, and lock the head into place on the axis, effectively clamping the atlas between the skull and C2. The alar ligaments are the cardinal restraint to axial rotation of the head (Dvorak et al., 1987). They are sufficiently strong to prevent anterior dislocation of the head even if the transverse ligament is completely severed (Fielding et al., 1974). Disruption of an alar ligament can result in rotatory instability of the head and atlas (Dvorak et al., 1987). Excessive axial rotation of the atlas can result in a lateral mass dislocating at the lateral atlantoaxial joint, causing fixed atlantoaxial deformity (Wortzman and Dewar, 1968). A less dramatic form of torticollis can arise if, after rotation of the head and atlas, a meniscoid of the lateral atlantoaxial joint fails to re-enter the joint space, catches under the capsule of the joint, and acts like a loose body to prevent derotation of the joint (Mercer and Bogduk, 1993).

FUNCTIONAL ANATOMY OF THE SPINE

Fig. 32.6. A sketch of a coronal view of how forces from the head are transmitted into the cervical spine. On each side, the weight of the head passes through the occipital condyle (oc), into the lateral mass (lm) of the atlas, and into the axis (C2) through the lateral atlantoaxial joint. From there, the forces diverge, partly into the posterior column of zygapophysial joints, and partly into the anterior column of vertebral bodies and discs. Half the load passes anteriorly and half posteriorly.

Increasing interest has been focused on the lateral atlantoaxial joints as a possible source of cervicogenic headache. This contention can be tested by controlled, intra-articular, diagnostic blocks of the putatively painful joint (Bogduk and Bartsch, 2008; Bogduk and Govind, 2009; Bogduk, 2014).

Lower transition zone The lower half of the C2 vertebra has the structure of a typical cervical vertebra (Figs 32.1 and 32.2). Centrally it presents a vertebral body, and laterally it presents paired inferior articular processes. Having received the lateral masses of the atlas, the axis transmits the load of the head along an anterior channel, through its vertebral body to the vertebral bodies below, and along paired posterior channels, through the zygapophysial joints (Fig. 32.6). Approximately half of the axial load is transmitted through the anterior channel, and half through the two posterior channels.

Typical cervical vertebrae The cardinal elements of a typical cervical vertebra are its vertebral body and two articular pillars (Figs 32.1

679

and 32.2). Secondarily, transverse processes project laterally from the articular pillars, and posteriorly the two pillars are united by a pair of laminae, which support a midline spinous process at their junction. The transverse processes and spinous processes serve as levers upon which act the muscles that control the position of the cervical vertebrae. Along its superior, posterolateral margin on each side, each vertebral body bears uncinate processes. Previously enigmatic, the uncinate processes underlie the nature of the joints between the cervical vertebral bodies and how they operate. Consecutive articular pillars are united by the zygapophysial joints (Figs 32.1 and 32.2), which are synovial joints formed by the inferior articular process of the vertebra above and the superior articular process of the vertebra below. Fibroadipose meniscoids intervene between the articular cartilages of these joints (Mercer and Bogduk, 1993). The zygapophysial joints are planar, and at typical cervical levels are oriented at about 40° to the coronal and transverse planes, so that they face backwards and upwards (Nowitzke et al., 1994). At the C2–3 level, however, the joints also face medially, such that the pair of joints depict an ellipsoid socket into which nestles the weight of the axis, and the load that it carries from the head (Figs 32.2 and 32.6). Consecutive vertebral bodies are united by intervertebral discs, and by the anterior and posterior longitudinal ligaments (Mercer and Bogduk, 1999). The anterior ligament connects only the typical cervical vertebrae, from C2 caudally. The posterior longitudinal ligament forms a carpet along the floor of the vertebral canal at typical cervical levels, but expands into the membrana tectoria to cover the back of the atlantoaxial region. In doing so, the ligament separates the dural sac and spinal cord from the mechanics of the median atlantoaxial joint. Posterior ligaments are lacking in the cervical spine. Interspinous ligaments are represented by only a sagittal layer of fascia (Mercer and Bogduk, 2003). The ligamentum nuchae lacks the structure of a ligament. It consists largely of a narrow, coronal raphe, anchored to the C7 spinous process, and formed by interlacing tendons of the splenius muscles and trapezius (Mercer and Bogduk, 2003). The intrinsic structure of the cervical intervertebral discs is unlike that of lumbar discs, and differs with age (Oda et al., 1988; Mercer and Bogduk, 1999). The nucleus pulposus of cervical discs is gelatinous only in children and young adults. By the age of 30 it dries out to form a fibrocartilaginous plate (Oda et al., 1988). Moreover, the nucleus is not surrounded by concentric lamellae of the anulus fibrosus (Mercer and Bogduk, 1999). The anulus fibrosus is largely deficient posteriorly, and consists of a thin, paramedian band of collagen fibers that run longitudinally between the

680

N. BOGDUK

Fig. 32.7. Sketches of various views of the internal structure of the cervical disc. In a front view, all fibers of the anterior anulus fibrosus pass towards a point on the inferior anterior surface of the vertebral body above. In a top view, the anulus fibrosus (af) is crescentic in shape, thick anteriorly but tapering at the uncinate processes (u). The nucleus pulposus (np) is a fibrocartilaginous plate. Posteriorly the anulus is restricted to a small bundle of paramedian, longitudinal fibers. A side view shows the fibers of the anulus fibrosus passing upwards and forwards. A transverse cleft runs from one uncinate process to the other.

vertebral bodies (Fig. 32.7). Posterolaterally, the nucleus is covered by the posterior longitudinal ligament, rather than by anulus fibrosus. Anteriorly, the anulus fibrosus is crescentic in shape, thin posteriorly near the uncinate processes, but thicker anteriorly towards the midline. All of its collagen fibers pass in a similar direction, effectively aiming to a median point on the lower anterior surface of the vertebral body above. This configuration endows the anulus fibrosus with the structure of a thick interosseous ligament that binds the anterior edges of consecutive vertebral bodies. The superior surface of each cervical vertebral body presents two curvatures: a slight convex curvature along the sagittal plane, and a deep concave curvature transversely between the uncinate processes (Fig. 32.8). These curvatures endow the vertebral body with the configuration of a saddle joint (Bogduk and Mercer, 2000). Consequently, the cervical interbody joints operate like a saddle joint, with motion restricted to two planes: the sagittal plane and an oblique coronal plane. In the sagittal plane, the vertebral bodies can rock and slide (rotate and translate), to provide for flexion and extension of the neck. From above downwards, the typical cervical vertebrae exhibit progressively less translation for each degree of rotation, during flexion or extension. This is reflected by the different locations of their axes of movement. At higher levels the axes

Fig. 32.8. Sketches of a posterolateral view of a typical cervical vertebral body, showing the two curvatures of its superior surface: a downward concavity along the sagittal plane, and a second concavity, facing upwards and forwards, between the uncinate processes (u). These two concavities endow the intervertebral disc with the features of a saddle joint.

lie in the vertebral body below the moving vertebra, but are progressively closer to their intervertebral disc at lower levels (Amevo et al., 1991) (Fig. 32.1). These differences correlate strongly with the height of the articular pillar at each segment (Nowitzke et al., 1994). Taller pillars provide less space into which the vertebra can translate once it has commenced sagittal rotation. Conversely, at segments with shorter pillars, sagittal rotation lifts the inferior articular processes of the moving vertebra off the supporting articular pillar, and provides a greater gap into which it can translate (Nowitzke et al., 1994). The second plane of movement of typical cervical vertebrae is set at 40° forwards of the coronal plane, and lies parallel to the plane of the zygapophysial joints

FUNCTIONAL ANATOMY OF THE SPINE

681

Fig. 32.9. A sketch of a typical cervical intervertebral joint, illustrating the mechanics of rotation of the vertebral body, across the oblique concavity of the uncinate processes, and around an oblique axis through the vertebral body. The motion is like that of a cone whose apex is fixed but whose bases nevertheless free to twist and spin, in the direction and plane indicated by the arrow.

Fig. 32.10. The mechanics of the early phase of whiplash injury. As a result of a thrust from below, the cervical spine undergoes a sigmoid deformation. Lowe segments, e.g., C5–6, undergo a posterior sagittal rotation around an abnormally high instantaneous axis of rotation (iar), which results in posterior elements being impacted and anterior elements being stretched.

(Bogduk and Mercer, 2000). Across this plane, each vertebra rotates like an inverted cone whose apex is fixed, but whose base can twist (Fig. 32.9). The apex of the cone corresponds to the anterior, median point on the vertebral body to which the fibers of the anulus fibrosus are directed; and the anulus fibrosus serves to hold this apex in place. Meanwhile the posterior, inferior edge of the vertebral body presents a convex surface that is cupped by the concave surface between the uncinate processes of the vertebra below. This latter geometry is that of an ellipsoid joint, and the posterior inferior margin of the vertebral body is free to spin, or swing, across this ellipsoid surface. Thus, while the anterior end of the vertebral body is fixed, its posterior end and its posterior elements are free to spin clockwise or counterclockwise across the oblique coronal plane. During this motion, the inferior articular processes of the zygapophysial joints simply glide laterally across the surfaces of their supporting superior articular processes. A consequence of this mode of operation is that the interbody joints of typical cervical vertebrae cannot tolerate a posterolateral anulus fibrosus, for it would impede the spin of the posterior vertebral body across the oblique coronal plane. Consequently, although a posterior anulus is present at birth and in young children, it gradually disappears as neck movements increase (Tondury, 1972). By about the age of 9 years, the posterolateral anulus tears, and clefts appear in the region of the uncinate processes. Progressively these clefts enlarge centrally, until they meet in the midline, at about the age of 30, to form a transverse cleft from one uncinate

process to the other. This cleft is not a degenerative change but a normal age change. The cleft effectively forms the “joint space” across the posterior intervertebral disc that allows axial rotation of the head to be accommodated and amplified in range by the typical cervical vertebrae. This structure and mechanics of the cervical spine are of relevance to the mechanisms of injury in whiplash. The early phase of whiplash injury involves a thrust from below (Bogduk and Yoganandan, 2001; Bogduk, 2006). This upward thrust deforms the cervical spine into a sigmoid shape, within which the lower cervical vertebrae – typically C5 and C6 – undergo an abnormal extension (Fig. 32.10). The vertebra rocks backwards but without translating. As a result, it rotates about an abnormally high axis of rotation (Kaneoka et al., 1999). During this motion, anterior elements are stretched while posterior elements are impacted. The anterior anulus fibrosus can be torn or avulsed, resulting in so-called rim lesions. Impaction in the zygapophysial joints can cause impaction fractures of the articular cartilages, or contusions of the intra-articular meniscoids. During later phases, the cervical spine rebounds into flexion, which can excessively strain the capsules of the zygapophysial joints (Curatolo et al., 2011). Physiologic studies in laboratory animals have shown that the capsule strains induced by whiplash injury result in persistent nociception from the injured joint, and persistent changes within the central nervous system characteristic of chronic pain (Winkelstein, 2011). Clinical studies have shown that the cervical zygapophysial joints

682

N. BOGDUK

are the single most common source of chronic neck pain after whiplash, accounting for between 50% and 60% of cases (Bogduk, 2011). Most commonly, neck pain – with referred pain to the shoulder girdle – stems from the C5–6 joint, while headache stems from the C2–3 zygapophysial joint. Less well understood is pain from the cervical intervertebral discs. Conspicuously, degenerative disc disease is not associated with neck pain. Furthermore, discogenic pain appears to be uncommon, once zygapophysial joint pain is taken into account (Yin and Bogduk, 2008). Perhaps discogenic pain is caused by strains of the interosseous ligament formed by the anterior anulus fibrosus, but diagnostic techniques by which to test this proposition have not been developed.

LUMBAR SPINE The cardinal role of the lumbar spine is to support the thorax and upper limbs – and any loads that they carry – and to transmit those loads to the pelvis and lower limbs (Bogduk, 2012a). Secondarily, the lumbar spine accommodates a modest range of movement between the thorax and pelvis. In order to subserve these functions, the essential elements of the lumbar spine are the vertebral bodies of the five lumbar vertebrae (Fig. 32.11). These are stacked into a strong column, and are united by intervertebral discs

and by the anterior and posterior longitudinal ligaments (Bogduk, 2012a). Bowing the column into a lordosis endows the lumbar spine with the ability to absorb dynamic axial loads (bouncing). Axial impulses deform the lordotic curve; the energy is absorbed by the elastic discs and longitudinal ligaments; and is returned to restore the more upright curve, once the axial impulse has passed (Bogduk, 2012a). The lumbar intervertebral discs are well designed to accommodate compression loads (Hickey and Hukins, 1980). Each consists of hydrated nucleus pulposus, surrounded by an anulus fibrosus, and capped superiorly and inferiorly by a vertebral endplate that joins the disc to the adjacent vertebral body (Fig. 32.12). The anulus fibrosus is formed by concentric layers of collagen fibers, in which the fibers in any one layer run in parallel, at about 60° to the long axis of the spine, but in successive layers that orientation alternates. Axial compression is resisted primarily by the concentric layers of the anulus fibrosus (Markolf and Morris, 1974) (Fig. 32.12). However, the tendency of the anulus under load is to buckle, both outwards and inwards. This buckling is resisted by the hydrostatic nucleus pulposus. When the nucleus is compressed it exerts a radial pressure that braces, and stiffens, the anulus, thereby preventing it from buckling. A small range of flexion-extension is accommodated by the discs (about 13° per segment), during which the anulus fibrosus on the side to which

Fig. 32.11. Sagittal magnetic resonance images of the lumbar spine. (A) Median scan showing the vertebral bodies and spinous processes (sp). The white dots mark the location of the axes of rotation of the vertebra above. (B) Lateral scan through the intervertebral foramina and the L3–4 to L5–S1 zygapophysial joints. ped, pedicle of L3; sap, superior articular process of L4; iap, inferior articular process of L5.

FUNCTIONAL ANATOMY OF THE SPINE

683

Fig. 32.12. Close-up views of a sagittal magnetic resonance image of an L3–4 intervertebral disc. (A) The components of the disc. np, nucleus pulposus; af, anulus fibrosus; vep, vertebral endplate. (B) The mechanics of the disc. Axial compression loads are primarily borne by the lamellae of collagen in the anulus fibrosus. When compressed, the nucleus pulposus exerts radial pressure to brace the anulus, and prevent it from buckling under load.

movement occurs is compressed slightly, while the anulus on the opposite side is stretched (Bogduk, 2012a). While strongly designed to resist compression, the lumbar discs are poorly designed to resist axial rotation. Because the collagen fibers of the anulus fibrosus alternate in direction in successive layers, only half are available to resist axial rotation in one direction or the other. For stability in axial rotation, the lumbar vertebral bodies and intervertebral discs rely on the posterior elements of the lumbar vertebrae (Bogduk, 2012a). The posterior elements are based on an arch (Bogduk, 2012a) (Fig. 32.13). The arch is supported by stout pedicles that emanate from the upper posterior surface of each vertebral body. The pedicles serve to transmit forces from the succeeding posterior elements to the vertebral bodies, which control the position or movements of the vertebral bodies. The arch is completed by left and right laminae that join in the midline. From the junction of the two laminae springs a large spinous process, and from the junction between the pedicle and lamina on each side arises a long transverse process. These processes serve as levers to which attach the muscles that control the movements of the lumbar vertebrae. At its superior and inferior lateral corners respectively, each lamina bears a superior and inferior articular process. Like large mittens, the paired superior articular processes reach cranially to grasp the inferior articular processes of the vertebra above, and form the zygapophysial joints. The plane of these joints is parallel to the longitudinal axis of the lumbar spine. Consequently, during flexion of the vertebral bodies, the inferior articular processes glide freely out of the sockets formed by the superior articular processes, until movement is arrested by tension in the joint capsules (Bogduk, 2012a). The axis of this movement typically lies in the disc below the moving vertebra (Pearcy and Bogduk, 1988) (Fig. 32.11A), which indicates only a small amount of translation for every degree of rotation of the moving

vertebra. As the inferior articular processes move, they lift away from the superior articular process, tantamount to partially subluxating the joint. Fibroadipose meniscoids protect the exposed surfaces of the articular cartilages during this displacement (Engel and Bogduk, 1982; Bogduk and Engel, 1984). In axial views, the lumbar zygapophysial joints variously present flat, C-shaped, or J-shaped appearances, which correspond to the primary functions of these joints (Horwitz and Smith, 1940). Flat joints essentially face medially and posteriorly. C-shaped joints have an anterior end that faces posteriorly, and a posterior end that faces medially. J-shaped joints have a small anterior lip facing posteriorly, and a larger surface facing medially. The medially facing surfaces serve to resist axial rotation of the vertebrae. Attempted axial rotation swings the inferior articular process laterally, but this movement is arrested by the opposing superior articular process. The range of motion is limited to about 2° or less per segment (Pearcy and Tibrewal, 1984), and is accommodated only by compression of the articular cartilage. The surfaces that face posteriorly serve to resist forward displacement (listhesis) of the vertebra. Impaction of an inferior articular process against its superior articular process tends to force the inferior process backwards, and lift the lamina from which it arises (like opening a hatchback). In turn this tendency stresses the junction between the lamina and its pedicle. Repeated impactions – particularly during repeated axial rotation – can cause stress fractures at this point, resulting in pars interarticularis defects. The lumbar zygapophysial joints can be a source of low-back pain, but its prevalence is uncertain. It appears to be uncommon or rare in injured workers, but is common in elderly patients (Bogduk, 2008, 2012b). The most common cause of chronic low-back pain is internal disc disruption (Bogduk et al., 2013). This condition is characterized by degradation of the nucleus

684

N. BOGDUK

Fig. 32.13. Sketches of the posterior elements of a lumbar vertebra. (A) Posterior view. The two laminae (la) form a quadrangular plate, from whose corners project the superior (sap) and inferior (iap) processes. From the junction of the two laminae projects the spinous process (sp). On each side, the inferior and superior articular processes of consecutive vertebrae form the zygapophysial joint (zj) (B) Axial (top) view. The posterior elements are connected to the vertebral body (vb) by the pedicles (p). The transverse process (tp) projects from the junction of the pedicle and lamina, on each side.

pulposus of the affected disc and the development of radial fissures into the posterior or posterolateral anulus. The condition has been produced in laboratory animals, and pursued in numerous clinical studies. Its cause is compression injuries that produce small fractures of the vertebral endplate. These result in degradation of the matrix of the nucleus pulposus. As the nucleus becomes less able to retain water, it is no longer able to pressurize and brace the anulus. Pressures in the nucleus drop, but rise in the posterior anulus. The unbraced anulus progressively delaminates, particularly in regions of high stress where the laminae are curved: at the posterolateral corners or the posterior paramedian sector. Pain arises as a result of chemical irritation of nociceptors in the anulus by degradation products from the nucleus, and as a result of the increased mechanical stresses on the surviving, intact laminae of anulus (Bogduk et al., 2013). To various degrees of certainty the condition can be diagnosed by characteristic features on magnetic resonance imaging, such as Modic lesions in the vertebral body or high-intensity zones in the anulus fibrosus, and by provocation discography (Bogduk et al., 2013). No treatment has been vindicated, but several minimally invasive interventions are being pursued, which encompass ablating nociceptors in the disc, injecting restorative agents such as stem cells, or injecting antagonists of inflammation.

MUSCLES The anatomy of muscles of the cervical and lumbar spine is made complex by the diversity of their numerous attachments. If those specifics are ignored, the anatomy becomes simpler.

Small muscles connect consecutive spinous processes and transverse processes. Too small to move their vertebrae effectively, these muscles serve as proprioceptors for the spine (Bastide et al., 1989). Prevertebral muscles are represented only in the cervical spine (Standring, 2008). The longus cervicis connects the vertebral bodies and transverse processes of the cervical vertebrae. It is covered by the longus capitis which anchors the skull to the cervical vertebrae. These muscles are weak flexors of the head and neck. Various suboccipital muscles control movements of the head in relation to the atlas and the axis. They are the rectus anterior and rectus lateralis anteriorly, and the rectus capitis posterior major and minor accompanied by obliquus inferior and obliquus superior, posteriorly (Standring, 2008). Collectively these muscles control the orientation of the head on the atlas and axis. The postvertebral muscles are aligned systematically, side by side and by layers (Standring, 2008). Multifidus is the deepest and most medial muscle. Its fascicles arise from a spinous process and descend to various insertions on articular processes and transverses processes one to several segments caudally. It is flanked by the longissimus system of muscles, which attach to transverse processes near their bases, and whose components are large at lumbar levels, but virtually miniscule at cervical levels. Further laterally runs the iliocostalis system, which attaches to transverse processes near their tips, and whose components are, likewise, large at lumbar levels but miniscule at cervical levels. A semispinalis system is vestigial at lumbar levels but well developed at cervical levels. Semispinalis cervicis arises from the cervical spinous processes, and covers the multifidus with fascicles longer than those of the latter muscle. Semispinalis capitis arises from the occiput, and is anchored to the cervical

FUNCTIONAL ANATOMY OF THE SPINE 685 transverse processes. It is the largest of the posterior lateral atlantoaxial joint before joining the cervical neck muscles. It is covered by the splenius muscle, which plexus (Lazorthes and Gaubert, 1956). passes cranially and laterally from the raphe of the ligaThe C3–7 cervical spinal nerves lie above their likementum nuchae to wrap around all the other posterior numbered vertebrae, enclosed in their respective intermuscles of the neck. Splenius cervicis reaches the upper vertebral foramina. They are joined by the C8 spinal cervical transverse processes, while splenius capitis nerve, which lies in the C7–T1 intervertebral foramen. reaches the superior nuchal line. Variously and collecThe ventral rami of C3 and C4 join the cervical plexus, tively, the cervical postvertebral muscles act to extend and the lower cervical ventral rami join the brachial the head and the cervical spine. plexus. The dorsal rami of the typical cervical spinal Other muscles use the vertebral column adventinerves form lateral branches that supply the splenius, tiously, as a base from which to act on nonspinal struclongissimus, and iliocostalis; and medial branches that tures. In the neck, these include the scalene muscles, supply the deeper and medial posterior neck muscles, which act on the ribs; and levator scapulae and trapezius, and the cervical zygapophysial joints (Bogduk, 1982). which act on the shoulder girdle. Sternocleidomastoid is The cervical medial branches have constant locations the principal flexor and rotator of the head and neck, but on the cervical articular pillars, which allow them to be passes directly from the manubrium and clavicle to the targeted for fluoroscopy-guided diagnostic blocks, by head, with no connection to the cervical spine. Being which pain from the zygapophysial joints can be diaglocked to the skull through the atlanto-occipital joints, nosed (Bogduk, 1982, 2011). the atlas is rotated when the sternocleidomastoid rotates Gray rami communicantes, from the stellate ganglion the head. and from the cervical ventral rami, form a plexus – In the lumbar spine, psoas major arises from the vercalled the vertebral nerve – that accompanies the vertetebral bodies, discs, and transverse processes to act on bral artery through the foramina transversaria of the femur, but does not move the lumbar spine the neck, and into the posterior cranial fossa (Bogduk (Bogduk et al., 1992). Quadratus lumborum attaches et al., 1981a). Although migraine cervicale, or the to the lumbar transverse processes but acts principally Barre´–Lieou syndrome, has been attributed to irritation on the 12th rib; its actions on the lumbar vertebrae are of these nerves, and spasm of the vertebral artery, laboeffectively trivial (Phillips et al., 2008). Transversus ratory studies have shown the vertebrobasilar system to abdominis stems from the lumbar transverse processes, be remarkable unresponsive to stimulation of the verteand has virtually no effect on the lumbar vertebrae bral nerve (Bogduk et al., 1981a; Lambert et al., 1984). (Macintosh et al., 1987). Likewise, latissimus dorsi gains The cervical sinuvertebral nerves are formed by some anchorage to the lumbar spinous processes but has somatic roots from the ventral rami and autonomic roots a negligible action on the lumbar spine (Bogduk from the rami communicantes in the vertebral nerve. As et al., 1998). recurrent meningeal branches they innervate the cervical dural sac, but also innervate the cervical discs and the posterior longitudinal ligament (Bogduk et al., 1988). The C1–3 sinuvertebral nerves innervate the ligaments INNERVATION of the median atlantoaxial joint before passing through The C1 spinal nerve is unlike other spinal nerves, which foramen magnum to supply the dura mater over the clireinforces the atlas being suboccipital rather than cervivus (Kimmel, 1960). cal in nature. This nerve lacks a typical dorsal root ganThe lumbar spinal nerves lie obliquely in their interglion, but ganglion cells can be found amongst the vertebral foramina, each below the like-numbered verterootlets of the spinal accessory nerve. The C1 dorsal bra. Their ventral rami enter the lumbar or lumbosacral ramus appears amongst the posterior suboccipital musplexus. Their dorsal rami form lateral and intermediate cles (Lazorthes and Gaubert, 1956). Sometimes it can branches that innervate the iliocostalis and longissimus have a cutaneous branch. The C1 ventral ramus crosses muscles respectively (Bogduk et al., 1982; Bogduk, the posterior arch of the atlas, behind the superior artic1983). Medial branches innervate the lumbar zygapophyular process. It innervates the atlanto-occipital joint sial joints and the multifidus (Bogduk et al., 1982; before entering the cervical plexus (Lazorthes and Bogduk, 1983). Where the medial branches cross the root Gaubert, 1956). of the superior articular process they can be targeted for The C2 spinal nerve lies behind the lateral atlantoaxial fluoroscopy-guided diagnostic blocks, by which pain joint, and forms a large dorsal ramus that supplies the from the lumbar zygapophysial joints can be diagnosed more superficial posterior neck muscles, and becomes (Bogduk, 1983, 2008, 2012b). cutaneous as the greater occipital nerve, over the occiput At each segmental level, the lumbar sinuvertebral (Bogduk, 1982). The C2 ventral ramus supplies the nerves arise from the ventral ramus and gray ramus

686

N. BOGDUK

Fig. 32.14. Sagittal magnetic resonance images of the thoracic spine. (A) Median section, through the vertebral bodies, spinal cord (sc), and spinous processes (sp) cord. (B) Paramedian section through the zygapophysial joints (zj). Intervertebral discs (ivd) are evidence in both sections. (Courtesy of Dr. Tim Maus, Mayo Clinic, Rochester MN.)

communicans. Each passes back into the intervertebral foramen to supply the dural sac, the posterior longitudinal ligament, and the posterior anulus fibrosus (Bogduk et al., 1981b; Bogduk, 1983). These nerves provide the sensory pathway for lumbar discogenic pain.

THORACIC SPINE There have been no substantial advances in the description of the anatomy of the thoracic spine since editions of anatomy textbooks of the 19th and 18th century. In parallel, there has been little advance in the understanding of thoracic spinal pain and its sources, let alone causes. No diagnostic or treatment procedures have been validated. Thoracic spinal pain essentially remains a mystery. Like cervical and lumbar vertebrae, the thoracic vertebrae have vertebral bodies that are connected by intervertebral discs and longitudinal ligaments, and posterior elements that are connected by zygapophysial joints (Fig. 32.14). The distinction of the thoracic spine is that it suspends the ribs. At typical thoracic levels, the head of the rib articulates with the intervertebral disc and demifacets on the edges of the vertebrae that bind that disc, and the articular tubercle of the rib articulates with the transverse process of the upper of the two vertebrae

Fig. 32.15. Axial magnetic resonance image of a typical thoracic spinal segment. vb, vertebral body; zj, zygapophysial joint; sp, spinous process; cvj, costovertebral joint; ctj, costotransverse joint. (Courtesy of Dr. Tim Maus, Mayo Clinic, Rochester, MN.)

(Fig. 32.15). Exceptions to this arrangement occur at T1 and at T11 and T12, where the head of the rib fully articulates with the like-numbered vertebrae. Few studies have explored the innervation of the thoracic spine (Bogduk, 2002). The thoracic sinuvertebral nerves are assumed to be homologous to those at cervical or lumbar levels. The courses of the thoracic dorsal rami appear to differ from those at cervical and lumbar levels, but are nevertheless homologous (Chua and Bogduk, 1995). Whereas the medial branches at cervical and lumbar levels wind around the base of the superior articular process at each segmental level, at thoracic levels the dorsal ramus stretches to the tip of the transverse process before dividing into medial and lateral branches. This difference is reconciled once it is realized that what are called the transverse processes at cervical and lumbar levels are embryologically costal elements (rudimentary ribs), whereas the embryologic transverse elements (or true transverse processes) are absorbed into the base of the superior articular process. Consequently, at cervical and lumbar levels, the medial branches cross the superior articular process because the true transverse processes also lie there. This distinction becomes pertinent for minimally invasive, diagnostic, and treatment procedures that target thoracic medial branches. The target lies on the transverse process, not on the superior articular process (Chua and Bogduk, 1995). A persisting curiosity pertains to the structure of thoracic intervertebral discs. Cervical discs differ greatly from lumbar discs, but undiscovered is the transition

FUNCTIONAL ANATOMY OF THE SPINE zone. Are thoracic discs like cervical discs, or do they have the structure of lumbar discs? Given that cervical uncinate processes are homologous to the heads of the ribs, unpublished observations suggest that discs change their structure where uncinate processes or their rib equivalent cease. Thoracic discs become lumbar in nature at T11, where the rib no longer articulates with the disc.

REFERENCES Amevo B, Worth D, Bogduk N (1991). Instantaneous axes of rotation of the typical cervical motion segments: a study in normal volunteers. Clin Biomech 6: 111–117. Bastide G, Zadeh J, Lefebvre D (1989). Are the ‘little muscles’ what we think they are? Surg Radiol Anat 11: 255–256. Bogduk N (1982). The clinical anatomy of the cervical dorsal rami. Spine 7: 319–330. Bogduk N (1983). The innervation of the lumbar spine. Spine 8: 286–293. Bogduk N (2002). Innervation and pain patterns of the thoracic spine. In: R Grant (Ed.), Physical therapy of the Cervical and Thoracic Spine, 3rd edn. Churchill Livingstone, New York, pp. 73–81. Bogduk N (2006). Whiplash injury. In: F Cervero, TS Jensen (Eds.), Handbook of Clinical Neurology Vol. 81: Pain, Elsevier, Amsterdam, pp. 791–801. Bogduk N (2008). Evidence-informed management of chronic back pain with facet injections and radiofrequency neurotomy. Spine J 8: 56–64. Bogduk N (2011). On cervical zygapophysial joint pain after whiplash. Spine 36: S194–S199. Bogduk N (2012a). Clinical Anatomy of the Lumbar Spine and Sacrum, 5th edn. Elsevier, Edinburgh. Bogduk N (2012b). Lumbar medial branch neurotomy. In: S Dagenais, S Haldeman (Eds.), Evidence-Based Management of Low Back Pain, Elsevier, St Louis, pp. 351–363. Bogduk N (2014). The neck and headaches. Neurol Clin 32: 471–487. Bogduk N, Bartsch T (2008). Cervicogenic headache. In: SD Silberstein, RB Lipton, DW Dodick (Eds.), Wolff’s Headache, 8th edn. Oxford University Press, New York, pp. 551–570. Bogduk N, Engel R (1984). The menisci of the lumbar zygapophyseal joints. A review of their anatomy and clinical significance. Spine 9: 454–460. Bogduk N, Govind J (2009). Cervicogenic headache: an assessment of the evidence on clinical diagnosis, invasive tests, and treatment. Lancet Neurol 8: 959–968. Bogduk N, Mercer SR (2000). Biomechanics of the cervical spine. I: Normal Kinematics. Clin Biomech 15: 633–648. Bogduk N, Yoganandan N (2001). Biomechanics of the cervical spine Part 3: minor injuries. Clin Biomech 16: 267–275. Bogduk N, Lambert G, Duckworth JW (1981a). The anatomy and physiology of the vertebral nerve in relation to cervical migraine. Cephalalgia 1: 11–24. Bogduk N, Tynan W, Wilson AS (1981b). The nerve supply to the human lumbar intervertebral discs. J Anat 132: 39–56.

687

Bogduk N, Wilson AS, Tynan W (1982). The human lumbar dorsal rami. J Anat 134: 383–397. Bogduk N, Windsor M, Inglis A (1988). The innervation of the cervical intervertebral discs. Spine 13: 2–8. Bogduk N, Pearcy M, Hadfield G (1992). Anatomy and biomechanics of psoas major. Clin Biomech 7: 109–119. Bogduk N, Johnson G, Spalding D (1998). The morphology and biomechanics of latissimus dorsi. Clin Biomech 13: 377–385. Bogduk N, Aprill C, Derby R (2013). Lumbar discogenic pain: state-of-the-art review. Pain Med 14: 813–836. Chua WH, Bogduk N (1995). The surgical anatomy of thoracic facet denervation. Acta Neurochir 136: 140–144. Curatolo M, Bogduk N, Ivancic PC et al. (2011). The role of tissue damage in whiplash-associated disorders. Spine 36: S309–S315. Dvorak J, Hayek J, Zehnder R (1987). CT-functional diagnostics of the rotatory instability of the upper cervical spine part 2. An evaluation on healthy adults and patients with suspected instability. Spine 12: 726–731. Engel R, Bogduk N (1982). The menisci of the lumbar zygapophysial joints. J Anat 135: 795–809. Fielding JW, Cochran G van B, Lawsing JF et al. (1974). Tears of the transverse ligament of the atlas. J Bone Joint Surg 56A: 1683–1691. Hickey DS, Hukins DWL (1980). Relation between the structure of the anulus fibrosus and the function and failure of the intervertebral disc. Spine 5: 100–116. Horwitz T, Smith RM (1940). An anatomical, pathological and roentgenological study of the intervertebral joints of the lumbar spine and of the sacroiliac joints. Am J Roentgenol 43: 173–186. Kaneoka K, Ono K, Inami S et al. (1999). Motion analysis of cervical vertebrae during whiplash loading. Spine 24: 763–770. Kimmel DL (1960). Innervation of the spinal dura mater and dura mater of the posterior cranial fossa. Neurology 10: 800–809. Koebke J, Brade H (1982). Morphological and functional studies on the lateral joints of the first and second cervical vertebrae in man. Anat Embryol 164: 265–275. Lambert GA, Duckworth JW, Bogduk N et al. (1984). Low pharmacological responsiveness of the vertebro-basilar circulation in Macaca nemestrina monkeys. Eur J Pharmacol 102: 451–458. Lazorthes G, Gaubert J (1956). L’innervation des articulations interapophysaire vertebrales. Comptes Rendues de l’Association des Anatomistes 43: 488–494. Macintosh JE, Bogduk N, Gracovetsky S (1987). The biomechanics of the thoracolumbar fascia. Clin Biomech 2: 78–83. Markolf KL, Morris JM (1974). The structural components of the intervertebral disc. J Bone Joint Surg 56A: 675–687. Mercer S, Bogduk N (1993). Intra-articular inclusions of the cervical synovial joints. Br J Rheumatol 32: 705–710. Mercer S, Bogduk N (1999). The ligaments and anulus fibrosus of human adult cervical intervertebral discs. Spine 24: 619–626.

688

N. BOGDUK

Mercer SR, Bogduk N (2003). Clinical anatomy of ligamentum nuchae. Clin Anat 16: 484–493. Nowitzke A, Westaway M, Bogduk N (1994). Cervical zygapophyseal joints: geometrical parameters and relationship to cervical kinematics. Clin Biomech 9: 342–348. Oda J, Tanaka H, Tsuzuki N (1988). Intervertebral disc changes with aging of human cervical vertebra from the neonate to the eighties. Spine 13: 1205–1211. Pearcy MJ, Bogduk N (1988). Instantaneous axes of rotation of the lumbar intervertebral joints. Spine 13: 1033–1041. Pearcy MJ, Tibrewal SB (1984). Axial rotation and lateral bending in the normal lumbar spine measured by threedimensional radiography. Spine 9: 582–587. Phillips S, Mercer S, Bogduk N (2008). Anatomy and biomechanics of quadratus lumborum. J Eng Med 222: 151–159.

Standring S (Ed.), (2008). Gray’s Anatomy, 40th edn. Churchill Livingstone, Edinburgh, pp. 736–743. Tondury G (1972). The behaviour of the cervical discs during life. In: C Hirsch, Y Zotterman (Eds.), Cervical pain, Pergamon Press, Oxford, pp. 59–66. Wasserman BR, Moskovitch R, Razi AE (2011). Rheumatoid arthritis of the cervical spine. Clinical considerations. Bull Hosp Joint Dis 68: 136–148. Winkelstein BA (2011). How can animal models inform on the transition to chronic symptoms in whiplash? Spine 36: S218–S225. Wortzman G, Dewar FP (1968). Rotary fixation of the atlantoaxial joint: rotational atlantoaxial subluxation. Radiology 90: 479–487. Yin W, Bogduk N (2008). The nature of neck pain in a private pain clinic in the United States. Pain Med 9: 196–203.

Handbook of Clinical Neurology, Vol. 136 (3rd series) Neuroimaging, Part II J.C. Masdeu and R.G. Gonza´lez, Editors © 2016 Elsevier B.V. All rights reserved

Chapter 33

Neuroimaging of spine tumors NANDOR K. PINTER, THOMAS J. PFIFFNER, AND LASZLO L. MECHTLER* Dent Neurologic Institute, Amherst, NY, USA

Abstract Intramedullary, intradural/extramedullary, and extradural spine tumors comprise a wide range of neoplasms with an even wider range of clinical symptoms and prognostic features. Magnetic resonance imaging (MRI), commonly used to evaluate the spine in patients presenting with pain, can further characterize lesions that may be encountered on other imaging studies, such as bone scintigraphy or computed tomography (CT). The advantage of the MRI is its multiplane capabilities, superior contrast agent resolution, and flexible protocols that play an important role in assessing tumor location, extent in directing biopsy, in planning proper therapy, and in evaluating therapeutic results. A multimodality approach can be used to fully characterize the lesion and the combination of information obtained from the different modalities usually narrows the diagnostic possibilities significantly. The diagnosis of spinal tumors is based on patient age, topographic features of the tumor, and lesion pattern, as seen at CT and MRI. The shift to high-end imaging incorporating diffusion-weighted imaging, diffusion tensor imaging, magnetic resonance spectroscopy, whole-body short tau inversion recovery, positron emission tomography, intraoperative and high-field MRI as part of the mainstream clinical imaging protocol has provided neurologists, neuro-oncologists, and neurosurgeons a window of opportunity to assess the biologic behavior of spine neoplasms. This chapter reviews neuroimaging of spine tumors, primary and secondary, discussing routine and newer modalities that can reduce the significant morbidity associated with these neoplasms.

INTRODUCTION The historic classification of spine tumors is based on the use of myelography with three main groups, as schematically depicted in Figure 33.1: (1) extradural extramedullary; (2) intradural extramedullary; and (3) intradural intramedullary. The incidence of metastatic disease involving the vertebrae, epidural space, and leptomeninges accounts for 97% of tumors involving the spine. Primary tumors of the spine, spinal cord, spinal meninges, and cauda equina are relatively rare (Duong et al., 2012). Data from national registries and improved imaging capabilities have allowed spine tumor specialists the opportunity to study and treat these unusual and rare tumors with more confidence and better results. The introduction of magnetic resonance imaging (MRI) to clinical practice has been one of the most important

advances in the care of patients with spine tumors. The characterization of spine tumors by MRI involves determining, in the context of patient’s age and sex, the location of the lesion and whether or not it enhances after gadolinium injection. Computed tomography (CT) best delineates osseous integrity while MRI is better at assessing soft-tissue involvement. The purpose of this chapter is to describe the neuroimaging findings of spine tumors based on the location of the tumor in its relationship to the dura and spinal cord (Mechtler and Nandigam, 2013).

METASTATIC TUMORS OF THE VERTEBRAL COLUMN The spine is the third most common site for metastatic disease and the most common site for bone metastasis

*Correspondence to: Laszlo L. Mechtler, MD, Professor of Neurology and Oncology, Dent Neurologic Institute, 3980 Sheridan Drive, Amherst NY 14226, USA. Tel: +1-716-250-2000, Fax: +1-716-250-2045, E-mail: [email protected]

690

N.K. PINTER ET AL.

Fig. 33.1. Historic classification of spine tumors based on computed tomography myelography. (A) Normal, (B) extradural extramedullary, (C) intradural extramedullary, and (D) intradural intramedullary. From Mechtler and Nandigam (2013).

(Shah and Salzman, 2011). Metastatic disease of the vertebral column is more frequent than primary neoplastic diseases. Approximately two-thirds of cancer patients will develop bone metastasis and symptomatic spinal metastasis will occur in almost 10% of cancer patients. The most common primary sites are the prostate, breast, kidney, lung, and thyroid. The incidence of skeletal metastases according to the primary tumor is as follows; breast 73% (47–85%), prostate 68% (33–85%), thyroid 42% (28–60%), lung 36% (30–55%), kidney 35% (33–40%), esophageal 6% (5–7%), and gastrointestinal 5% (3–11%) (Maccauro et al., 2011) The most common cause of metastatic spine disease is breast cancer in women; however, in men, prostate cancer is most common. The thoracic spine is the most commonly involved. The majority of the lesions are extradural in location, consisting of lesions which are localized to the epidural space and those which are nested in the vertebral body. Prostate, breast, and lung cancer are again the leading cause of spinal cord compression, each accounting for about 15–20% of the cases. The remaining cancers stem from renal cell, non-Hodgkin’s lymphoma, multiple myeloma, colorectal cancers, sarcomas, and unknown tumors. Pain, the most common initial feature, occurs in 95% of adults and 80% of children. Pain is usually localized to the site of metastasis and is caused by stretching the pain-sensitive bony periosteum. Radicular pain is less frequent but is also localizing. Nocturnal pain upon lying down is typical. Three types of bone metastasis are distinguished: osteolytic, osteoblastic, and mixed; 71% are osteolytic, 8% are osteoblastic, and 21% are mixed. Osteolytic metastases typically develop in cancers of the breast, lung, kidney, thyroid, oropharyngeal cancers (Shah and Salzman, 2011) and in melanoma (Sun et al., 2013). This is a result of osteoclast activation, rather than a direct invasion of bone tissue by tumor cells. In osteoblastic metastases the balance of bone metabolism is shifted to the benefit of bone production as a result of pathologic activation of osteoblasts. Osteoblastic lesions usually occur in prostate, bladder and

nasopharyngeal cancer, medulloblastoma, neuroblastomas, and bronchial carcinoid (Long et al., 2010).

Imaging of vertebral metastases In today’s clinical practice MRI is the most important modality in imaging of metastatic spine disease. Plain film is no longer the routine diagnostic toolbar due to its low sensitivity and specificity (Salvo et al., 2009; Shah and Salzman, 2011). Nuclear medicine studies have a well-defined role in metastasis imaging. Bone scans have been used for screening, since the tracer accumulates in metastatic sites with high sensitivity, thus reflecting the increased bone turnover. The sensitivity and specificity of bone scans were improved with singlephoton emission computed tomography (SPECT) scans (Ryan and Fogelman, 1995). Flurodeoxyglucose (F18FDG) positron emission tomography (PET) alone and PET CT can discover spinal metastases with a sensitivity of 74% and 98%, respectively (Metser et al., 2004). F18FDG PET has been reported to be more sensitive in detecting osteolytic metastases (Cook and Fogelman, 2000). CT has a lower sensitivity in detecting osseous metastases and an inferior diagnostic accuracy compared to MRI (Buhmann Kirchhoff et al., 2009). In fact, CT is less accurate in detecting paraspinal soft tissue, bone edema, and bone metastases that may be missed if destruction is not present (Shah and Salzman, 2011). Therefore, CT has a rather complementary role in first-line imaging of spinal metastases and owns priority only in those cases when the integrity and fine structure of the trabecular and cortical bone are a question, preoperative planning is required, or when MRI is contraindicated. MRI is superior to CT in all other cases. Metastatic lesions are most commonly focal or multifocal and the diffuse involvement of the vertebral bodies is less common. Focal abnormalities are hypointense on T1 and hyperintense on T2 and short tau inversion recovery (STIR) sequences. In general, metastases will enhance with contrast, although it is important to always acquire a noncontrast study for comparison. The diffuse

NEUROIMAGING OF SPINE TUMORS

691

Fig. 33.2. Diffuse metastatic involvement of the cervical and upper thoracic spine. The T1-weighted image (A) shows heterogeneous signal intensity in all vertebral bodies, with mostly isointense to hypointense signal accompanied by focal hyperintense areas (open arrow) compared to intervertebral discs. On T2-weighted image (B), similar heterogeneity can be seen with hyperintense and isointense areas. On short tau inversion recovery (STIR) sequence (C) diffuse hyperintense signal correlates with the T1 hypointensity, typical of metastatic disease of the bone marrow. The involvement of spinous processes is highly suspicious on T2 and becomes unquestionable on STIR (arrows). The focal hyperintensities on T1 are associated with fatty marrow and represented as hypointense signal on STIR.

marrow involvement can be difficult to assess, because a generally low signal intensity appearance can be misleading, giving the false impression of normal marrow. It is helpful to compare the marrow’s signal intensity to that of the discs and muscles. In adults it can be regarded as abnormal if the marrow has lower signal intensity than discs or muscles (Fig. 33.2) (Long et al., 2010). Metastases tend to occur in the posterior part of the vertebral body, involving the pedicles. Most often metastases are destructive and can be expansive. In osteolytic metastases, the cancellous bone is replaced by tumor tissue and acceleration of bone resorption occurs due to an imbalance in metabolism. The focal disequilibrium in calcified elements will create the characteristic image of circumscribed translucency on radiographs and a generally low-density area framed by intact, partly intact, or sclerotic osseous components on CT. On MRI hypointense T1, hyperintense T2, and STIR signals will represent the structural and biochemical (e.g., edema) changes in the bone marrow (Long et al., 2010). Posterior cortical and pedicle destruction is not infrequent and best visualized by CT. During the course of antitumor therapy progressive sclerosis may be visible on follow-up imaging, indicating positive response to therapy. Fluid levels may occur in lytic metastases (Jarraya et al., 2013) and might even mimic primary tumors (Colangeli et al., 2010). The term osteoblastic refers to the biologic behavior of the lesions, but from the imaging point of view, sclerotic or osteosclerotic may sound more practical, as they reflect the imaging findings. Osteoblastic metastases are usually focal or mottled and in most cases are multifocal. Diffuse involvement of the vertebral body occurs occasionally.

On MRI the lesions show T1 and T2 hypointensity. On T2 and STIR images a bright rim surrounding the sclerotic lesions may be depicted, reflecting bone marrow edema. This is known as the “halo” sign. Edematous signal can often be observed on STIR, expanding from the body to the pedicles, without a well-defined lesion on T1 or T2 sequences. The enhancement pattern in MRI may vary depending on the degree of sclerosis. On CT and plain film these lesions show high density resulting from excessive calcification. The epidural expansion of the tumor may create the “draped curtain sign,” which can be depicted on axial MRI slices (Fig. 33.3).

Vertebral compression deformity Vertebral compression deformity is frequent in elderly patients. The cause of this abnormality can be benign or malignant. Distinguishing between benign and pathologic vertebral body compression fractures is usually possible on MRI. Chronic benign fractures typically have marrow signal intensity that is isointense with normal vertebrae on all sequences. In addition, there is no involvement of the posterior elements, absence of paravertebral or epidural mass, and preservation of posterior cortex. Pathologic fractures show comparatively low signal intensity on T1- and high signal intensity on T2-weighted sequences (Fig. 33.4) (Griffith and Guglielmi, 2010). Pathologic compression fractures typically enhance with contrast; however, conventional MR techniques cannot always be used to differentiate benign from malignant lesions due to their similar appearances. For example, osteoporotic compression fracture can be confused with metastatic compression

692

N.K. PINTER ET AL.

Fig. 33.3. A 48-year-old man with multiple myeloma. Sagittal (A) and axial (B) contrast-enhanced T1-weighted scans with fatsuppression shows a large enhancing mass occupying the T12 vertebral body, extending into the spinal canal and the left pedicle (arrows). The yellow arrows on the axial image mark as the tumor invades the anterior epidural space bilaterally and compression on the thecal sac, creating the “draped curtain sign.” Reformatted computed tomography in the coronal and sagittal planes (C and D) shows the osteolytic nature of this tumor. This was thought to be a solitary plasmacytoma. Further imaging revealed “punched-out” osteolytic lesions (black arrows) characteristic of multiple myeloma.

Fig. 33.4. Compression fracture resulting from metastatic disease. T1 hypointense (A), T2 intermediate-hyperintense (B) and short tau inversion recovery STIR hyperintense signal (C) is demonstrated in the compressed T9 vertebral body. The body shows a vertebral plana deformity. Posteriorly focal sclerotic changes can be observed on the sagittal reconstruction of CT scan (D).

in the acute phase. Edema in an acute benign compression fracture replaces the normal marrow, resulting in hypointensity on T1-weighted images and hyperintensity on T2-weighted images. The vertebral body with benign fracture may enhance with contrast. These MR signal intensity characteristics are similar to those of metastases and cause ambiguity, especially when only a single lesion is present. Pathologic fractures are often multiple; other key features of pathologic compression fractures include loss of the posterior body height, pedicle involvement, epidural and paraspinal mass. Diffusion-weighted imaging (DWI) is emerging as a powerful clinical tool for directing the care of patients with cancer (Padhani et al., 2011). The basic biologic premise for the use of DWI is the characteristic increased cellular content of malignant tissue, and increased water content. These features result in higher signal intensity of malignant disease on high b-value images with corresponding low apparent diffusion coefficient (ADC)

values (Khoo et al., 2001). This is especially helpful in patients with neuroblastomas, leukemia, lymphoma, rhabdomyosarcoma, Ewing sarcoma, and metastatic primitive neuroectodernal tumors. DWI can also be a useful tool for distinguishing acute benign osteoporotic from malignant vertebral compression fractures, resulting in low or isointense signal on DWI and high ADC values on ADC maps (Duvauferrier et al., 2013; Rumpel et al., 2013). Whole-body STIR has also been used to monitor disease progression in patients with lymphoma, multiple myeloma, metastatic primitive neuroectodernal tumors, and neurofibromatosis (Fayad et al., 2013).

PRIMARY TUMORS OF THE SPINE Benign primary spine tumors Hemangioma is the most common primary tumor of the spine; the incidence from autopsy studies is estimated to

NEUROIMAGING OF SPINE TUMORS be 10% (Ropper et al., 2011). Histologically they consist of multiple small, thin-walled vessels interspersed with trabeculae and infiltrating the medullary cavity. Hemangiomas have a characteristic, polka-dot appearance on the axial CT images, which represent the cross-section of the prominent vertical trabeculae. On MRI lesions show high signal intensity on T1 and T2 images, often with a heterogeneous structure, which may also contain flow voids as signs of vascular elements. Osteoid osteomas occur in young adults and their most common symptom is a painful scoliosis that worsens at night (Theodorou et al., 2008). They usually arise from the posterior elements. Lumbar spine involvement is typical. On radiographs osteoid osteomas have the appearance of a lytic lesion surrounded by sclerotic rim. The same finding is present on CT scans. The sclerotic changes may affect the central area or nidus. On T1 images the nidus displays intermediate signal intensity and can have signal voids as a result of calcification. On T2 images the nidus appears as a low-intensity area surrounded by high-intensity rim representing edema. Osteoblastomas have a nonspecific MRI appearance, with T1 and T2 prolongation depending on the degree of matrix mineralization. Osteoblastomas have a reactive rim similar to osteoid osteomas, which can lead to overestimation of lesion (Chai and Cho, 2013). Aneurysmal bone cysts are expansile lesions containing blood-filled cysts separated by walls of bone (Long et al., 2010). The spine is affected in 12–30% of all aneurysmal bone cyst cases. On CT scans ballooning, lobulated cystic lesions can be found occasionally with

693

fluid–fluid levels (Ropper et al., 2011). The latter is indicative of hemorrhage and may be more sensitively visualized by MRI (Fig. 33.5). Signal changes, representing blood degradation products of different age, can be appreciated (Vidal and Murphey, 2007). Giant cell tumor is a benign, locally aggressive tumor that accounts for approximately 4–5% of primary bone tumors. Most commonly, giant cell tumors are found in long bones. CT and MRI may reveal an associated soft-tissue component, and contrast administration may help to differentiate the bony and soft-tissue elements. On MRI, they appear as low to intermediate signal intensity on T1- and T2-weighted imaging. Giant cell tumors often cause the destruction of sacral foramina. Osteochondroma is an uncommon tumor, accounting for about 4% of solitary spinal tumors. Osteoschondromas seem to demonstrate distinct predilection for certain parts of the spine, specifically C2. It is often difficult to discover the lesion on radiograph or MRI images due to its small size and the continuity with the bone it originates from. Thin-slice CT scans may be needed to confirm the diagnosis (Vidal and Murphey, 2007).

Malignant primary spine tumors Chondrosarcoma is the second most common nonlymphoproliferative malignant tumor of the spine in adults (Rodallec et al., 2008). The mineralized matrix is best visualized by CT, and areas of calcification are represented as signal voids on MRI. The noncalcified tumor shows low attenuation on CT, low to intermediate signal

Fig. 33.5. Aneurysmal bone cyst (ABC). Sagittal and axial T2-weighted images (A and B respectively) show a well-defined, multilobulated lesion in the L3 vertebral body. The lesion consists of several small cystic compartments which demonstrate sharply separated areas of T2 hyperintense and intermediate signal, representing fluid–fluid levels due to gravitational forces. This is best demonstrated on the axial T2 scan due to the supine position of the patient (white open arrow, B). The lesion extends to the spinal canal and causes the compression of the dural sac (white arrows). Computed tomography (CT) scan (C) shows an almost complete replacement of osseous tissue by the lesion with only a thin layer of cortical bone remaining on the right side (black open arrow). It also depicts the ballooning character of the lesion and some sclerotic changes on the margins. The invasive nature of the ABC is also highlighted by the partial destruction of the upper endplate as displayed by the sagittal T2 scan (asterisk). On sagittal reconstruction of the CT scan (D) it is also clear that the lesion extends posteriorly into the pedicle (black arrow).

694

N.K. PINTER ET AL.

intensity on T1, and high signal intensity on T2 images due to the water and hyaline content. Spinal osteosarcomas are the rarest tumors of the spine, accounting for 4% of all osteosarcomas and 5% of primary malignant spine tumors. Each level may be involved, but the thoracic and lumbar are most common. In the majority of cases, the lesions arise from posterior elements. Osteosarcomas are osteoblastic tumors, hence their main imaging characteristics. The mineralized matrix of these tumors allows for detection by plain film or CT, occasionally creating the image of “ivory vertebra.” MRI can assess the extent of the tumor, the soft-tissue component, and nerve root involvement, if present. On bone scintigraphy osteosarcomas show marked tracer uptake. Ewing sarcoma of the spine occurs in 5% of patients with Ewing’s tumor, predominantly involving the sacrum, half of which have extraosseous soft-tissue mass. Ninety percent occur before the age of 20, although a second peak is observed at age 50. STIR is the most sensitive sequence. Ewing’s sarcoma accounts for nearly 20% of spinal cord compression in children. Neuroblastoma is the most common cause of spinal cord compression in children less than 5 years of age. Typical imaging findings include T1-weighted hypointensity with epidural or paraspinal mass, and T2-weighted hypointensity due to hypercellularity (Rossi et al., 2007; Simon et al., 2012). Chordoma accounts for 1–4% of all primary malignant bone tumors and is the most common nonlymphoproloferative primary malignant tumor of the osseous spine in adults (Vidal and Murphey, 2007; Rodallec et al., 2008). Chordomas are almost always located in the midline; the most common sites are the sacrum (50%) and the clivus (35%), followed by the cervical and thoracolumbar spine. CT aids in assessment of the extent of the tumor and distinguishes soft tissue from calcified areas (Atlas, 2009). Imaging from MR reveals a destructive, lobulated mass that exhibits low to intermediate signal intensity on T1-weighted imaging and high signal intensity on T2 and STIR sequences. MRI is the best to evaluate the soft-tissue and probable epidural component of the tumor. Chordomas display heterogeneous contrast enhancement with multiple septa and may invade the intervertebral disc space (Long et al., 2010).

Hematologic malignancies of the spine Plasmacytomas are focal proliferations of malignant plasma cells and are thought to represent the early stage of multiple myeloma. A second lesion is found in 33% of cases presumed to be solitary plasmacytoma. They tend to be located in the thoracic spine without predilection of anterior or posterior elements. In about two-thirds of cases, a characteristic appearance of lytic and expansile replacement of cancellous tissue surrounded by partly

intact, or even sclerotic, cortical bone is present (Long et al., 2010). This may produce the image of a “minibrain” on axial CT images (Rodallec et al., 2008). In one-third of cases the image is less characteristic, with a bubble-like cystic appearance. Purely sclerotic plasmacytomas are rare. On MRI low T1 and variable or high T2 signal with varying degrees of contrast enhancement is demonstrated. Multiple myeloma is the most common primary neoplasm of bone (Long et al., 2010). Through the secretion of osteoclast-activating factors it causes excessive osteolysis, resulting in multiple osteolytic lesions along the skeleton and skull. Conventional radiographs show the classic appearance of osteopenia with punched-out lesions, accompanied by compression fractures. Bone scintigraphy is not reliable for staging or monitoring. On MRI the pattern of multiple myeloma is characterized by numerous rounded foci of low T1 and high T2 and STIR signal, with mild to moderate enhancement (Long et al., 2010). Lymphoma of the spine may present in the bone as primary disease or as metastases. Thirty percent of systemic non-Hodgkin’s lymphoma have skeletal involvement. In addition, epidural, leptomeningeal, and intramedullary forms have been well described. T2-weighted imaging and ADC show relative hypointensity due to the hypercellular nature of the disease (“blue cell tumor”). Primary tumors of the spine are summarized in Table 33.1.

INTRADURAL EXTRAMEDULLARY TUMORS The most common intradural extramedullary lesions are schwannomas, meningiomas, and neurofibromas (AbulKasim et al., 2008). The range of tumors found in the intradural extramedullary space is few in comparison with bony or central cord tumors. Most of these masses originate from a systemic/congenital disorder such as the phakomatoses or are metastases from systemic cancers. The current standard diagnostic study for a spinal tumor is MRI (Setzer et al., 2007). An MR image provides precise positioning on the extent of spinal cord compression, and further information about the spinal cord and tumor itself. The imaging protocol should include sagittal and axial T1-weighted and T2-weighted sequences, including contrast-enhanced sagittal and axial T1-weighted sequences, and, if needed, coronal images. The STIR sequence is excellent for evaluating intramedullary cord lesions as well as marrow and soft-tissue edema. Contrast-enhanced images are important to define the extent of the lesion and are useful in distinguishing associated cysts or syrinx from neoplastic involvement and are important in postoperative followup (Bloomer et al., 2006; Abul-Kasim et al., 2008).

Table 33.1 Primary tumors of the spine MRI, magnetic resonance imaging; CT, computed tomography; STIR, short tau inversion recovery

Primary benign

Primary malignant

Location

MRI

CT

Other

Hemangioma

Usually confined to vertebral body, but may extend to pedicle

T1 and T2 hyperintense; often heterogeneous with flow voids; STIR: different degrees of suppressed signal depending on fat content

White polka-dot appearance

Osteoid osteoma

Posterior elements; 1.5–2 cm Neural arch Cervical most common

Similar to osteoid osteoma; T1 and T2 signals depend on degree of calcification

Well-circumscribed expansile

Aneurysmal bone cyst

Neural arch extending into vertebral body

Multiple cystic component; fluid–fluid levels due to bleeding and sedimentation; rim and septal enhancement

Ballooning, osteolytic lesion with sclerotic rim

Giant cell tumor

Most common in long bones; most frequent spinal location is sacrum

Low-intermediate signal; CE scan to assess soft tissue Heterogeneous enhancement

May reveal the destruction of sacral foramina

Osteochondroma

Common in cervical spine (C2)

May miss a small tumor

Modality of choice

Chondrosarcoma

Calcified and noncalcified part 5% in spine

Noncalcified part is hypointense or intermediate on T1 and hyperintense on T2; peripheral and septal enhancement

Best to visualize mineralized matrix

Most common tumor of the spine: 10–12% 25–30% multiple Negative bone scan. Enhances intensely T1-weighted hypointense“aggressive” variant Marked tracer uptake in nidus on technetium scans Cause of painful scoliosis in child or young adult Focal scoliosis concave on side of tumor Second decade Scoliosis 50–60% Bone scan positive Variable enhancement Second and third decade X-ray: lucent, expansile lesion Nocturnal back pain Absent pedicle sign No malignant degeneration Locally aggressive Bone scan positive Lytic, expansile lesion Third decade Spinous/transverse process Spinal lesions more common in males Low-grade, but third most common primary malignant bone tumor Bone scan positive Cortical disruption Fifth decade Continued

Table 33.1 Continued

Lymphoproliferative

Location

MRI

CT

Other

Osteosarcoma

Thoracic and lumbar most common; arises in posterior elements

Soft-tissue component T1-hypointense, T2 hypointense

Osteoblastic lesion; occasionally “ivory vertebra”

Chordoma

Midline; most common is sacrum and clivus; discs usually involved

Destructive lobulated mass; hypointense or intermediate on T1; hyperintense on T2 and STIR; heterogeneous enhancement with septa

Bone destruction; soft-tissue mass with amorphous calcification

Neuroblastoma

Most common cause of cord compression in children < 5 years

T1-weighted hypointensity with epidural or paraspinal mass; T2-weighted hypointensity

Ewing sarcoma

5% in spine Mostly sacrum; 50%

STIR most sensitive Moderate enhancement T1 hypointense, T2 intermediate /hyperintense

Enhancing paraspinal mass with stippled calcification Tiny perforations of the cortex

Lymphoma

Osseous, epidural, cord or leptomeningeal

Diffuse enhancement, epidural extension Whole-body MRI with T1-weighted imaging, STIR, and diffusion-weighted imaging

Lytic, paraspinal mass

Plasmacytoma

Most often thoracic spine

T1 hypointensity; T2 signal varies; variable enhancement STIR hyperintense

In two-thirds of cases lytic expansile replacement of spongiosa

Multiple myeloma

Along the whole skeleton and the skull

Multiple rounded foci; T1 hypointense; T2 and STIR hyperintense

Multiple osteolytic lesions

Bone scan: marked tracer uptake Metastases in lung, bone, liver Peak incidence in fourth decade Invades spinal canal May be seen in patients with Paget’s disease 90% local recurrence Enhance heterogeneously T2-weighted imaging very high signal with disc involvement Bone scan: normal or decreased uptake Bone scan positive Associated with opsomyoclonus (2–3%) “Dumbbell” appearance 20% of spinal cord compression in children Male-to-female ratio is 2:1 Bone scan positive Fever, leukcytosis elevated sedimentation rate 30% of systemic lymphoma have skeletal involvement Most are non-Hodgkin’s lymphoma Primary osseous lymphoma – good prognosis Poor prognostic factors are large size and M protein persistence after radiotherapy Male-to-female ratio 2:1 No/low level M protein Bone scan positive X-ray: osteopenia with punchedout lesions; bone scan not reliable, but positron emission tomography is

NEUROIMAGING OF SPINE TUMORS

Leptomeningeal metastases Leptomeningeal metastasis (LM) occurs by tumor cells infiltrating into the arachnoid and the pia mater (leptomeninges), causing focal or multifocal infiltration. This metastatic condition can be observed in solid, hematologic, and primary brain tumors (Bruna et al., 2009). LM is diagnosed in 4–15% of patients with solid tumors. The most involved neoplasms are breast (12–35%), lung cancer (10–26%), melanoma (5–25%), gastrointestinal cancer (4–14%), and cancers of unknown primary origin (1–7%) (Kesari and Batchelor, 2003; Bruna et al., 2009; Clarke et al., 2010; Le Rhun et al., 2013). However, LM is likely underestimated, considering that autopsy studies reveal that the frequency of LM is 19–40% in patients with cancer. Common primary central nervous system (CNS) tumors that may spread to the leptomeninges, so-called “drop metastasis,” are maliganant astrocytomas, ependymomas, or medulloblastomas. LM from CNS neoplasms occurs in younger patients, whereas metastases from lung or breast carcinomas occur in older patients. Abnormalities of the standard cerebrospinal fluid (CSF) analysis are observed in more than 90% of cases of LM. These abnormalities are nonspecific and include increased opening pressure (>200 mm H2O) in 46%, increased leukocytes (>4/mm3) in 57%, elevated protein (>50 mg/dL) in 76%, and decreased glucose (100 kph, rollover, ejection) ● Motorized recreational vehicles ● Bicycle struck or collision Paresthesias in extremities

Sitting position in Emergency Department Simple rear-end motor vehicle accident Ambulatory at any time Delayed onset of neck pain Absence of midline cervicalspine tenderness

Adapted from Hoffman et al. (1998).

Adapted from Stiell et al. (2001).

Table 37.1 Low-risk criteria, National Emergency X-Radiography Utilization Study (NEXUS) NEXUS low-risk criteria

750

V.M. ZOHRABIAN AND A.E. FLANDERS

Table 37.3 Canadian C-spine rule. The rule is not applicable in nontrauma cases, Glasgow coma scale (GCS) 15, unstable vital signs, age 4 mm in length on sagittal images often indicative of a complete neurologic injury (Ramon et al., 1997). Moreover, the location of cord hemorrhage has been shown to closely correspond to the neurologic level of injury, with frank hemorrhage correlating with a poor neurologic recovery (Kulkarni et al., 1987; Cotler et al., 1988; Bondurant et al., 1990; Flanders et al., 1990; Silberstein et al., 1992; Marciello et al., 1993).

Fig. 37.19. Hemorrhagic spinal cord injury. (A) Sagittal and (B) axial GRE images in a patient with a C5 flexion teardrop fracture clearly demonstrate a focus of hypointensity compatible with hemorrhage in the deoxyhemoglobin state in the right aspect of the spinal cord at C5 (yellow circles).

IMAGING OF TRAUMA OF THE SPINE

EDEMA In response to injury, focal accumulation of intracellular and interstitial fluid presents as abnormal hyperintensity on T2-weighted images (Figs 37.20 and 37.21). Spinal cord edema, colloquially referred to as a cord contusion, can occur with or without cord hemorrhage. Edema involves a variable length of spinal cord above and below the level of injury, with the length of spinal cord shown to be proportional to the degree of initial neurologic deficit. Spinal cord hemorrhage always coexists with spinal cord edema. Cord edema alone usually confers a more favorable prognosis than cord hemorrhage. Subacute progressive ascending myelopathy (SPAM) is a condition in which there is progression or ascent of the initial neurologic level of injury, frequently within days or weeks of injury. The phenomenon was first described by Frankel (1969). The etiology of this complication remains elusive, with a number of proposed etiologies including, but not limited to, vascular thrombosis, altered CSF flow, apoptosis, and infection (Yablon et al., 1989; Belanger et al., 2000; Visocchi et al., 2003;

761

Al-Ghatany et al., 2005; Schmidt, 2006). MRI shows ascent of the edema beyond the initial injury site with expansion and a rim of peripheral sparing (Belanger et al., 2000; Planner et al., 2008) (Fig. 37.22). With aggressive therapy (e.g., steroids and elevated mean arterial pressure), recovery from SPAM is generally good, although the chance of mortality is increased if the brainstem is involved. Follow-up MRI shows nearnormalization of cord changes with myelomalacia and/or atrophy. Spinal cord injury without radiographic abnormality (SCIWORA) was first defined in young children with neurologic deficits in the absence of abnormalities on plain radiographs, flexion-extension radiographs, and/or CT. SCIWORA can be found in up to 30–40% of children with SCI (Pang, 2004). SCIWORA is most commonly encountered in the cervical spine. In adults, it is believed to be secondary to hyperextension dislocation or hyperflexion sprain injury in the setting of cervical spondylosis and pre-existing stenosis (Regenbogen et al., 1986; Davis et al., 1991). The pathologic basis of

Fig. 37.20. Cord edema. (A) Sagittal STIR and (B) axial T2-weighted images in a patient with an L1 burst fracture with retropulsion demonstrates compression of the conus medullaris associated with T2-hyperintense signal, compatible with edema.

Fig. 37.21. Brown-Se´quard syndrome from stab wound to the neck. (A) Coronal STIR and (B) axial T2-weighted images demonstrate a thin, linear high-signal-intensity cleft traversing the left half of the spinal cord (yellow arrows).

762

V.M. ZOHRABIAN AND A.E. FLANDERS

Fig. 37.22. Subacute progressive ascending myelopathy. (A) Sagittal STIR image at the time of injury shows the cranial extent of cord edema at the C5–6 interspace. (B) Repeat MRI 3 days after admission demonstrates extension of the edema cephalad to approximately the C3–4 interspace and caudad to the T2–3 interspace. There is also a disc herniation at the C6–7 level (dotted arrow). (C) Sagittal T2-weighted image 10 days after steroid administration shows reduction in the spinal cord lesion, improved from the initial study. (Reproduced from Schwartz and Flanders, 2007: Figure 8.66.)

SCIWORA is compression of the thecal sac between the edge of the dorsally displaced vertebral body and the buckled ligamentum flavum. MRI is used to detect separation of the intervertebral disc, disruption of the ALL and annulus fibrosus, prevertebral hemorrhage, and parenchymal SCI (Goldberg et al., 1989; Davis et al., 1991). It has been recommended that children presenting with neurologic deficits, even if transient, undergo further evaluation with MRI (Pang, 2004).

SWELLING Normally, the spinal cord is relatively uniform in caliber, with the exception of a slight increase in diameter in the lower cervical and lower thoracic regions due to the transmission of the brachial and lumbar plexuses, respectively. Spinal cord swelling is a nondescript imaging finding, although is characterized as a focal increase in caliber center at the level of injury. These changes are best demonstrated on the T1-weighted sequences. In trauma, the focal enlargement tapers gradually cranially and caudally from the epicenter of injury, although, in some instances, it may progress cranially only. Swelling is difficult to ascertain when there is superimposed cord compression or spinal canal stenosis that effaces the surrounding subarachnoid space. Spinal cord swelling alone does not predict the extent of cord parenchymal injury.

CHRONIC SPINAL CORD INJURY Imaging is indicated in any patient suffering new loss of neurologic function, new spasticity or loss of tone, ascending neurologic deficit, or pain. The combination of plain radiography and MRI usually suffices in obtaining the needed information, although CT myelography may be performed where MRI is contraindicated or when metal artifact precludes sufficient MRI evaluation. In SCI patients, delayed or late deterioration of neurologic function has been referred to as posttraumatic progressive myelopathy, usually in the setting of spinal cord cysts or myelomalacia (Barnett et al., 1966).

Cysts The pathogenesis of the development and growth of spinal cord cysts remains controversial, although it is likely multifactorial. Imaging often falls short in accurately distinguishing between hydromyelia (ependymal lined cavity), syringomyelia (glial lined cavity), and syringohydromyelia (combined or indeterminate cyst), and as such, the general term spinal cord cyst may be used. On MRI, spinal cord cysts follow CSF signal intensity on all pulse sequences, and there may also be CSF-related flow artifacts in larger cysts (Quencer et al., 1986). These cysts often demonstrate well-defined borders with the surrounding cord parenchyma, although this distinction may sometimes be lost as a result of distortion due to

IMAGING OF TRAUMA OF THE SPINE prior hemorrhage, gliosis, or scarring (Fig. 37.23). The cysts may be simple or complex, with varying numbers of septations, and may occur above, below, or at the site of initial injury.

Myelomalacia Myelomalacia, or “soft cord,” is characterized by an absence of confluent spinal cord cysts. Histologically, myelomalacia is characterized by microcysts, reactive astrocytosis, and thickening of the pia arachnoid. Noncystic, nonenhancing signal abnormality that is hypointense to normal cord on T1-weighted images, yet greater in intensity than CSF, as well as hyperintense on T2-weighted images, likely represents myelomalacia in the appropriate clinical setting. Unlike in spinal cord cysts, myelomalacia will not parallel CSF signal intensity and its margins will usually be irregular and ill defined (Falcone et al., 1994). The cord may be normal in size, although it is frequently atrophic at the site of myelomalacia (Fig. 37.24). Spinal cord tethering is a coexistent

763

feature of myelomalacia, with loss of the subarachnoid space most commonly in the dorsal compartment, and expansion of the cord parenchyma by fibrous adhesions (Falcone et al., 1994). Uncommonly, a tear in the anterior dural margin from bony fragments may result in spinal cord herniation and also cause the cord to appear tethered.

ADVANCED IMAGING IN SPINAL CORD INJURY Although conventional MRI sequences provide macroscopic information about the spinal cord parenchyma, including degree of cord compression and amount of hemorrhage, they are limited in the functional assessment of axonal integrity. Advanced MRI techniques, most importantly diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI), have the potential to reveal structural information regarding the integrity of spinal cord white-matter tracts.

Fig. 37.23. Posttraumatic cyst. (A) Sagittal T1-weighted image demonstrates a low-intensity lesion at the T6–7 level with cord atrophy. (B) Sagittal T2-weighted and (C, D) axial T2-weighted images demonstrate a cystic lesion occupying the central portion of the spinal cord. The sagittal image demonstrates a small focus of noncystic hyperintensity at the superior aspect of the lesion likely representing myelomalacia. (E) Postcontrast fat-suppressed T1-weighted image demonstrates mild peripheral enhancement without evidence of nodularity.

764

V.M. ZOHRABIAN AND A.E. FLANDERS

Fig. 37.24. Myelomalacia. (A) Sagittal T1-weighted image in a patient with cervical decompression and anterior/posterior spinal fusion. Based on a prior CT scan (not shown), the fusion is solid. (B) Sagittal T2 and (C) sagittal STIR images demonstrate noncystic T2 hyperintensity in the spinal cord between C3 and C5 with cord thinning, compatible with myelomalacia. There is no corresponding signal abnormality in the cord on the T1-weighted image, unlike in the previously seen case of posttraumatic cyst.

DWI measures the movement rate of water molecules resulting from random (Brownian) motion. Stationary water molecules retain high signal, while molecules that move between a paired set of diffusion-sensitizing gradients lose signal as a function of the magnitude of displacement. The diffusion of water in a specific direction can be measured (apparent diffusion coefficient, or ADC), and, as such, can be classified in biologic tissues as isotropic (water diffusion at equal rates in all directions) or anisotropic (water diffuses preferentially in a particular direction). DTI is an application of DWI in which diffusion coefficients are obtained in multiple different directions (generally six or greater). Quantitative measures of diffusional anisotropy can then be calculated, the most common of which is fractional anisotropy (FA). FA ranges from 0 (purely isotropic) to 1 (highly anisotropic). With DWI or DTI, the preferred direction of water diffusion in spinal cord white matter tracts is parallel, or longitudinal, to the long axis of the axons; therefore, intact axons are highly anisotropic (high FA) (Doran and Bydder, 1990; Hajnal et al., 1991; Barkovich, 2000). Following SCI, DWI has shown alterations in ADC values even when conventional sequences have been normal (Ford et al., 1994) (Figs 37.25 and 37.26). Also, FA values in DTI have been shown to decrease in SCI in relation to the severity of neurologic injury (Shanmuganathan et al., 2008). Measures of anisotropy may even be used to evaluate injury severity and potentially treatment effects (Hauben et al., 2000). Unfortunately, unlike in the brain, diffusion imaging of

the spinal cord is technically challenging given the intrinsically small size of the cord, susceptibility artifacts from surrounding bony structures, motion from CSF pulsations, carotid/vertebral artery pulsation artifact, and respiratory motion (Ries et al., 2000). MR spectroscopy (MRS), which examines the resonance frequencies of protons on specific metabolites, has been infrequently applied to evaluation of the spinal cord given technical limitations. Patients with SCI and chronic neuropathic pain have been shown to exhibit decreased N-acetyl aspartate and increased myo-inositol in their thalami (Pattany et al., 2002). Additional experimental SCI studies have identified decreases in N-acetyl aspartate and increases in lactate (Vink et al., 1987, 1989; Falconer et al., 1996). Functional MRI (fMRI) is most commonly based on the blood oxygen level-dependent (BOLD) technique, which exploits differences in susceptibility effects from oxyhemoglobin and deoxyhemoglobin. Following neuronal activation, there is an increase in BOLD signal as cerebral blood flow, and consequently, oxyhemoglobin, increases in the activated central nervous system region. This technique is limited by small size of the spinal cord, spinal cord motion, and nearby location of large pial vessels, although studies of BOLD fMRI of brain activation following SCI have shown reorganization, expansion, and shifting of motor cortical representations in nonaffected limbs (Foltys et al., 2000; Mikulis et al., 2002; Turner et al., 2003; Stroman et al., 2004).

IMAGING OF TRAUMA OF THE SPINE

765

Fig. 37.25. Diffusion tensor imaging (DTI) of the normal human cervical spinal cord in vivo at 1.5 T demonstrates (A) expected rostral-caudal anisotropy, indicated by blue. (B) Axial color-coded DTI in the upper cervical cord shows greater diffusional anisotropy in the white matter as compared to the gray matter. (C) Fiber tractography shows expected continuity of fibers. (Reproduced from Schwartz and Flanders, 2007: Figure 13.13.)

Fig. 37.26. (A) In vivo spinal cord fiber tractography at 4.7 T shows disruption of white matter in a rat with cord hemisection (arrow). (B) The corresponding axial diffusion tensor image at the level of injury shows decreased anisotropy on the side of the hemisection (arrow) when compared with the intact side (arrowhead). (Reproduced from Schwartz and Flanders, 2007: Figure 13.12.)

SUMMARY Imaging is critically important in the appropriate management of patients presenting with spine trauma. Plain radiographs have essentially been replaced by multidetector CT, the mainstay in the diagnosis of bony abnormalities in the acute setting. CT also offers information that is useful in determining stability of the spine. MRI is advantageous in the delineation of spinal cord and

soft-tissue/ligamentous injuries, and also offers prognostic information after SCI. Advanced MRI studies have shifted the focus from macroscopic to microscopic, with diffusion-weighted techniques, such as DTI, offering structural information on the integrity of whitematter tracts in the spinal cord. A working knowledge of the imaging manifestations of injury to the spinal column and spinal cord is essential in delivering appropriate care and guiding optimal clinical outcomes.

766

V.M. ZOHRABIAN AND A.E. FLANDERS

REFERENCES Al-Ghatany M, Al-Shraim M, Levi AD et al. (2005). Pathological features including apoptosis in subacute posttraumatic ascending myelopathy. Case report and review of the literature. J Neurosurg Spine 2: 619–623. Allen Jr BL, Ferguson RL, Lehmann TR et al. (1982). A mechanistic classification of closed, indirect fractures and dislocations of the lower cervical spine. Spine (Phila Pa 1976) 7: 1–27. Anderson LD, D’Alonzo RT (1974). Fractures of the odontoid process of the axis. J Bone Joint Surg Am 56: 1663–1674. Baker LL, Goodman SB, Perkash I et al. (1990). Benign versus pathologic compression fractures of vertebral bodies: assessment with conventional spin-echo, chemical-shift, and STIR MR imaging. Radiology 174: 495–502. Barkovich AJ (2000). Concepts of myelin and myelination in neuroradiology. AJNR Am J Neuroradiol 21: 1099–1109. Barnett HJ, Botterell EH, Jousse AT et al. (1966). Progressive myelopathy as a sequel to traumatic paraplegia. Brain 89: 159–174. Belanger E, Picard C, Lacerte D et al. (2000). Subacute posttraumatic ascending myelopathy after spinal cord injury. Report of three cases. J Neurosurg 93: 294–299. Biffl WL, Moore EE, Offner PJ et al. (2001). Blunt carotid and vertebral arterial injuries. World J Surg 25: 1036–1043. Bondurant FJ, Cotler HB, Kulkarni MV et al. (1990). Acute spinal cord injury. A study using physical examination and magnetic resonance imaging. Spine (Phila Pa 1976) 15: 161–168. Cotler HB, Kulkarni MV, Bondurant FJ (1988). Magnetic resonance imaging of acute spinal cord trauma: preliminary report. J Orthop Trauma 2: 1–4. Davis SJ, Teresi LM, Bradley Jr WG et al. (1991). Cervical spine hyperextension injuries: MR findings. Radiology 180: 245–251. Denis F (1984). Spinal instability as defined by the threecolumn spine concept in acute spinal trauma. Clin Orthop Relat Res: 65–76. Doran M, Bydder GM (1990). Magnetic resonance: perfusion and diffusion imaging. Neuroradiology 32: 392–398. Edeiken-Monroe B, Wagner LK, Harris Jr JH (1986). Hyperextension dislocation of the cervical spine. AJR Am J Roentgenol 146: 803–808. Elijovich L, Kazmi K, Gauvrit JY et al. (2006). The emerging role of multidetector row CT angiography in the diagnosis of cervical arterial dissection: preliminary study. Neuroradiology 48: 606–612. Falcone S, Quencer RM, Green BA et al. (1994). Progressive posttraumatic myelomalacic myelopathy: imaging and clinical features. AJNR Am J Neuroradiol 15: 747–754. Falconer JC, Liu SJ, Abbe RA et al. (1996). Time dependence of N-acetyl-aspartate, lactate, and pyruvate concentrations following spinal cord injury. J Neurochem 66: 717–722. Flanders AE, Schaefer DM, Doan HT et al. (1990). Acute cervical spine trauma: correlation of MR imaging findings with degree of neurologic deficit. Radiology 177: 25–33.

Flanders AE, Tartaglino LM, Friedman DP et al. (1992). Magnetic resonance imaging in acute spinal injury. Semin Roentgenol 27: 271–298. Flanders AE, Spettell CM, Tartaglino LM et al. (1996). Forecasting motor recovery after cervical spinal cord injury: value of MR imaging. Radiology 201: 649–655. Foltys H, Kemeny S, Krings T et al. (2000). The representation of the plegic hand in the motor cortex: a combined fMRI and TMS study. Neuroreport 11: 147–150. Ford JC, Hackney DB, Alsop DC et al. (1994). MRI characterization of diffusion coefficients in a rat spinal cord injury model. Magn Reson Med 31: 488–494. Frankel HL (1969). Ascending cord lesion in the early stages following spinal injury. Paraplegia 7: 111–118. Friedman DP, Flanders AE (1992). Unusual dissection of the proximal vertebral artery: description of three cases. AJNR Am J Neuroradiol 13: 283–286. Friedman D, Flanders A, Thomas C et al. (1995). Vertebral artery injury after acute cervical spine trauma: rate of occurrence as detected by MR angiography and assessment of clinical consequences. AJR Am J Roentgenol 164: 443–447, discussion 448–449. Goldberg AL, Rothfus WE, Deeb ZL et al. (1989). Hyperextension injuries of the cervical spine. Magnetic resonance findings. Skeletal Radiol 18: 283–288. Hackney DB, Finkelstein SD, Hand CM et al. (1994). Postmortem magnetic resonance imaging of experimental spinal cord injury: magnetic resonance findings versus in vivo functional deficit. Neurosurgery 35: 1104–1111. Hajnal JV, Doran M, Hall AS et al. (1991). MR imaging of anisotropically restricted diffusion of water in the nervous system: technical, anatomic, and pathologic considerations. J Comput Assist Tomogr 15: 1–18. Harris JH, Edeiken-Monroe B (1987). The radiology of acute cervical spine trauma, Williams & Wilkins, Baltimore. Harris Jr JH, Carson GC, Wagner LK (1994a). Radiologic diagnosis of traumatic occipitovertebral dissociation: 1. Normal occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 162: 881–886. Harris Jr JH, Carson GC, Wagner LK et al. (1994b). Radiologic diagnosis of traumatic occipitovertebral dissociation: 2. Comparison of three methods of detecting occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 162: 887–892. Hauben E, Nevo U, Yoles E et al. (2000). Autoimmune T cells as potential neuroprotective therapy for spinal cord injury. Lancet 355: 286–287. Hoffman JR, Wolfson AB, Todd K et al. (1998). Selective cervical spine radiography in blunt trauma: methodology of the National Emergency X-Radiography Utilization Study (NEXUS). Ann Emerg Med 32: 461–469. Hoffman JR, Mower WR, Wolfson AB et al. (2000). Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency X-Radiography Utilization Study Group. N Engl J Med 343: 94–99.

IMAGING OF TRAUMA OF THE SPINE Jung HS, Jee WH, Mccauley TR et al. (2003). Discrimination of metastatic from acute osteoporotic compression spinal fractures with MR imaging. Radiographics 23: 179–187. Kerslake RW, Jaspan T, Worthington BS (1991). Magnetic resonance imaging of spinal trauma. Br J Radiol 64: 386–402. Kulkarni MV, Mcardle CB, Kopanicky D et al. (1987). Acute spinal cord injury: MR imaging at 1.5 T. Radiology 164: 837–843. Kulkarni MV, Bondurant FJ, Rose SL et al. (1988). 1.5 tesla magnetic resonance imaging of acute spinal trauma. Radiographics 8: 1059–1082. Marciello MA, Flanders AE, Herbison GJ et al. (1993). Magnetic resonance imaging related to neurologic outcome in cervical spinal cord injury. Arch Phys Med Rehabil 74: 940–946. McArdle CB, Crofford MJ, Mirfakhraee M et al. (1986). Surface coil MR of spinal trauma: preliminary experience. AJNR Am J Neuroradiol 7: 885–893. Metz GA, Curt A, Van De Meent H et al. (2000). Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J Neurotrauma 17: 1–17. Mikulis DJ, Jurkiewicz MT, Mcilroy WE et al. (2002). Adaptation in the motor cortex following cervical spinal cord injury. Neurology 58: 794–801. Mirvis SE, Geisler FH, Jelinek JJ et al. (1988). Acute cervical spine trauma: evaluation with 1.5-T MR imaging. Radiology 166: 807–816. Pang D (2004). Spinal cord injury without radiographic abnormality in children, 2 decades later. Neurosurgery 55: 1325–1342. discussion 1342–1343. Pattany PM, Yezierski RP, Widerstrom-Noga EG et al. (2002). Proton magnetic resonance spectroscopy of the thalamus in patients with chronic neuropathic pain after spinal cord injury. AJNR Am J Neuroradiol 23: 901–905. Planner AC, Pretorius PM, Graham A et al. (2008). Subacute progressive ascending myelopathy following spinal cord injury: MRI appearances and clinical presentation. Spinal Cord 46: 140–144. Pongpornsup S, Wajanawichakorn P, Danchaivijitr N (2009). Benign versus malignant compression fracture: a diagnostic accuracy of magnetic resonance imaging. J Med Assoc Thai 92: 64–72. Pratt ES, Green DA, Spengler DM (1990). Herniated intervertebral discs associated with unstable spinal injuries. Spine (Phila Pa 1976) 15: 662–666. Quencer RM, Sheldon JJ, Post MJ et al. (1986). MRI of the chronically injured cervical spinal cord. AJR Am J Roentgenol 147: 125–132. Ramon S, Dominguez R, Ramirez L et al. (1997). Clinical and magnetic resonance imaging correlation in acute spinal cord injury. Spinal Cord 35: 664–673. Regenbogen VS, Rogers LF, Atlas SW et al. (1986). Cervical spinal cord injuries in patients with cervical spondylosis. AJR Am J Roentgenol 146: 277–284. Ries M, Jones RA, Dousset V et al. (2000). Diffusion tensor MRI of the spinal cord. Magn Reson Med 44: 884–892.

767

Schaefer DM, Flanders AE, Osterholm JL et al. (1992). Prognostic significance of magnetic resonance imaging in the acute phase of cervical spine injury. J Neurosurg 76: 218–223. Schmidt BJ (2006). Subacute delayed ascending myelopathy after low spine injury: case report and evidence of a vascular mechanism. Spinal Cord 44: 322–325. Schwartz E, Flanders A (2007). Spinal Trauma: Imaging, Diagnosis, and Management. Wolters Kluwer Health/ Lippincott, Willliams & Wilkins, Philadelphia. Shamoun JM, Riddick L, Powell RW (1999). Atlanto-occipital subluxation/dislocation: a “survivable” injury in children. Am Surg 65: 317–320. Shanmuganathan K, Gullapalli RP, Zhuo J et al. (2008). Diffusion tensor MR imaging in cervical spine trauma. AJNR Am J Neuroradiol 29: 655–659. Silberstein M, Tress BM, Hennessy O (1992). Prediction of neurologic outcome in acute spinal cord injury: the role of CT and MR. AJNR Am J Neuroradiol 13: 1597–1608. Sixta S, Moore FO, Ditillo MF et al. (2012). Screening for thoracolumbar spinal injuries in blunt trauma: an Eastern Association for the Surgery of Trauma practice management guideline. J Trauma Acute Care Surg 73: S326–S332. Spinal Cord Injury: Facts and Figures at a Glance. Available at: www.nscisc.uab.edu. Accessed April 10, 2015. Stiell IG, Wells GA, Vandemheen KL et al. (2001). The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA 286: 1841–1848. Stiell IG, Clement CM, Mcknight RD et al. (2003). The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med 349: 2510–2518. Stroman PW, Kornelsen J, Bergman A et al. (2004). Noninvasive assessment of the injured human spinal cord by means of functional magnetic resonance imaging. Spinal Cord 42: 59–66. Turner JA, Lee JS, Schandler SL et al. (2003). An fMRI investigation of hand representation in paraplegic humans. Neurorehabil Neural Repair 17: 37–47. Vink R, Knoblach SM, Faden AI (1987). 31P magnetic resonance spectroscopy of traumatic spinal cord injury. Magn Reson Med 5: 390–394. Vink R, Noble LJ, Knoblach SM et al. (1989). Metabolic changes in rabbit spinal cord after trauma: magnetic resonance spectroscopy studies. Ann Neurol 25: 26–31. Visocchi M, Di Rocco F, Meglio M (2003). Subacute clinical onset of postraumatic myelopathy. Acta Neurochir (Wien) 145: 799–804. discussion 804. Wagner A, Albeck MJ, Madsen FF (1992). Diagnostic imaging in fracture of lumbar vertebral ring apophyses. Acta Radiol 33: 72–75. White AA, Panjabi MM (1991). Clinical Biomechanics of the Spine, J.B. Lippincott Co, Philadelphia. Yablon IG, Ordia J, Mortara R et al. (1989). Acute ascending myelopathy of the spine. Spine (Phila Pa 1976) 14: 1084–1089.

Handbook of Clinical Neurology, Vol. 136 (3rd series) Neuroimaging, Part II J.C. Masdeu and R.G. Gonza´lez, Editors © 2016 Elsevier B.V. All rights reserved

Chapter 38

Hereditary and metabolic myelopathies PETER HEDERA* Department of Neurology, Vanderbilt University, Nashville, TN, USA

Abstract Hereditary and metabolic myelopathies are a heterogeneous group of neurologic disorders characterized by clinical signs suggesting spinal cord dysfunction. Spastic weakness, limb ataxia without additional cerebellar signs, impaired vibration, and positional sensation are hallmark phenotypic features of these disorders. Hereditary, and to some extent, metabolic myelopathies are now recognized as more widespread systemic processes with axonal loss and demyelination. However, the concept of predominantly spinal cord disorders remains clinically helpful to differentiate these disorders from other neurodegenerative conditions. Furthermore, metabolic myelopathies are potentially treatable and an earlier diagnosis increases the likelihood of a good clinical recovery. This chapter reviews major types of degenerative myelopathies, hereditary spastic paraplegia, motor neuron disorders, spastic ataxias, and metabolic disorders, including leukodystrophies and nutritionally induced myelopathies, such as vitamin B12, E, and copper deficiencies. Neuroimaging studies usually detect a nonspecific spinal cord atrophy or demyelination of the corticospinal tracts and dorsal columns. Brain imaging can be also helpful in myelopathies caused by generalized neurodegeneration. Given the nonspecific nature of neuroimaging findings, we also review metabolic or genetic assays needed for the specific diagnosis of hereditary and metabolic myelopathies.

Hereditary and metabolic myelopathies are a heterogeneous group of neurologic disorders characterized by a clinical syndrome indicating predominant spinal cord involvement, even though many times the underlying pathology may be more widespread. Phenotypic features suggesting spinal cord dysfunction can be attributed to variable degrees of degeneration of corticospinal (pyramidal) and spinocerebellar tracts or dorsal columns (Hedera, 2011). Additionally, selective degeneration of lower motor neurons in the anterior horns can be occasionally present. Progressive and gradual spasticity with a wide range of weakness severity points to pyramidal tract damage that can be combined with flaccid weakness and hypotonia if degeneration of lower motor neurons is also present. Limb ataxia without central cerebellar signs, such as oculomotor abnormalities and dysarthria, is caused by neurodegeneration of the spinocerebellar tracts, while relatively selective impairment of vibration and positional sensation, resulting in sensory

ataxia with positive Romberg test, reflects the involvement of dorsal columns. Neurodegeneration is most pronounced in the terminal segments of the longest axons, accounting for a preferential involvement of the distal parts of the dorsal column in the cervical cord and the distal parts of the corticospinal tract in the lumbar spinal cord (Fink and Hedera, 1999). Thus, motor and sensory signs in hereditary and metabolic myelopathies are mostly limited to the lower extremities, encompassing a syndrome of spastic paraparesis with deep-tendon hyperreflexia, positive Babinski signs, and bladder hyperactivity due to detrusor sphincter dyssynergia. Variable involvement of the upper extremities with upper motor neuron signs can be classified as spastic quadriparesis. The peripheral nervous system is commonly affected in a similar fashion, especially in metabolic myelopathies with coexisting distal peripheral nerve damage. The combination of central and peripheral nervous system involvement

*Correspondence to: Peter Hedera, M.D., Ph.D., Department of Neurology, Vanderbilt University, 465 21st Avenue South, 6140 MRB III, Nashville TN 37232-8552, USA. E-mail: [email protected]

770 P. HEDERA comprises a syndrome of myeloneuropathy, charactermild with a full range of passive motions. Examination ized by a relatively rare association of hypo/areflexia of upper extremities may show a mild hyperreflexia with with upgoing toes. Tr€omner and Hoffmann signs, but other hallmarks of Pathologic changes in hereditary and metabolic pyramidal involvement, most notably loss of dexterity, myelopathies are relatively stereotypical with a variable should be absent. Stumbling caused by difficulties with combination of diffuse axonal degeneration and demyfoot dorsiflexion is the most common heralding sympelination. Axonal loss can be either a primary neuropathtom of spastic gait and usually progresses into a characologic process of dying-back axonopathy or a secondary teristic scissoring spastic gait. Some individuals are phenomenon due to severe myelin loss (Fink, 2013). Neuunaware of the disease even if unequivocal signs of spasroimaging studies usually detect a nonspecific spinal ticity can be detected on a clinical examination. Muscle cord atrophy and additional molecular or biochemical weakness is less evident and tends to be limited to distal diagnostic tests are necessary to determine the etiology leg muscles. More widespread weakness can be present of myelopathy. Myelin loss can selectively affect either in advanced stages of the disease and affect also proxidescending corticospinal tracts or ascending dorsal colmal muscles. HSP is progressive and severe spasticity umns, resulting in usually symmetric demyelination of and weakness can lead to wheelchair dependency. Howthese structures. Demyelination is more common in ever, there is a significant intrafamilial and interfamilial acquired metabolic myelopathies or inherited metabolic variability in the degree of disability (Fink and leukoencephalopathies, but also these neuroimaging Hedera, 1999). findings are rather nonspecific. Neurodegenerative myeHSP can be classified according to the mode of inherlopathies are frequently associated with additional strucitance or clinically, based on the presence of additional tural changes in the central nervous system and magnetic neurologic and systemic signs (Finsterer et al., 2012). resonance imaging (MRI) of the brain may provide furHSP is categorized as pure or uncomplicated when only ther diagnostic clues, such as abnormalities of the corpus spastic paraparesis is detected. The presence of addicallosum (CC) or leukoencephalopathy, that can help the tional neurologic signs, such as ataxia, amyotrophy, diagnostic process (Finsterer et al., 2012; Fink, 2013). optic nerve atrophy, deafness, cognitive impairment Hereditary myelopathies can be classified based on a due to developmental delay, leukodystrophy, or parkincombination of clinical, genetic, and pathologic features. sonism is typical for complex or complicated HSP. Although some overlap exists, the most commonly More than 50 disease-causing genes have been idenused categorization includes four major types: hereditified as a cause of HSP, making it one of the most genettary spastic paraplegia (HSP), motor neuron disorders, ically heterogeneous neurodegenerative syndromes spastic ataxias, and metabolic disorders, including (Finsterer et al., 2012; Fink, 2013). However, mutations leukodystrophies. Inherited metabolic disorders have in these multiple genes result in a quite stereotypical a considerable overlap with metabolic myelopathies clinical phenotype with only very limited genotype– caused by nutrient deficiencies. For example, myelopaphenotype correlation. The prevalence of HSP may thy caused by insufficient vitamin E levels can be both differ in different ethnic groups but it has been estimated acquired and inherited as an inborn metabolic defect. to affect up to 1 per 10 000 individuals (Erichsen et al., Toxic myelopathies are also classified with metabolic 2009). All HSP cases are caused by inherited or de novo types and similarly, toxic accumulation of certain commutations. All three Mendelian modes of inheritance, pounds can induce deficiency of several nutritional autosomal-dominant (AD), autosomal-recessive (AR), factors. and X-linked, have been identified as a potential cause of HSP (Finsterer et al., 2012). A detailed review of various genetic types of HSP is beyond the scope of this HEREDITARY SPASTIC PARAPLEGIAS chapter and we will only focus on the most common The typical clinical presentation of HSP is a gradual and genetic causes of HSP. progressive spastic weakness of the lower extremities, Up to 80% of all cases of HSP are caused by mutations associated with variable degrees of impaired vibration in genes inherited in AD fashion that includes vertical sensation and autonomic dysfunction with bladder transmission with multiple generations affected by hyperactivity (Fink and Hedera, 1999; Finsterer et al., HSP, the occurrence of male-to-male transmission, a 2012). Spastic paraparesis is mostly symmetric, but a 50% risk for offspring in each successive generation, slightly asymmetric onset can be seen in some patients. and equal frequency of the disease between males and A striking asymmetry of spastic paraparesis should females. AD inheritance may not be obvious in small prompt further investigation to exclude structural and kindreds, especially when parents are not available for compressive causes. Spasticity in HSP tends to be accenclinical examination. De novo mutations or reduced pentuated by walking and resting spasticity can be relatively etrance are other possible common causes of absent

HEREDITARY AND METABOLIC MYELOPATHIES 771 family history. Approximately 25% of patients with HSP correlation with the disability or duration of the disease caused by mutations in AD genes have apparently spo(Hedera et al., 2005; Rezende et al., 2015). Furthermore, radic disease and the possibility of AD HSP cannot be different genetic types of AD HSP appear to differ in dismissed, even if a family history of HSP is lacking. the amount of spinal cord atrophy and SPG3A patients Uncomplicated HSP is a typical phenotype associated may have only borderline changes in spinal cord size. with AD forms. Mutations in spast (spastin) gene are The degree of spinal cord atrophy is most prominent responsible for about 40% of all HSP cases known as in patients with SPG6 and SPG8 who have signs of severe SPG4 (SPG refers to spastic gait and 4 refers to the order cord atrophy in the cervical segments and especially of identification of this genetic locus), followed by atlasthoracic segments (Hedera, 2013) (Fig. 38.1). tin-1 mutations causing SPG3A (Finsterer et al., 2012). Quantitative neuroimaging may be helpful in the The age of disease onset in patients with SPG4 ranges further diagnostic evaluations of HSP and especially from infancy to the seventh decade, even though the diffusion tensor imaging (DTI) can provide additional majority of patients develop the disease in the second information about white-matter fiber integrity and third decades. Contrary to this, SPG3A accounts (Aghakhanyan et al., 2014). This MR technique can demfor the most cases with an early onset of symptoms, onstrate the diffusion properties of water molecules defined as less than 10 years of age, and 10–15% of all in vivo, indirectly reflecting the integrity changes of AD HSP (Finsterer et al., 2012). Other causative genes axons and myelin sheath. Fractional anisotropy (FA) for AD HSP are considerably less frequent and SPG6, and apparent diffusion coefficient (ADC) values reflect caused by mutations in the nonimprinted Prader– the integrity of the white matter and it can be used to Willi/Angelman syndrome 1/NIPA1 gene, SPG8 caused evaluate the degeneration of white-matter tracts, includby strumpellin mutations and SPG31, caused by mutaing descending corticospinal tracts. DT MRI studies in a tions in the receptor accessory protein 1/REEP1 gene case series of patients with spastin (SPG4) mutations are responsible for the majority of non-SPG4 and nonshowed widespread white-matter abnormalities, includSPG3A AD HSP cases. The rate of progression varies ing axons forming the CC, subcortical white matter of considerably among different genetic types of HSP the motor system and limbic tracts, while conventional and also within families, but certain clinical features volumetric analysis in these patients did not identify allows some degree of genotype–phenotype correlation. any macroscopic abnormalities in the brain of spinal Patients with SPG3A typically experience very slow discord. FA of the CC and pyramidal tracts correlated with ability progression with maintained walking into the disease severity. These findings strongly support the sixth or seventh decades, while patients with SPG6 and concept that AD HSP is a dying-back axonopathy. Thus, SPG8 tend to have a more aggressive course, with wheelalthough the bodies of the affected neurons are in the chair dependency early in the course of the disease. Howcortex, on MRI HSP is characterized by relative sparing ever, definite diagnosis can be achieved only by of the cortical mantle and marked damage to the distal molecular genetic testing. portions of the corticospinal tracts, extending into the Conventional spinal cord and brain MRI findings can spinal cord (Aghakhanyan et al., 2014). be normal or very nonspecific in patients with AD HSP AR HSPs are considerably less common than AD who have an uncomplicated HSP type (Hedera et al., HSP and, in general, they tend to have an early age of 2005). Neuroimaging of the neuroaxis is warranted in onset and a complicated phenotype that frequently patients who have an apparently sporadic disease or if includes mental retardation, seizures, dystonia, and the formal diagnosis of HSP has not been established ataxia (Finsterer et al., 2012). AR inheritance is supin their families to exclude compressive and inflammaported by multiple affected children born to unaffected tory causes of myelopathy. MRI has been used to meaparents. This type of HSP is more prevalent in consansure spinal cord size in AD HSP patients. Most studies guineous populations, but common types of AR HSP included SPG4 patients and smaller anteroposterior are frequently encountered in nonconsanguineous mardiameter of the thoracic spinal cord was found in riages. Pure AR HSP is relatively rare and SPG7, caused 16 SPG4 patients. This pattern is characterized by preferby mutations in the paraplegin gene and accounting for ential damage to the anterolateral portions of the spinal about 5% of all AR cases, is the most common type of cord, which harbors the corticospinal tracts. However, uncomplicated AR HSP. SPG7 patients are otherwise qualitative analysis of spinal cord MRI studies in these undistinguishable from other forms of HSP and neuropatients is essentially normal in the majority of these imaging is unrevealing (Finsterer et al., 2012). Moreover, patients. Quantitative analysis of various genetic types clinical presentation in some patients with SPG7 may of HSP, including SPG3A, SPG4, SPG6, and SPG8, sugoverlap with a complex phenotype, further blurring the gested that spinal cord atrophy is a common feature of boundaries among various genetic types of AR HSP. these genetic types of HSP but there is no close The vast majority of complicated AR HSP have

772

P. HEDERA

Fig. 38.1. Magnetic resonance imaging of spine in a patient with mutation in the NIPA1 (SPG6) gene with severe atrophy of the cervical (A, B) and thoracic spinal cord (C, D) (arrows). No signal changes within the spinal cord are present.

characteristic neuroradiologic abnormalities, but they are almost exclusively detected on brain imaging studies. Spinal cord imaging in general is either normal or shows a nonspecific spinal cord atrophy that does not help to differentiate from other types of HSP. Mutations in SPG11, caused by mutations in the spastacsin gene, are the most common cause of complicated AR HSP and their frequency increases in up to 80% of the patients when considering SPG cases with thin or absent CC and mental retardation (Paisan-Ruiz et al., 2008) (Fig. 38.2A, B). However, hypoplastic CC is not entirely specific for SPG11 and other forms of AR HSP are associated with the same CC abnormalities. The other common types of AR HSP with thin CC include SPG15, caused by mutations in the spastizin gene, and SPG35, caused by mutations in the fatty acid 2-hydroxylase (FA2H) gene. Additional radiographic abnormalities in AR HSP include various degrees of increased signal on T2-weighted or fluid-attenuated inversion recovery (FLAIR) images in subcortical white matter, suggesting radiologic overlap with other genetic leukodystrophies. Increased white-matter signal in the forceps minor area of CC, resembling “ears of the lynx,” has been suggested as another typical neuroimaging feature of SPG11 (Fig. 38.2C) (Riverol et al., 2009). Widespread leukodystrophy can be seen in SPG5A, caused by mutations in the cytochrome P450, family 7, subfamily B, polypeptide 1 gene, CYP7B1, and in SPG35, caused by mutations in

the FA2H gene. FAH2-associated HSP patients have a high incidence of seizures, parkinsonism, and dystonia with radiologic evidence of thin CC, white-matter disease, abnormal signal from basal ganglia, and cerebellar atrophy (Cao et al., 2013) (Fig. 38.2D–F). Interestingly, SPG35 can also present as neurodegeneration with brain iron accumulation without any other neuroradiologic abnormalities. X-linked disorders are the rarest form of HSP. They are inherited in a recessive X-linked mode affecting predominantly only males; however, female carriers can rarely manifest a mild disease because of unfavorable X-chromosome inactivation. SPG1 is caused by mutations in the L1 cell adhesion molecule (L1CAM) gene and this condition is allelic with X-linked aqueductal stenosis and hydrocephalus, and MASA syndrome (mental retardation, aphasia, shuffling gait and adducted thumbs). Thus, the detection of obstructive hydrocephalus is highly suggestive of this diagnosis (Takahashi et al., 1997) (Fig. 38.3). Spasticity in these patients is not a secondary problem due to the increased intracranial pressure, because successfully shunted patients still experience a progressive gait disorder caused by neurodegeneration of the corticospinal tracts. SPG2 is caused by mutations in the proteolipid protein 1 (PLP1) gene. Different mutations in the same gene are associated with Pelizaeus–Merzbacher disease, which is characterized by severe hypomyelination of the

Fig. 38.2. (A, B) Atrophy of the corpus callosum (arrows) in two different patients with mutations in the spatacsin gene (SPG11). (C) “Ears of the lynx” appearance of the forceps minor region in SPG11 patient. Imaging characteristics of SPG35 caused by mutations in FAH gene with atrophy of corpus callosum (D, yellow arrowhead) and cortical atrophy (D, blue arrowhead), atrophy of the cerebellum (E, yellow arrowhead) and cervical spinal cord (E, blue arrowhead), hypointensity of the globus pallidus (F, yellow arrowheads) and white-matter hyperintensity (F, blue arrowheads). (D–F reproduced from Cao et al., 2013, with permission.)

Fig. 38.3. Cranial magnetic resonance imaging of a patient with SPG1 after a ventriculoperitoneal shunt. T1-weighted axial images demonstrated a small fourth ventricle (A), an obstructed aqueduct of Sylvius (B), a fused thalamus (C), marked dilatation of the lateral ventricles, hypoplastic white matter, thin cortical mantle, and agenesis of the corpus callosum and septum pellucidum (D). (Reproduced from Takahashi et al., 1997, with permission.)

774

P. HEDERA

Fig. 38.4. Cranial magnetic resonance imaging of a patient with PLP1 mutation and clinical phenotype of hereditary spastic paraplegia. Signs of hypomyelination with posterior predominance are present (arrows).

white matter in the cerebral hemispheres, cerebellum, and brainstem, and devastating neurologic deficits, including congenital nystagmus, progressive spasticity, and a profound developmental delay. SPG2 can present as both complicated and pure HSP (Finsterer et al., 2012). Neuroimaging of SPG2 is nonspecific and white-matter changes vary from patchy abnormalities to more diffuse leukoencephalopathy on T2-weighted or FLAIR images. However, the T2 signal hypointensity of the white matter is milder in hypomyelination than in demyelination and other white-matter lesions (Fig. 38.4). Definite diagnosis can be established only by molecular testing.

MOTOR NEURON DISORDERS Degeneration of both upper and lower motor neurons with sparing of cerebellar, extrapyramidal, and sensory neuronal circuits is the hallmark of motor neuron disorders (Strong and Gordon, 2005). Pathologic changes are more diffuse, with a variable involvement of motor cortex, brainstem motor nuclei, and spinal cord. Spinal cord pathology in motor neuron disorders consists of a combination of degeneration of the descending corticospinal tracts and a neuronal loss in the anterior horns of the spinal cord. Thus, motor neuron disorders are not pure myelopathies but may commonly mimic isolated myelopathies because of a relative paucity of bulbar signs and lower motor neuron signs. Amyotrophic lateral sclerosis (ALS) is a prototypical motor neuron disorder (Borasio and Miller, 2001). The clinical picture of ALS reflects the combined neurodegeneration of both types of motor neurons, with spasticity, loss of dexterity with slow movements, and hyperreflexia caused by dysfunction of upper motor neurons, and weakness, atrophy, and fasciculations

caused mostly by lower motor neuron loss. These clinical findings, together with a focal onset of symptoms, and a gradual but relentless progression should raise a high degree of clinical suspicion that points to ALS. Electromyography is currently the diagnostic standard to confirm a suspected case of ALS. Some patients with ALS may experience symptoms limited to spasticity of the lower extremities. Moreover, certain genetic types of ALS have a higher tendency to present as an ascending spasticity. Overall, approximately 10% of ALS cases are genetically based and inherited types of ALS tend to have an earlier age of onset. ALS type 2 is an AR condition caused by mutations in the alsin gene (Chandran et al., 2007). ALS2 has been associated with three different phenotypes: juvenile ALS with age of onset between 3 and 20 years; juvenile primary lateral sclerosis (PLS), where there is no degeneration of lower motor neurons; and infantile onset of HSP with spasticity affecting only the lower extremities. Progressive spasticity of the lower extremities, otherwise indistinguishable from HSP, is a typical presenting symptom, and the upper limbs are usually affected after the first decade of disease duration. The disease progression is slower than with other types of ALS, with survival for many decades. ALS4 is another type of ALS that can have an isolated spasticity with absent or minimal lower motor neuron signs. This type of ALS is caused by mutations in the senataxin gene. Mutations in the same gene are also associated with ataxia with oculomotor apraxia type II, characterized by severe gait ataxia, dystonia, and oculomotor apraxia in one-half of patients, and a mixed motor and sensory neuropathy with distal weakness. Neuroimaging studies are crucial to rule out compressive or inflammatory myelopathies in these patients.

HEREDITARY AND METABOLIC MYELOPATHIES Spinal cord atrophy was suggested to correlate with the progression of disease but considerable variability limits usefulness of quantitative MRI spinal cord imaging because of a diverse clinical phenotype of ALS (Branco et al., 2014). Advanced neuroimaging techniques, such as DTI, are a promising way to evaluate disease progression, especially stemming from degeneration of upper motor neurons. DTI abnormalities, such as reduced FA and increased ADC, may be detected in patients with ALS who do not have any abnormal structural findings in the spinal cord on conventional MRI (Wang et al., 2014) (Fig. 38.5). PLS is defined by slowly progressive corticospinal dysfunction that is strictly isolated to upper motor

775

neurons and, in contrast to ALS, neurodegeneration of lower motor neuron must be absent (Tartaglia et al., 2007). The relationship between PLS and ALS remains controversial and PLS has been seen as a variant of ALS or a separate entity (Strong and Gordon, 2005). Most of the proposed diagnostic criteria for the differentiation of PLS from ALS require the absence of denervation for 5 years after the onset of spasticity. Most PLS is considered sporadic and, in contrast to ALS, survival is much longer, with less relentless disease progression. The onset of spasticity is typically asymmetric, including a spastic lower-extremity monoparesis that has to be differentiated from other forms of myelopathy. Most patients later develop severe pseudobulbar symptoms

Fig. 38.5. Normally appearing C3 segment of the cervical spine of a patient with amyotrophic lateral sclerosis (ALS) (1a) and agematched control (1b). ALS (2a) causes increased fractional anisotropy compared to control (2b), and abnormal apparent diffusion coefficient (3a ALS and 3b normal control). (Reproduced from Wang et al., 2014, with permission.)

776

P. HEDERA

Fig. 38.6. Cervical spine magnetic resonance imaging in a patient with primary lateral sclerosis with signal hyperintensity on (A) T2-weighted and (B, C) fluid-attenuated inversion recovery images corresponding to corticospinal tracts (arrows).

affecting speech and swallowing, clearly delineating the clinical picture of PLS (Tartaglia et al., 2007). Routine imaging in PLS is also unremarkable. Cortical thinning in the primary motor cortex can be seen in both ALS and PLS but the magnitude of thinning is greater in PLS patients. In PLS patients, cortical thinning is highly localized to the precentral gyrus and adjacent paracentral region, with only one very small cluster in the lateral orbitofrontal cortex. Spinal cord imaging may show a nonspecific spinal cord atrophy and DTI can reveal a widespread degeneration of pyramidal tracts. Occasionally, signal hyperintensity on T2-weighted and FLAIR images can be seen in PLS patients, likely reflecting axonal degeneration with secondary demyelination of these white-matter tracts (Fig. 38.6). However, they cannot be considered diagnostic as similar MRI findings have been reported in some ALS patients and this finding alone does not allow ALS to be differentiated from PLS (Nobue et al., 2011).

SPASTIC ATAXIAS Clinical diagnosis of ataxia, coupled with dysmetria, dysdiadochokinesis, scanning dysarthria, and oculomotor

abnormalities, is usually straightforward and indicates cerebellar pathology or disruption of cerebellar outflow circuits. Neurodegenerative ataxias can be classified as “pure” cerebellar ataxias, with symptoms limited to cerebellar dysfunction and spinocerebellar or spastic ataxias that are associated with additional neurologic problems, including spasticity (de Bot et al., 2012). Several inherited types of neurodegenerative ataxias may occasionally present as myelopathies because of severe spasticity affecting the lower extremities, and the differentiation between these entities is sometimes virtually impossible on clinical grounds alone. Spinocerebellar ataxias (SCA) are degenerative ataxias involving the cerebellum as well as the brainstem, spinal cord, and cerebrum. Although SCAs involve several systems in the central nervous system, the distribution of the pathologic lesions is unique to each degenerative ataxia. The estimated prevalence of AD SCAs can be up to 3.0 per 100 000 individuals worldwide (Erichsen et al., 2009). Prominent spasticity caused by corticospinal tract involvement is particularly common in SCA types 1, 2, and 3, accounting for approximately 40% of all AD SCA. All three types of SCA are caused by the expansion of CAG repeats and expansion in

HEREDITARY AND METABOLIC MYELOPATHIES

777

Fig. 38.7. (A, B) Cranial magnetic resonance imaging (MRI) of a patient with SCA3 who presented with spastic paraplegia phenotype: MRI showed cerebellar atrophy and a mild cervical spine atrophy.

successive generations may lead to genetic anticipation with an earlier age of onset and more severe clinical phenotypes in children of affected parents. Motor phenotypes vary considerably among SCAs and SCA3, also known as Machado–Joseph disease, most commonly overlaps with HSP (Hedera, 2009) (Fig. 38.7). Several SCA3 patients manifesting only a spastic gait without any obvious cerebellar signs have been diagnosed with HSP rather than SCA. The emergence of oculomotor abnormalities should prompt the diagnosis of SCA rather than HSP because cerebellar signs are extremely rare in AD HSP. Spasticity can be a feature of other types of SCA, such as SCA8 and SCA17, and molecular diagnosis is necessary for the definitive differentiation of HSP and SCA in these patients. Early and severe spasticity is also a feature of spastic ataxias, which are a heterogeneous group of hereditary ataxias with both AD and AR inheritance (de Bot et al., 2012). Spastic ataxia of Charlevoix–Saguenay (SACS) is most frequent among French-Canadians. It is an AR condition caused by mutations in the sacsin gene and progressive spasticity with nystagmus is the most common initial finding. SACS is progressive, and the emergence of profound ataxia and cerebellar dysarthria is typical. SACS may easily be confused with several types of AR HSP. Friedreich’s ataxia (FRDA) is the most common AR ataxia, with a prevalence of 1 per 50 000 individuals (Pandolfo, 2009). It is much more common in some ethnic isolates. The most common disease causing molecular mechanism is a GAA trinucleotide repeat expansion in intron 1 of the frataxin (FXN) gene. The diseasecausing range of GAA repeats is quite broad, ranging from 70 to more than 1000, and patients with a higher number of repeats tend to have an earlier onset and more systemic complications, including scoliosis, diabetes mellitus, and hypertrophic cardiomyopathy. The

spinocerebellar tracts, dorsal columns, and pyramidal tracts are mostly affected, whereas the cerebellum and lower brainstem are involved to a lesser degree. The disorder usually manifests before adolescence, and most patients report impaired coordination with staggering gait and dysarthria. The combination of ataxia with bilateral Babinski sign, lower-extremity areflexia, and impaired vibration sensation is a characteristic finding in FRDA. Since the discovery of the genetic basis of FRDA, atypical and late-onset type of the disease has also been identified. Late-onset FRDA is defined as an onset after age 25 years and is usually caused by smaller GAA repeat expansions or point mutations in the frataxin gene (Pandolfo, 2009). Patients more often have isolated lower-limb spasticity and retained or even brisk reflexes, resembling spastic paraparetic clinical picture. Compared to classic FRDA, patients with the late-onset form tend to have a milder and more slowly evolving disease with fewer nonneurologic manifestations of the disease. Imaging of the spinal cord in spastic ataxias is variable and ranges from normal appearance to nonspecific spinal cord atrophy. Typical FRDA patients tend to have some degree of spinal cord atrophy and the cerebellum has an essentially normal appearance, while atypical, late-onset FRDA has more radiologic evidence of cerebellar atrophy (Pandolfo, 2009). Overall, the presence of cerebellar atrophy is the most helpful finding in this type of severe spasticity and its detection is an important diagnostic clue.

LEUKODYSTROPHIES Leukodystrophies are a heterogeneous group of inborn errors of metabolism characterized by disruption of myelin integrity leading to progressive demyelination and secondary axonal degeneration. Their neuroradiologic

778 P. HEDERA hallmark is diffuse or patchy white-matter disease or leuareflexia. Spinal cord axonopathy is the major pathologic koencephalopathy. Most metabolic defects causing leusubstrate of this condition and peripheral sensorimotor kodystrophies belong to families of lipid storage or neuropathy is both axonal and demyelinating. Spastic peroxisomal disorders (Hedera, 2011). Leukodystrophies paraparesis is progressive, and some patients may have a broad spectrum of clinical severity, ranging from develop dementia and visual problems later in the course a rapidly progressive course in the pediatric patients with of the disease. Very-long-chain fatty acid assay is also devastating cognitive, behavioral, visual, and motor diagnostic in this form of adrenoleukodystrophy. problems, to a more slowly progressive course with preFemale carriers can occasionally develop clinical probsentation in early or late adulthood (Tillema and Renaud, lems that are usually milder and have much later onset 2012). Residual enzymatic activity is the most important (Engelen et al., 2012). determinant of the age of onset and subsequent progresMRI findings in adrenomyelopathy range from sion, even though additional modifying factors also play essentially normal brain imaging to white-matter disa role. Some of the late-onset leukodystrophies comease. Similarly to the classic form, white-matter MRI monly mimic myelopathies because progressive spasticabnormalities are initially found in the occipital regions. ity is the most significant clinical problem and they will be However, adrenomyelopathy patients tend to have a the main focus of this chapter (Mu¨ller vom Hagen much lower extent of leukodystrophy and contrast et al., 2014). enhancement, a typical finding in the classic form Diagnosis of adult onset of leukodystrophies is often caused by inflammatory demyelination, is not present. a challenging clinical problem and here we will review The demyelination progresses anteriorly with the conditions that need to be differentiated from other involvement of the splenium of the CC and posterior types of myelopathies. The specific diagnosis is typically limbs of the internal capsule (Fig. 38.8A, B). Spinal cord made by biochemical assays of enzymatic activity and imaging may show a variable degree of atrophy, but this direct mutational analysis of causative genes has an is nonspecific, and brain MRI should raise a high degree increasing role in the diagnosis algorithms. The availabilof suspicion for this diagnosis (Fatemi et al., 2005). ity of several enzyme replacement therapies increases Krabbe’s disease (globoid cell leukodystrophy) is an the impetus for a prompt diagnosis of these disorders. AR neurogenetic disorder caused by deficiency in galacLeukodystrophies are commonly misdiagnosed as multitocerebrosidase (galactosylceramide b-galactosidase) ple sclerosis (MS) or vascular changes (leukoaraiosis). activity, with an estimated incidence rate of 1:100 000 Even though adult-onset leukodystrophies frequently (Tillema and Renaud, 2012). Similarly to other leukodyspresent as myelopathy, spinal cord imaging is usually trophies, classic early-onset variant and atypical forms unrevealing and, in contrast to MS, areas of demyelinwith an adult age of onset of the disease exist. ation are not detected in the spinal cord. Small-vessel A progressive demyelinating disease of the central nercerebrovascular disease can also be confused with leukovous system and peripheral nervous system is a hallmark dystrophies but the absence of vascular risk factors and of Krabbe’s disease. A classic form shows severe and of basal ganglia or brainstem involvement should favor rapidly fatal cognitive and motor deficits, seizures, an alternative diagnosis. Additionally, peripheral nerve and visual loss. The adult-onset form, accounting for involvement is another common feature of metabolic approximately 10% of all patients, typically presents as leukodystrophies that can facilitate the correct diagnosis spastic paraparesis and visual problems; ataxia and poly(Tillema and Renaud, 2012). neuropathy develop later. MRI signal abnormalities can Adrenoleukodystrophy is an X-linked disorder resulthave nonspecific appearance and affect deep white mating in severe white-matter disease (Engelen et al., 2012). ter, cerebellar and long tracts, including an isolated This is the most frequent leukodystrophy, with incidence hyperintensity of corticospinal tracts. White-matter ranging from 1:27 000 to 1:40 000. It is caused by mutachanges tend to have posterior predominance, but are tions in the ABCD1 gene that is involved in peroxisomal not as exclusive as in adrenomyelopathy and spare the import of fatty acids and the disruption of its function U fibers (Fig. 38.8C, D). The spinal cord may be atrophic, leads to the accumulation of very-long-chain fatty acids especially in the cervical segment (Debs et al., 2013). in the nervous system and adrenal cortex. Typical clinical Metachromatic leukodystrophy is an AR disorder presentation includes behavioral problems in affected affecting about 1:40 000 individuals. It is caused by a boys in the first decade, followed by progressive spasticdeficiency of arylsulfatase A enzyme or, very rarely, ity, visual disturbances, and severe cognitive decline. by mutations in the specific sphingolipid activator proAdrenomyeloneuropathy is an important variant of tein (saposin-B) gene (Tillema and Renaud, 2012). The adrenoleukodystrophy, with a progressive spastic paratypical clinical presentation of this entity is similar to paresis and peripheral polyneuropathy, resulting in a other early-onset leukodystrophies. Adult forms are characteristic combination of upgoing toes and rare, and slowly progressive spastic paraparesis,

HEREDITARY AND METABOLIC MYELOPATHIES

779

Fig. 38.8. (A, B) Adrenomyelopathy is characterized by white-matter abnormalities in the occipital regions (arrows). (C, D) Krabbe’s disease tends to have posterior predominance of demyelination with sparing of the U fibers (arrows). (E) Cerebrotendinous xanthomatosis is typical for white-matter disease in the cerebellum (arrows). (F) Adult-onset leukoencephalopathy with axonal spheroids and pigmented glia is characterized by frontal dominance of white matter disease (arrows).

combined with peripheral polyneuropathy, can be seen in these patients. They also frequently develop psychosis and other psychiatric problems. MRI abnormalities consist of a periventricular predominant leukoencephalopathy but, in contrast to Krabbe’s disease or adrenomyeloneuropathy, frontal predominance of white-matter signal changes is more typical. In early stages, subcortical U fibers are typically spared. This MRI pattern is the same for both early and adult forms of the disease. Cerebrotendinous xanthomatosis is an AR lipid storage disease caused by mutations in the CYP27A1 gene encoding the sterol 27-hydroxylase. This enzymatic deficit leads to impaired bile acid synthesis and the accumulation of cholestanol in the brain, tendons, and eyes, accounting for hallmark clinical features, such as

juvenile cataracts, tendon xanthomas, premature atherosclerosis, and progressive neurologic deficits. Elevated cholestanol plasma levels and elevated urinary bile alcohols further support the diagnosis, together with tendon xanthoma, typically on the Achilles tendon; however, some patients do not develop xanthomas. Pyramidal and extrapyramidal signs, ataxia, seizures, psychiatric disorders, dementia, and peripheral neuropathy are common and some patients may experience relatively isolated spastic myelopathy rather than the more complex neurologic phenotype. Increased T2 signal due to leukodystrophy can be seen in the periventricular regions, similarly to other causes of white-matter disease. However, additional involvement of the brainstem and especially leukodystrophy in the cerebellum surrounding the dentate nucleus are more specific MRI findings of

780

P. HEDERA

cerebrotendinous xanthomatosis (Fig. 38.8E). Patients with the spastic paraparesis variant have also MRI signal abnormalities in the lateral and dorsal columns of the spinal cord. Alexander’s disease and adult onset of leukoencephalopathy with neuroaxonal spheroids and pigmented glia are two relatively common leukodystrophies that can present with a diffuse spasticity that can be occasionally confused with possible myelopathy. However, the most common presenting symptoms include cognitive and behavioral problems pointing towards a diffuse brain disorder rather than myelopathy. Alexander’s disease is an AD condition caused by mutations in the glial fibrillary acidic protein gene and many patients have de novo mutations without a family history. Additional symptoms associated with Alexander’s disease include cerebellar ataxia, dysarthria, dysphagia, seizures, and a global cognitive decline. Late forms of the disease may have predominant spasticity with relatively preserved intellectual performance. The typical MRI findings include symmetric T2 hyperintensities with a frontal predominance, a periventricular rim of T2 hypointensity and T1 hyperintensity, and signal changes in the basal ganglia, thalamus, and brainstem. Spinal cord may be atrophic. The late-onset form may have a relative paucity of MRI changes. Adult-onset leukoencephalopathy with axonal spheroids and pigmented glia is also an AD disease caused by mutations in the colony-stimulating factor 1 receptor (CSF1R) gene. Gradual cognitive decline is the initial symptom in many patients and, very rarely, isolated spasticity is detected on the clinical examination. The common radiologic features include diffuse whitematter T2 hyperintensities with frontal dominance (Kleinfeld et al., 2013) (Fig. 38.8F). Additional involvement of periventricular and deep white matter and CC is also frequently seen. MRI of patients presenting with motor symptoms without any cognitive problems may show T2 hyperintensities involving the entire course of pyramidal tracts down to the brainstem and spinal cord.

METABOLIC MYELOPATHIES Metabolic myelopathies are caused either by deficiency of various nutritional factors playing crucial roles in myelin and axon integrity and maintenance or by direct toxic effects of various toxins and xenobiotics. Most of them have a subacute presentation, with the exception of a few toxic myelopathies that may have a more acute course. Preferential involvement of the dorsal columns with a variable damage of the corticospinal tracts is a typical clinical finding. Similarly to many leukodystrophies, peripheral nerves are also commonly affected as well, with a preferential damage to the dorsal root

ganglion cells, resulting in axon loss in the peripheral nerves and ascending dorsal column tracts. Dorsal column involvement may be the most clinically conspicuous finding, with severely impaired position and vibration sensations causing a syndrome of sensory ataxia. Additionally, optic nerve damage may be a part of the clinical presentation.

VITAMIN B12 DEFICIENCY (SUBACUTE COMBINED DEGENERATION) AND RELATED CONDITIONS Vitamin B12, also known as cobalamin, is a water-soluble compound found in many animal proteins, including eggs and milk. Additionally, several foods, such as breakfast cereals, are now fortified by vitamin B12. One possible cause of vitamin B12 deficiency is an insufficient intake due to dietary choices in a strict vegan diet without supplementation. Moreover, its absorption is complex and abnormalities in any step of this cascade are the other possible causes of deficiency. Cobalamin is cleaved from proteins in the stomach and an acidic milieu and the presence of pepsin are necessary for isolation of vitamin B12. It then binds to R-proteins secreted by salivary gland and gastric mucosa and this complex is cleaved in the small intestine. The final step includes the binding of free vitamin B12 to the cobalamin-binding protein called the intrinsic factor that is secreted by parietal cells in the gastric mucosa. This complex is actively absorbed in the ileal enterocytes through specific receptors. A passive absorption of free vitamin B12 also takes place, but it is very ineffective and accounts for about of 1% of absorbed amounts. Atrophic gastritis with achlorhydria or pepsin insufficiency, autoimmune gastritis with antiparietal antibodies, known as pernicious anemia that results in insufficient intrinsic factor levels, gastric resection, or bariatric procedures all impair vitamin B12 absorption and may be responsible for significant deficiency (Savage and Lindenbaum, 1995). Many of these processes have a higher prevalence with age and 15–20% of elderly people may have deficient levels of vitamin B12. Reduced acidity by proton pump inhibitors, H2 blockers, and metformin may also interfere with its absorption. Vitamin B12 is also stored in the liver and it may take up to 5 years to deplete hepatic storage after vitamin B12 malabsorption develops. Clinical symptoms of B12 deficiency are complex and neurologic and hematologic abnormalities are the core signs of this condition. Hematologic abnormalities include megaloblastic anemia with macrocytosis, thrombocytopenia, hypersegmented polymorphonuclear cells and anisocytosis, and poikilocytosis. Severe anemia may be associated with fatigue, dyspnea, and other constitutional symptoms. Gastric pain and weight loss may be seen in malabsorptive conditions.

HEREDITARY AND METABOLIC MYELOPATHIES Neurologic manifestation is also variable but myelopathic changes constituting subacute combined degeneration with pyramidal signs and dorsal column dysfunction are classic presentations of vitamin B12 deficiency (Kumar, 2012). Paresthesias are the most common initial symptom in patients with subacute combined degeneration. The myelopathy can be accompanied by a mild, predominantly demyelinating peripheral neuropathy. Autonomic neuropathy with prominent orthostatic hypotension and autonomic failure can be rarely seen as well. However, myelopathy is not the sole manifestation of B12 hypovitaminosis and nonspecific cognitive and neuropsychiatric problems may occur without spinal cord dysfunction. The diagnosis of B12 deficiency is laboratory-based and abnormally low plasma levels are diagnostic. However, plasma levels alone may be insufficient for the diagnosis because many patients with neurologic sequelae of B12 hypovitaminosis may actually have values within the low-normal range. Laboratory diagnosis is further supported by accumulation of compounds that are dependent on B12 catalytic activity for enzymes converting methylmalonyl coenzyme A to succinyl coenzyme A and the conversion of homocysteine to methionine. Plasma levels of methylmalonic acid and homocysteine levels should be checked when cobalamin levels are inconclusive and their elevation can confirm the metabolic consequences of insufficient vitamin B12 levels. The presence of macrocytic anemia should alert

781

clinicians to check vitamin B12 status, but definitive hematologic abnormalities may be absent in some patients with neurologic symptoms. Antibodies to intrinsic factor and antiparietal cell antibodies have low sensitivity and specificity. The Schilling test could be used to determine whether vitamin B12 absorption is intact, but this test has fallen out of favor and is not commonly used. Abnormal somatosensory evoked potentials may support the dysfunction of dorsal columns and visual evoked potentials may detect a subclinical optic neuropathy that can be seen in almost half of these patients. Replacement therapy is effective and intramuscular or sublingual administration of cobalamin can overcome gastrointestinal malabsorption. Supplementation can reverse hematologic abnormalities within 1 month but neurologic improvement may take up to 1 year. Moreover, many patients do not experience a complete recovery, but the neurologic deficit should stabilize on effective replacement therapy. Neurologic prognosis depends on the duration of untreated deficiency and the extent of neurologic deficits before the therapy. Monitoring of methylmalonic acid and homocysteine levels can also help to determine the efficacy of this therapy. Radiologic hallmarks reflect present pathology with signal changes seen in the dorsal columns and lateral corticospinal and spinothalamic tracts, with an inverted-V shape of affected areas of the dorsal spinal cord (Fig. 38.9A–D). Lower cervical and thoracic segments

Fig. 38.9. (A–D) Magnetic resonance imaging (MRI) changes seen in vitamin B12 deficiency are characterized by T2 signal changes seen in the dorsal columns and lateral corticospinal and spinothalamic tracts with an inverted V-shape of affected areas of the thoracic dorsal spinal cord (arrow). (E, F) Vitamin E deficiency MRI findings (a patient with inherited a-tocopherol transfer protein deficiency is shown) are similar, with an increased T2 signal in the posterior columns (arrow).

782 P. HEDERA of the spinal cord are most susceptible to vacuolar et al., 2006). However, neurologic deficit caused by degeneration and demyelination. Subacute demyelinisolated folate deficiency, especially resembling myelopation may cause reduced T1 signal and enhancement athy due to vitamin B12 deficiency, are very rare. Plasma folic levels are necessary to establish the diagnosis. after gadolinium contrast. Central myelin changes can Concomitant supplementation of deficient folate and be also seen in advanced cases, but spinal cord MRI is vitamin B12 should be avoided because correction of much more revealing in these conditions. folate before the rectification of cobalamin body Vitamin B12 levels can be affected by several other processes. Nitrous oxide, known as laughing gas, is an supplies may worsen neurologic deficit by diverting inhalational anesthetic commonly used in outpatient surmethylation toward DNA synthesis. This will correct gery and dentistry. It can be also abused as a street drug hematologic problems but aggravate myelin disruption. because of it euphoriant effects. Subacute combined degeneration can be induced by chronic exposure to nitrous oxide (Kinsella and Green, 1995). Moreover, VITAMIN E DEFICIENCY even a single exposure to this anesthetic in patients with subclinical and previously unrecognized vitamin B12 a-Tocopherol is the most active member of vitamin E lipid-soluble compounds. After absorption in the intesdeficiency can trigger symptoms of cobalamin defitine it forms chylomicrons that are taken up in the liver. ciency. The term anesthesia paraesthetica has been proposed for this mechanism. Nitrous oxide inactivates Hepatic processing of vitamin E includes binding to the methylcobalamin by oxidizing the cobalt moiety to an a-tocopherol transport protein and incorporation into inactive form. Plasma levels of vitamin B12 may be norvery-low-density lipoproteins (VLDL) that transport mal and the assay of methylmalonic acid is critical to vitamin E to target tissues. Most vitamin E is stored in diagnose this condition. Methionine supplementation the adipose tissue. It can be found in nuts, sunflower may be beneficial because this corrects the inhibition seeds, and whole grains and regular diet should provide of methionine synthetase. Imaging findings are similar sufficient recommended daily intake. Vitamin E defito a classic vitamin B12 deficiency. However, with the ciency usually results from malabsorption due to gastrofast onset of symptoms after exposure, MRI may be intestinal, pancreatic, or hepatic diseases (Kumar, 2012). unremarkable if patients are imaged in the early course Additionally, vitamin E deficiency may be caused by of the disease. various genetic defects affecting lipoprotein formation Genetic causes of vitamin B12 deficiency are very rare or a-tocopherol active transport. Bassen–Kornzweig and they present as severe megaloblastic anemia and syndrome (abetalipoproteinemia) is associated with the myelopathy, otherwise indistinguishable from acquired absence of VLDL and subsequent malabsorption synforms of cobalamin deficiency. Imerslund–Gra¨sbeck drome. Genetic defects in the microsomal triglyceride syndrome is a rare AR disease characterized by vitamin transfer protein gene and defects in chylomicron syntheB12 deficiency due to selective malabsorption of the sis and secretion are other examples of secondary causes vitamin. Mutations in the cubilin and amnionless genes, of vitamin E deficiency. a-Tocopherol transfer protein coding for the subunits of receptors absorbing vitamin deficiency is an AR disease resulting in a severe deficit B12, are the causes. Another phenocopy is caused by of vitamin E with essentially undetectable plasma levels mutations in the gastric intrinsic factor gene. Parenteral and without any other signs of malabsorption (Gordon, supplementation of vitamin B12 is necessary. 2001). Plasma vitamin E levels also depend on serum Folic acid is contained in green, leafy vegetables, citlipids, cholesterol, and VLDL levels, and serum vitamin rus fruits, and some animal products. Furthermore, it is E levels can be corrected for these factors by dividing added to many processed foods. Folate is absorbed in the serum vitamin E levels by the sum of serum triglycerides proximal small intestine and its deficiency is mostly and cholesterol. Genetically based disorders of vitamin caused by malabsorptive conditions. Other vitamins, E deficiency tend to have an earlier age of onset and such as B12, are also commonly deficient. The active more severe course. Vitamin E supplementation may form of folate, tetrahydrofolic acid, is essential in the partially reverse the clinical phenotype but high doses synthesis of purine, thymidine, and amino acids. Addiand the parenteral route may be necessary to bypass tionally, methionine synthetase requires methyletetrahydefects in vitamin E metabolism. drofolate as the methyl donor in the synthesis of Clinical symptoms of vitamin E deficiency include methionine. Hematologic problems of folate deficiency neurologic problems, retinitis pigmentosa, hematologic are very similar to vitamin B12 deficiency. Neurologic changes with acanthocytosis in the peripheral blood manifestations may resemble subacute combined degensmear, cardiomyopathy, and steatorrhea caused by eration, and peripheral polyneuropathy, nonspecific cogmalabsorption (Gordon, 2001). Clinical signs of hypovinitive and psychiatric problems can be also seen (Kumar taminosis A and K can be also seen in these patients.

HEREDITARY AND METABOLIC MYELOPATHIES Neurologic phenotype varies and, in patients with an early onset, especially due to a-tocopherol transfer, protein deficiency spastic ataxia is present and can be virtually undistinguishable from FRDA based on clinical signs alone. The clinical presentation of patients with a later age of onset resembles subacute combined degeneration with more pronounced peripheral neuropathy. Structural changes induced by vitamin E deficiency consist of axonal degeneration of large axons in the spinal cord with the predilection for dorsal columns with secondary demyelination. MRI is also very similar to vitamin B12 deficiency, with an increased signal in the posterior columns, especially in the lower cervical segments (Fig. 38.9E, F). Cerebellar atrophy can be seen in approximately one-half of patients with a-tocopherol transfer protein deficiency, but it is relatively rare in other causes of vitamin B12 deficiency.

COPPER DEFICIENCY Copper is a ubiquitous trace element present in many foods and its deficiency is caused by reduced absorption that takes place in the small intestine. ATPase-7A, which is mutated in Menkes’ disease, is required for an active transport to enterocytes. Part of the absorbed copper may be bound to intracellular metallothionein and the rest is taken up by hepatocytes. ATPase-7B, which is mutated in Wilson’s disease, plays a crucial role in copper trafficking. Copper homeostasis is maintained by biliary excretion and the rest of copper is secreted to plasma, where it is bound to ceruloplasmin. Gastric surgeries, including bariatric procedures are one common cause of copper deficiency (Kumar, 2006). The other mechanism is due to blocked absorption because

783

of elevated intracellular levels of metallothionein in enterocytes. Zinc is the most potent inducer of metallothionein and increased zinc intake in supplements or denture creams can induce copper deficiency (Hedera et al., 2009). Copper trapped in the enterocytes is lost through intestinal epithelial turnover into the stool, and is not effectively absorbed into portal venous blood. Plasma copper levels are very low and cerulopasmin usually reflects the amount of copper in plasma. Copper deficiency needs to be differentiated from Wilson’s disease that has also very low cerulopasmin. Twentyfour-hour urine copper excretion is abnormally high in Wilson’s disease and very low in copper deficiency. Copper deficiency is associated with hematologic and neurologic abnormalities (Kumar, 2006). Microcytic anemia and neutropenia reflect bone marrow suppression in acquired copper deficiency. A progressive myeloneuropathy, resembling subacute combined degeneration, is a result of a prolonged copper deficiency. Paresthesias, sensory ataxia, and variable degree of weakness are characteristic. However, neurologic problems may occur without obvious hematologic abnormalities. Effective copper supplementation usually promptly reverses blood cell count abnormalities and may arrest the progression of neurologic deficit. If zinc overload is present, cessation of excessive zinc consumption is also required. The degree of neurologic recovery is often incomplete with significant residual deficits. Imaging of the spinal cord is similar to other metabolic myelopathies, with areas of increased signal on T2-weighted images in the posterior and paramedian cord (Kumar et al., 2006). Demyelination and axonal degeneration may appear as areas of reduced signal on T1-weighted images (Fig. 38.10A–C).

Fig. 38.10. (A, B) Copper deficiency has a similar appearance to other metabolic myelopathies with increased signal on T2-weighted images in the posterior and paramedian cord (arrows). (C) Areas of demyelination have reduced signal on T1 images (arrow). (D) Magnetic resonance imaging of a patient from northwestern Zambia who developed acute paralysis caused by konzo 16 years ago with increased T2 signal in the central segments of cervical spinal cord (arrows).

784

P. HEDERA

TOXIC MYELOPATHIES Several food-derived toxins and adverse effects of some medications can cause toxic myelopathies. Their clinical presentation is very similar to metabolic causes, even though sometimes a more acute course can be seen. Nutritional toxic myelopathies are very rare in developed countries. Lathyrism is caused by consumption of food derived from Lathyrus sativus (grass pea or chickling pea) and an excitotoxic amino acid b-N-oxalyl-aminoL-alanine. Konzo is a similar disease but the onset of spastic paraparesis may be quite acute. It is caused by Manihot esculenta (cassava) consumption and toxicity is related to the amount of ingested cyanide. Neuroimaging studies are very limited and the appearance of the spinal cord may vary from normal to nonspecific signal abnormalities (Fig. 38.10D). Subacute myelo-optic neuropathy is a myeloneuropathy associated with optic nerve damage (Kumar, 2012). It was mostly diagnosed in Japan and toxicity from the antiparasitic drug clioquinol has been proposed. Clioquinol chelates copper and the pathophysiology of this condition may be similar to acquired copper deficiency. Indeed, MRI of spinal cord is undistinguishable from that of patients with copper deficiency. Myelopathy is an adverse effect of intrathecally administered cytarabine and methotrexate chemotherapy and increased signal on T2-weighted images in the central and posterior columns can be seen in these patients.

REFERENCES Aghakhanyan G, Martinuzzi A, Frijia F et al. (2014). Brain white matter involvement in hereditary spastic paraplegias: analysis with multiple diffusion tensor indices. AJNR Am J Neuroradiol 35: 1533–1538. Borasio G, Miller R (2001). Clinical characteristics and management of ALS. Semin Neurosci 21: 155–166. Branco LM, De Albuquerque M, De Andrade HM et al. (2014). Spinal cord atrophy correlates with disease duration and severity in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Frontotemporal Degener 15: 93–97. Cao L, Huang XJ, Chen CJ et al. (2013). A rare family with hereditary spastic paraplegia type 35 due to novel FA2H mutations: a case report with literature review. J Neurol Sci 329: 1–5. Chandran J, Ding J, Cai H (2007). Alsin and the molecular pathways of amyotrophic lateral sclerosis. Mol Neurobiol 36: 224–231. de Bot ST, Willemsen MA, Vermeer S et al. (2012). Reviewing the genetic causes of spastic ataxias. Neurology 79: 1507–1514. Debs R, Froissart R, Aubourg P et al. (2013). Krabbe disease in adults: phenotypic and genotypic update from a series of 11 cases and a review. J Inherit Metab Dis 36: 859–868. Engelen M, Kemp S, de Visser M et al. (2012). X-linked adrenoleukodystrophy (X-ALD): clinical presentation and

guidelines for diagnosis, follow-up and management. Orphanet J Rare Dis 7: 51. Erichsen AK, Koht J, Stray-Pedersen A et al. (2009). Prevalence of hereditary ataxia and spastic paraplegia in southeast Norway: a population-based study. Brain 132: 1577–1588. Fatemi A, Smith SA, Dubey P et al. (2005). Magnetization transfer MRI demonstrates spinal cord abnormalities in adrenomyeloneuropathy. Neurology 64: 1739–1745. Fink JK (2013). Hereditary spastic paraplegia: clinicopathologic features and emerging molecular mechanisms. Acta Neuropathol 126: 307–328. Fink JK, Hedera P (1999). Hereditary spastic paraplegia: heterogeneity and genotype-phenotype correlation. Semin Neurol 19: 301–309. Finsterer J, L€ oscher W, Quasthoff S et al. (2012). Hereditary spastic paraplegias with autosomal dominant, recessive, X-linked, or maternal trait of inheritance. J Neurol Sci 318: 1–18. Gordon N (2001). Hereditary vitamin-E deficiency. Dev Med Child Neurol 43: 133–135. Hedera P (2009). Hereditary spastic paraplegia or spinocerebellar ataxia? Not always as easy as it seems. Eur J Neurol 16: 887–888. Hedera P (2011). Hereditary myelopathies. Continuum 17: 800–815. Hedera P (2013). Recurrent de novo c.316G > A mutation in NIPA1 hotspot. J Neurol Sci 335: 231–232. Hedera P, Eldevik OP, Maly P et al. (2005). Spinal cord magnetic resonance imaging in autosomal dominant hereditary spastic paraplegia. Neuroradiology 47: 730–734. Hedera P, Peltier A, Fink JK et al. (2009). Myelopolyneuropathy and pancytopenia due to copper deficiency and high zinc levels of unknown origin II. The denture cream is a primary source of excessive zinc. Neurotoxicology 30: 996–999. Kinsella LJ, Green R (1995). ‘Anesthesia paresthetica’: nitrous oxide-induced cobalamin deficiency. Neurology 45: 1608–1610. Kleinfeld K, Mobley B, Hedera P et al. (2013). Adult-onset leukoencephalopathy with neuroaxonal spheroids and pigmented glia: report of five cases and a new mutation. J Neurol 260: 558–571. Kumar N (2006). Copper deficiency myelopathy (human swayback). Mayo Clin Proc 81: 1371–1384. Kumar N (2012). Metabolic and toxic myelopathies. Semin Neurol 32: 123–136. Kumar N, Ahlskog JE, Klein CJ et al. (2006). Imaging features of copper deficiency myelopathy: a study of 25 cases. Neuroradiology 48: 78–83. Mu¨ller vom Hagen J, Karle KN, Schu¨le R et al. (2014). Leukodystrophies underlying cryptic spastic paraparesis: frequency and phenotype in 76 patients. Eur J Neurol 21: 983–988. Nobue K, Iwata, Justin Y et al. (2011). White matter alterations differ in primary lateral sclerosis and amyotrophic lateral sclerosis. Brain 134: 2642–2655. Paisan-Ruiz C, Dogu O, Yilmaz A et al. (2008). SPG11 mutations are common in familial cases of complicated hereditary spastic paraplegia. Neurology 70: 1384–1389.

HEREDITARY AND METABOLIC MYELOPATHIES Pandolfo M (2009). Friedreich ataxia: the clinical picture. J Neurol 256 (suppl 1): 3–8. Rezende TJ, de Albuquerque M, Lamas GM et al. (2015). Multimodal MRI-based study in patients with SPG4 mutations. PLoS One 10: e0117666. Riverol M, Samaranch L, Pascual B et al. (2009). Forceps minor region signal abnormality “ears of the lynx”: an early MRI finding in spastic paraparesis with thin corpus callosum and mutations in the spatascin gene (SPG11) on chromosome 15. J Neuroimaging 19: 52–60. Savage DG, Lindenbaum J (1995). Neurological complications of acquired cobalamin deficiency: clinical aspects. Baillieres Clin Haematol 8: 657–678. Strong MJ, Gordon PH (2005). Primary lateral sclerosis, hereditary spastic paraplegia and amyotrophic lateral

785

sclerosis: discrete entities or spectrum? Amyotroph Lateral Scler Other Motor Neuron Disord 6: 8–16. Takahashi S, Makita Y, Okamoto N et al. (1997). L1CAM mutation in a Japanese family with X-linked hydrocephalus: a study for genetic counseling. Brain Dev 19: 559–562. Tartaglia MC, Rowe A, Findlater K et al. (2007). Differentiation between primary lateral sclerosis and amyotrophic lateral sclerosis: examination of symptoms and signs at disease onset and during follow-up. Arch Neurol 64: 232–236. Tillema JM, Renaud DL (2012). Leukoencephalopathies in adulthood. Semin Neurol 32: 85–94. Wang Y, Liu L, Ma L et al. (2014). Preliminary study on cervical spinal cord in patients with amyotrophic lateral sclerosis using MR diffusion tensor imaging. Acad Radiol 21: 590–596.

Handbook of Clinical Neurology, Vol. 136 (3rd series) Neuroimaging, Part II J.C. Masdeu and R.G. Gonza´lez, Editors © 2016 Elsevier B.V. All rights reserved

Chapter 39

The degenerative spine: pattern recognition and guidelines to image interpretation P.M. PARIZEL*, A.J.L. VAN HOYWEGHEN, A. BALI, J. VAN GOETHEM, AND L. VAN DEN HAUWE Department of Radiology, Antwerp University Hospital and University of Antwerp, Antwerp, Belgium

Abstract Degenerative disease of the spine, in the form of intervertebral disc degeneration and bony growth, causing osteophytes and impinging upon the spinal canal and neural foramina, is the most frequent disorder affecting the spine. In this chapter we first discuss briefly the indications for computed tomography or magnetic resonance imaging in suspected degenerative spine disease. We then describe changes of disc height, signal intensity, and disc contour with aging and repeated microtrauma, as well as the imaging techniques most appropriate to image them. A grading system for lumbar disc changes is provided. Stenosis of the canal and neural foramina is reviewed next, concluding with a description of degenerative changes affecting the vertebral endplates and bone marrow.

INTRODUCTION Imaging studies play an important role in the diagnostic workup of the patient with back pain (Maus, 2010). Accurate and consistent interpretation of the degenerative spine remains a daunting challenge, for radiologists and clinicians alike. Unfortunately, radiologic examinations are often used arbitrarily and without consideration of the clinical signs and symptoms. Moreover, in middle-aged and elderly patients, it can be excruciatingly difficult to separate abnormalities, which are clinically relevant, from physiologic changes, which are agerelated. There is no clear definition of what constitutes “normal aging.” The process of aging differs significantly among individuals; it depends on the genetic constitution but it is heavily influenced by acquired factors (e.g., lifestyle, obesity, smoking, physically strenuous work, psychosocial factors) (Skovron et al., 1994; Katz, 2006). When reporting radiologic studies, especially in elderly individuals, the objective should not be to provide an encyclopedic enumeration of all the abnormalities, but rather to identify changes that may bear on the clinical picture and to disclose or rule out a systemic disease,

such as metastases, as a cause of back or limb pain (Maus, 2010). A serious systemic etiology is found in only a small fraction of patients presenting with back pain to a primary care setting, and, almost all of these patients have risk factors or other symptoms (Chou et al., 2011). Therefore, it is recommended to avoid computed tomography (CT) or magnetic resonance imaging (MRI) in patients with nonspecific low-back pain (Chou et al., 2011), unless a serious underlying condition is suspected. The so-called “red flags” which warrant imaging include, but are not limited to, progressive neurologic deficit, fever, infection, age (>50 years or presacral fat and < CSF)
hypointense intranuclear cleft Hyperintense though < presacral fat Mildly hyperintense (slightly > outer fibers of annulus) Hypointense (¼ outer fibers of annulus fibrosus) Hypointense

7

Hypointense

Indistinct

8

Hypointense

Indistinct

3 4

CSF, cerebrospinal fluid.

but this should be discouraged because the term tear could be misinterpreted as implying a traumatic etiology, which is not always the case (Fardon et al., 2014). The relationship between the presence of radial fissures and aging remains controversial. In autopsy studies, radial fissures are common and their prevalence increases with age: 40% for the age group 50–60 years and 75% for the age group 60–70 years (Kieffer et al., 1969). Other authors have stated that radial fissures are found only in a minority of postmortem examinations in individuals older than 40 years, and are to be considered independent of age and degeneration (Moneta et al., 1994). Fissures can be classified by their orientation into three categories: 1.

2.

3.

Concentric fissures (Fig. 39.4) are circumferential separations or delaminations of the annular fibers parallel to the outer contour of the disc. Radial fissures originate from the nucleus pulposus and extend peripherally through the annulus. Transverse fissures (also known as “rim lesions”) (Fig. 39.5) are peripheral, horizontally oriented fissures limited to the outer annulus, near the attachment of annular fibers into the apophyseal bone (Sharpey’s fibers).

Relatively wide annular fissures, with stretching of the residual annular margin (sometimes including avulsion of an annular fragment), have been called annular gaps; this term is relatively new and not yet accepted as standard (Fardon et al., 2014).

State-of-the art high-resolution MRI of the lumbar spine is considered to be sufficiently accurate for the diagnosis of tears of the annulus fibrosus. Annular fissures induce a change in signal intensity on T2-weighted MR images. A peripheral focus of high signal intensity is sometimes present, and is called the high-intensity zone (HIZ) (Fig. 39.6). This is a descriptive term, which avoids the semantic controversy between fissures and tears as to the etiology of the lesion. The HIZ is brighter than the relatively hypointense nucleus pulposus of the degenerated disc and is clearly distinguishable from the annulus with its natural low signal. The clinical significance of an HIZ is still under debate. Some studies have found that the HIZ is correlated to histopathologic changes of the disc and pain on discography, indicating clinical significance (Aprill and Bogduk, 1992; Lam et al., 2000; Peng et al., 2006), whilst others state that, though HIZs occur more frequently in symptomatic patients, their prevalence in asymptomatic patients is too high to be of clinical use (Carragee et al., 2000; Pande et al., 2009). In the majority of cases, after intravenous injection of a gadolinium-based contrast, enhancement of annular tears occurs (96% of cases), both in symptomatic and in asymptomatic individuals (Stadnik et al., 1998). Annular tears enhance as a result of their vascularized granulation tissue, which may be part of the healing process (Ross et al., 1990). Though there is a statistically significant correlation between annular tears and disc protrusions, annular defects are frequently found in asymptomatic individuals (Jensen et al., 1994) and are not a reliable predictor

Fig. 39.4. Concentric annular fissure in a 50-year-old man. Sagittal (A) and axial (B) turbo spin echo T2-weighted images. A concentric fissure (arrow) represents a circumferential separation or delamination of the annular fibers parallel to the outer contour of the disc.

Fig. 39.5. Rim lesion in a 31-year-old man. Sagittal turbo spin echo T2-weighted image. A transverse fissure, also known as “rim lesion,” is a peripheral, linear, T2-hyperintense fissure, near the attachment of the annular fibers into the apophyseal bone (Sharpey’s fibers).

Fig. 39.6. A high-intensity zone (HIZ) in the L4–L5 intervertebral disc. The patient is a 46-year-old man with chronic back pain. A peripheral focus of high signal intensity is seen on this sagittal turbo spin echo T2-weighted image in the outer margin of the L4–L5 disc. Such a lesion is frequently referred to as a “high-intensity zone,” which is a purely descriptive term.

THE DEGENERATIVE SPINE: PATTERN RECOGNITION AND GUIDELINES of low-back pain in the future (Borenstein et al., 2001), so careful clinical correlation has to be performed. Other investigators have found a clear relationship between the presence of annular fissures and the lifetime occurrence of a history of low-back pain (Videman and Nurminen, 2004).

DISC CONTOUR CHANGES Radiologic reporting of contour changes of the intervertebral disc has suffered greatly from the use of incoherent and nonuniform nomenclature. Several attempts at classification have been made, either based on the anatomy and pathology, or on the cross-sectional morphology of the disc contour on imaging studies (Fig. 39.7). Recently, an updated version of the Lumbar Disc Nomenclature Recommendations (version 2.0) by the combined task forces of the North American Spine Society, the American Society of Spine Radiology, and the American Society of Neuroradiology has been published (Fardon et al., 2014). In Table 39.2, we provide a summary of these recommendations; however, in order to really understand the subtle differences, we strongly

793

recommend interested readers to study the full text of this article, which is freely available via PubMed. Nonfocal displacement of disc material, beyond the adjacent vertebral endplates and ring apophysis, and involving more than 25% of the disc circumference, is called “bulging” (Fig. 39.8). Disc bulging can be symmetric (circumferential disc displacement in all directions, involving up to 100% of the disc circumference) or asymmetric (with one side of the disc more afflicted than the other, as frequently seen in association with scoliosis). Bulging may or may not represent pathologic change, physiologic variant, or normal aging. The word bulging is merely a descriptive term, referring to a morphologic characteristic, and is not correlated with specific symptoms or etiology. Therefore, bulging should not be regarded as a type of hernia.

Herniation Herniation is a broad term used to describe a focal displacement of disc material beyond the contours of the intervertebral space, as delineated by the endplates (not including osteophytes). Focal is defined as a

Fig. 39.7. Nomenclature of disc contour changes. The blue line indicates the disc contour; the blue dotted line indicates the nucleus pulposus; the red lines indicate the contours of the vertebra. (A) Normal intervertebral disc; the disc does not extend beyond the vertebral endplate contour. (B) Asymmetrically bulging disc; the disc tissue extends beyond the edges of the ring apophysis. (C) Broad-based protruding disc; (D) extruded disc.

Table 39.2 Lumbar disc nomenclature recommendations (Fardon et al., 2014) General terminology Bulging

Herniation

Definition

Specification

A general displacement over >25% of the periphery of the disc A localized displacement,